Abstract
Background
Diabetes mellitus is a metabolic disorder resulting from a defect in insulin secretion, function, or both. Hyperglycaemia in non‐critically ill hospitalised people is associated with poor clinical outcomes (infections, prolonged hospital stay, poor wound healing, higher morbidity and mortality). In the hospital setting people diagnosed with diabetes receive insulin therapy as part of their treatment in order to achieve metabolic control. However, insulin therapy can be provided by different strategies (sliding scale insulin (SSI), basal‐bolus insulin, and other modalities). Sliding scale insulin is currently the most commonly used method, however there is uncertainty about which strategy provides the best patient outcomes.
Objectives
To assess the effects of SSI for non‐critically ill hospitalised adults with diabetes mellitus.
Search methods
We identified eligible trials by searching MEDLINE, Embase, LILACS, and the Cochrane Library. We searched the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) and ClinicalTrials.gov trial registers. The date of the last search for all databases was December 2017. We also examined reference lists of identified randomised controlled trials (RCTs) and systematic reviews, and contacted trial authors.
Selection criteria
We included RCTs comparing SSI with other strategies for glycaemic control in non‐critically ill hospitalised adult participants of any sex with diabetes mellitus.
Data collection and analysis
Two review authors independently extracted data, assessed trials for risk of bias, and evaluated the overall certainty of evidence utilising the GRADE instrument. We synthesised data using a random‐effects model meta‐analysis with 95% prediction intervals, if possible, or descriptive analysis, as appropriate.
Main results
Of 720 records screened, we included eight trials that randomised 1048 participants with type 2 diabetes (387 SSI participants and 615 participants in comparator groups were available for final analysis). We included non‐critically ill medical and surgical adults with the diagnosis of diabetes mellitus. The mean follow‐up time was measured by the mean length of hospital stay and ranged between five and 24 days. The mean age of participants was 44.5 years to 71 years.
Overall, we judged the risk of bias on the trial level as unclear for selection bias, high for outcome‐related performance and detection bias with regard to hypoglycaemic episodes, other adverse events, and mean glucose levels, and low for all‐cause mortality and length of hospital stay. Attrition bias was low for all outcome measures.
Six trials compared SSI with a basal‐bolus insulin scheme, three of which investigating 64% of all participants in this category also applying an SSI approach in the bolus comparator part. One trial had a basal insulin‐only comparator arm, and the remaining trial used continuous insulin infusion as the comparator. For our main comparison of SSI versus basal‐bolus insulin, the results were as follows. Four trials reported mortality data. One out of 268 participants in the SSI group (0.3%) compared with two out of 334 participants in the basal‐bolus group (0.6%) died (low‐certainty evidence). Severe hypoglycaemic episodes, defined as blood glucose levels below 40 mg/dL (2.2 mmol/L), showed a risk ratio (RR) of 0.22, 95% confidence interval (CI) 0.05 to 1.00; P = 0.05; 5 trials; 667 participants; very low‐certainty evidence. The 95% prediction interval ranged between 0.02 and 2.57. All nine severe hypoglycaemic episodes were observed among the 369 participants on basal‐bolus insulin (2.4%). The mean length of hospital stay was 0.5 days longer for the SSI group, 95% CI ‐0.5 to 1.4; P = 0.32; 6 trials; 717 participants; very low‐certainty evidence. The 95% prediction interval ranged between ‐1.7 days and 2.7 days. Adverse events other than hypoglycaemic episodes, such as postoperative infections, showed a RR of 1.16, 95% CI 0.25 to 5.37; P = 0.85; 3 trials; 481 participants; very low‐certainty evidence. The mean blood glucose levels ranged across basal‐bolus groups from 156 mg/dL (8.7 mmol/L) to 221 mg/dL (12.3 mmol/L). The mean blood glucose level in the SSI groups was 14.8 mg/dL (0.8 mmol/L) higher (95% CI 7.8 (0.4) to 21.8 (1.2); P < 0.001; 6 trials; 717 participants; low‐certainty evidence). The 95% prediction interval ranged between ‐3.6 mg/dL (‐0.2 mmol/L) and 33.2 mg/dL (1.8 mmol/L). No trial reported on diabetes‐related mortality or socioeconomic effects.
Authors' conclusions
We are uncertain which insulin strategy (SSI or basal‐bolus insulin) is best for non‐critically hospitalised adults with diabetes mellitus. A basal‐bolus insulin strategy in these patients might result in better short‐term glycaemic control but could increase the risk for severe hypoglycaemic episodes. The certainty of the body of evidence comparing SSI with basal‐bolus insulin was low to very low and needs to be improved by adequately performed, well‐powered RCTs in different hospital environments with well‐educated medical staff using identical short‐acting insulins in both intervention and comparator arms to compare the rigid SSI approach with flexible insulin application strategies.
Plain language summary
Sliding scale insulin for non‐critically ill hospitalised adults with diabetes mellitus
Review question
What are the effects of sliding scale insulin compared with other insulin therapies for non‐critically ill hospitalised adults with diabetes mellitus?
Background
There are several options to treat people with diabetes who have been admitted to the hospital. Both low glucose levels (hypoglycaemia) and high glucose levels (hyperglycaemia) might increase the risk of death and complications such as infections and longer hospital stay and therefore should be avoided. The most common method to treat hospitalised people with diabetes is sliding scale insulin therapy. The term 'sliding scale' refers to the increasing administration of the pre‐meal insulin dose based on the blood sugar level before the meal. For example, if a person has a blood glucose level between 140 mg/dL to 180 mg/dL, a short‐acting usual insulin dose could be 4 units insulin, and for a blood glucose level between 181 mg/dL and 220 mg/dL, it could be 6 units insulin. This kind of rigid insulin application usually fixes the amount of carbohydrate to be eaten at each meal but does not deliver insulin in a physiologic manner. It is therefore doubtful whether good glycaemic control (satisfactory blood glucose levels) can be achieved by sliding scale insulin in hospitalised diabetic people and whether this approach results in better outcomes in the long term. There are other insulin strategies like background (basal) insulin doses combined with flexible pre‐meal insulin (bolus insulin) according to what and how much people want to eat. Sliding scale insulin means strict adherence to a regular schedule of meals and physical activity, and people must follow a prescribed diet. Intensified insulin therapy (basal‐bolus strategy) allows flexible insulin doses according to physical activity, stress, and meal preferences, and under hospital conditions demands well‐trained medical staff. Sliding scale insulin is still widely used, and it remains unclear which insulin strategy is best to treat hospitalised diabetic people.
Study characteristics
We found eight randomised controlled trials (studies in which participants are assigned to one of two or more treatment groups in a random manner) where 387 people on sliding scale insulin therapy and 615 on mainly basal‐bolus insulin treatment with type 2 diabetes could be analysed. The average length of hospital stay was between five and 24 days. Six studies compared sliding scale insulin with basal‐bolus insulin; one study compared sliding scale insulin with basal insulin only; and the remaining study used continuous insulin infusion as the control group. The average age of participants was 44.5 years to 71 years.
The evidence is up to date as of December 2017.
Key results
The main comparison was between sliding scale insulin and basal‐bolus insulin therapy with the following results. Of the four studies reporting deaths, one out of 268 participants in the SSI group compared with two out of 334 participants in the basal‐bolus group died. Severe hypoglycaemic episodes, defined as blood glucose levels below 40 mg/dL, occurred in 5 per 1000 people in the sliding scale insulin groups compared with 24 per 1000 people in the basal‐bolus insulin groups. These data are uncertain because further analyses showed neither a positive nor a negative effect comparing both insulin strategies. The length of hospital stay in the sliding scale insulin groups compared with the basal‐bolus insulin groups was 0.5 days longer, again with further analyses indicating that this information is uncertain. The results for adverse events other than hypoglycaemic episodes, such as postoperative infections, did not indicate an advantage or disadvantage of either strategy. The average blood glucose level during hospital stay in the sliding scale groups was 14.8 mg/dL higher compared with the basal‐bolus groups. We are uncertain about these data because analyses showed neither a positive nor a negative effect comparing both insulin strategies. No trial reported on deaths caused by diabetes as such or socioeconomic effects like costs of the interventions or absence from work.
Certainty of the evidence
Overall, our confidence in the results for all analysed outcomes was low or very low mainly due to the low number of studies and participants and because the results were not precise, that is they could change in any direction once new studies are published.
Summary of findings
Summary of findings for the main comparison. Sliding scale insulin compared to basal‐bolus insulin for non‐critically ill hospitalised adults with diabetes mellitus.
| Sliding scale insulin compared to basal‐bolus insulin for non‐critically ill hospitalised adults with diabetes mellitus | ||||||
| Patient or population: non‐critically ill hospitalised adults with diabetes mellitus Setting: medical and surgical inpatients Intervention: sliding scale insulin Comparison: basal‐bolus insulin | ||||||
| Outcomes | Risk with basal‐bolus insulin | Risk with sliding scale insulin | Relative effect (95% CI) | No. of participants (trials) | Certainty of the evidence (GRADE) | Comments |
|
All‐cause mortality (N) Follow‐up: 5 to 9 days |
6 per 1000 | 3 per 1000 (0 to 29) | Peto OR 0.50 (0.05 to 4.85) | 602 (4) | ⊕⊕⊝⊝ lowa | 4/8 trials reported data (very low number of events) |
| Diabetes‐related morbidity | Not reported | |||||
|
Severe hypoglycaemic episodes (N) Follow‐up: 5 to 9 days |
24 per 1000 | 5 per 1000 (1 to 24) | RR 0.22 (0.05 to 1.00) | 667 (5) | ⊕⊝⊝⊝ very lowb | The 95% prediction interval ranged between 0.02 and 2.57 All 9 severe hypoglycaemic episodes were observed among 369 participants (2.4%) in the basal‐bolus insulin groups |
|
Length of hospital stay (days) Follow‐up: 5 to 24 days |
The mean length of hospital stay ranged across control groups from 5.3 days to 24.2 days | The mean length of hospital stay in the intervention groups was 0.5 days longer (0.5 days shorter to 1.4 days longer) | ‐ | 717 (6) | ⊕⊝⊝⊝ very lowc | The 95% prediction interval ranged between ‐1.7 days and 2.7 days |
|
Adverse events other than hypoglycaemic episodes (postoperative complications, mainly infectious) (N) Follow‐up: 6 to 24 days |
62 per 1000 | 72 per 1000 (15 to 332) | RR 1.16 (0.25 to 5.37) | 481 (3) | ⊕⊝⊝⊝ very lowd | |
|
Mean glucose level (mg/dL (mmol/L)) Follow‐up: 5 to 24 days |
The mean glucose level ranged across control groups from 156 mg/dL (8.7 mmol/L) to 221 mg/dL (12.3 mmol/L) | The mean glucose level in the intervention groups was 14.8 mg/dL (0.8 mmol/L) higher (7.8 mg/dL (0.4 mmol/L) higher to 21.8 mg/dL (1.2 mmol/L) higher) | ‐ | 717 (6) | ⊕⊕⊝⊝ lowe | The 95% prediction interval ranged between ‐3.6 mg/dL (‐0.2 mmol/L) and 33.2 mg/dL (1.8 mmol/L) |
| Socioeconomic effects | Not reported | |||||
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio | ||||||
| GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect. | ||||||
aDowngraded by two levels because of serious imprecision (small number of participants, small number of trials, uncommon event) (see Appendix 14). bDowngraded by one level because of trial limitations (risk of performance and detection bias) and by two levels because of serious imprecision (small number of participants, confidence interval consistent with benefit and harm) (see Appendix 14). cDowngraded by one level because of inconsistency (non‐consistent direction of effects) and by two levels because of serious imprecision (small median number of participants, confidence interval consistent with benefit and harm) (see Appendix 14). dDowngraded by one level because of trial limitations (risk of performance and detection bias); by one level because of inconsistency (non‐consistent direction of effects); and by one level because of imprecision (small number of trials and participants, confidence interval consistent with benefit and harm) (see Appendix 14). eDowngraded by one level because of trial limitations (risk of performance and detection bias) and by one level because of imprecision (small median number of participants) (see Appendix 14).
Background
Description of the condition
Diabetes mellitus (DM) is a metabolic disorder resulting from a defect in insulin secretion, function, or both. The consequence of DM is chronic hyperglycaemia, with subsequent effects on carbohydrate, fat, and protein metabolism. Long‐term DM complications include retinopathy, nephropathy, neuropathy, and an increased risk of cardiovascular disease (e.g. myocardial infarction, cerebrovascular events). Adequate glycaemic control in hospitalised people with DM represents a clinical challenge due to the wide range of treatment strategies and the morbidity/mortality risk that these participants present (Umpierrez 2002).
Diabetes mellitus is the most common chronic disease in low‐, middle‐, and high‐income countries. In 2000, the worldwide diabetes prevalence was 2.8% (for all age groups); in 2011, the global prevalence in the adult population was 8.3%. It is estimated that in 2030 diabetes prevalence will be 4.4% (for all age groups), affecting approximately 366 million people worldwide (Whiting 2011; Wild 2004).
People diagnosed with DM are frequently admitted for hospital care, with an estimated admission rate by year of about 25% to 30% (Moss 1999). At hospital admission, glycaemic control becomes a priority; there is an expected increase in serum glucose because of the metabolic stress (caused by the present illness or procedures) resulting in hyperglycaemia, increased infection risk, and poor wound healing. Insulin therapy regimens are currently under discussion (Braithwaite 2008; Buchleitner 2012). Most of the data comes from people with critical care illness; however, even in these patients the best treatment regimens are unclear (Farrokhi 2012; Friedrich 2010; Griesdale 2009; Klonoff 2011).
Description of the intervention
Four strategies are used to achieve glycaemic control in hospitalised diabetic people: basal insulin, prandial insulin, correction‐dose insulin, and sliding scale insulin (SSI). These strategies are used to maintain optimal glucose levels (fasting and postprandial). The main difference between these strategies is the intervention purpose: basal insulin administration inhibits hepatic glucose production and is mainly responsible for controlling fasting glucose; prandial (bolus or mealtime) insulin aims to control postprandial glucose levels by promoting glucose disposal into the muscles; and correction dose insulin is administered in addition to prandial insulin using an algorithm‐based treatment to provide an adjustment to carbohydrate intake (Lien 2011).
Sliding scale insulin (sometimes called correctional insulin) is poorly defined and is used in different ways: traditionally, the insulin prescription is administered according to the measured blood glucose levels and therefore constitutes a reactive response to hyperglycaemia; SSI is an intervention usually characterised by a subcutaneous administration (regular insulin or rapid‐acting insulin analogues), regardless of the timing of food intake, presence or absence of pre‐existing insulin administration, and sensitivity to insulin, without the acute illness consideration. The SSI scheme aims to maintain the glucose levels between 7.7 mmol/L (140 mg/dL) and 10 mmol/L (180 mg/dL) (Lien 2011; Umpierrez 2012). Some physicians even employ this approach using long‐acting insulin analogues or neutral protamine Hagedorn (NPH) insulin.
Sliding scale insulin is regularly prescribed as part of the medical treatment in hospitalised diabetic people, despite the lack of physiological congruence and strong evidence to support it.
Adverse effects of the intervention
There are many considerations regarding the use of SSI in the treatment of diabetic inpatients. This strategy has been widely criticised when used to obtain glycaemic control. Previous studies have revealed longer days to achieve metabolic control, greater variability of glucose levels (hypoglycaemia and hyperglycaemia), higher infection rate, and increased mortality (Baldwin 2005; Childs 2003; Umpierrez 2012).
Why it is important to do this review
Sliding scale insulin is one of the most frequently used therapies in the metabolic control of diabetic inpatients. However, this strategy may be associated with poor glycaemic control causing more complications (higher infection rate, increased mortality, prolonged hospital stay). Several non‐systematic review articles have addressed the SSI scheme (Miller 2011; Umpierrez 2007b). Given the potential public health relevance, we have systematically collected and examined all randomised controlled trials on SSI treatments for non‐critical diabetic hospitalised people in order to establish the risk and benefits of this therapy.
Objectives
To assess the effects of sliding scale insulin for non‐critically ill hospitalised adults with diabetes mellitus.
Methods
Criteria for considering studies for this review
Types of studies
We included randomised controlled clinical trials (RCTs).
Types of participants
Non‐critically ill hospitalised adult (18 years or older) diabetic people. We expected that most of the trials would have type 2 diabetes participants, because people with type 1 diabetes mellitus require a basal‐bolus insulin treatment. We excluded nursing home residents and people treated in intensive care units.
Diagnostic criteria
In order to be consistent with changes in the classification of and diagnostic criteria for diabetes over the years, the diagnosis had to be established using the standard criteria valid at the time of the trial commencing (e.g. ADA 1999; ADA 2008; WHO 1998). Ideally, the diagnostic criteria should have been described. Where necessary, we used the trial authors' definition of diabetes mellitus and planned to subject diagnostic criteria to a subgroup analysis.
Types of interventions
We planned to investigate the following comparisons of intervention versus comparator.
Intervention
Sliding scale insulin.
Comparators
Basal insulin only.
Basal‐bolus insulin (delivered by multiple subcutaneous injections or an insulin pump).
Correction dose insulin.
Other concomitant interventions had to be the same in the intervention and comparator groups to establish fair comparisons.
Minimum duration of intervention and follow‐up
We accepted any duration of intervention and follow‐up.
We defined extended follow‐up periods (also called open‐label extension studies) as follow up of participants once the original trial, as specified in the trial protocol, has been terminated. However, such studies are frequently of an observational nature, and we only planned to evaluate them for adverse events (Buch 2011; Megan 2012).
Summary of specific exclusion criteria
Nursing home residents and people treated in intensive care units.
Children and adolescents up to 18 years of age.
We did not exclude trials because one or several of our primary or secondary outcome measures were not reported in the publication. If no additional data were available, we planned to provide some basic information in a supplementary table.
Types of outcome measures
Primary outcomes
All‐cause mortality.
Diabetes‐related mortality.
Hypoglycaemic episodes.
Secondary outcomes
Length of hospital stay.
Adverse events other than hypoglycaemic episodes.
Mean blood glucose.
Socioeconomic effects.
Method and timing of outcome measurement
All‐cause mortality: defined as the total number of deaths from any cause, measured as in‐hospital mortality and as 30‐day all‐cause mortality.
Diabetes‐related mortality: defined as the total number of deaths due to myocardial infarction, ischaemic heart disease, stroke or sequelae of stroke, heart failure, cardiac arrest, peripheral vascular disease, kidney disease, hyperglycaemia, and hypoglycaemia, measured as in‐hospital mortality and as 30‐day mortality.
Hypoglycaemic episodes: defined as the number of blood glucose values less than 70 mg/dL or according to trial authors' definitions.
Length of hospital stay: defined as the duration of a single episode of hospitalisation and measured by subtracting the day of admission from the day of discharge.
Adverse events other than hypoglycaemia, e.g. hypokalaemia defined as a serum potassium concentration less than 3.5 mEq/L or injection site reactions, measured at any time during the hospital stay.
Mean blood glucose: as measured during participant hospital stay.
Socioeconomic effects, e.g. costs of the intervention, absence from work, medication consumption.
Search methods for identification of studies
Electronic searches
We searched the following sources from the inception of each database to the date given, placing no restrictions on language of publication.
Cochrane Central Register of Controlled Trials (CENTRAL) via the Cochrane Library (Issue 11 of 12, November 2017; searched on 22 December 2017).
MEDLINE Ovid SP (Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE; 1946 to present; searched on 22 December 2017).
Embase Ovid SP (1974 to 21 December 2017; searched on 22 December 2017).
LILACS (Latin American and Caribbean Health Science Information database; last update 8 December 2017; searched on 22 December 2017).
US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov) (searched on 22 December 2017).
World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) (www.who.int/trialsearch/); searched on 22 December 2017).
We applied a continuous MEDLINE (via Ovid SP) email alert service to identify newly published trials using the same search strategy as described for MEDLINE (Appendix 1).
Searching other resources
We attempted to identify other potentially eligible trials or ancillary publications by searching the reference lists of included trials, (systematic) reviews, meta‐analyses, and health technology assessment reports.
Data collection and analysis
Selection of studies
Two review authors (LECL, NDF) independently scanned the abstract, title, or both, of every record retrieved to determine which trials should be assessed further. We evaluated the full texts of all potentially relevant articles for inclusion in the review. Any discrepancies were resolved through consensus or by recourse to a third review author (AVH). We prepared a flow diagram of the number of trials identified and excluded at each stage in accordance with the PRISMA flow diagram of trial selection (Liberati 2009).
Data extraction and management
Two review authors (LECL, DGP) independently extracted key participant and intervention characteristics from the included trials. We reported data on efficacy outcomes and adverse events using standard data extraction forms from Cochrane Metabolic and Endocrine Disorders. Any disagreements were resolved by discussion or by consultation with a third review author (AVH) where necessary (for details see Table 2; Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8; Appendix 9; Appendix 10; Appendix 11; Appendix 12).
1. Overview of trial populations.
|
Trial ID (trial design) |
Intervention(s) and comparator(s) | Description of power and sample size calculation | Screened/eligible (N) | Randomised (N) | Analysed (N) | Finishing trial (N) | Randomised finishing trial (%) | Follow‐up |
|
Said 2013 (parallel RCT) |
I: SSI (regular insulin) | ‐ | ‐/63 | 22 | 22 | 22 | 100 | 6 days |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 21 | 21 | 21 | 100 | ||||
| C2: basal‐bolus regimen (glargine + glulisine) | 20 | 19 | 19 | 95 | ||||
| Total: | 63 | 62 | 62 | 98 | ||||
|
Umpierrez 2013 (parallel non‐inferiority RCT) |
I: SSI (regular insulin) | "...reasonable to assume that the SD of mean daily BG is bounded by 50 mg/dL. On the basis of two‐sample t‐test or Wilcoxon test (one sided, α = 0.05), 147 subjects were required for both the basal plus and the basal‐bolus groups to ensure 80% power to reject the non‐inferiority hypothesis, with Bonferroni correction applied to adjust for multiple comparisons across four time points" | 375 (210 medical and 165 surgical patients) | ‐ | 74 | ‐ | N/A | 6 days |
| C1: basal‐bolus regimen (glargine once daily + glulisine before meals + corrective doses of glulisine by SSI) | ‐ | 146a | ‐ | N/A | ||||
| C2: basal‐plus regimen (daily glargine + corrective doses of glulisine by SSI before meals) | ‐ | 133 | ‐ | N/A | ||||
| Total: | 375b | 353 | ‐ | N/A | ||||
|
Schroeder 2012 (cluster RCT) |
I: SSI (short‐acting human insulin) | ‐ | 65 | 30 | 30 | 30 | 100 | 7 days |
| C: basal‐bolus regimen (NPH and regular insulin) | 35 | 35 | 35 | 100 | ||||
| Total: | 65 | 65 | 65 | 100 | ||||
|
Umpierrez 2011 (parallel RCT) |
I: SSI (regular insulin) | "Assuming a within‐group standard deviation of 40 mg/dL and a‐error rate of 5% and a <10% attrition rate, we estimated that 104 subjects per group were needed to achieve 90% power" | 1405/297 | 116 | 107 | 107 | 92 | 9 days |
| C: basal‐bolus regimen (glargine + glulisine) | 118 | 104 | 104 | 88 | ||||
| Total: | 297/234c | 211 | 211 | 90 | ||||
|
Korytkowski 2009 (parallel RCT) |
I: SSI (regular insulin) + NPH for persistent glucose levels > 10 mmol/L | ‐ | 50 | 25 | 25 | 25 | 100 | 24 days |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | 25 | 25 | 25 | 100 | ||||
| Total: | 50 | 50 | 50 | 100 | ||||
|
Umpierrez 2007a (parallel RCT) |
I: SSI (regular insulin) | ‐ | 130 | 65 | 65 | 65 | 100 | 5 days |
| C: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | 65 | 65 | 65 | 100 | ||||
| Total: | 130 | 130 | 130 | 100 | ||||
|
Datta 2007 (parallel RCT) |
I: SSI (regular insulin) | ‐ | 81 | 39 | 39 | 39 | 100 | 5 days |
| C: basal regimen (glargine) + regular insulin if BG > 200 mg/dL | 42 | 42 | 42 | 100 | ||||
| Total: | 81 | 81 | 81 | 100 | ||||
|
Gustafson 2002d (parallel RCT) |
I: SSI (regular insulin) | ‐ | 50 | 25 | 25 | 25 | 100 | 1 day |
| C: continuous insulin infusion (regular insulin) | 25 | 25 | 25 | 100 | ||||
| Total: | 50 | 50 | 50 | 100 | ||||
| Grand total | All interventions | N/A | 387 | |||||
| All comparators | N/A | 615 | ||||||
| All interventions and comparators | 1048 | 1002 | ||||||
‐: denotes not reported
aAccording to table 1 (clinical characteristics of study patients) N = 144 participants. b22/375 (5.9%) participants were excluded from further analysis because they received < 24 hours of insulin treatment, were transferred to the intensive care unit, or received corticosteroid therapy. c23/234 (9.8%) participants were excluded after randomisation because they received < 24 hours of insulin treatment (N = 14); were treated with continuous insulin infusion during parenteral nutrition (N = 3); consent was withdrawn (N = 3); or surgery was cancelled (N = 3). 63/297 participants refused to participate at or after randomisation (information from trial flow diagram). dParticipants were subdivided into two groups categorised by the type of cardioplegia administration and mode of insulin treatment. Group I participants received a 5% dextrose‐based cardioplegic solution and continuous insulin infusion for glucose management. Group II participants received normal saline‐based cardioplegia and sliding scale insulin for glucose management.
BG: blood glucose; C: comparator; I: intervention; N/A: not applicable; NPH: neutral protamine Hagedorn insulin; RCT: randomised controlled trial; SD: standard deviation; SSI: sliding scale insulin
We provided information about potentially relevant ongoing trials, including trial identifier, in the Characteristics of ongoing studies table and in 'Matrix of study endpoints (publications and trial documents)' (Appendix 6). We attempted to retrieve the protocol for each included trial and planned to report in Appendix 6 primary, secondary, and other outcomes in comparison with data in publications.
We emailed the authors of all included trials to enquire as to whether they would be willing to answer questions regarding their trials. We presented the results of this survey in Appendix 13. We sought relevant missing information on the trial from the primary author(s) of the article, if possible.
Dealing with duplicate and companion publications
In the event of duplicate publications, companion documents, or multiple reports of a primary trial, we maximised the information yield by collating all available data, and used the most complete data set aggregated across all known publications. We listed duplicate publications, companion documents, multiple reports of a primary trial, and trial documents of included trials (such as trial registry information) as secondary references under the study ID of the included trial. Furthermore, we listed duplicate publications, companion documents, multiple reports of a trial, and trial documents of excluded trials (such as trials registry information) as secondary references under the study ID of the excluded trial.
Data from clinical trials registers
If data from included trials were available as study results in clinical trials registers such as ClinicalTrials.gov or similar sources, we made full use of this information and extracted the data. If there was also a full publication of the trial, we collated and critically appraised all available data. If an included trial was marked as a completed study in a clinical trials register but no additional information (study results, publication, or both) was available, we added this trial to the Characteristics of studies awaiting classification table.
Assessment of risk of bias in included studies
Two review authors (LECL, DGP) independently assessed the risk of bias of each included trial. Any disagreements were resolved by consensus or by consultation with a third review author (AVH) where necessary. If adequate information was unavailable from the trials, trial protocols, or both, we contacted the trial authors for missing data on 'Risk of bias' items.
We used the Cochrane 'Risk of bias' assessment tool to assess each 'Risk of bias' domain as low, high, or unclear risk of bias as described in the Cochrane Handbook for Systematic Reviews of Interventions (for details see Appendix 2) (Higgins 2011a; Higgins 2017).
Summary assessment of risk of bias
We presented a 'Risk of bias' graph and a 'Risk of bias' summary figure.
We distinguished between self reported, investigator assessed, and adjudicated outcome measures.
We considered the following self reported outcomes.
Adverse events other than hypoglycaemia.
Hypoglycaemic episodes.
We considered the following investigator‐assessed outcomes.
Adverse events other than hypoglycaemia.
All‐cause mortality.
Diabetes‐related mortality.
Hypoglycaemic episodes.
Length of hospital stay.
Mean blood glucose.
Socioeconomic effects.
Risk of bias for a trial across outcomes
Some 'Risk of bias' domains such as selection bias (sequence generation and allocation sequence concealment) affect the risk of bias across all outcome measures in a trial. In case of high risk of selection bias, we marked all endpoints investigated in the associated trial as being at high risk. Otherwise, we did not performed a summary assessment of the risk of bias across all outcomes for a trial.
Risk of bias for an outcome within a trial and across domains
We assessed the risk of bias for an outcome measure by including all entries relevant to that outcome (i.e. both trial‐level entries and outcome‐specific entries). We considered a judgement of low risk of bias to denote a low risk of bias for all key domains; unclear risk of bias to denote an unclear risk of bias for one or more key domains; and high risk of bias to denote a high risk of bias for one or more key domains.
Risk of bias for an outcome across trials and across domains
These are the main summary assessments that we incorporated into our judgements about the certainty of evidence in the 'Summary of findings' tables. We defined outcomes as at low risk of bias when most information came from trials at low risk of bias; unclear risk of bias when most information came from trials at low or unclear risk of bias; and high risk of bias when a sufficient proportion of information came from trials at high risk of bias.
Measures of treatment effect
When at least two trials were available for a comparison of a given outcome, we expressed dichotomous data as risk ratio (RR) with 95% confidence intervals (CIs). We planned to calculate time‐to‐event data as hazard ratio (HR) with 95% CI with the generic inverse variance method. We expressed continuous data reported on the same scale as mean difference (MD) with 95% CIs. For trials addressing the same outcome but using different outcome measure scales, we planned to use standardised mean differences (SMD) with 95% CI.
Unit of analysis issues
We took into account the level at which randomisation occurred, such as cross‐over trials, cluster‐randomised trials, and multiple observations for the same outcome. If more than one comparison from the same trial was eligible for inclusion in the same meta‐analysis, we either combined groups to create a single, pairwise comparison, or we appropriately reduced the sample size so that the same participants did not contribute multiply times (splitting the 'shared' group into two or more groups). Although the latter approach offers some solution for adjusting the precision of the comparison, it does not account for correlation arising from inclusion of the same set of participants in multiple comparisons (Higgins 2011b).
We planned to re‐analyse cluster‐randomised trials that did not appropriately adjust for potential clustering of participants in their analysis. Variance of the intervention effects would have been inflated by a design effect (DEFF). Calculation of a DEFF involves estimation of an intracluster correlation (ICC). We planned to obtain estimates of ICCs through contact with trial authors, or by imputing ICC values using either estimates from other included trials that reported ICCs or external estimates from empirical research (e.g. Bell 2013). We planned to examine the impact of clustering by performing sensitivity analyses.
Dealing with missing data
We attempted to obtain missing data from the authors of included trials. We carefully evaluated important numerical data such as screened, randomly assigned participants, as well as intention‐to‐treat and as‐treated and per‐protocol populations. We investigated attrition rates (e.g. dropouts, losses to follow‐up, withdrawals), and we critically appraised issues concerning missing data and imputation methods (e.g. last observation carried forward).
When included trials did not report means and standard deviations for outcomes and we were unable to obtain the requested information from trial authors, we imputed these values by estimating the mean and the variance from the median, the range, and the size of the sample (Hozo 2005).
We planned to investigate the impact of imputation on meta‐analyses by performing sensitivity analyses.
Assessment of heterogeneity
In the event of substantial clinical or methodological heterogeneity, we planned not to report trial results as the pooled effect estimate in a meta‐analysis.
We identified heterogeneity (inconsistency) by visually inspecting the forest plots and by using a standard Chi² test with a significance level of α = 0.1 (Deeks 2017). In view of the low power of this test, we also considered the I² statistic, which quantifies inconsistency across trials, to assess the impact of heterogeneity on the meta‐analysis (Higgins 2002; Higgins 2003).
Assessment of reporting biases
If we included 10 or more trials investigating a particular outcome, we used funnel plots to assess small‐trial effects. Several explanations can account for funnel plot asymmetry, including true heterogeneity of effect with respect to trial size, poor methodological design (and hence bias of small trials), and publication bias (Sterne 2017). We therefore interpreted results carefully (Sterne 2011).
Data synthesis
We planned to undertake (or display) a meta‐analysis only if we judged the participants, interventions, comparisons, and outcomes to be sufficiently similar to ensure an answer that would be clinically meaningful. Unless good evidence showed homogeneous effects across trials of different methodological quality, we primarily summarised data judged as at low risk of bias using a random‐effects model (Wood 2008). We interpreted random‐effects meta‐analyses with due consideration for the whole distribution of effects and presented a 95% prediction interval (Borenstein 2017a; Borenstein 2017b; Higgins 2009). A prediction interval requires at least three trials to be calculated and specifies a predicted range for the true treatment effect in an individual trial (Riley 2011). For rare events such as event rates below 1%, we planned to use the Peto odds ratio (OR) method, provided there was no substantial imbalance between intervention and comparator group sizes, and intervention effects were not exceptionally large. In addition, we performed statistical analyses according to the statistical guidelines presented in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2017).
Subgroup analysis and investigation of heterogeneity
We expected the following characteristics to introduce clinical heterogeneity and performed subgroup analyses to investigate interactions (Altman 2003).
Medical or surgical intervention.
Comorbid conditions.
Age, depending on data.
Time of follow‐up.
Type of insulin used.
Sensitivity analysis
We planned to perform sensitivity analyses to explore the influence of the following factors (when applicable) on effect sizes by restricting the analysis to the following.
Published trials.
Taking into account risk of bias, as specified in the Assessment of risk of bias in included studies section.
Use of the following filters: imputation, language of publication, source of funding (industry versus other), or country.
We planned to test the robustness of results by repeating analysis using different measures of effect size (i.e. RR, OR, etc.) and different statistical models (fixed‐effect and random‐effects models).
Certainty of evidence
We presented the overall certainty of the evidence for each outcome specified below according to the GRADE approach, which takes into account issues related not only to internal validity (risk of bias, inconsistency, imprecision, publication bias) but also to external validity, such as directness of results. Two review authors (LECL, DGP) independently rated the certainty of evidence for each outcome. Differences in assessment were solved by discussion or by consultation with a third review author (AVH).
Appendix 14 'Checklist to aid consistency and reproducibility of GRADE assessments' helps with standardisation of the 'Summary of findings' tables (Meader 2014). Alternatively, we planned to use GRADEpro GDT software and wanted to present evidence profile tables as an appendix (GRADEpro GDT 2015). We presented results for outcomes as described in the Types of outcome measures section. If meta‐analysis was not possible, we presented the results in a narrative format in the 'Summary of findings' table. We justified all decisions to downgrade the certainty of the evidence with footnotes, and made comments to aid the reader's understanding where necessary.
'Summary of findings' table
We presented a summary of the evidence in the Table 1. This provides key information about the best estimate of the magnitude of effect, in relative terms and as absolute differences for each relevant comparison of alternative management strategies, numbers of participants and trials addressing each important outcome, and a rating of overall confidence in effect estimates for each outcome. We created the Table 1 according to the methods described in the Cochrane Handbook for Systematic Reviews of Interventions and using the Review Manager 5 table editor (RevMan 2014; Schünemann 2017). We reported the following outcomes, listed according to priority.
All‐cause mortality.
Diabetes‐related mortality.
Severe hypoglycaemic episodes.
Length of hospital stay.
Adverse events other than hypoglycaemia.
Mean blood glucose.
Socioeconomic effects.
Results
Description of studies
For a detailed description of trials, see Table 2 and Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; and Characteristics of ongoing studies sections.
Results of the search
The electronic search strategies retrieved a total of 939 articles. After exclusion of duplicates, two review authors (LECL, DGP) independently assessed 720 titles, abstracts, or both. We obtained the full text of 45 articles/records, of which eight completed trials were included in our review. One potentially relevant trial is ongoing (NCT01143948), and two trials are awaiting classification (IRCT201311118307N2; ISRCTN55845280) An adapted PRISMA flow diagram of trial selection is presented in Figure 1.
1.
Trial flow diagram.
Included studies
For a detailed description of the characteristics of the included trials, see Characteristics of included studies and Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8; Appendix 9; Appendix 10; Appendix 11; Appendix 12). The following is a succinct overview.
Source of data
Eight trials met our inclusion criteria. These were published as peer‐reviewed articles in English.
Comparisons
All included trials investigated SSI therapy against other strategies for glycaemic control. Six trials used a basal‐bolus regimen (Korytkowski 2009; Said 2013; Schroeder 2012; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). Two were three‐arm trials (Said 2013; Umpierrez 2013). Umpierrez 2013 also employed a basal‐bolus correction regimen, and Said 2013 also used a mixed correction regimen (70/30 regular insulin plus neutral protamine Hagedorn (NPH) by means of SSI). One trial employed a basal insulin regimen by means of glargine plus regular insulin in 6% of cases (Datta 2007), and another trial used a continuous insulin infusion strategy (Gustafson 2002).
Overview of trial population
A total of 1048 participants with type 2 diabetes were randomised to SSI intervention and comparators; 387 participants in the SSI groups and 615 in the comparator groups were finally analysed. The sample size of randomised individuals per trial ranged from 50 to 375 participants.
Trial design
Seven trials had a parallel RCT design, and one trial was a cluster RCT (Schroeder 2012). The trials were published between 2002 and 2013. One trial was self funded (Said 2013), and five trials were partly or entirely sponsored by the pharmaceutical industry (Datta 2007; Korytkowski 2009; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013).
Settings
Six trials were conducted in the USA (Datta 2007; Gustafson 2002; Korytkowski 2009; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013), one trial in Israel (Schroeder 2012), and one in Egypt (Said 2013). Four trials evaluated surgical participants (Datta 2007; Gustafson 2002; Schroeder 2012; Umpierrez 2011); three trials evaluated medical inpatients (Korytkowski 2009; Said 2013; Umpierrez 2007a); and one trial evaluated both surgical and medical participants (Umpierrez 2013).
Participants
Two trials specified the ethnic background of their participants (Umpierrez 2007a; Umpierrez 2011). Two trials reported the duration of diabetes mellitus (Umpierrez 2011; Umpierrez 2013). The percentage of female participants per intervention/comparator group ranged between 26% and 86%. The mean age of participants ranged between 44.5 years and 71 years. The mean participants' glycated haemoglobin ranged between 7.4% and 8.9%. The mean body mass index (BMI) ranged between 26.9 kg/m² and 55 kg/m². Major exclusion criteria within the included trials were no history of diabetes, type 1 DM, corticosteroid therapy during hospitalisation, hyperglycaemic crisis, admission to an intensive care unit, severe heart failure or cardiogenic shock, clinically relevant hepatic disease, impaired renal function, gastrointestinal bleeding, pregnancy, inability to give informed consent, and any mental condition.
Interventions
The average daily insulin dose of SSI ranged between 8 units and 28 units. The average daily dose of basal‐bolus insulin ranged between 12 units and 44 units, and the average daily dose of basal‐plus insulin was 27 units. The mean duration of hospital stay ranged between 5 and 24 days.
Outcomes
Five trials explicitly stated a primary endpoint in the publication (Said 2013; Schroeder 2012; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). The most commonly defined primary outcomes in the publications were differences in glycaemic control, as measured by mean daily blood glucose concentration. Five trials were registered in ClinicalTrials.gov (Korytkowski 2009; Said 2013; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013).
Excluded studies
We excluded 29 of 45 articles/records after evaluation of the full publication/record. The main reasons for exclusion were wrong trial design, wrong population, wrong setting, and wrong comparison (for further details see Characteristics of excluded studies).
Risk of bias in included studies
For an overview of review authors' judgements about each 'Risk of bias' item for individual trials and across all trials, see Figure 2 and Figure 3.
2.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials (blank cells indicate that the particular outcome was not investigated in some trials).
3.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial (blank cells indicate that the trial did not report that particular outcome).
Allocation
Two trials had a low risk of selection bias (Umpierrez 2011; Umpierrez 2013), and six trials did not provide enough information to adequately judge selection bias (Datta 2007; Gustafson 2002; Korytkowski 2009; Said 2013; Schroeder 2012; Umpierrez 2007a).
Blinding
The majority of trials showed a high risk of performance and detection bias for the outcome measures hypoglycaemic episodes, adverse events other than hypoglycaemic episodes, and mean glucose levels.
Incomplete outcome data
We rated all trials as at low risk of attrition bias if all data for the associated outcome measures were reported. Umpierrez 2011 excluded 10% of randomised participants from the analysis, and Umpierrez 2013 excluded 6% of randomised participants from the analysis (for details see Table 2).
Selective reporting
We rated one trial as at high risk of reporting bias (Schroeder 2012).
Other potential sources of bias
We rated one trial as at unclear risk of other bias because it did not use adequate statistical analyses for its cluster‐randomised design (Schroeder 2012).
Effects of interventions
See: Table 1
Most trials compared an SSI regimen with a basal‐bolus insulin regimen (Korytkowski 2009; Said 2013; Schroeder 2012; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). Three of these trials, investigating 64% of all participants for this comparison, also applied the SSI approach in the bolus comparator part (Korytkowski 2009; Umpierrez 2007a; Umpierrez 2013), thus essentially reducing the comparison to the evaluation of additional basal insulin administration. Umpierrez 2013 had a second comparator arm ("basal‐plus"), which did not differ in a clinically relevant way from the basal‐bolus arm of the trial. We used only the results of the basal‐bolus comparator part of this trial in our review.
Baseline characteristics
For details of baseline characteristics, see Appendix 4 and Appendix 5.
SSI versus basal‐bolus insulin
Primary outcomes
All‐cause mortality
Four trials reported data on death (Said 2013; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). One of 268 participants in the SSI group (0.3%) compared with two of 334 participants in the basal‐bolus group (0.6%) died (Peto odds ratio 0.50, 95% confidence interval (CI) 0.05 to 4.85; 4 trials; 602 participants; low‐certainty evidence; Analysis 1.1).
1.1. Analysis.
Comparison 1 Sliding scale insulin versus basal‐bolus insulin regimen, Outcome 1 All‐cause mortality.
Diabetes‐related mortality
No trial reported on diabetes‐related mortality.
Hypoglycaemic episodes
Non‐severe hypoglycaemia
Non‐severe hypoglycaemia was usually defined by trial authors as a blood glucose level below 60 mg/dL (3.3 mmol/L) or 70 mg/dL (3.9 mmol/L). Schroeder 2012 reported that "no significant difference was noted in hypoglycaemic rates (blood glucose level below 70 mg/dL) between the two groups (P = 0.6)". Korytkowski 2009 reported hypoglycaemia frequency as the percentage of patient‐days with one or more event (SSI compared with basal‐bolus insulin: 4.8 days versus 2.7 days, with P = 0.34) and percentage of all glucose values less than 3.9 mmol/L (70 mg/dL) (SSI compared with basal‐bolus insulin: 1.1 blood glucose measures versus 1.3 blood glucose measures, P = 0.35).
The risk ratio (RR) for non‐severe hypoglycaemic episodes comparing SSI with basal‐bolus insulin was 0.25 (95% CI 0.11 to 0.59 in favour of SSI; P = 0.001; 3 trials; 561 participants; Analysis 1.2). Overall, nine non‐severe hypoglycaemic episodes occurred in 246 participants on SSI (3.7%) compared with 49 non‐severe hypoglycaemic episodes in 315 participants on basal‐bolus insulin (15.6%). Calculation of the 95% prediction interval did not reveal a meaningful interval. Performance bias and detection bias were high in the majority of the included trials reporting this outcome.
1.2. Analysis.
Comparison 1 Sliding scale insulin versus basal‐bolus insulin regimen, Outcome 2 Non‐severe hypoglycaemic episodes.
Comparing the subgroup of medical with surgical inpatients did not show a statistically significant interaction effect of subgroup differences. There were not enough data to analyse the other prespecified subgroups.
Severe hypoglycaemia
Five trials reported on severe hypoglycaemic episodes (Said 2013; Schroeder 2012; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). Severe hypoglycaemia was defined by trial authors as blood glucose levels below 40 mg/dL (2.2 mmol/L), Said 2013; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013, or as hypoglycaemia requiring intravenous glucose administration (Schroeder 2012). The RR for severe hypoglycaemic episodes comparing SSI with basal‐bolus insulin was 0.22 (95% CI 0.05 to 1.00; P = 0.05; 5 trials; 667 participants; very low‐certainty evidence; Analysis 1.3). All nine severe hypoglycaemic episodes were observed among the 369 participants on basal‐bolus insulin (2.4%). The 95% prediction interval ranged between 0.02 and 2.57. The fixed‐effect model showed a RR of 0.20 (95% CI 0.05 to 0.87). Elimination of the cluster RCT, Schroeder 2012, showed a statistically non‐significant effect estimate for both the random‐effects and the fixed‐effect model (RR 0.22, 95% CI 0.04 to 1.24 and RR 0.20, 95% CI 0.04 to 1.03, respectively). Performance bias and detection bias were high in the majority of the included trials reporting this outcome.
1.3. Analysis.
Comparison 1 Sliding scale insulin versus basal‐bolus insulin regimen, Outcome 3 Severe hypoglycaemic episodes.
Secondary outcomes
Length of hospital stay
Six trials reported data on length of hospital stay (Korytkowski 2009; Said 2013; Schroeder 2012; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). The mean length of hospital stay ranged across basal‐bolus groups from 5.3 days to 24.2 days. The mean length of hospital stay in the SSI groups was 0.5 days longer (95% CI ‐0.5 to 1.4; P = 0.32; 6 trials; 717 participants; very low‐certainty evidence; Analysis 1.4). The 95% prediction interval ranged between ‐1.7 days and 2.7 days. Elimination of the cluster RCT, Schroeder 2012, showed a mean difference between intervention and comparator groups of ‐0.06 days (95% CI ‐0.91 to 0.80).
1.4. Analysis.
Comparison 1 Sliding scale insulin versus basal‐bolus insulin regimen, Outcome 4 Length of hospital stay.
Comparing the subgroup of medical with surgical inpatients did not show a statistically significant interaction effect of subgroup differences. There were not enough data to analyse the other prespecified subgroups.
Adverse events other than hypoglycaemic episodes (such as postoperative infections)
Four trials reported on adverse events other than hypoglycaemic episodes (Korytkowski 2009; Schroeder 2012; Umpierrez 2011Umpierrez 2013).
Schroeder 2012 reported that "there was no difference in the postoperative complications between the groups. No difference in operative‐related and unrelated complications between groups was noted".
The RR for postoperative complications comparing SSI with basal‐bolus insulin was 1.16 (95% CI 0.25 to 5.37; P = 0.85; 3 trials; 481 participants; very low‐certainty evidence; Analysis 1.5). Overall, 20 complications among 206 participants on SSI (9.7%) occurred compared with 17 complications among 275 participants on basal‐bolus insulin (6.2%). Calculation of the 95% prediction interval did not reveal a meaningful interval. Performance bias and detection bias were high in the majority of the included trials reporting this outcome, and selection bias was high for Schroeder 2012.
1.5. Analysis.
Comparison 1 Sliding scale insulin versus basal‐bolus insulin regimen, Outcome 5 Adverse events other than hypoglycaemic episodes (infectious complications).
Mean blood glucose levels
Six trials reported on mean glucose level during hospital stay (Korytkowski 2009; Said 2013; Schroeder 2012; Umpierrez 2007a; Umpierrez 2011; Umpierrez 2013). The mean blood glucose level ranged across basal‐bolus groups from 156 mg/dL (8.7 mmol/L) to 221 mg/dL (12.3 mmol/L). The mean blood glucose level in the SSI groups was 14.8 mg/dL (0.8 mmol/L) higher (95% CI 7.8 (0.4) to 21.8 (1.2); P < 0.001; 6 trials; 717 participants; low‐certainty evidence; Analysis 1.6). The 95% prediction interval ranged between ‐3.6 mg/dL (‐0.2 mmol/L) and 33.2 mg/dL (1.8 mmol/L). The cluster RCT, Schroeder 2012, did not have a substantial impact on the effect size (data not shown).
1.6. Analysis.
Comparison 1 Sliding scale insulin versus basal‐bolus insulin regimen, Outcome 6 Mean glucose levels.
Comparing the subgroup of medical with surgical inpatients did not show a statistically significant interaction effect of subgroup differences. There were not enough data to analyse the other prespecified subgroups.
Four trials reported on hyperglycaemia, which was variably defined (Datta 2007; Said 2013; Umpierrez 2007a; Umpierrez 2011). Datta 2007 reported that 16% of SSI‐treated participants compared with 6% of basal‐bolus insulin‐treated participants had blood glucose levels greater than 200 mg/dL. Said 2013 reported that 68% of SSI‐treated participants compared to 53% of basal‐bolus insulin‐treated participants had blood glucose levels greater than 300 mg/dL. Blood glucose levels greater than 240 mg/dL were reported in 9% of SSI‐treated participants in Umpierrez 2007a and in 12% of SSI‐treated participants in Umpierrez 2011.
All trials reported that the mean total daily dose of insulin in the basal‐bolus group was greater than in the SSI group.
Socioeconomic effects
No trial reported on socioeconomic effects.
SSI versus basal insulin
Only one trial compared an SSI regimen with a basal insulin regimen (Datta 2007). This trial, which involved 81 participants, did not provide information on all‐cause mortality, diabetes‐related mortality, or socioeconomic effects. One of 39 participants in the SSI group (2.6%) compared with two of 42 participants in the basal insulin group (4.8%) reported a non‐severe hypoglycaemic episode (Analysis 2.1). Mean length of hospital stay was the same in both the intervention and the comparator group (3.8 days; Analysis 2.2). Adverse events other than hypoglycaemic episodes (postoperative infections) were reported in two participants in both the intervention and the comparator group (Analysis 2.3). The mean blood glucose level in the SSI group was 20 mg/dL higher than in the basal insulin group (Analysis 2.4).
2.1. Analysis.
Comparison 2 Sliding scale insulin versus basal insulin regimen, Outcome 1 Non‐severe hypoglycaemic episodes.
2.2. Analysis.
Comparison 2 Sliding scale insulin versus basal insulin regimen, Outcome 2 Length of hospital stay.
2.3. Analysis.
Comparison 2 Sliding scale insulin versus basal insulin regimen, Outcome 3 Adverse events other than hypoglycaemic episodes.
2.4. Analysis.
Comparison 2 Sliding scale insulin versus basal insulin regimen, Outcome 4 Mean glucose levels.
One trial compared SSI with a continuous insulin infusion during coronary artery bypass surgery in 50 participants (Gustafson 2002). The only outcome measure of relevance for this review was mean blood glucose levels, which were 110 mg/dL less in the post‐cardiopulmonary bypass period in the SSI group compared with the continuous insulin infusion group (Analysis 3.1). There was also a difference in the cardioplegic solution between the groups: the continuous insulin infusion group received a dextrose 5%‐based cardioplegic solution, and the SSI group received normal 0.9% saline, limiting the interpretation of the results of this trial.
3.1. Analysis.
Comparison 3 Sliding scale insulin versus continuous insulin infusion, Outcome 1 Mean glucose levels.
Subgroup analyses
With the exception of medical versus surgical inpatients for some outcome measures, we did not perform subgroup analyses as data were insufficient to estimate effects in various subgroups.
Sensitivity analyses
With the exception of comparing the random‐effects model with the fixed‐effect model for some outcome measures, we did not perform sensitivity analyses due to insufficient data.
Assessment of reporting bias
We did not draw funnel plots due to the limited number of trials (N = 8).
Ongoing studies
We found one ongoing trial, NCT01143948, and two trials awaiting classification, IRCT201311118307N2; ISRCTN55845280 (for details see Characteristics of ongoing studies and Characteristics of studies awaiting classification sections).
The ongoing trial is investigating an SSI regimen compared with a basal‐bolus insulin regimen in approximately 45 participants with type 2 diabetes and liver cirrhosis.
Two trials awaiting classification, with a total of approximately 320 participants, compared an SSI regimen with a basal‐bolus insulin regimen in diabetic people with cardiovascular diseases and diabetic people with acute medical and surgical conditions.
Discussion
Summary of main results
This systematic review analysed the evidence from published RCTs of SSI therapy for the treatment of non‐critically ill adults with diabetes mellitus. We included eight trials randomising 1048 participants with type 2 diabetes. The main comparison with SSI was basal‐bolus insulin therapy. A basal‐bolus insulin strategy might result in better short‐term glycaemic control, but might also result in an increase in severe hypoglycaemic episodes. We are uncertain about the effects of SSI and basal‐bolus insulin on length of hospital stay, postoperative complications, and mean blood glucose levels. No studies reported on diabetes‐related mortality or socioeconomic effects.
Overall completeness and applicability of evidence
Most trials analysed in our review were conducted in the USA, and younger adult diabetic participants were underrepresented in the trials. Our inclusion criteria specified surgical and medical patients, and both were represented, reflecting the general practice in the non‐critical hospital environment. However, most trials did not include people with newly diagnosed diabetes, people who were treated with insulin before hospitalisation, or people on corticosteroid therapy, who all form an important part of the hospitalised population with hyperglycaemia. The total daily dose of insulin was lower in the SSI groups compared with the basal‐bolus insulin groups, probably reflecting that the SSI regimen on admission remained unchanged throughout the hospitalisation period. It is therefore unclear whether the results of this review are associated with the SSI regimen per se or a suboptimal insulin dosing in the SSI intervention groups, which might be an attribute of the rigid SSI treatment approach.
Quality of the evidence
We assessed the certainty of the evidence as low to very low mainly because of trial limitations (especially performance bias and detection bias), imprecision (low number of participants and trials, wide prediction intervals), and inconsistency (non‐consistent direction of effects) for most of the outcome measures.
Potential biases in the review process
We attempted to contact the trial authors to obtain additional information (Appendix 13), as well as information about potential admissions of non‐critically ill participants to intensive care units, but did not receive an answer from the authors of three included trials. We therefore cannot exclude the possibility that in some trials some participants were also treated in critical care. Furthermore, we did not systematically search the grey literature.
Agreements and disagreements with other studies or reviews
Lee 2015 published a meta‐analysis of SSI treatment for blood glucose control. The review authors included 11 trials in their meta‐analysis but did not discriminate between critical and non‐critical care participants. Gómez Cuervo 2016 published a meta‐analysis of adverse events related to basal‐bolus insulin therapy compared with SSI. The review authors included five trials combining observational studies and RCTs in their meta‐analysis. Christensen 2017 published a meta‐analysis of basal‐bolus insulin regimens compared to other insulin regimens. The review authors included five trials using basal‐bolus insulin as the main intervention in their meta‐analysis. The overall results of these reviews were similar to our results, indicating that SSI therapy resulted in lower hypoglycaemia rates and higher mean glucose levels compared with non‐SSI therapies.
Authors' conclusions
Implications for practice.
Based on low‐ to very low‐certainty evidence, our results for non‐critically ill hospitalised adults with diabetes mellitus in the medical and surgical in‐hospital setting showed that a basal‐bolus insulin strategy that applied higher total daily doses of insulin than the sliding scale insulin regimen might result in better short‐term glycaemic control but might also cause an increase in severe hypoglycaemic episodes. We are uncertain whether sliding scale insulin increases or decreases mortality, length of hospital stay, and adverse events other than hypoglycaemic episodes (such as infectious complications).
Implications for research.
Future randomised controlled trials should adequately define the clinical conditions of hospitalised people with diabetes mellitus and the associated appropriate treatment choice. Multinational trials would improve the external validity of trial results. There is a need for adequately performed, well‐powered randomised controlled trials in different hospital environments with well‐educated medical staff. Identical short‐acting insulins in both intervention and comparator arms should be used to compare the rigid sliding scale insulin approach with flexible insulin application strategies.
Notes
Portions of the Background and Methods sections, Appendices, Additional tables, and Figures 1 to 3 of this review are based on a standard template established by the Cochrane Metabolic and Endocrine Disorders Group.
Acknowledgements
We thank our families for their support, our friends for their company, every member of this review author team and especially the editorial team, without their support this review would not be possible. Thanks to them we have learned and grown as people. The search strategies where developed by Maria‐Inti Metzendorf.
Appendices
Appendix 1. Search strategies
| Cochrane Central Register of Controlled Trials (CENTRAL) (Cochrane Library) |
| 1. [mh "Diabetes Mellitus"] 2. diabet*:ti,ab,kw 3. ("IDDM" or "NIDDM" or "MODY" or T1D* or T2D*):ti,ab,kw 4. ((("non insulin*" or noninsulin*) next depend*) or "non insulindepend*" or noninsulindepend*):ti,ab,kw 5. ((insulin* next depend*) or insulindepend*):ti,ab,kw 6. [mh "Diabetes Insipidus"] 7. (diabet* next "insipidus"):ti,ab,kw 8. {or #1‐#5} 9. #6 or #7 10. #8 not #9 11. (correct* near/4 insulin*):ti,ab,kw 12. ("sliding" near/4 scale*):ti,ab,kw 13. #11 or #12 14. #10 and #13 |
| MEDLINE (Ovid SP) |
| 1. exp Diabetes Mellitus/ 2. diabet*.tw. 3. (IDDM or NIDDM or MODY or T1D* or T2D*).tw. 4. (non insulin* depend* or noninsulin* depend* or non insulindepend* or noninsulindepend*).tw. 5. (insulin* depend* or insulindepend*).tw. 6. exp Diabetes Insipidus/ 7. diabet* insipidus.tw. 8. or/1‐5 9. 6 or 7 10. 8 not 9 11. (correct* adj3 insulin*).tw. 12. (sliding adj3 scale*).tw. 13. 11 or 12 14. 10 and 13 [15‐25: Cochrane Handbook 2008 RCT filter ‐ sensitivity maximizing version] 15. randomized controlled trial.pt. 16. controlled clinical trial.pt. 17. randomi?ed.ab. 18. placebo.ab. 19. drug therapy.fs. 20. randomly.ab. 21. trial.ab. 22. groups.ab. 23. or/15‐22 24. exp animals/ not humans/ 25. 23 not 24 26. 14 and 25 |
| Embase (Ovid SP) |
| 1. exp Diabetes Mellitus/ 2. diabet*.tw. 3. (IDDM or NIDDM or MODY or T1D* or T2D*).tw. 4. (non insulin* depend* or noninsulin* depend* or non insulindepend* or noninsulindepend*).tw. 5. (insulin* depend* or insulindepend*).tw. 6. exp Diabetes Insipidus/ 7. diabet* insipidus.tw. 8. or/1‐5 9. 6 or 7 10. 8 not 9 11. (correct* adj3 insulin*).tw. 12. (sliding adj3 scale*).tw. 13. 11 or 12 14. 10 and 13 [15:Wong 2006"sound treatment studies" filter – BS version] 15. random*.tw. or clinical trial*.mp. or exp health care quality/ 16. 14 and 15 17. limit 16 to embase |
| LILACS (IAHx) |
| ((diabet$ OR insulin$) AND (scale$ OR esquema$ OR escala$) AND (movil$ OR deslizante$ OR sliding)) + Filter "Controlled Clinical Trial" |
| WHO ICTRP (standard search) |
| diabet* AND sliding |
| ClinicalTrials.gov (advanced search) |
|
Search terms: (diabetes OR diabetic) AND sliding Age Group: Adult (18‐65), Senior (66+) |
Appendix 2. Assessment of risk of bias
| 'Risk of bias' domains |
|
Random sequence generation Selection bias due to inadequate generation of a randomised sequence ‐ assessment at trial level We described for each included trial the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups.
Allocation concealment Selection bias due to inadequate concealment of allocations prior to assignment ‐ assessment at trial level We described for each included trial the method used to conceal allocation to interventions prior to assignment and assessed whether intervention allocation could have been foreseen in advance of, or during recruitment, or changed after assignment.
Blinding of participants and study personnel Performance bias due to knowledge of the allocated interventions by participants and personnel during the trial ‐ assessment at outcome level We evaluated the risk of performance bias separately for each outcome (Hróbjartsson 2013). We noted whether outcomes were self reported, investigator assessed, or adjudicated outcome measures (see below).
Blinding of outcome assessment Detection bias due to knowledge of the allocated interventions by outcome assessment ‐ assessment at outcome level We evaluated the risk of detection bias separately for each outcome (Hróbjartsson 2013). We noted whether outcomes were self reported, investigator assessed, or adjudicated outcome measures (see below).
Incomplete outcome data Attrition bias due to amount, nature, or handling of incomplete outcome data ‐ assessment at outcome level We described for each included trial, and for each outcome, the completeness of data including attrition and exclusions from the analysis. We investigated whether attrition and exclusions were reported and the number included in the analysis at each stage (compared with the number of randomised participants per intervention/comparator groups), if reasons for attrition or exclusion were reported, and whether missing data were balanced across groups or were related to outcomes. We considered the implications of missing outcome data per outcome, such as high dropout rates (e.g. above 15%) or disparate attrition rates (e.g. difference of 10% or more between trial arms).
Selective reporting Reporting bias due to selective outcome reporting ‐ assessment at trial level We assessed outcome reporting bias by integrating the results of Appendix 7, 'Matrix of trial endpoints (publications and trial documents)' (Mathieu 2009), with those of Appendix 8, 'High risk of outcome reporting bias according to ORBIT classification' (Kirkham 2010). This analysis formed the basis for the judgement of selective reporting.
Other bias Bias due to problems not covered elsewhere ‐ assessment at trial level Other risk of bias reflects other circumstances that may threaten the validity of the trials, e.g. funding bias and academic bias.
Summary assessment of risk of bias Risk of bias for a trial across outcomes Some 'Risk of bias' domains like selection bias (sequence generation and allocation sequence concealment) affected the risk of bias across all outcome measures in a trial. Otherwise, we did not perform a summary assessment of the risk of bias across all outcomes for a trial. In case of high risk of selection bias, we excluded the trial. Risk of bias for an outcome within a trial and across domains We assessed the risk of bias for an outcome measure including all of the entries relevant to that outcome, i.e. both trial‐level entries and outcome‐specific entries. We defined low risk of bias as low risk of bias for all key domains; unclear risk of bias as unclear risk of bias for one or more key domains; and high risk of bias as high risk of bias for one or more key domains. Risk of bias for an outcome across trials and across domains These were our main summary assessments that were to be incorporated in our judgements about the certainty of evidence in the 'Summary of finding' table(s). We defined low risk of bias as most information coming from trials at low risk of bias; unclear risk of bias as most information coming from trials at low or unclear risk of bias; and high risk of bias as a sufficient proportion of information coming from trials at high risk of bias. Abbreviation: ORBIT: Outcome Reporting Bias In Trials |
Appendix 3. Description of interventions
| Trial ID | Sliding scale insulin | Comparator(s) |
| Said 2013 |
Sliding scale insulin Regular insulin 3 times daily SC approx. 30 min before meal for BG > 150 mg/dL (or every 8 hours if a participant was not eating) If BG < 150 mg/dL give no insulin If BG 151 to 200 mg/dL give 5 IU insulin If BG 201 to 250 mg/dL give 10 IU insulin If BG 251 to 300 mg/dL give 15 IU insulin If BG 301 to 350 mg/dL give 20 IU insulin If BG 351 to 400 mg/dL give 25 IU insulin If BG > 400 mg/dL give 30 IU insulin and call the internal medicine resident |
C1 ‐ basal‐bolus regimen: Glargine once every night
+ Glulisine: give glulisine 3 times before each meal for BG ≥ 150 mg/dL according to scale provided as SSI, hold glulisine if a participant is not able to eat to prevent hypoglycaemia. C2 ‐ premixed 70/30 insulin regimen (30% regular insulin and 70% isophane NPH insulin): premixed 70/30 + supplemental regular insulin at lunch time according to SSI
In both cases: give 2/3 of TDD before breakfast and 1/3 of TDD before dinner. Insulin adjustment:
|
| Umpierrez 2013 |
Sliding scale insulin for BG > 140 mg/dL In participants who were not eating or with intermittent nutritional intake, regular insulin was administered every 6 hours according to the recommended dose from the protocol. In participants who were eating, regular insulin was administered before each meal and at bedtime, following the “usual” according to the dose of the sliding scale protocol. If fasting and pre‐meal glucose was > 140 mg/dL, the SSI with regular insulin could be increased; if participant developed hypoglycaemia, the SSI was decreased; glucose was measured before each meal and at bedtime. |
C1 ‐ basal‐bolus regimen: glargine once daily and glulisine before meals plus corrective doses of glulisine as SSI for BG > 140 mg/dL (TDD of 0.5 units/kg, half as glargine once daily and half as glulisine before meals) C2 ‐ basal‐plus regimen: daily dose of glargine and corrective doses of glulisine as SSI before meals for BG > 140 mg/dL (0.25 units/kg of glargine plus corrective doses of glulisine before meals) In participants > 70 years of age and/or with creatinine > 2 mg/dL, starting TDD in the basal‐bolus group was reduced to 0.3 units/kg, and to 0.15 units/kg in the basal‐plus group. |
| Schroeder 2012 | Glucose levels were monitored 4 times a day (before each meal and once at 10 p.m.); short‐acting human insulin was administered SC according to the participant's blood glucose level: between 150 and 200 mg/dL, 4 units were administered; for each 50 mg/dL increase in glucose levels, 2 more units of insulin were added. |
Basal‐bolus regimen: subcutaneous insulin was administered for all uncontrolled diabetic participants (glucose > 200 mg/dL 2 or more times); after the initial dose, insulin doses were changed daily based on previous measurements; initial TDD was 0.25 U/kg divided equally in 4 doses; regular insulin was administered 3 times per day before each meal and NPH only at 10 p.m.; modified pre‐meal regular insulin based on pre‐meal glucose levels, increasing dose as needed by 2 to 4 units/injection; NPH doses were modified based on the previous morning fasting levels. Participants were educated regarding the importance of a proper diet and glucose control. |
| Umpierrez 2011 | Participants received regular insulin 4 times daily if BG > 140 mg/dL; doses were adjusted according to a prespecified protocol; if the mean daily BG level was 240 mg/dL, or if 3 consecutive values were > 240 mg/dL on the maximal SSI dose, participants were switched to basal‐bolus regimen starting TDD of 0.5 units/kg. | Basal‐bolus regimen: started at a TDD of 0.5 units/kg divided half as glargine once daily and the other half as glulisine given before meals; if a participant was not able to eat, glargine was given but glulisine was held until meals were resumed; the TDD was reduced to 0.3 units/kg in participants > 70 years of age and/or with a creatinine > 2.0 mg/dL. |
| Korytkowski 2009 | Sliding scale insulin (regular insulin) was administered SC every 4 to 6 h according to results of bedside blood glucose; if > 130 mg/dL twice, SSI started at doses of 2, 4, or 6 units depending on participant's characteristics; if ≥ 180 mg/dL, low‐dose SSI every 4 to 6 h for 24 to 36 h; if it continued > 180 mg/dL, moderate‐dose SSI every 4 to 6 h for 24 to 36 h; if 2x > 180 mg/dL, high‐dose SSI for 24 to 36 h; if 2x > 180 mg/dL, NPH every 12 h for enteral nutrition or once daily in fasting participants. | Basal‐bolus regimen: if blood glucose > 130 mg/dL, insulin glargine 10 units SC once a day and SSI regular insulin; if ≥ 180 mg/dL, NPH dose was increased by adding 50% of SSI insulin required in prior 24 h |
| Umpierrez 2007a |
Regular insulin 4 times daily for BG > 140 mg/dL If participant was eating, regular insulin was given before each meal and at bedtime, beginning with 4 U if > 140 to 180 mg/dL and adding 2 U for each "level cohort" 181 to 220 mg/dL, 221 to 260 mg/dL, 261 to 300 mg/dL, 351 to 400 mg/dL, > 400 mg/dL ("usual" column of the sliding‐scale protocol); if participant was not eating, regular insulin was given every 6 h, beginning with 2 U if > 140 mg/dL and adding 2 U for each "level cohort" ("insulin‐sensitive" column of the sliding scale protocol); if fasting and pre‐meal glucose were persistently > 140 mg/dL, SSI was increased, beginning with 6 U if > 140 mg/dL and adding 2 U for each "level cohort" (insulin resistant); if a participant developed hypoglycaemia, regular insulin dose was decreased in order of 2 U for each "level cohort". |
Basal‐bolus regimen: total daily dose of 0.4 U/kg/day if admission blood glucose was 140 to 200 mg/dL or 0.5 U/kg/day if 201 to 400 mg/dL; 1/2 of TDD was given as glargine once daily and 1/2 as glulisine; glulisine was given in 3 divided doses before each meal; glulisine was stopped if participant was not able to eat. Supplementary glulisine following the SSI protocol if BG > 140 mg/dL: if a participant was eating, glulisine was given before each meal and at bedtime following the “usual” scheme; if a participant was not eating, supplemental glulisine was given every 6 h following the “insulin‐sensitive” scheme. If fasting and pre‐meal BG > 140 mg/dL, glargine was increased by 20%; if participant developed hypoglycaemia (< 70 mg/dL), glargine dose was decreased by 20%. |
| Datta 2007 | Glucose was measured every 6 h, and participants were given a SC dose of regular insulin based on glucose levels as follows: 140 to 160 mg/dL, 4 U; 161 to 180 mg/dL, 6 U; 181 to 200 mg/dL, 8 U; 201 to 220 mg/dL, 10 U; and 221 mg/dL or more, 12 U. | Basal regimen: subcutaneous insulin glargine dose was given at dose of 0.3 U/kg; if glucose 140 to 180 mg/dL, glargine was increased by 30%; if glucose was > 180 mg/dL, glargine was increased by 60%; if glucose > 200 mg/dL, participants were treated with an additional dose of regular insulin equal to 25% of the previous glargine dose (this was the case in 6% of blood glucose measurements only). |
| Gustafson 2002 | SSI with regular insulin as follows: 200 to 250 mg/dL, 10 U; 251 to 300 mg/dL, 15 U; 301 to 350 mg/dL, 20 U; 351 to 400 mg/dL, 25 U; 401 to 500 mg/dL, 30 U; > 501 mg/dL, 35 U |
Continuous insulin infusion with regular insulin: on the day of surgery, obtain glucose level 5:30 p.m. and 8 p.m.; for glucose < 180 mg/dL, check every 2 h; for glucose > 180 mg/dL, begin infusion at 0.5 U/h and check hourly. Infusion maintenance: for glucose < 80 mg/dL, give 25 cm³ 50% dextrose IV, reduce infusion by 1 U/h; for glucose < 80 mg/dL 2 consecutive times, reduce by 1/2 current rate, check every 15 min until glucose > 80 mg/dL, then resume hourly monitoring; resume infusion when glucose is > 180 mg/dL, begin at 0.5 U/h and check glucose in 1 h; for glucose 80 to 120 mg/dL, reduce infusion by 0.5 U and check hourly; for glucose 121 to 180 mg/dL, keep infusion rate and check hourly, if glucose remains check every 2 h, if level remains in this range check in 4 h, for glucose > 180 mg/dL, if glucose has not decreased since last checked, increase infusion by 0.5 U/h, if glucose decreased since last checked, keep infusion rate the same. For glucose > 250 mg/dL during open‐heart surgery, adjust infusion rate to 0.1 U/kg/h; stop infusion if tube feeding or meal is stopped at any time and check hourly, resume infusion when glucose is > 180 mg/dL, begin infusion at 0.5 U/h and follow maintenance of the infusion. |
| BG: blood glucose; IU: international units; IV: intravenous; NPH: neutral protamine Hagedorn insulin; SC: subcutaneous; SSI: sliding scale insulin; TDD: total daily dose; U: units | ||
Appendix 4. Baseline characteristics (I)
| Trial ID | Intervention(s) and comparator(s) | Duration of intervention (follow‐up) | Description of participants | Trial period (year to year) | Country | Setting | Ethnic groups (%) | Duration of diabetes (mean years (SD)) |
| Said 2013 | I: SSI (regular insulin) | 6 days (8 days) | Non‐surgical patients with a known history of type 2 diabetes mellitus for > 3 months admitted to general medicine wards | 2010 to 2014 | Egypt | Medical unit | ‐ | ‐ |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 4 days (62 days) | ‐ | ||||||
| C2: basal‐bolus regimen (glargine + glulisine) | 5 days (72 days) | ‐ | ||||||
| Umpierrez 2013 | I: SSI (regular insulin) | 6 days (6 days) | People with known history of type 2 diabetes and a blood glucose before randomisation of 140 to 400 mg/dL or with a known history of type 2 diabetes for > 3 months, admitted to general medicine and surgery services | 2010 to 2012 | USA | Medical and surgical units | ‐ | 8.4 (7) |
| C1: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | ‐ | 9.9 (8) | ||||||
| C2: basal‐plus regimen (glargine + SSI glulisine) | ‐ | 7.6 (7) | ||||||
| Schroeder 2012 | I: SSI (short‐acting human insulin) | 7 days (7 days) | All individuals with previously diagnosed diabetes or suffering from recurrent hyperglycaemia (≥ 2 measurements of blood glucose levels > 180 mg/dL) admitted to an orthopaedic surgery department | 2007 to 2008 | Israel | Orthopaedic unit | ‐ | ‐ |
| C: basal‐bolus regimen (NPH + regular insulin) | ‐ | ‐ | ||||||
| Umpierrez 2011 | I: SSI (regular insulin) | 9 days (9 days) | People with a blood glucose level of 140 to 400 mg/dL with a history of diabetes for > 3 months, admitted to general elective or emergency surgery and not expected to require intensive care unit admission | 2008 to 2009 | USA | Surgical unit | Other: 7 Black: 55 White: 38 | 6.8 (6.3) |
| C: basal‐bolus regimen (glargine + glulisine) | Other: 10 Black: 41 White: 49 | 6.3 (6.2) | ||||||
| Korytkowski 2009 | I: SSI (regular insulin) + NPH | 8 days (24 days) | People diagnosed with diabetes (≥ 2 blood glucose measures > 7.2 mmol/L (139 mg/dL) during enteral nutrition therapy) | 2005 to 2006 | USA | Surgical and medical unit | ‐ | ‐ |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | ‐ | ‐ | ||||||
| Umpierrez 2007a | I: SSI (regular insulin) | 5 days (5 days) | Non‐surgical, insulin‐naive individuals with a known history of type 2 diabetes for 3 months, with a blood glucose level of 140 to 400 mg/dL admitted to medical general services | 2005 to 2006 | USA | Medical unit | Hispanic: 21 Black: 74 White: 5 | ‐ |
| C: basal‐bolus regimen (glargine and glulisine + SSI glulisine) | Hispanic: 27 Black: 66 White: 7 | ‐ | ||||||
| Datta 2007 | I: SSI (regular insulin) | 4 days (30 days) | People with a bariatric surgical procedure and blood glucose levels > 144 mg/dL measured postoperatively in the recovery unit (67% in the sliding scale group and 62% in the glargine group had a history of diabetes). 32/81 participants had an open abdominal operation and were admitted to the surgical intensive care unit; when they were ready for transfer to the general surgical ward they were randomised. 49/81 participants underwent a laparoscopic procedure and were admitted to a general surgical unit; they were randomised on admission. |
2006 | USA | Surgical unit | ‐ | ‐ |
| C: basal regimen (glargine) + regular insulin | ‐ | ‐ | ||||||
| Gustafson 2002 | I: SSI (regular insulin) | 1 day (1 day) | Participants with type 2 diabetes undergoing coronary artery bypass surgery | 2001 | USA | Coronary care unit | ‐ | ‐ |
| C: continuous insulin infusion (regular insulin) | ‐ | ‐ | ||||||
| ‐: denotes not reported C: comparator; I: intervention; NPH: neutral protamine Hagedorn insulin; SD: standard deviation; SSI: sliding scale insulin | ||||||||
Appendix 5. Baseline characteristics (II)
| Trial ID | Intervention(s) and comparator(s) | Sex (female %) | Age (mean years (SD)) | Glycated haemoglobin (% (SD)) | BMI (mean kg/m² (SD)) | Comedications/Cointerventions (% of participants) |
| Said 2013 | I: SSI (regular insulin) | 31.8 | 55 (7) | 8.8 (2.6) | 31 (5) | ‐ |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 42.9 | 55 (7) | 9.4 (2.3) | 31 (5) | ‐ | |
| C2: basal‐bolus regimen (glargine + glulisine) | 26.3 | 54 (11) | 9.1 (3.0) | 31 (6) | ‐ | |
| Umpierrez 2013 | I: SSI (regular insulin) | 41 | 58.7 (12) | 8.0 (2.1) | 32.6 (9) | ‐ |
| C1: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | 40 | 58.7 (11) | 8.7 (2.5) | 32.6 (8) | ‐ | |
| C2: basal‐plus regimen (glargine + SSI glulisine) | 41 | 58.6 (13) | 8.3 (2.3) | 33 (9) | ‐ | |
| Schroeder 2012 | I: SSI (short‐acting human insulin) | 50 | 71 (‐) | ‐ | ‐ | ‐ |
| C: basal‐bolus regimen (NPH + regular insulin) | 50 | 70 (‐) | ‐ | ‐ | ‐ | |
| Umpierrez 2011 | I: SSI (regular insulin) | 50 | 57 (10) | 7.4 (1.9) | 32.2 (8.5) | ‐ |
| C: basal‐bolus regimen (glargine + glulisine) | 48 | 58 (11) | 8.1 (2.4) | 30.3 (7.4) | ‐ | |
| Korytkowski 2009 | I: SSI (regular insulin) + NPH | 44 | 63 (10) | ‐ | 26.9 (1.7) | ‐ |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | 36 | 67 (10) | ‐ | 29.1 (5.8) | ‐ | |
| Umpierrez 2007a | I: SSI (regular insulin) | 67 | 56 (11) | 8.7 (2.5) | 32 (9) | ‐ |
| C: basal‐bolus regimen (glargine and glulisine + SSI glulisine) | 35 | 56 (13) | 8.9 (2) | 32 (8) | ‐ | |
| Datta 2007 | I: SSI (regular insulin) | 82 | 44.5 (‐) | 8.3 (‐) | 53 (12) | Bariatric surgery: 100 |
| C: basal regimen (glargine) + regular insulin | 86 | 45.4 (‐) | 7.7 (‐) | 55 (16) | Bariatric surgery: 100 | |
| Gustafson 2002 | I: SSI (regular insulin) | 40 | 69.6 (8.2) | ‐ | ‐ | CABG: 100 |
| C: continuous insulin infusion (regular insulin) | 44 | 67.4 (7.4) | ‐ | ‐ | CABG: 100 | |
| ‐: denotes not reported BMI: body mass index; C: comparator; CABG: coronary artery bypass graft; I: intervention; NPH: neutral protamine Hagedorn insulin; SD: standard deviation; SSI: sliding scale insulin | ||||||
Appendix 6. Matrix of trial endpoints (publications and trial documents)
| Trial ID | Endpoints quoted in trial document(s) (ClinicalTrials.gov, FDA/EMA document, manufacturer's website, published design paper)a | Trial results available in trial register | Endpoints quoted in publication(s)b,c | Endpoints quoted in abstract of publication(s)b,c |
| Said 2013 |
Source:NCT01855243 Primary outcome measures: differences in glycaemic control as measured by mean daily BG concentration during hospital stay (time frame: during hospital stay, which was expected to be on average 3 weeks); the results of BG values were measured as pre‐meal glucose and mean daily BG during the hospital stay from the day of admission |
Uncleard | Primary outcome measure: differences in glycaemic control as measured by mean daily BG concentration during hospital stay | Primary outcome measures: ‐ |
| Secondary outcome measures: number of hypoglycaemic events, hypoglycaemic episodes classified as major (BG ≤ 40 mg/dL or associated with impaired mental status or loss of consciousness) or minor (BG 40 to 59 mg/dL), severe hyperglycaemia episodes defined as BG > 300 mg/dL, mortality rate (time frame for all outcomes: during hospital stay, which was expected to be on average 3 weeks) | Secondary outcome measures: number of hypoglycaemic events, number of episodes of severe hyperglycaemia, length of hospital stay, mortality rate | Secondary outcome measures: ‐ | ||
| Other outcome measures: ‐ | Other outcome measures: ‐ | Other outcome measures: mean daily BG after first day of hospitalisation | ||
| Umpierrez 2013 |
Source:NCT00979628 Primary outcome measure: mean BG levels after first day of treatment |
Yes | Primary outcome measure: difference in glycaemic control, as measured by mean daily BG concentration | Primary outcome measures: ‐ |
| Secondary outcome measures: number of participants with hypoglycaemic events (BG < 70 mg/dL) during hospital stay (time frame: up to 12 days) | Secondary outcome measures: differences in any of the following measures: number of hypoglycaemic events (BG < 70 mg/dL and < 40 mg/dL) after the first day of treatment, number of episodes of hyperglycaemia (BG > 200 mg/dL) after the first day of treatment, total daily dose of insulin, hospital stay, hospital complications | Secondary outcome measures: ‐ | ||
| Other outcome measures: ‐ | Other outcome measures: differences in glycaemic control between medical and surgical participants | Other outcome measures: improvement in mean daily BG after the first day of therapy, mean daily BG, treatment failure (defined as > 2 consecutive BG > 240 mg/dL or a mean daily BG > 240 mg/dL) | ||
| Schroeder 2012 | Source: N/T | Primary outcome measure: mean blood glucose levels during hospitalisation | Primary outcome measure: ‐ | |
| Secondary outcome measures: ‐ | Secondary outcome measures: ‐ | |||
| Other outcome measures: % glucose measurements > 400 mg/dL, daily difference between the mean levels in both departments, frequency of hypoglycaemic episodes (BG < 70 mg/dL) | Other outcome measures: mean BG throughout hospitalisation, number of severe hyperglycaemic events (BG > 400 mg/dL), number of hypoglycaemic events | |||
| Umpierrez 2011 |
Source:NCT00596687 Primary outcome measure: mean BG concentration (time frame: hospital stay days 2 to 10) |
Yes | Primary outcome measures: differences in mean daily BG concentration and a composite of postoperative complications including wound infection, pneumonia, bacteraemia, respiratory failure and acute renal failure | Primary outcome measure: ‐ |
| Secondary outcome measures: number of participants with hypoglycaemic events (< 70 mg/dL) (time frame: hospital stay days 2 to 10) | Secondary outcome measures: differences in any of the following measures: occurrence of mild and severe hypoglycaemia (< 70 mg/dL and < 40 mg/dL), length of hospital stay, surgical complications (wound infection and dehiscence, bacteraemia, pneumonia, acute renal failure (defined as an increase in serum creatinine > 50% of baseline or a serum creatinine > 2.5 mg/dL, or both), admission to the ICU, death | Secondary outcome measures: ‐ | ||
| Other outcome measures: ‐ | Other outcome measures: ‐ | Other outcome measures: differences in daily BG, composite of postoperative complications including wound infection, pneumonia, bacteraemia, respiratory and acute renal failure | ||
| Korytkowski 2009 |
Source:NCT00177398 Primary outcome measures: frequency and severity of hyperglycaemia and hypoglycaemia; time to mean 24‐hour plasma glucose < 180 mg/dL |
No | Primary outcome measure: ‐ | Primary outcome measure: ‐ |
| Secondary outcome measures: triglyceride level at study entry and conclusion; hospital length of stay; inpatient mortality; number of days of antibiotic therapy; number of days with recorded temperature ≥ 100.4 °F; occurrence of any of the following: myocardial infarction, heart failure, cerebrovascular accident, pneumonia, renal impairment (defined as serum creatinine > 2 mg/dL); quality of life measures; outpatient glycaemic control and self management practices | Secondary outcome measures: ‐ | Secondary outcome measures: ‐ | ||
| Other outcome measures: ‐ | Other outcome measures: BG levels during all postoperative days: mean BG, % BG 80 to 140 mg/dL, % BG 141 to 200 mg/dL; total doses of insulin administered; frequency of hypoglycaemia (BG < 60 mg/dL); total inpatient length of stay; frequency of postoperative infections | Other outcome measures: mean BG levels, frequency of hypoglycaemia, total adverse events, length of hospital stay | ||
| Umpierrez 2007a |
Source:NCT00394407 Primary outcome measure: BG control (time frame: during hospitalisation) |
No | Primary outcome measures: differences in glycaemic control as measured by mean daily BG concentration | Primary outcome measure: ‐ |
|
Secondary outcome measures: frequency of hypoglycaemia, frequency of severe hyperglycaemia (BG > 400 mg/dL), length of hospital stay, inflammatory markers (C‐reactive protein, interleukin‐6, tumour necrosis factor) Time frame for all outcomes: during hospitalisation |
Secondary outcome measures: number of hypoglycaemic events, number of episodes of severe hyperglycaemia, length of stay, mortality rate | Secondary outcome measures: ‐ | ||
| Other outcome measures: ‐ | Other outcome measures: ‐ | Other outcome measures: BG target < 140 mg/dL, mean daily BG, rate of hypoglycaemia, length of hospital stay | ||
| Datta 2007 | Source: N/T | Primary outcome measure: ‐ | Primary outcome measure: ‐ | |
| Secondary outcome measures: ‐ | Secondary outcome measures: ‐ | |||
| Other outcome measures: mean BG levels, % of BG values between 80 and 140 mg/dL, % of BG values between 141 and 200 mg/dL, % of BG values > 200 mg/dL, total doses of insulin administered, frequency of hypoglycaemia (BG < 60 mg/dL), total inpatient length of hospital stay, frequency of postoperative infections | Other outcome measures: mean BG levels | |||
| Gustafson 2002 | Source: N/T | Primary outcome measure: ‐ | Primary outcome measure: ‐ | |
| Secondary outcome measures: ‐ | Secondary outcome measures: ‐ | |||
| Other outcome measures: maintenance of BG within 180 and 200 mg/dL (intraoperative BG levels) | Other outcome measures: BG levels | |||
| ‐: denotes not reported aTrial document(s) refer to all available information from published design papers and sources other than regular publications (e.g. FDA/EMA documents, manufacturer's websites, trial registers). bPublication(s) refers to trial information published in scientific journals (primary reference, duplicate publications, companion documents, or multiple reports of a primary trial). cOther outcome measures refer to all outcomes not specified as primary or secondary outcome measures. dResults information has been submitted to ClinicalTrials.gov by the sponsor or investigator, but is not yet publicly available (or 'posted') on ClinicalTrials.gov. BG: blood glucose; EMA: European Medicines Agency; FDA: US Food and Drug Administration; ICU: intensive care unit; N/T: no trial document available | ||||
Appendix 7. High risk of outcome reporting bias according to ORBIT classification
| Trial ID | Outcome | High risk of bias (category A)a | High risk of bias (category D)b | High risk of bias (category E)c | High risk of bias (category G)d |
| Said 2013 | N/D | ||||
| Umpierrez 2013 | N/D | ||||
| Schroeder 2012 | Hypoglycaemic events, other adverse events | Yes | Yes | No | No |
| Umpierrez 2011 | N/D | ||||
| Korytkowski 2009 | N/D | ||||
| Umpierrez 2007a | N/D | ||||
| Datta 2007 | N/D | ||||
| Gustafson 2002 | N/D | ||||
|
aClear that outcome was measured and analysed; trial report states that outcome was analysed but reports only that result was not significant (Classification 'A', table 2, Kirkham 2010).
bClear that outcome was measured and analysed; trial report states that outcome was analysed but reports no results (Classification 'D', table 2, Kirkham 2010).
cClear that outcome was measured but was not necessarily analysed; judgement says likely to have been analysed but not reported due to non‐significant results (Classification 'E', table 2, Kirkham 2010).
dUnclear whether outcome was measured; not mentioned, but clinical judgement says likely to have been measured and analysed but not reported due to non‐significant results (Classification 'G', table 2, Kirkham 2010). N/D: none detected; ORBIT: Outcome Reporting Bias In Trials | |||||
Appendix 8. Definition of endpoint measurementa
| Trial ID | All‐cause mortality | Diabetes‐related mortality | Length of hospital stay | Glycated haemoglobin (%) | Socioeconomic effects | All hypoglycaemic events | Severe/serious hypoglycaemia | Severe/serious adverse events other than hypoglycaemic episodes |
| Said 2013 | Mortality rate (IO) | N/R | Mean hospital stay (days) (IO) | NR | NR | Hypoglycaemic episodes were classified as major (BG ≤ 40 mg/dL or associated with impaired mental status or loss of consciousness) or minor (BG between 40 and 59 mg/dL) events (IO). | Hypoglycaemic episodes were classified as major (BG ≤ 40 mg/dL or associated with impaired mental status or loss of consciousness) (IO). | NR |
| Umpierrez 2013 | Mortality rate (IO) | N/R | Mean hospital stay (days) (IO) | Mean admission glycated haemoglobin (IO) | N/R | BG < 70 mg/dL (IO) | BG < 40 mg/dL (IO) |
Postoperative wound infection, pneumonia, bacteraemia, acute renal and respiratory failure (IO) |
| Schroeder 2012 | N/R | N/R | In‐hospital length of stay (days) (IO) | N/R | N/R | BG < 70 mg/dL (IO) | Requiring intravenous glucose administration | Postoperative complications (IO) |
| Umpierrez 2011 | Mortality rate (IO) | N/R | Mean hospital stay (days) (IO) | N/R | N/R | "Mild": BG < 70 mg/dL (IO) | BG < 40 mg/dL (IO) | Postoperative complications, wound infection, pneumonia, bacteraemia, respiratory failure, acute renal failure (IO) |
| Korytkowski 2009 | N/R | N/R | Mean hospital stay (days) during enteral nutrition therapy (IO) | N/R | N/R | % of patient‐days with ≥ 1 event and % of all glucose values < 3.9 mmol/L (IO) |
N/R | Body temperature 100.4 °F, antibiotic use, arrhythmias, pulmonary emboli, deep venous thrombosis, wound infection, respiratory symptoms, cardiac arrest, liver abscess (IO) |
| Umpierrez 2007a | Mortality rate (IO) | N/R | Mean hospital stay (days) (IO) |
Mean admission glycated haemoglobin | N/R | BG < 60 mg/dL or < 70 mg/dL (IO) | BG < 40 mg/dL (IO) | N/R |
| Datta 2007 | N/R | N/R | Mean hospital stay (days) (IO) | Mean admission glycated haemoglobin (IO) | N/R | BG < 60 mg/dL (IO) | N/R | Postoperative infections (IO) |
| Gustafson 2002 | N/R | N/R | N/R | N/R | N/R | N/R | N/R | N/R |
|
aIn addition to definition of endpoint measurement, description of who measured the outcome (AO: adjudicated outcome measurement; IO: investigator‐assessed outcome measurement; SO: self reported outcome measurement). BG: blood glucose; N/D: not defined; N/R: not reported | ||||||||
Appendix 9. Adverse events (I)
| Trial ID | Intervention(s) and comparator(s) | Participants included in analysis (N) | Deaths (N) | Deaths (% of participants) | Participants with at least 1 adverse event (N) | Participants with at least 1 adverse event (%) | Participants with at least 1 severe/serious adverse event (N) | Participants with at least 1 severe/serious adverse event (%) |
| Said 2013 | I: SSI (regular insulin) | 22 | 0 | ‐ | 0 | ‐ | 0 | ‐ |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 21 | 0 | ‐ | 0 | ‐ | 0 | ‐ | |
| C2: basal‐bolus regimen (glargine + glulisine) | 19 | 0 | ‐ | 0 | ‐ | 0 | ‐ | |
| Umpierrez 2013 | I: SSI (regular insulin) | 74 | 0 | 0 | 7 | 9.4 | 6 | 8.1 |
| C1: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | 144 | 0 | 0 | 21 | 14.5 | 17 | 11.8 | |
| C2: basal‐plus regimen (glargine + SSI glulisine) | 133 | 3 | 2 | 20 | 15 | 17 | 12.7 | |
| Schroeder 2012 | I: SSI (short‐acting human insulin) | 30 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| C: basal‐bolus regimen (NPH + regular insulin) | 35 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | |
| Umpierrez 2011 | I: SSI (regular insulin) | 107 | 1 | 0.9 | 26 | 24.2 | 19 | 17.7 |
| C: basal‐bolus regimen (glargine + glulisine) | 104 | 1 | 0.9 | 9 | 8.6 | 7 | 6.7 | |
| Korytkowski 2009 | I: SSI (regular insulin) + NPH | 25 | ‐ | ‐ | 11 | 44 | 4 | 16 |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | 25 | ‐ | ‐ | 6 | 24 | 3 | 12 | |
| Umpierrez 2007a | I: SSI (regular insulin) | 65 | 0 | 0 | 2 | 3 | 0 | ‐ |
| C: basal‐bolus regimen (glargine and glulisine + SSI glulisine) | 65 | 1 | 1.5 | 2 | 3 | 1 | 1.5 | |
| Datta 2007 | I: SSI (regular insulin) | 39 | ‐ | ‐ | 2 | 5.1 | 0 | ‐ |
| C: basal regimen (glargine) + regular insulin | 42 | ‐ | ‐ | 2 | 4.7 | 0 | ‐ | |
| Gustafson 2002 | I: SSI (regular insulin) | 25 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| C: continuous insulin infusion (regular insulin) | 25 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | |
| ‐: denotes not reported C: comparator; I: intervention; NPH: neutral protamine Hagedorn insulin; SSI: sliding scale insulin | ||||||||
Appendix 10. Adverse events (II)
| Trial ID | Intervention(s) and comparator(s) | Participants included in analysis (N) | Participants discontinuing trial due to an adverse event (N) | Participants discontinuing trial due to an adverse event (%) |
| Said 2013 | I: SSI (regular insulin) | 22 | 0 | ‐ |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 21 | 0 | ‐ | |
| C2: basal‐bolus regimen (glargine + glulisine) | 19 | 0 | ‐ | |
| Umpierrez 2013 | I: SSI (regular insulin) | 74 | 0 | ‐ |
| C1: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | 144 | 0 | ‐ | |
| C2: basal‐plus regimen (glargine + SSI glulisine) | 133 | 0 | ‐ | |
| Schroeder 2012 | I: SSI (short‐acting human insulin) | 30 | 0 | ‐ |
| C: basal‐bolus regimen (NPH + regular insulin) | 35 | 0 | ‐ | |
| Umpierrez 2011 | I: SSI (regular insulin) | 107 | ‐ | ‐ |
| C: basal‐bolus regimen (glargine + glulisine) | 104 | ‐ | ‐ | |
| Korytkowski 2009 | I: SSI (regular insulin) + NPH | 25 | 0 | ‐ |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | 25 | 0 | ‐ | |
| Umpierrez 2007a | I: SSI (regular insulin) | 65 | 0 | ‐ |
| C: basal‐bolus regimen (glargine and glulisine + SSI glulisine) | 65 | 0 | ‐ | |
| Datta 2007 | I: SSI (regular insulin) | 39 | 0 | ‐ |
| C: basal regimen (glargine) + regular insulin | 42 | 0 | ‐ | |
| Gustafson 2002 | I: SSI (regular insulin) | 25 | 0 | ‐ |
| C: continuous insulin infusion (regular insulin) | 25 | 0 | ‐ | |
| ‐: denotes not reported C: comparator; I: intervention; NPH: neutral protamine Hagedorn insulin; SSI: sliding scale insulin | ||||
Appendix 11. Adverse events (III)
| Trial ID | Intervention(s) and comparator(s) | Participants included in analysis (N) | Participants with a specific adverse event (description) | Participants with at least 1 specific adverse event (N) | Participants with at least 1 specific adverse event (%) |
| Said 2013 | I: SSI (regular insulin) | 22 | ‐ | ‐ | ‐ |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 21 | ‐ | ‐ | ‐ | |
| C2: basal‐bolus regimen (glargine + glulisine) | 19 | ‐ | ‐ | ‐ | |
| Umpierrez 2013 | I: SSI (regular insulin) | 74 | (1) Wound infections (2) Acute renal failure (3) Bacteraemia | (1) 1 (2) 3 (3) 3 | (1) 1.3 (2) 4 (3) 4 |
| C1: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | 144 | (1) Wound infections (2) Pneumonia (3) Acute respiratory failure (4) Acute renal failure (5) Bacteraemia | (1) 1 (2) 3 (3) 1 (4) 9 (5) 7 | (1) 0.6 (2) 2 (3) 0.6 (4) 6.2 (5) 4.8 | |
| C2: basal‐plus regimen (glargine + SSI glulisine) | 133 | (1) Wound infections (2) Acute respiratory failure (3) Acute renal failure (4) Bacteraemia (5) Mortality | (1) 3 (2) 2 (3) 8 (4) 4 (5) 3 | (1) 2.2 (2) 1.5 (3) 6 (4) 3 (5) 2.2 | |
| Schroeder 2012 | I: SSI (short‐acting human insulin) | 30 | ‐ | ‐ | ‐ |
| C: basal‐bolus regimen (NPH + regular insulin) | 35 | ‐ | ‐ | ‐ | |
| Umpierrez 2011 | I: SSI (regular insulin) | 107 | (1) Wound infections (2) Pneumonia (3) Acute respiratory failure (4) Acute renal failure (5) Bacteraemia | (1) 11 (2) 3 (3) 5 (4) 11 (5) 2 | (1) 10.2 (2) 2.8 (3) 4.6 (4) 10.2 (5) 1.8 |
| C: basal‐bolus regimen (glargine + glulisine) | 104 | (1) Wound infections (2) Acute respiratory failure (3) Acute renal failure (4) Bacteraemia | (1) 3 (2) 1 (3) 4 (4) 1 | (1) 2.8 (2) 0.9 (3) 3.8 (4) 0.9 | |
| Korytkowski 2009 | I: SSI (regular insulin) + NPH | 25 | (1) Body temperature 100.4 °F (2) Arrhythmias (3) Pulmonary emboli (4) Respiratory symptoms (5) Cardiac arrest | (1) 8 (2) 2 (3) 1 (4) 2 (5) 1 | (1) 32 (2) 8 (3) 4 (4) 8 (5) 4 |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | 25 | (1) Arrhythmias (2) Pulmonary emboli (3) Deep venous thrombosis (4) Wound infection (5) Respiratory symptoms (6) Liver abscess | (1) 1 (2) 2 (3) 2 (4) 1 (5) 2 (6) 1 | (1) 4 (2) 8 (3) 8 (4) 4 (5) 8 (6) 4 | |
| Umpierrez 2007a | I: SSI (regular insulin) | 65 | ‐ | ‐ | ‐ |
| C: basal‐bolus regimen (glargine and glulisine + SSI glulisine) | 65 | ‐ | ‐ | ‐ | |
| Datta 2007 | I: SSI (regular insulin) | 39 | Postoperative infections | 2 | 5.1 |
| C: basal regimen (glargine) + regular insulin | 42 | Postoperative infections | 2 | 4.7 | |
| Gustafson 2002 | I: SSI (regular insulin) | 25 | ‐ | ‐ | ‐ |
| C: continuous insulin infusion (regular insulin) | 25 | ‐ | ‐ | ‐ | |
| ‐: denotes not reported C: comparator; I: intervention; NPH: neutral protamine Hagedorn insulin; SSI: sliding scale insulin | |||||
Appendix 12. Adverse events (IV)
| Trial ID | Intervention(s) and comparator(s) | Participants included in analysis (N) | Participants with at least 1 hypoglycaemic episode (N) | Participants with at least 1 hypoglycaemic episode (%) | Participants with at least 1 nocturnal hypoglycaemic episode (N) | Participants with at least 1 nocturnal hypoglycaemic episode (% participants) | Participants with at least 1 severe/serious hypoglycaemic episode (N) | Participants with at least 1 severe/serious hypoglycaemic episode (%) |
| Said 2013 | I: SSI (regular insulin) | 22 | 1 | 4.5 | ‐ | ‐ | ‐ | ‐ |
| C1: 70/30 insulin SSI (regular insulin + NPH) | 21 | 0 | 0 | ‐ | ‐ | ‐ | ‐ | |
| C2: basal‐bolus regimen (glargine + glulisine) | 19 | 2 | 10.5 | ‐ | ‐ | 2 | 10.5 | |
| Umpierrez 2013 | I: SSI (regular insulin) | 74 | 2 | 2.7 | ‐ | ‐ | 0 | 0 |
| C1: basal‐bolus regimen (glargine + glulisine + SSI glulisine) | 74 | 2 | 2.7 | ‐ | ‐ | 0 | 0 | |
| C2: basal‐plus regimen (glargine + SSI glulisine) | 144 | 23 | 16 | ‐ | ‐ | 1 | 4.3 | |
| Schroeder 2012 | I: SSI (short‐acting human insulin) | 133 | 17 | 13 | ‐ | ‐ | 1 | 5.9 |
| C: basal‐bolus regimen (NPH + regular insulin) | 30 | 0 | ‐ | ‐ | ‐ | 0 | ‐ | |
| Umpierrez 2011 | I: SSI (regular insulin) | 35 | 2 | 5.7 | ‐ | ‐ | 0 | ‐ |
| C: basal‐bolus regimen (glargine + glulisine) | 107 | 6 | 5.6 | ‐ | ‐ | 0 | ‐ | |
| Korytkowski 2009 | I: SSI (regular insulin) + NPH | 12 | 0 | ‐ | ‐ | ‐ | ‐ | ‐ |
| C: basal‐bolus regimen (glargine + SSI regular insulin) | 25 | 5 | 20 | ‐ | ‐ | ‐ | ‐ | |
| Umpierrez 2007a | I: SSI (regular insulin) | 25 | 3 | 12 | ‐ | ‐ | ‐ | ‐ |
| C: basal‐bolus regimen (glargine and glulisine + SSI glulisine) | 65 | 2 | 3 | ‐ | ‐ | 0 | ‐ | |
| Datta 2007 | I: SSI (regular insulin) | 65 | 2 | 3 | ‐ | ‐ | 0 | ‐ |
| C: basal regimen (glargine) + regular insulin | 39 | 1 | 2.6 | ‐ | ‐ | ‐ | ‐ | |
| Gustafson 2002 | I: SSI (regular insulin) | 42 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ |
| C: continuous insulin infusion (regular insulin) | 25 | ‐ | ‐ | ‐ | ‐ | ‐ | ‐ | |
| ‐: denotes not reported C: comparator; I: intervention; NPH: neutral protamine Hagedorn insulin; SSI: sliding scale insulin | ||||||||
Appendix 13. Survey of authors providing information on included trials
| Trial ID | Date trial author contacted | Date trial author replied | Date trial author was asked for additional information (short summary) | Date trial author provided data (short summary) |
| Said 2013 | N/A | N/A | N/A | N/A |
| Umpierrez 2013 | 22 October 2015 | 22 October 2015 | 22 October 2015 We asked if blinding of outcome assessment was performed. |
22 October 2015 Trial authors did not have an independent adjudication board. |
| Schroeder 2012 | 25 July 2015 | 25 July 2015 | 25 July 2015 We asked if the statistical data in Figure 2 of the publication were correct. |
25 July 2015 Trial authors agreed with our observations. |
| Umpierrez 2011 | 22 October 2015 | 22 October 2015 | 22 October 2015 We asked if blinding of outcome assessment was performed. |
22 October 2015 Trial authors did not have an independent adjudication board. |
| Korytkowski 2009 | 21 October 2015 | No reply | N/A | N/A |
| Umpierrez 2007a | 22 October 2015 | 22 October 2015 | 22 October 2015 We asked if blinding of outcome assessment was performed. |
22 October 2015 Trial authors did not have an independent adjudication board. |
| Datta 2007 | 21 October 2015 | No reply | N/A | N/A |
| Gustafson 2002 | 21 October 2015 16 June 2016 |
No reply | N/A | N/A |
| N/A: not applicable | ||||
Appendix 14. Checklist to aid consistency and reproducibility of GRADE assessments
| (1) All‐cause mortality | (2) Diabetes‐related mortality | (3) Severe hypoglycaemic episodes | (4) Length of hospital stay | (5) Adverse events other than hypoglycaemic episodes (postoperative complications) | (6) Mean blood glucose levels | (7) Socioeconomic effects | ||
| Trial limitations (risk of bias)a | Was random sequence generation used (i.e. no potential for selection bias)? | Unclear | N/R | Unclear | Unclear | Unclear | Unclear | N/R |
| Was allocation concealment used (i.e. no potential for selection bias)? | Unclear | Unclear | Unclear | Unclear | Unclear | |||
| Was there blinding of participants and personnel (i.e. no potential for performance bias) or was outcome not likely to be influenced by lack of blinding? | Yes | No (↓) | Yes | No (↓) | No (↓) | |||
| Was there blinding of outcome assessment (i.e. no potential for detection bias) or was outcome measurement not likely to be influenced by lack of blinding? | Yes | No (↓) | Yes | No (↓) | No (↓) | |||
| Was an objective outcome used? | Yes | Unclear | Yes | Yes | Yes | |||
| Were more than 80% of participants enrolled in trials included in the analysis (i.e. no potential reporting bias)?e | Yes | Yes | Yes | Yes | Yes | |||
| Were data reported consistently for the outcome of interest (i.e. no potential selective reporting)? | Yes | Yes | Yes | Yes | Yes | |||
| No other biases reported (i.e. no potential of other bias)? | Yes | Yes | Yes | Yes | Yes | |||
| Did the trials end as scheduled (i.e. not stopped early)? | Yes | Yes | Yes | Yes | Yes | |||
| Inconsistencyb | Point estimates did not vary widely? | Yes | Yes | Yes | No (↓) | Yes | ||
| To what extent did confidence intervals overlap (substantial: all confidence intervals overlap at least 1 of the included studies point estimate; some: confidence intervals overlap but not all overlap at least 1 point estimate; no: at least 1 outlier: where the confidence interval of some of the studies does not overlap with those of most included studies)? | Substantial | Substantial | Substantial | Substantial | Substantial | |||
| Was the direction of effect consistent? | Yes | Yes | No (↓) | No (↓) | Yes | |||
| What was the magnitude of statistical heterogeneity (as measured by I²: low (I² < 40%), moderate (I² 40% to 60%), high (I² > 60%))? | Low | Low | Low | High (↓) | Moderate | |||
| Was the test for heterogeneity statistically significant (P < 0.1)? | Not statistically significant | Not statistically significant | Not statistically significant | Statistically significant (↓) | Not Statistically significant | |||
| Indirectness | Were the populations in the included studies applicable to the decision context? | Applicable | Applicable | Applicable | Applicable | Applicable | ||
| Were the interventions in the included studies applicable to the decision context? | Applicable | Applicable | Applicable | Applicable | Applicable | |||
| Was the included outcome not a surrogate outcome? | Yes | Unclear | Yes | Yes | Yes | |||
| Was the outcome time frame sufficient? | Sufficient | Sufficient | Sufficient | Sufficient | Sufficient | |||
| Were the conclusions based on direct comparisons? | Yes | Yes | Yes | Yes | Yes | |||
| Imprecisionc | Was the confidence interval for the pooled estimate not consistent with benefit and harm? | No (↓) | No (↓) | No (↓) | Yes (↓) | Yes | ||
| What is the magnitude of the median sample size (high: 300 participants, intermediate: 100 to 300 participants, low: < 100 participants)?e | Low (↓) | Low (↓) | Low (↓) | Low (↓) | Low (↓) | |||
| What is the magnitude of the number of included studies (large: > 10 studies, moderate: 5 to 10 studies, small: < 5 studies)?e | Small (↓) | Moderate | Moderate | Small (↓) | Moderate | |||
| Was the outcome a common event (e.g. occurs more than 1/100)? | No (↓) | Yes | N/A | Yes | N/A | |||
| Publication biasd | Was a comprehensive search conducted? | Yes | Yes | Yes | Yes | Yes | ||
| Was grey literature searched? | No (↓) | No (↓) | No (↓) | No (↓) | No (↓) | |||
| Were no restrictions applied to study selection on the basis of language? | Yes | Yes | Yes | Yes | Yes | |||
| There was no industry influence on studies included in the review? | No (↓) | No (↓) | No (↓) | No (↓) | No (↓) | |||
| There was no evidence of funnel plot asymmetry? | N/A | N/A | N/A | N/A | N/A | |||
| There was no discrepancy in findings between published and unpublished trials? | N/A | N/A | N/A | N/A | N/A | |||
|
aQuestions on risk of bias are answered in relation to the majority of the aggregated evidence in the meta‐analysis rather than to individual trials.
bQuestions on inconsistency are based primarily on visual assessment of forest plots and the statistical quantification of heterogeneity based on I². cWhen judging the width of the confidence interval it is recommended to use a clinical decision threshold to assess whether the imprecision is clinically meaningful. dQuestions address the comprehensiveness of the search strategy, industry influence, funnel plot asymmetry, and discrepancies between published and unpublished trials. eDepends on the context of the systematic review area. (↓): key item for potential downgrading the certainty of the evidence (GRADE) as shown in the footnotes of the 'Summary of findings' table(s) N/A: not applicable; N/R: not reported | ||||||||
Data and analyses
Comparison 1. Sliding scale insulin versus basal‐bolus insulin regimen.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|
| 1 All‐cause mortality | 4 | 602 | Peto Odds Ratio (Peto, Fixed, 95% CI) | 0.50 [0.05, 4.85] |
| 2 Non‐severe hypoglycaemic episodes | 3 | 561 | Risk Ratio (M‐H, Random, 95% CI) | 0.25 [0.11, 0.59] |
| 2.1 Medical participants | 1 | 130 | Risk Ratio (M‐H, Random, 95% CI) | 1.0 [0.15, 6.89] |
| 2.2 Surgical participants | 1 | 211 | Risk Ratio (M‐H, Random, 95% CI) | 0.20 [0.08, 0.51] |
| 2.3 Medical/surgical participants | 1 | 220 | Risk Ratio (M‐H, Random, 95% CI) | 0.17 [0.04, 0.71] |
| 3 Severe hypoglycaemic episodes | 5 | 667 | Risk Ratio (M‐H, Random, 95% CI) | 0.22 [0.05, 1.00] |
| 4 Length of hospital stay | 6 | 717 | Mean Difference (IV, Random, 95% CI) | 0.47 [‐0.46, 1.40] |
| 4.1 Medical participants | 3 | 221 | Mean Difference (IV, Random, 95% CI) | 0.23 [‐0.94, 1.39] |
| 4.2 Surgical participants | 2 | 276 | Mean Difference (IV, Random, 95% CI) | 1.18 [‐0.97, 3.34] |
| 4.3 Medical/surgical participants | 1 | 220 | Mean Difference (IV, Random, 95% CI) | ‐0.40 [‐1.80, 1.00] |
| 5 Adverse events other than hypoglycaemic episodes (infectious complications) | 3 | 481 | Risk Ratio (M‐H, Random, 95% CI) | 1.16 [0.25, 5.37] |
| 6 Mean glucose levels | 6 | 717 | Mean Difference (IV, Random, 95% CI) | 14.79 [7.75, 21.82] |
| 6.1 Medical participants | 3 | 221 | Mean Difference (IV, Random, 95% CI) | 9.52 [‐14.38, 33.43] |
| 6.2 Surgical participants | 2 | 276 | Mean Difference (IV, Random, 95% CI) | 16.20 [9.95, 22.44] |
| 6.3 Medical/surgical participants | 1 | 220 | Mean Difference (IV, Random, 95% CI) | 16.0 [4.98, 27.02] |
Comparison 2. Sliding scale insulin versus basal insulin regimen.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|
| 1 Non‐severe hypoglycaemic episodes | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2 Length of hospital stay | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3 Adverse events other than hypoglycaemic episodes | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4 Mean glucose levels | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected |
Comparison 3. Sliding scale insulin versus continuous insulin infusion.
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
|---|---|---|---|---|
| 1 Mean glucose levels | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected |
Characteristics of studies
Characteristics of included studies [ordered by study ID]
Datta 2007.
| Methods | Parallel randomised controlled trial | |
| Participants |
Inclusion criteria: patients who underwent bariatric surgical procedure, Roux‐en‐Y gastric bypass, between 21 to 60 years old, glucose level of more than 144 mg/dL, measured postoperatively in the recovery unit Exclusion criteria: not described Diagnostic criteria: medical history of diabetes and blood glucose levels exceeding 144 mg/dL, measured postoperatively in the recovery unit Setting: hospital Age group: adults Gender distribution: females and males Country where trial was performed: USA |
|
| Interventions |
Treatment before study: insulin, sulphonylurea, metformin, and thiazolidinedione Titration period: none Intervention: sliding scale insulin: BG level measured every 6 hours, and participants were given a predetermined SC dose of regular insulin based on the current BG level, as follows: less than 140 mg/dL, no insulin; between 140 and 160 mg/dL, 4 U; between 161 and 180 mg/dL, 6 U; between 181 and 200 mg/dL, 8 U; between 201 and 220 mg/dL, 10 U; and 221 mg/dL or more, 12 U. Comparator: basal insulin, initial dose of 0.3 U of insulin glargine per kg of body weight. Blood glucose was monitored every 6 hours thereafter by point‐of‐care reflectance meters. The initial dose of insulin glargine was repeated every 24 hours as long as the BG level remained between 80 and 140 mg/dL. If the BG level was between 140 and 180 mg/dL, the next dose of insulin glargine was increased by 30%; if the glucose level was more than 180 mg/dL, the next dose of insulin glargine was increased by 60%. All BG values greater than 200 mg/dL were also treated with an additional compensatory dose of SC regular insulin equal to 25% of the previous dose of insulin glargine. Duration of intervention: 4 days Duration of follow‐up: 30 days Number of study centres: 1 Run‐in period: no description |
|
| Outcomes | Outcomes reported in abstract of publication: mean blood glucose levels (postoperative hyperglycaemia) | |
| Study registration |
Trial identifier: N/T Trial terminated early: no |
|
| Publication details |
Language of publication: English Commercial funding: Sanofi‐Aventis Publication status: peer review journal |
|
| Stated aim for study | Quote from publication: "We hypothesized that once‐daily insulin glargine may be more effective than SSRI in achieving glycaemic control and also may be safer (that is, associated with less chance of occurrence of hypoglycaemia)" | |
| Notes | ‐ | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Quote from publication: "Patients were randomised to receive insulin glargine or SSRI" Comment: no further description, page 226, "study protocol" and no more information provided by the authors |
| Allocation concealment (selection bias) | Unclear risk | Comment: not described |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk |
Quote from publication: "Treatment was not blinded" Comment: treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: treatment was not blinded, outcome unlikely to be affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Adverse events other than hypoglycaemic episodes | High risk |
Quote from publication: "Treatment was not blinded" Comment: treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk |
Quote from publication: "Treatment was not blinded" Comment: treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk |
Quote from publication: "Treatment was not blinded" Comment: treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Severe hypoglycaemic episodes | High risk | Comment: outcome assessor not reported, treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Non‐severe hypoglycaemic episodes | High risk | Comment: outcome assessor not reported, treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk | Comment: outcome not reported, outcome not likely to have been influenced by blinding |
| Blinding of outcome assessment (detection bias) Adverse events other than hypoglycaemia | High risk | Comment: outcome assessor not reported, treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Mean glucose level | High risk | Comment: outcome assessor not reported, treatment was not blinded, outcome likely affected by lack of blinding techniques |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Adverse events other than hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: no missing data |
| Selective reporting (reporting bias) | Low risk | Comment: the study protocol was not available, no ORBIT criteria detected |
| Other bias | Low risk | Comment: none detected |
Gustafson 2002.
| Methods | Parallel randomised controlled trial | |
| Participants |
Inclusion criteria: patients diagnosed with diabetes mellitus type 2 undergoing coronary artery bypass surgery Exclusion criteria: not described Diagnostic criteria: medical history of diabetes Setting: hospital Age group: elderly people Gender distribution: females and males Country where trial was performed: USA |
|
| Interventions |
Intervention: sliding scale insulin plus saline‐based cardioplegic solution. Mix 50 U of regular human insulin in 100 cm³ of normal saline (concentration is 0.5 U = 1 cm³) 200 to 250 mg/dL 10 IU insulin, 251 to 300 mg/dL 15 IU insulin, 301 to 350 mg/dL 20 IU insulin, 351 to 400 mg/dL 25 IU insulin, 401 to 500 mg/dL 30 IU insulin, > 501 mg/dL 35 IU insulin Comparator: continuous insulin infusion plus 5% dextrose‐based cardioplegic solution Maintenance of the insulin infusion:
Duration of intervention: 1 day Duration of follow‐up: 1 day Run‐in period: not reported Number of study centres: 1 Treatment before study: not reported |
|
| Outcomes | Outcomes reported in abstract of publication: blood glucose levels | |
| Study registration |
Trial identifier: N/T Study terminated early: no |
|
| Publication details |
Language of publication: English Funding: not reported Publication status: peer review journal |
|
| Stated aim for study | "Continuous insulin infusion was not an effective mode of treatment in maintaining safe blood glucose levels (< 200 mg/dL) during the intraoperative period of diabetic patients requiring open‐heart surgery. The two modifications investigated to gain better control of the blood glucose were a change in the base solution of the cardioplegia and the use of a sliding insulin scale" | |
| Notes | Trial was conducted in a coronary unit | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk |
Quote from publication: "The patients were then randomly divided into two groups" (abstract section) Comment: insufficient detail |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: no missing data |
| Selective reporting (reporting bias) | Low risk | Comment: the study protocol was not available, no ORBIT criteria detected |
| Other bias | Low risk | Comment: none detected |
Korytkowski 2009.
| Methods | Parallel randomised controlled trial | |
| Participants |
Inclusion criteria
Exclusion criteria
Diagnostic criteria: medical history of diabetes (2 or more blood glucose measurements > 7.2 mmol/L during enteral nutrition therapy) Setting: hospital Age group: adults/elderly people Gender distribution: females and males Country where trial was performed: USA |
|
| Interventions |
Information provided in the online appendix Regular insulin sliding scale ‐ standard orders Recommended indications:
Regimens:
Intervention: treatment algorithm (Appendix C) glargine group: BG > 7.2 mmol/L (130 mg/dL) x 1: Start glargine 10 units once a day with SSRI If ≥ 2 BG > 10 mmol/L (180 mg/dL) add 50% of SSRI given in previous 24 hours to glargine dose; or if ≤ 1 BG > 10 mmol/L (180 mg/dL) or all BG 3.9 to 10 mmol/L (70 to 180 mg/dL) with ≤ 20 units SSRI per day continue current regimen Comparator: treatment algorithm (Appendix B) if BG > 7.2 mmol/L (130 mg/dL) x 2: Start low‐dose SSRI, then participants were divided into the following 2 strategies
Duration of intervention: 8 days Duration of follow‐up: 23 days Run‐in period: February 2005 to August 2006 Number of study centres: 1 Treatment before study: not reported Titration period: none |
|
| Outcomes | Outcomes reported in abstract of publication: glycaemic control (mean glucose levels), frequency of hypoglycaemia, total adverse events, length of stay | |
| Study registration |
Trial identifier: ClinicalTrials.gov: NCT00177398 Study terminated early: no |
|
| Publication details |
Language of publication: English (Commercial and non‐commercial) funding: Sanofi‐Aventis, University of Pittsburgh Publication status: peer review journal |
|
| Stated aim for study | Quote from publication: "The purpose of this study is to determine if the early initiation of a long acting insulin (i.e. glargine) with supplemental doses of short acting regular (SSR) insulin in hospitalised people with diabetes who are fed using tube feedings reduces the frequency of high and low blood sugar levels when compared to use of SSR insulin alone" | |
| Notes | Some data were extracted from ClinicalTrials.gov | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk | Quote from publication: "Randomized clinical trial" Comment: insufficient detail |
| Allocation concealment (selection bias) | Unclear risk | Comment: insufficient detail |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk |
Quote from publication: "open label" Comment: outcome likely influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: open‐label trial, outcome unlikely to be influenced by blinding |
| Blinding of participants and personnel (performance bias) Adverse events other than hypoglycaemic episodes | High risk |
Quote from publication: "open label" Comment: outcome likely influenced by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk |
Quote from publication: "open label" Comment: outcome influenced by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk |
Quote from publication: "open label" Comment: outcome likely influenced by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Severe hypoglycaemic episodes | High risk |
Quote from publication: "open label" Comment: outcome likely influenced by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk | Comment: outcome assessor not reported, outcome unlikely to be influenced by blinding |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Adverse events other than hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: no missing data |
| Selective reporting (reporting bias) | Low risk | Comment: not detected by ORBIT criteria |
| Other bias | Low risk | Comment: not detected |
Said 2013.
| Methods | Parallel randomised controlled trial | |
| Participants |
Inclusion criteria: non‐surgical patients, 18 to 64 years old with a known history of type 2 DM for longer than 3 months admitted to general medicine wards Exclusion criteria: hyperglycaemia without any know history of DM, presence of diabetic ketoacidosis, patients admitted to intensive care unit or undergoing surgery during their hospitalisation, patients with relevant hepatic disease, impaired renal function (serum creatinine > 3.0 mg/dL), systemic infections, or pregnancy, and patients on medication known to interfere with blood glucose level Diagnostic criteria: known history of type 2 DM Setting: hospital Age group: adults Gender distribution: females and males Country where trial was performed: Egypt |
|
| Interventions |
Intervention:
Basal‐bolus scheme Glargine
Glulisine
In both cases: give 2/3 of TDD before breakfast and 1/3 of TDD before dinner Supplemental insulin (regular insulin) given as SSI protocol Insulin adjustment
Comparator: sliding scale insulin protocol: BG (mg/dL) < 150 mg/dL = no insulin, 151 to 200 mg/dL = 5 IU, 201 to 250 mg/dL = 10 IU, 251 to 300 mg/dL = 15 IU, 301 to 350 mg/dL = 20 IU, 351 to 400 mg/dL = 25 IU, and > 400 mg/dL = 30 IU and call the internal medicine resident Duration of intervention: 6 days, 4 days, 5 days Duration of follow‐up: 8 days, 62 days, 72 days Run‐in period: March 2010 to May 2013 Number of study centres: 1 Treatment before study: insulin plus oral antidiabetic agents, oral antidiabetic agent as monotherapy, multiple oral antidiabetic agents and non‐pharmacological agent Titration period: none |
|
| Outcomes |
Primary outcome: glycaemic control by mean daily BG concentration during participant hospital stay Secondary outcomes: hypoglycaemic events, length of hospital stay, and mortality rate |
|
| Study registration |
Trial identifier:NCT01855243 Study terminated early: no |
|
| Publication details |
Language of publication: English Funding: Cairo University Publication status: peer review journal |
|
| Stated aim for study | To compare efficacy and safety of traditional sliding scale insulin versus modified 70/30 insulin versus modified basal‐plus supplemental scale/bolus insulin regimens for glycaemic control in hospitalised diabetic patients | |
| Notes | Some data were extracted from ClinicalTrials.gov | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk |
Quote from publication: "open‐label parallel, randomized study" Comment: insufficient detail |
| Allocation concealment (selection bias) | Unclear risk | Comment: no description |
| Blinding of participants and personnel (performance bias) All‐cause mortality | Low risk | Comment: no events, outcome unlikely to be affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk | Comment: open‐label, outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: outcome unlikely to be affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk | Comment: open‐label, outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk | Comment: open‐label, outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) All‐cause mortality | Low risk | Comment: no events, outcome not likely to be affected by potential lack of blinding |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk | Comment: not reported, outcome unlikely to be affected by potential lack of blinding |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) All‐cause mortality | Low risk | Comment: no events |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: no missing data |
| Selective reporting (reporting bias) | Low risk | Comment: not detected |
| Other bias | Low risk | Comment: not detected |
Schroeder 2012.
| Methods | Cluster‐randomised controlled trial (2 orthopaedic wards) | |
| Participants |
Inclusion criteria: diabetes or hyperglycaemia, admitted to the orthopaedic surgery department Exclusion criteria: no description Diagnostic criteria: all individuals with previously diagnosed diabetes or suffering from recurrent hyperglycaemia (2 or more measurements of blood glucose levels > 180 mg/dL) Setting: hospital Age group: adults Gender distribution: females and males Country where trial was performed: Israel |
|
| Interventions |
Intervention(s): sliding scale insulin (Actrapid) Regular insulin was given according to participant's blood glucose level (i.e. for levels between 150 mg and 200 mg of blood glucose, 4 units of short‐acting human insulin (Actrapid) were given; for each 50 mg increase in blood glucose levels, 2 more units of insulin were added). Comparator(s): basal‐bolus insulin The initial total daily dose of insulin was 0.25 U/kg body weight divided equally between 4 doses, unless the participant was on a higher insulin dose prior to admission. Regular insulin was administered 3 times per day before each meal, and NPH was given once daily at 10:00 p.m. Duration of intervention: 7 days Duration of follow‐up: 7 days Run‐in period: October 2007 to March 2008 Number of study centres: 1 Treatment before study: not reported Titration period: none |
|
| Outcomes | Outcomes reported in abstract of publication: mean blood glucose levels during hospitalisation, percentage of glucose measurements > 400 mg/dL, the daily difference between the mean levels in both departments, and the frequency of hypoglycaemic episodes (blood glucose level < 70 mg/dL) | |
| Study registration |
Trial identifier: N/T Study terminated before regular end: no |
|
| Publication details |
Language of publication: English Funding: none stated Publication status: peer review journal |
|
| Stated aim for study | Quote from publication: "We performed a randomised controlled prospective study to test the efficacy and safety of this intervention and the ability to maintain glucose levels in the desired range as a result of the treatment protocol" | |
| Notes | ‐ | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk |
Quote from publication: "We performed a randomized controlled prospective study" Comment: insufficient detail |
| Allocation concealment (selection bias) | Unclear risk | Comment: no details |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk |
Quote from publication: "The first two steps were to introduce the concept of intensive glucose monitoring to both the physicians and nurses in the orthopaedic department and appoint one of the nurses as a ‘diabetes’ nurse; the third step was to implement a scheduled predetermined basal bolus insulin regimen for all uncontrolled diabetic patients (blood glucose levels > 200 mg/dL on two or more tests)." Comment: no description, cluster design (possible recruitment bias), outcome likely affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: outcome unlikely to be affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk | Comment: no description, cluster design (personnel not blinded ‐ training session about the intervention), outcome likely affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk | Comment: no description, cluster design (possible recruitment bias), outcome likely affected by lack of blinding |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk | Comment: not reported, outcome unlikely to be influenced by potential lack of blinding |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: no missing data |
| Selective reporting (reporting bias) | High risk |
Quote from publication: "There was no difference in the postoperative complications between the groups". No difference in operative‐related and ‐unrelated complications between groups was noted. Comment: clear that outcome was measured and analysed; trial report states that outcome was analysed, but reports only that result was not significant |
| Other bias | Unclear risk | Comment: lack of detail related to the cluster design (possibility of incorrect analysis) |
Umpierrez 2007a.
| Methods | Parallel randomised controlled trial | |
| Participants |
Inclusion criteria: non‐surgical, insulin‐naive people with a known history of diabetes for 3 months, aged 18 to 80 years, and admitted to medical general services with a blood glucose level between 140 mg/dL and 400 mg/dL. Diabetes treatment with either diet alone or any combination of oral antidiabetic agents and the absence of diabetic ketoacidosis Exclusion criteria: people without a known history of diabetes, intensive care unit patients, the use of corticosteroid therapy, patients expected to undergo surgery during the hospitalisation course, people with clinically relevant hepatic disease, serum creatinine 3.0 mg/dL, pregnancy, and any mental condition rendering the person unable to understand the scope and possible consequences of the study Diagnostic criteria: medical history of diabetes and with a blood glucose level between 140 and 400 mg/dL Setting: hospital Age group: 18 to 80 years Gender distribution: females and males Country where trial was performed: USA |
|
| Interventions |
Intervention(s): sliding scale insulin Discontinue oral antidiabetic drugs on admission
Comparator(s): basal bolus insulin (glargine and glulisine) Discontinue oral antidiabetic drugs on admission Start total daily insulin dose: 0.4 units/kg body weight day 1 when the admission blood glucose concentration is 140 to 200 mg/dL; 0.5 units/kg body weight day 1 when the admission blood glucose concentration is between 201 and 400 mg/dL. Give 1/2 of total daily dose as insulin glargine and 1/2 as insulin glulisine Give insulin glargine once daily at the same time of day Give insulin glulisine in 3 equally divided doses before each meal. Hold insulin glulisine if participant is not able to eat. Supplemental insulin: give supplemental insulin glulisine following the “sliding‐scale” protocol for blood glucose 140 mg/dL
Duration of intervention: 5 days Duration of follow‐up: 5 days Run‐in period: September 2005 to June 2006 Number of study centres: 1 Treatment before study: no description Titration period: no description |
|
| Outcomes | Outcomes reported: the primary endpoint was to determine differences in glycaemic control between treatment groups as measured by the mean daily blood glucose concentration. Secondary outcomes included differences between treatment groups in number of hypoglycaemic events, number of episodes of severe hyperglycaemia, length of hospital stay, and mortality rate | |
| Study registration |
Trial identifier:NCT00394407 Study terminated early: no |
|
| Publication details |
Language of publication: English (Commercial and non‐commercial) funding: Sanofi‐Aventis, American Heart Association (0555306B), the National Institutes of Health (R03 DK07319001), and the General Clinical Research Center (M01 RR‐00039) Publication status: peer review journal |
|
| Stated aim for study | Quote from publication: "This study will compare how well regular insulin will compare to glargine (Lantus) once daily plus glulisine (Apidra) insulin before meals in hospitalised people with type 2 diabetes and elevated blood glucose (sugar) levels. Lantus is a long‐acting insulin which is given subcutaneously (under the skin) once daily. Apidra is a rapid‐acting insulin which is given subcutaneously several times a day and frequently before meals. Regular insulin is a short‐acting insulin in clinical use for more than 20 years that is also given subcutaneously several times per day. Lantus, Apidra and regular insulins are approved for use in the treatment of individuals with diabetes by the FDA" | |
| Notes | ‐ | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Unclear risk |
Quote from publication: "In this multicenter, prospective, open‐label, randomized study ..." Comment: insufficient detail |
| Allocation concealment (selection bias) | Unclear risk | Comment: not described |
| Blinding of participants and personnel (performance bias) All‐cause mortality | Low risk |
Quote from publication: "open‐label" Comment: outcome unlikely to be affected by lack of blinding |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk | Comment: open‐label, outcome likely influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: open‐label, outcome unlikely to be influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk | Comment: open‐label, outcome measure likely influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk | Comment: open‐label, outcome likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Severe hypoglycaemic episodes | High risk | Comment: author reply: "We did not have an independent adjudication board or member in our studies"; outcome measure likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) All‐cause mortality | Low risk | Comment: author reply: "We did not have an independent adjudication board or member in our studies"; outcome measure unlikely to be influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Non‐severe hypoglycaemic episodes | High risk | Comment: author reply: "We did not have an independent adjudication board or member in our studies"; outcome measure likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk | Comment: author reply: "We did not have an independent adjudication board or member in our studies"; outcome measure unlikely to be influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Mean glucose level | High risk | Comment: author reply: "We did not have an independent adjudication board or member in our studies"; outcome measure likely influenced by lack of blinding |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) All‐cause mortality | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: no missing data |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: no missing data |
| Selective reporting (reporting bias) | Low risk | Comment: none detected by ORBIT criteria |
| Other bias | Low risk | Comment: none detected. |
Umpierrez 2011.
| Methods | Parallel randomised controlled trial | |
| Participants |
Inclusion criteria: general elective or emergency surgery and not expected to require ICU admission were eligible for inclusion. Glucose (BG) level between 140 mg/dL and 400 mg/dL, had a history of diabetes for more than 3 months, aged 18 to 80 years old, treated with diet alone, any combination of oral antidiabetic agents, or low‐dose insulin therapy at a daily dose of 0.4 units/kg before admission Exclusion criteria: hyperglycaemia without a known history of diabetes, cardiac surgery, clinically relevant hepatic disease or impaired renal function (serum creatinine > 3.0 mg/dL), history of diabetic ketoacidosis (21), pregnancy, or any mental condition rendering the participant unable to give informed consent Diagnostic criteria: people with a BG level between 140 mg/dL and 400 mg/dL with a history of diabetes mellitus for more than 3 months Setting: hospital Age group: adults Gender distribution: females and males Country where trial was performed: USA |
|
| Interventions |
Intervention(s): sliding scale insulin Participants received 4 times daily for BG 140 mg/dL. The doses of insulin were adjusted according to a prespecified protocol. For participants receiving SSI, if the mean daily BG level was > 240 mg/dL, or if 3 consecutive values were 240 mg/dL on the maximal sliding scale dose, participants were switched to basal‐bolus regimen starting at a TDD of 0.5 units/kg Comparator(s): basal bolus insulin Participants treated with basal‐bolus therapy were started at a TDD of 0.5 units/kg divided half as insulin glargine once daily and the other half as insulin glulisine given before meals. The insulin TDD was reduced to 0.3 units/kg in participants ≥ 70 years of age and/or with a serum creatinine ≥ 2.0 mg/dL Duration of intervention: 9 days Duration of follow‐up: 9 days Run‐in period: February 2008 to October 2009 Number of study centres: 3 Treatment before study: diet alone, oral agents, insulin alone, insulin plus oral agents Titration period: none |
|
| Outcomes | Outcomes reported: differences between treatment groups in mean daily blood glucose concentration, composite of postoperative complications including wound infection, pneumonia, bacteraemia, respiratory failure, and acute renal failure, differences between treatment groups in any of the following measures: occurrence of mild and severe hypoglycaemia (70 mg/dL and 40 mg/dL, respectively), length of hospital stay, surgical complications (wound infection and dehiscence, bacteraemia, pneumonia, and acute renal failure defined as an increase in serum creatinine 50% of baseline or a serum creatinine 2.5 mg/dL, or both), admission to the ICU, death | |
| Study registration |
Trial identifier:NCT00596687 Study terminated early: no |
|
| Publication details |
Language of publication: English (Commercial and non‐commercial) funding: Sanofi‐Aventis, Research Grant 7‐03‐CR‐35, National Institutes of Health (NIH) Grant R03‐DK‐073190‐01 and General Clinical Research Center Grant M01 RR‐00039. DS is supported by NIH Research Grant K12‐RR‐017643 Publication status: peer review journal |
|
| Stated aim for study | To evaluate the optimal management of hyperglycaemia in patients undergoing general surgery | |
| Notes | The primary outcomes in the protocol from ClinicalTrials.gov and in the published article were not the same, and 13 participants failed treatment with SSI and were switched to the basal‐bolus scheme | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote from publication: "Computer generated block randomisation table" |
| Allocation concealment (selection bias) | Low risk | Quote from publication: "Treatment assignment was coordinated by a research pharmacist at each institution" |
| Blinding of participants and personnel (performance bias) All‐cause mortality | Low risk | Comment: open‐label, outcome unlikely to be influenced by blinding |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Comment: outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: outcome unlikely to be influenced by blinding |
| Blinding of participants and personnel (performance bias) Adverse events other than hypoglycaemic episodes | High risk | Comment: outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk | Comment: outcome likely affected by lack of blinding techniques |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk | Comment: outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) Severe hypoglycaemic episodes | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Quote from author survey: "We did not have an independent adjudication board or member in our studies" Comment: outcome likely affected by lack of blinding techniques |
| Blinding of outcome assessment (detection bias) All‐cause mortality | Low risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Quote from author survey: "We did not have an independent adjudication board or member in our studies" Comment: outcome unlikely to be affected by lack of blinding technique |
| Blinding of outcome assessment (detection bias) Non‐severe hypoglycaemic episodes | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Quote from author survey: "We did not have an independent adjudication board or member in our studies" Comment: outcome likely affected by lack of blinding technique |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Quote from author survey: "We did not have an independent adjudication board or member in our studies" Comment: outcome unlikely to be affected by lack of blinding technique |
| Blinding of outcome assessment (detection bias) Adverse events other than hypoglycaemia | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Quote from author survey: "We did not have an independent adjudication board or member in our studies" Comment: outcome likely affected by lack of blinding technique |
| Blinding of outcome assessment (detection bias) Mean glucose level | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Quote from author survey: "We did not have an independent adjudication board or member in our studies" Comment: outcome likely affected by lack of blinding technique |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk | Comment: low attrition rate not likely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) All‐cause mortality | Low risk | Comment: low attrition rate not likely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: low attrition rate not likely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: low attrition rate not likely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Adverse events other than hypoglycaemic episodes | Low risk | Comment: low attrition rate not likely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: low attrition rate not likely to have influenced outcome measure |
| Selective reporting (reporting bias) | Low risk | Comment: none detected by ORBIT criteria |
| Other bias | Low risk | Comment: none detected |
Umpierrez 2013.
| Methods |
Parallel randomised controlled trial, non‐inferiority design 2:2:1 ratio to 1 of 3 regimens: a basal‐bolus regimen with insulin glargine given once daily and glulisine before meals plus corrective doses of glulisine by sliding scale for BG > 140 mg/dL; a basal‐plus regimen with a daily dose of glargine and corrective doses of glulisine by sliding scale before meals for BG > 140 mg/dL; and SSI for BG > 140 mg/dL |
|
| Participants |
Inclusion criteria: people with a known history of type 2 diabetes (T2DM) and with a BG before randomisation between 140 mg/dL and 400 mg/dL, a known history of T2DM for > 3 months, age between 18 and 80 years, and treatment at home with diet alone, any combination of oral antidiabetic agents, or low‐dose insulin therapy at a daily dose of 0.4 units/kg before admission to general medicine and surgery services Exclusion criteria: individuals with any BG > 400 mg/dL before randomisation or with a history of hyperglycaemic crises; individuals with hyperglycaemia without a known history of diabetes; people admitted to or expected to require ICU admission, undergoing cardiac surgery, receiving corticosteroid therapy, clinically relevant hepatic disease, impaired renal function (serum creatinine > 3.0 mg/dL), history of diabetic ketoacidosis, pregnant women, or with any mental condition rendering them unable to give informed consent Diagnostic criteria: medical history of diabetes mellitus for 3 months or more and with a blood glucose before randomisation between 140 and 400 mg/dL Setting: hospital Age group: adults Gender distribution: females and males Country where trial was performed: USA |
|
| Interventions |
Intervention: sliding scale insulin, 4 times daily Comparator(s):
Duration of intervention: 6 days Duration of follow‐up: 6 days Run‐in period: January 2010 to June 2012 Number of study centres: 5 Treatment before study: diet alone, oral agents, insulin alone, insulin with oral agents Titration period: none |
|
| Outcomes | Outcomes reported in abstract of publication: difference in glycaemic control, as measured by mean daily blood glucose concentration, differences between treatment groups in any of the following measures: number of hypoglycaemic events (BG 70 and 40 mg/dL) after the first day of treatment, number of episodes of hyperglycaemia (BG 200 mg/dL) after the first day of treatment, TDD of insulin, hospital stay, hospital complications | |
| Study registration |
Trial identifier:NCT00979628 Study terminated early: no |
|
| Publication details |
Language of publication: English Commercial funding: Sanofi‐Aventis Publication status: peer review journal |
|
| Stated aim for study | Quote from publication: "The present study aims to determine which insulin treatment is best for glucose control in hospitalised people with diabetes admitted to general medicine wards" | |
| Notes | The number of participants reported in ClinicalTrials.gov differs from the number of participants reported in the publication | |
| Risk of bias | ||
| Bias | Authors' judgement | Support for judgement |
| Random sequence generation (selection bias) | Low risk | Quote from publication: "Computed‐generated randomization table to coordinate the randomization and treatment assignment" |
| Allocation concealment (selection bias) | Low risk | Quote from publication: "A research pharmacist at each institution followed a computer‐generated randomization..." |
| Blinding of participants and personnel (performance bias) All‐cause mortality | Low risk |
Quote from publication: "Open label" Comment: outcome unlikely to be influenced by blinding |
| Blinding of participants and personnel (performance bias) Non‐severe hypoglycaemic episodes | High risk | Comment: outcome likely influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Length of hospital stay | Low risk | Comment: outcome unlikely to be influenced by blinding |
| Blinding of participants and personnel (performance bias) Adverse events other than hypoglycaemic episodes | High risk | Comment: outcome likely influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Mean glucose level | High risk | Comment: outcome likely influenced by lack of blinding |
| Blinding of participants and personnel (performance bias) Severe hypoglycaemic episodes | High risk | Comment: outcome likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Severe hypoglycaemic episodes | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Author survey: "Trial authors did not have an independent adjudication board" Comment: outcome likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) All‐cause mortality | Low risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Author survey: "Trial authors did not have an independent adjudication board" Comment: outcome unlikely to be influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Non‐severe hypoglycaemic episodes | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Author survey: "Trial authors did not have an independent adjudication board" Comment: outcome likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Length of hospital stay | Low risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Author survey: "Trial authors did not have an independent adjudication board" Comment: outcome unlikely to be influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Adverse events other than hypoglycaemia | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Author survey: "Trial authors did not have an independent adjudication board" Comment: outcome likely influenced by lack of blinding |
| Blinding of outcome assessment (detection bias) Mean glucose level | High risk |
Quote from ClinicalTrials.gov: "masking: none (open label)" Author survey: "Trial authors did not have an independent adjudication board" Comment: outcome likely influenced by lack of blinding |
| Incomplete outcome data (attrition bias) Severe hypoglycaemic episodes | Low risk |
Quote from publication: "A total of 375 patients with T2D gave consent (210 medical and 165 surgical patients); of them, 22 patients were excluded from further analysis because they received 24 h of insulin treatment, were transferred to the ICU, or received corticosteroid therapy" Comment: low attrition rate unlikely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) All‐cause mortality | Low risk | Comment: low attrition rate unlikely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Non‐severe hypoglycaemic episodes | Low risk | Comment: low attrition rate unlikely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Length of hospital stay | Low risk | Comment: low attrition rate unlikely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Adverse events other than hypoglycaemic episodes | Low risk | Comment: low attrition rate unlikely to have influenced outcome measure |
| Incomplete outcome data (attrition bias) Mean glucose level | Low risk | Comment: low attrition rate unlikely to have influenced outcome measure |
| Selective reporting (reporting bias) | Low risk | Comment: none detected |
| Other bias | Low risk | Comment: none detected |
‐: denotes not reported
Note: where the judgement is 'unclear' and the description is blank, the trial did not report that particular outcome.
ADA: American Diabetes Association; BG: blood glucose; DM: diabetes mellitus; ICU: intensive care unit; IU: international units; IV: intravenous NPH: neutral protamine Hagedorn insulin; N/T: no trial document available; ORBIT: Outcome Reporting Bias In Trials; SC: subcutaneous; SSI: sliding scale insulin; SSRI: sliding scale regular insulin; TDD: total daily dose; TPN: total parenteral nutrition; U: units
Characteristics of excluded studies [ordered by study ID]
| Study | Reason for exclusion |
|---|---|
| Alfonso 2006 | Wrong study design: quasi‐randomised trial (participants were randomised to receive lispro or regular insulin based on whether the last digit of their Social Security number was odd or even); both groups were treated by the sliding scale scheme |
| Attia 1998 | Wrong population: individuals with ketoacidosis |
| Choi 2016 | Wrong study design: modelling study |
| Christensen 2017 | Wrong study design: systematic review |
| Connor 2015 | Wrong population: not diabetic participants |
| Dickerson 2003 | Wrong comparator: the trial compared insulin plus routine medications versus routine medications |
| Emam 2010 | Wrong setting: critical care setting |
| Fiore 2011 | Data not extractable. Bayesian model study |
| Furnary 1999 | Wrong study design: cohort study |
| Gerards 2016 | Wrong comparator: sliding scale plus other treatment |
| Gómez Cuervo 2016 | Wrong study design: systematic review |
| Hagelberg 2008 | Wrong population: critical care setting |
| Hakeam 2017 | Wrong comparator: no sliding scale in 1 of the intervention arms |
| Jackson 2014 | Wrong study design: systematic review |
| Lee 2015 | Wrong study design: systematic review |
| Mader 2014 | Wrong comparator: the trial compared insulin plus routine medications versus routine medications (combination of oral agents, insulin, or both) |
| NCT00700362 | Withdrawn trial |
| NCT01081938 | Wrong comparator: the trial compared SSI versus SSI plus basal insulin |
| NCT01594060 | Wrong study design: non‐randomised trial |
| NTC00670228 | Wrong population: only 39% of participants had diabetes at baseline, no subgroup analysis available for diabetic participants |
| Okabayashi 2009a | Wrong comparator: artificial pancreas trial |
| Okabayashi 2009b | Wrong comparator: artificial pancreas trial |
| Perera 2011 | Wrong comparator: the trial evaluated routine oral medications |
| Sabouhi 2015 | Wrong setting: critical care unit |
| Schnipper 2010 | Wrong study design: trial authors randomised the team work, not the intervention |
| Schoeffler 2005 | Wrong study design: cohort study |
| Vellanki 2015 | Wrong comparator: no sliding scale insulin |
| Walters 2006 | Wrong comparator: no comparison with sliding scale insulin |
| Zaman Huri 2014 | Wrong study design: non‐randomised trial |
SSI: sliding scale insulin
Characteristics of studies awaiting assessment [ordered by study ID]
IRCT201311118307N2.
| Methods |
Type of study: interventional Allocation: randomised using permuted block Intervention model: parallel assignment Masking: not blinded Primary purpose: treatment |
| Participants |
Condition: diabetic people with cardiovascular diseases Enrolment: 70 Inclusion criteria: people with cardiovascular diseases, with a history of diabetes of more than 3 months, blood glucose 140 to 400 mg/dL, glycated haemoglobin 6.5%, aged 18 to 80 years, who can eat, receiving oral hypoglycaemic agent and insulin with doses of 0.4 IU/kg/day or lower Exclusion criteria: hyperglycaemia without known diabetes, coronary artery bypass surgery, severe liver diseases, renal dysfunction (creatinine > 3 mg/dL), history of diabetic ketoacidosis, pregnancy, inability to fill and understand the consent form, consented patients who want to discontinue the study at any time |
| Interventions |
Intervention(s): sliding scale insulin Comparator(s): basal‐bolus insulin |
| Outcomes |
Primary outcome(s): fasting blood level of glucose in each group, blood glucose level before each meal, intensive care unit admission, mortality rate, severe hypoglycaemic episodes, recurrent myocardial infarction, heart failure Secondary outcome(s): coronary care unit admission, mild hypoglycaemic episodes, length of hospitalisation Other outcome(s): ‐ |
| Study details | IRCT201311118307N2 |
| Publication details | ‐ |
| Stated aim of study | This study will be performed to compare basal‐bolus insulin with sliding scale insulin in glycaemic control of diabetic patients with cardiovascular diseases. |
| Notes | Recruitment status: complete Investigator: Dr Taher Entezari‐maleki, Tabriz University of Medical Science, sDaneshgah Street Tabriz (Islamic Republic of Iran), tentezari@gmail.com We contacted the authors but have received no response. |
ISRCTN55845280.
| Methods |
Type of study: interventional Allocation: single‐centre randomised Intervention model: parallel assignment Masking: open trial Primary purpose: treatment |
| Participants |
Condition: acute medical or surgical condition Enrolment: 250 (target sample size) Inclusion criteria: adult inpatients (aged > 18 years, men and women) with an acute medical or surgical condition, and any of the following criteria:
Exclusion criteria:
|
| Interventions |
Group 1: intensive insulin therapy with basal NPH and nutritional regular insulin consisting of:
Group 2: intensive insulin therapy with basal glargine and nutritional regular insulin consisting of:
Group 3: standard care consisting of regular insulin 4 times a day or NPH twice daily, according to a sliding scale |
| Outcomes |
Primary outcome measures: number of measurements of blood glucose levels between 70 and 130 mg/dL; all measurements will be recorded at day 7 and at the moment of discharge Secondary outcome measures: mean blood glucose level, length of hospitalisation, healthcare‐associated infections, number of measurements of blood glucose levels < 70 mg/dL, time to achieve good glycaemic control (≥ 2 measurements between 70 and 130 mg/dL) |
| Study details | ISRCTN55845280 |
| Publication details | ‐ |
| Stated aim of study | This trial will assess whether an adjustable insulin scheme which simulates the physiological secretion of this hormone is superior to standard treatment in medical inpatients with hyperglycaemia |
| Notes | Overall trial status: completed Investigator: Dr Luis Rojas, Marcoleta 367, Quinto Piso, Santiago, Chile, lrojas@med.puc.cl We contacted the authors but have received no response |
NPH: neutral protamine Hagedorn insulin
Characteristics of ongoing studies [ordered by study ID]
NCT01143948.
| Trial name or title | Acronym: VIRDICT (Various Insulin Regimens for Diabetic Inpatients with Cirrhosis Trial) |
| Methods |
Type of study: interventional Allocation: randomised Intervention model: parallel assignment Masking: open‐label Primary purpose: treatment |
| Participants |
Condition: type 2 diabetes, liver cirrhosis Enrolment: 45 Inclusion criteria: hospitalised type 2 diabetes mellitus with liver cirrhosis and entry blood glucose (fasting or random) > 180 mg/dL Exclusion criteria: type 1 diabetes mellitus, pregnancy, steroids: prednisone > 7.5 mg/day or equivalent, serum creatinine > 3 mg/dL, people in intensive care units |
| Interventions |
Intervention(s): regular insulin Comparator(s): NPH and bolus regular insulin and basal glargine and bolus glulisine |
| Outcomes |
Primary outcome(s): optimal glycaemic control in inpatients with cirrhosis (time frame: 3 to 7 days) Secondary outcome(s): incidence of hypoglycaemic episodes (time frame: 3 to 7 days) Other outcome(s): ‐ |
| Starting date |
Study start date: January 2011 Estimated study completion date: May 2011 |
| Contact information | Responsible party/principal investigator: Mostafa M Alfishawy, MB BCh; mostafa.alfishawy@yahoo.com |
| Study identifier | NCT01143948 |
| Official title | Insulin therapy in the inpatient management of cirrhotic patients with type 2 diabetes |
| Stated purpose of study | Best management of hyperglycaemia in non‐intensive care unit cirrhotic patients with type 2 diabetes by different insulin regimens: either sliding scale regular insulin or basal‐bolus insulin using NPH and regular insulin or glargine and glulisine |
| Notes | Recruitment status: unknown Investigators: Ibtisam Z Eissa, MD, Mary N Rizk, MD, Ahmad A Khairy, MD; all Cairo University We contacted the authors but have received no response |
NPH: neutral protamine Hagedorn (isophane) insulin
Differences between protocol and review
It was not possible to evaluate the outcome measure glycated haemoglobin because no trial followed participants over the prolonged period of time necessary to establish reliable glycated haemoglobin measures (three months). We instead used mean blood glucose values during the hospital stay of the participants. In addition to blood glucose measurements we also reported hyperglycaemic ranges, where provided, and the total daily insulin dose applied.
For the outcome hypoglycaemic episodes, we have reported only severe hypoglycaemic episodes in the Abstract, Plain language summary, and 'Summary of findings' table because this outcome measure is less prone to performance and detection bias compared to non‐severe hypoglycaemic episodes.
Contributions of authors
Luis E Colunga‐Lozano (LECL): protocol draft, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation, review draft, and future review updates.
Franscisco Javier Gonzalez Torres (FGT): data extraction, data analysis, review draft.
Netzahualpilli Delgado‐Figueroa (NDF): protocol draft, acquiring trial reports, trial selection, data extraction, data analysis, data interpretation, review draft.
Daniel A Gonzalez‐Padilla (DGP): trial selection, data extraction, data analysis, data interpretation, review draft.
Adrian V Hernandez (AVH): protocol draft, trial selection, data extraction, data analysis, data interpretation, review draft.
Yuani Roman (YR): protocol draft, trial selection, data extraction, data analysis, data interpretation, review draft.
Carlos A Cuello‐García (CCG): protocol draft, trial selection, data extraction, data analysis, data interpretation, review draft.
Sources of support
Internal sources
-
Universidad de Guadalajara, Centro Universitario de Ciencias de la Salud., Mexico.
Library services to obtain the articles
-
McMaster University, Canada.
Library services to obtain the articles
External sources
No sources of support supplied
Declarations of interest
Luis E Colunga‐Lozano: none known
Franscisco Javier Gonzalez Torres: none known
Netzahualpilli Delgado‐Figueroa: none known
Daniel A Gonzalez‐Padilla: none known
Adrian V Hernandez: none known
Yuani Roman: none known
Carlos A Cuello‐García: The American Society of Hematology and the American College of Rheumatology paid for the institution and I worked as a policy development consultant. I received a grant from the Cochrane Methods Innovation Fund, which was paid to my institution.
New
References
References to studies included in this review
Datta 2007 {published data only}
- Datta S, Qaadir A, Villanueva G. Sliding scale regular insulin for control of hyperglycemia after a bariatric surgical procedure. Endocrine Practice 2007;13(3):225‐31. [DOI] [PubMed] [Google Scholar]
Gustafson 2002 {published data only}
- Gustafson PA, Zarro DL, Palanzo DA, Manley NJ, Montesano RM, Quinn M, et al. Conventional approach to glucose management for diabetic patients undergoing coronary artery bypass surgery. Perfusion 2002;17(2):141‐4. [DOI: 10.1191/0267659102pf539oa] [DOI] [PubMed] [Google Scholar]
Korytkowski 2009 {published data only}
- Korytkowski MT, Salata RJ, Koerbel GL, Selzer F, Karslioglu E, Idriss AM, et al. Insulin therapy and glycemic control in hospitalized patients with diabetes during enteral nutrition therapy: a randomized controlled clinical trial. Diabetes Care 2009;32(4):594‐6. [DOI: 10.2337/dc08-1436] [DOI] [PMC free article] [PubMed] [Google Scholar]
- NCT00177398. Effect of Glargine insulin on glucose control in hospitalized patients who receive tube feedings. clinicaltrials.gov/ct2/show/NCT00177398 (accessed 15 June 2018).
Said 2013 {published data only}
- NCT01855243. Comparison on efficacy and safety of three inpatient insulin therapy in type2 DM. clinicaltrials.gov/ct2/show/NCT01855243 (accessed 15 June 2018).
- Said E, Farid S, Sabry N, Fawzi M. Comparison on efficacy and safety of three inpatient insulin regimens for management of non‐critical patients with type 2 diabetes. Pharmacology & Pharmacy 2013;4:556‐65. [DOI: 10.4236/pp.2013.47080] [DOI] [Google Scholar]
Schroeder 2012 {published data only}
- Schroeder JE, Liebergall M, Raz I, Egleston R, Ben Sussan G, Peyser A, et al. Benefits of a simple glycaemic protocol in an orthopaedic surgery ward: a randomized prospective study. Diabetes/Metabolism Research and Reviews 2012;28(1):71‐5. [DOI: 10.1002/dmrr.1217] [DOI] [PubMed] [Google Scholar]
Umpierrez 2007a {published data only}
- NCT00394407. Basal/bolus versus sliding scale insulin in hospitalized patients with type 2 diabetes. clinicaltrials.gov/ct2/show/NCT00394407 (accessed 15 June 2018).
- Umpierrez GE, Smiley D, Zisman A, Prieto LM, Palacio A, Ceron M, et al. Randomized study of basal‐bolus insulin therapy in the inpatient management of patients with type 2 diabetes (RABBIT 2 trial). Diabetes Care 2007;30(9):2181‐6. [DOI: 10.2337/dc07-0295] [DOI] [PubMed] [Google Scholar]
Umpierrez 2011 {published data only}
- NCT00596687. Basal bolus insulin versus SSRI in type 2 diabetes undergoing general surgery (RABBIT 2‐SX). clinicaltrials.gov/ct2/show/NCT00596687 (accessed 15 June 2018).
- Umpierrez GE, Smiley D, Jacobs S, Peng L, Temponi A, Mulligan P, et al. Randomized study of basal‐bolus insulin therapy in the inpatient management of patients with type 2 diabetes undergoing general surgery (RABBIT 2 surgery). Diabetes Care 2011;34(2):256‐61. [DOI: 10.2337/dc10-1407] [DOI] [PMC free article] [PubMed] [Google Scholar]
Umpierrez 2013 {published data only}
- NCT00979628. Basal bolus versus basal insulin in type 2 diabetes mellitus (T2DM) (Basal‐Plus). clinicaltrials.gov/ct2/show/NCT00979628 (accessed 15 June 2018).
- Umpierrez GE, Smiley D, Hermayer K, Khan A, Olson DE, Newton C, et al. Randomized study comparing a basal bolus with a basal plus correction insulin regimen for the hospital management of medical and surgical patients with type 2 diabetes: basal plus trial. Diabetes Care 2013;36:2169–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
References to studies excluded from this review
Alfonso 2006 {published data only}
- Alfonso A, Koops MK, Mong DP, Vigersky RA. Glycemic control with regular versus lispro insulin sliding scales in hospitalized type 2 diabetics. Journal of Diabetes and Its Complications 2006;20:153‐7. [DOI: 10.1016/j.jdiacomp.2005.06.009] [DOI] [PubMed] [Google Scholar]
Attia 1998 {published data only}
- Attia N, Holcombe J, Jones TW, Tamborlane WV. Comparison of human regular and lispro insulins after interruption of continuous subcutaneous insulin infusion and in the treatment of acutely decompensated IDDM. Diabetes Care 1998;21:818‐21. [DOI] [PubMed] [Google Scholar]
Choi 2016 {published data only}
- Choi K, Oh TJ, Lee JC, Kim M, Kim HC, Cho YM, et al. In‐silico trials for glucose control in hospitalized patients with type 2 diabetes. Journal of Korean Medical Science 2016;31(2):231‐9. [DOI: 10.3346/jkms.2016.31.2.231] [DOI] [PMC free article] [PubMed] [Google Scholar]
Christensen 2017 {published data only}
- Christensen MB, Gotfredsen A, Nørgaard K. Efficacy of basal‐bolus insulin regimens in the inpatient management of non‐critically ill patients with type 2 diabetes: a systematic review and meta‐analysis. Diabetes/Metabolism Research and Reviews 2017;33:e2885. [DOI: 10.1002/dmrr.2885] [DOI] [PubMed] [Google Scholar]
Connor 2015 {published data only}
- Connor JT, Broglio KR, Durkalski V, Meurer WJ, Johnston KC. The stroke hyperglycemia insulin network effort (SHINE) trial: an adaptive trial design case study. Trials 2015;4(16):72. [DOI: 10.1186/s13063-015-0574-8] [DOI] [PMC free article] [PubMed] [Google Scholar]
Dickerson 2003 {published data only}
- Dickerson LM, Ye X, Sack JL, Hueston WJ. Glycemic control in medical inpatients with type 2 diabetes mellitus receiving sliding scale insulin regimens versus routine diabetes medications: a multicenter randomized controlled trial. Annals of Family Medicine 2003;1(1):29‐35. [DOI: 10.1370/afm.2] [DOI] [PMC free article] [PubMed] [Google Scholar]
Emam 2010 {published data only}
- Emam IA, Allan A, Eskander K, Dhanraj K, Farag el‐S, El‐Kadi Y, et al. Our experience of controlling diabetes in the peri‐operative period of patients who underwent cardiac surgery. Diabetes Research and Clinical Practice 2010;88(3):242‐6. [DOI] [PubMed] [Google Scholar]
Fiore 2011 {published data only}
- Fiore LD, Brophy M, Ferguson RE, D'Avolio L, Hermos JA, Lew RB, et al. A point‐of‐care clinical trial comparing insulin administered using a sliding scale versus a weight‐based regimen. Clinical Trials 2011;8:183‐95. [DOI: 10.1177/1740774510395635] [DOI] [PMC free article] [PubMed] [Google Scholar]
Furnary 1999 {published data only}
- Furnary AP, Zerr KJ, Grunkemeier GL, Starr A. Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Annals of Thoracic Surgery 1999;67:352‐62. [DOI: 10.1016/S0003-4975(99)00014-4] [DOI] [PubMed] [Google Scholar]
Gerards 2016 {published data only}
- Gerards MC, Maar JS, Steenbruggen TG, Hoekstra JB, Vriesendorp TM, Gerdes VE. Add‐on treatment with intermediate‐acting insulin versus sliding‐scale insulin for patients with type 2 diabetes or insulin resistance during cyclic glucocorticoid‐containing antineoplastic chemotherapy: a randomized crossover study. Diabetes, Obesity & Metabolism 2016;18(10):1041‐4. [DOI] [PubMed] [Google Scholar]
Gómez Cuervo 2016 {published data only}
- Gómez Cuervo C, Sánchez Morla A, Pérez‐Jacoiste Asín MA, Bisbal Pardo O, Pérez Ordoño L, Vila Santos J. Effective adverse event reduction with bolus‐basal versus sliding scale insulin therapy in patients with diabetes during conventional hospitalization: systematic review and meta‐analysis. Endocrinologia y Nutricion 2016;63(4):145‐56. [DOI] [PubMed] [Google Scholar]
Hagelberg 2008 {published data only}
- Hagelberg A, Ivert T, Efendić S, Ohrvik J, Anderson RE. Insulin glargine improves glycaemic control after coronary surgery in patients with diabetes or pre‐diabetes. Scandinavian Cardiovascular Journal: SCJ 2008;42(1):71‐6. [DOI: 10.1080/14017430701721756] [DOI] [PubMed] [Google Scholar]
Hakeam 2017 {published data only}
- Hakeam HA, Mulia HA, Azzam A, Amin T. Glargine insulin use versus continuous regular insulin in diabetic surgical noncritically ill patients receiving parenteral nutrition: randomized controlled study. JPEN: Journal of Parenteral and Enteral Nutrition 2017;41(7):1110‐8. [DOI: 10.1177/014860711664471] [DOI] [PubMed] [Google Scholar]
- NCT02216799. Glargine insulin vs. continuous regular insulin in diabetic surgical patients receiving parenteral nutrition (GLUCOSE‐in‐PN). clinicaltrials.gov/ct2/show/NCT02216799 (accessed 1 February 2018).
Jackson 2014 {published data only}
- Jackson B, Grubbs L. Basal‐bolus insulin therapy and glycemic control in adult patients with type 2 diabetes mellitus: a review of the literature. Journal of the American Association of Nurse Practitioners 2014;26(6):348‐52. [DOI] [PubMed] [Google Scholar]
Lee 2015 {published data only}
- Lee YY, Lin YM, Leu WJ, Wu MY, Tseng JH, Hsu MT, et al. Sliding‐scale insulin used for blood glucose control: a meta‐analysis of randomized controlled trials. Metabolism: Clinical and Experimental 2015;64(9):1183‐92. [DOI] [PubMed] [Google Scholar]
Mader 2014 {published data only}
- Mader JK, Neubauer KM, Shaupp L, Augustin T, Beck P, Spat S, et al. Efficacy, usability and sequence of operations of a work flow‐integrated algorithm for basal–bolus insulin therapy in hospitalized type 2 diabetes patients. Diabetes, Obesity and Metabolism 2014;16:137‐46. [DOI: 10.1111/dom.12186] [DOI] [PubMed] [Google Scholar]
NCT00700362 {published data only}
- NCT00700362. Insulin infusion and infectious diabetes foot ulcers (IIIFU). clinicaltrials.gov/ct2/show/NCT00700362 (accessed 22 October 2018).
NCT01081938 {published and unpublished data}
- NCT01081938. Insulin glargine for diabetes metabolism (DM) type II patients under enteral nutrition. clinicaltrials.gov/ct2/show/NCT01081938 (accessed 22 October 2018).
NCT01594060 {unpublished data only}
- NCT01594060. Basal bolus vs. sliding scale for treatment of diabetic patients in medical wards. clinicaltrials.gov/ct2/show/NCT01594060 (accessed 1 February 2018).
NTC00670228 {unpublished data only}
- NCT00670228. Effect of Lantus and Apidra in patients with acute ST elevation myocardial infarction (INTENSIVE). clinicaltrials.gov/ct2/show/NCT00670228 (accessed 1 February 2018).
Okabayashi 2009a {published data only}
- Okabayashi T, Nishimori I, Maeda H, Yamashita K, Yatabe T, Hanazaki K. Effect of intensive insulin therapy using a closed‐loop glycemic control system in hepatic resection patients: a prospective randomized clinical trial. Diabetes Care 2009;32(8):1425‐7. [DOI: 10.2337/dc08-2107.] [DOI] [PMC free article] [PubMed] [Google Scholar]
Okabayashi 2009b {published data only}
- Okabayashi T, Nishimori I, Yamashita K, Sugimoto T, Maeda H, Yatabe T, et al. Continuous postoperative blood glucose monitoring and control by artificial pancreas in patients having pancreatic resection: a prospective randomized clinical trial. Archives of Surgery 2009;144(10):933‐7. [DOI: 10.1001/archsurg.2009.176.] [DOI] [PubMed] [Google Scholar]
Perera 2011 {published data only}
- Perera NJ, Harding AJ, Constantino MI, Molyneaux L, McGill M, Chua EL, et al. Triple‐B (basal‐bolus‐booster) subcutaneous insulin regimen: a pragmatic approach to managing hospital inpatient hyperglycaemia. Practical Diabetes 2011;28(6):266‐9. [DOI: 10.1002/pdi.1612] [DOI] [Google Scholar]
Sabouhi 2015 {published data only}
- Sabouhi F, Maleki M, Amini M, Kerdegari M. Impacts of insulin infusion protocol on blood glucose level and outcomes in acute coronary syndrome patients with diabetes mellitus. Iranian Journal of Nursing and Midwifery Research 2015;20(3):304‐8. [PMC free article] [PubMed] [Google Scholar]
Schnipper 2010 {published data only}
- Schnipper JL, Lian CL, Ndumele CD, Pendergrass ML. Effects of a computerized order set on the inpatient management of hyperglycemia: a cluster‐randomized controlled trial. Endocrine Practice 2010;16(2):209‐19. [DOI: 10.4158/EP09262.OR] [DOI] [PubMed] [Google Scholar]
Schoeffler 2005 {published data only}
- Schoeffler JM, Rice DA, Gresham DG. 70/30 insulin algorithm versus sliding scale insulin. Annals of Pharmacotherapy 2005;39(10):1606‐10. [DOI] [PubMed] [Google Scholar]
Vellanki 2015 {published data only}
- Vellanki P, Bean R, Oyedokun FA, Pasquel FJ, Smiley D, Farrokhi F, et al. Randomized controlled trial of insulin supplementation for correction of bedtime hyperglycemia in hospitalized patients with type 2 diabetes. Diabetes Care 2015;38(4):568‐74. [DOI] [PMC free article] [PubMed] [Google Scholar]
Walters 2006 {published data only}
- Walters MR, Weir CJ, Lees KR. A randomised, controlled pilot study to investigate the potential benefit of intervention with insulin in hyperglycaemic acute ischaemic stroke patients. Cerebrovascular Diseases 2006;22(2‐3):116‐22. [DOI: 10.1159/000093239] [DOI] [PubMed] [Google Scholar]
Zaman Huri 2014 {published data only}
- Zaman Huri H, Permalu V, Vethakkan SR. Sliding‐scale versus basal‐bolus insulin in the management of severe or acute hyperglycaemia in type 2 diabetes patients: a retrospective study. PLOS ONE 2014;2(9):e106505. [DOI: 10.1371/journal.pone.0106505] [DOI] [PMC free article] [PubMed] [Google Scholar]
References to studies awaiting assessment
IRCT201311118307N2 {unpublished data only}
- IRCT201311118307N2. Basal/bolus versus sliding scale insulin in glycaemic control of diabetic patients with cardiovascular diseases. apps.who.int/trialsearch/Trial3.aspx?trialid=IRCT201311118307N2 (accessed 1 February 2018).
ISRCTN55845280 {published data only}
- ISRCTN55845280. Intensive versus conventional insulin therapy for inpatient medical patients: a single centre randomized open trial [Insulinoterapia intensificada versus habitual en pacientes hospitalizador por patologia medica]. www.isrctn.com/ISRCTN55845280 (accessed 22 October 2018).
References to ongoing studies
NCT01143948 {unpublished data only}
- NCT01143948. Various insulin regimens for diabetic inpatients with cirrhosis trial (VIRDICT). clinicaltrials.gov/ct2/show/NCT01143948 (accessed 1 February 2018).
Additional references
ADA 1999
- The Expert Committee on the Diagnosis and Classification of Diabetes Mellitus, the American Diabetes Association. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1999;22(Suppl 1):S5‐19. [DOI] [PubMed] [Google Scholar]
ADA 2008
- American Diabetes Association. Standards of medical care in diabetes ‐ 2008. Diabetes Care 2008;31(Suppl 1):S12‐54. [PUBMED: 18165335] [DOI] [PubMed] [Google Scholar]
Altman 2003
- Altman DG, Bland JM. Interaction revisited: the difference between two estimates. BMJ 2003;326(7382):219. [PUBMED: 12543843] [DOI] [PMC free article] [PubMed] [Google Scholar]
Baldwin 2005
- Baldwin D, Villanueva G, McNutt R, Bhatnagar S. Eliminating inpatient sliding‐scale insulin: a reeducation project with medical house staff. Diabetes Care 2005;28:1008‐11. [DOI: 10.2337/diacare.28.5.1008; PUBMED: 15855558] [DOI] [PubMed] [Google Scholar]
Bell 2013
- Bell ML, McKenzie JE. Designing psycho‐oncology randomised trials and cluster randomised trials: variance components and intra‐cluster correlation of commonly used psychosocial measures. Psycho‐Oncology 2013;22:1738‐47. [DOI] [PubMed] [Google Scholar]
Borenstein 2017a
- Borenstein M, Higgins JP, Hedges LV, Rothstein HR. Basics of meta‐analysis: I² is not an absolute measure of heterogeneity. Research Synthesis Methods 2017;8(1):5‐18. [DOI] [PubMed] [Google Scholar]
Borenstein 2017b
- Borenstein M. Prediction intervals. www.meta‐analysis.com/prediction (accessed 3 July 2017).
Braithwaite 2008
- Braithwaite SS. Inpatient insulin therapy. Current Opinion in Endocrinology, Diabetes & Obesity 2008;15(2):159‐66. [DOI: 10.1097/MED.0b013e3282f827e7; PUBMED: 18316952] [DOI] [PubMed] [Google Scholar]
Buch 2011
- Buch MH, Aletaha D, Emery P, Smolen JS. Reporting of long‐term extension studies: lack of consistency calls for consensus. Annals of the Rheumatic Diseases 2011;70(6):886‐90. [DOI] [PubMed] [Google Scholar]
Buchleitner 2012
- Buchleitner AM, Martínez‐Alonso M, Hernández M, Solà I, Mauricio D. Perioperative glycaemic control for diabetic patients undergoing surgery. Cochrane Database of Systematic Reviews 2012, Issue 9. [DOI: 10.1002/14651858.CD007315.pub2] [DOI] [PubMed] [Google Scholar]
Childs 2003
- Childs BP. Death to the sliding scale!. Diabetes Spectrum 2003;16(2):68‐9. [DOI: 10.2337/diaspect.16.2.68] [DOI] [Google Scholar]
Deeks 2017
- Deeks JJ, Higgins JPT, Altman DG (editors), on behalf of the Cochrane Statistical Methods Group. Chapter 9: Analysing data and undertaking meta‐analyses. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors). Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017). The Cochrane Collaboration, 2017. Available at www.training.cochrane.org/handbook.
Farrokhi 2012
- Farrokhi F, Klindukhova O, Chandra P, Peng L, Smiley D, Newton C, et al. Risk factors for inpatient hypoglycemia during subcutaneous insulin therapy in non‐critically ill patients with type 2 diabetes. Journal of Diabetes Science and Technology 2012;6:1022‐9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Friedrich 2010
- Friedrich JO, Chant C, Adhikari NK. Does intensive insulin therapy really reduce mortality in critically ill surgical patients? A reanalysis of meta‐analytic data. Critical Care 2010;14(5):324. [DOI: 10.1186/cc9240; PUBMED: 21062514] [DOI] [PMC free article] [PubMed] [Google Scholar]
GRADEpro GDT 2015 [Computer program]
- McMaster University (developed by Evidence Prime). GRADEpro GDT. Version accessed 17 September 2017. Hamilton (ON): McMaster University (developed by Evidence Prime), 2015.
Griesdale 2009
- Griesdale DE, Souza RJ, Dam RM, Heyland DK, Cook DJ, Malhotra A, et al. Intensive insulin therapy and mortality among critically ill patients: a meta‐analysis including NICE‐SUGAR study data. Canadian Medical Association Journal 2009;180(8):821‐7. [DOI: 10.1503/cmaj.090206; PUBMED: 19318387] [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2002
- Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta‐analysis. Statistics in Medicine 2002;21:1539‐58. [DOI] [PubMed] [Google Scholar]
Higgins 2003
- Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta‐analysis. BMJ 2003;327:557‐60. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2009
- Higgins JPT, Thompson SG, Spiegelhalter DJ. A re‐evaluation of random‐effects meta‐analysis. Journal of the Royal Statistical Society: Series A (Statistics in Society) 2009;172(1):137‐59. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2011a
- Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. [DOI] [PMC free article] [PubMed] [Google Scholar]
Higgins 2011b
- Higgins JPT, Deeks JJ, Altman DG (editors). Chapter 16: Special topics in statistics. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from handbook.cochrane.org.
Higgins 2017
- Higgins JPT, Altman DG, Sterne JAC (editors). Chapter 8: Assessing risk of bias in included studies. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017). The Cochrane Collaboration, 2017. Available at www.training.cochrane.org/handbook.
Hozo 2005
- Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Medical Research Methodology 2005;5:13. [DOI: 10.1186/1471-2288-5-13] [DOI] [PMC free article] [PubMed] [Google Scholar]
Hróbjartsson 2013
- Hróbjartsson A, Thomsen AS, Emanuelsson F, Tendal B, Hilden J, Boutron I, et al. Observer bias in randomized clinical trials with measurement scale outcomes: a systematic review of trials with both blinded and non blinded assessors. Canadian Medical Association Journal 2013;185(4):E201‐11. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kirkham 2010
- Kirkham JJ, Dwan KM, Altman DG, Gamble C, Dodd S, Smyth R, et al. The impact of outcome reporting bias in randomised controlled trials on a cohort of systematic reviews. BMJ 2010;340:c365. [DOI: 10.1136/bmj.c365] [DOI] [PubMed] [Google Scholar]
Klonoff 2011
- Klonoff DC. Intensive insulin therapy in critically ill hospitalized patients: making it safe and effective. Journal of Diabetes Science and Technology 2011;5:755‐67. [DOI] [PMC free article] [PubMed] [Google Scholar]
Liberati 2009
- Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, et al. The PRISMA statement for reporting systematic and meta‐analyses of studies that evaluate interventions: explanation and elaboration. PLoS Medicine 2009;6(7):1‐28. [DOI: 10.1371/journal.pmed.1000100] [DOI] [PMC free article] [PubMed] [Google Scholar]
Lien 2011
- Lien LF, Cox ME, Feinglos MN, Corsino L. Glycemic control in the hospitalized patient. In: Lien LF, Cox ME, Feinglos MN, Corsino L editor(s). Springer Science + Business Media, LLC. 2011. Vol. 1, New York: Springer New York, 2011:1‐145. [DOI: 10.1007/978-1-60761-006-9; 978‐1‐60761‐006‐9] [DOI] [Google Scholar]
Mathieu 2009
- Mathieu S, Boutron I, Moher D, Altman DG, Ravaud P. Comparison of registered and published primary outcomes in randomized controlled trials. JAMA 2009;302:977‐84. [DOI] [PubMed] [Google Scholar]
Meader 2014
- Meader N, King K, Llewellyn A, Norman G, Brown J, Rodgers M, et al. A checklist designed to aid consistency and reproducibility of GRADE assessments: development and pilot validation. Systematic Reviews 2014;3:82. [DOI] [PMC free article] [PubMed] [Google Scholar]
Megan 2012
- Megan B, Pickering RM, Weatherall M. Design, objectives, execution and reporting of published open‐label extension studies. Journal of Evaluation in Clinical Practice 2012;18(2):209‐15. [DOI] [PubMed] [Google Scholar]
Miller 2011
- Miller DB. Why won't the sliding scale go away?. Canadian Journal of Diabetes 2011;35(4):340‐3. [DOI: 10.1016/S1499-2671(11)54010-9] [DOI] [Google Scholar]
Moss 1999
- Moss SE, Klein R, Klein BE. Risk factors for hospitalization in people with diabetes. Archives of Internal Medicine 199;159(17):2053‐7. [DOI: 10.1001/archinte.159.17.2053; PUBMED: 10510991] [DOI] [PubMed] [Google Scholar]
RevMan 2014 [Computer program]
- The Nordic Cochrane Centre, The Cochrane Collaboration. Review Manager (RevMan). Version 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014.
Riley 2011
- Riley RD, Higgins JP, Deeks JJ. Interpretation of random effects meta‐analyses. BMJ 2011;342:d549. [DOI] [PubMed] [Google Scholar]
Schünemann 2017
- Schünemann HJ, Oxman AD, Higgins JPT, Vist GE, Glasziou P, Akl E, et al. on behalf of the Cochrane GRADEing Methods Group and the Cochrane Statistical Methods Group. Chapter 11: Completing ‘Summary of findings’ tables and grading the confidence in or quality of the evidence. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017). The Cochrane Collaboration, 2017. Available at www.training.cochrane.org/handbook.
Sterne 2011
- Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau J, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta‐analyses of randomised controlled trials. BMJ 2011;343:d4002. [DOI] [PubMed] [Google Scholar]
Sterne 2017
- Sterne JAC, Egger M, Moher D, Boutron I (editors). Chapter 10: Addressing reporting biases. In: Higgins JPT, Churchill R, Chandler J, Cumpston MS (editors), Cochrane Handbook for Systematic Reviews of Interventions version 5.2.0 (updated June 2017). The Cochrane Collaboration, 2017. Available at www.training.cochrane.org/handbook.
Umpierrez 2002
- Umpierrez GE, Isaacs SD, Bazargan N, You X, Thaler LM, Kitabchi AE. Hyperglycemia: an independent marker of in‐hospital mortality in patients with undiagnosed diabetes. Journal of Clinical Endocrinology and Metabolism 2002;87(3):978‐82. [DOI: 10.1210/jcem.87.3.8341; PUBMED: 11889147] [DOI] [PubMed] [Google Scholar]
Umpierrez 2007b
- Umpierrez GE, Palacio A, Smiley D. Sliding scale insulin use: myth or insanity?. American Journal of Medicine 2007;120(7):563‐7. [DOI: 10.1016/j.amjmed.2006.05.070; PUBMED: 17602924] [DOI] [PubMed] [Google Scholar]
Umpierrez 2012
- Umpierrez GE, Hellman R, Korytkowski MT, Kosiborod M, Maynard GA, Montori VM, et al. Management of hyperglycemia in hospitalized patients in non‐critical care setting: an Endocrine Society clinical practice guideline. Journal of Clinical Endocrinology and Metabolism 2012;97(1):16‐38. [DOI: 10.1210/jc.2011-2098; PUBMED: 22223765] [DOI] [PubMed] [Google Scholar]
Whiting 2011
- Whiting DR, Guariguata L, Weil C, Shaw J. IDF diabetes atlas: global estimates of the prevalence of diabetes for 2011 and 2030. Diabetes Research and Clinical Practice 2011;94(3):311‐21. [DOI: 10.1016/j.diabres.2011.10.029; PUBMED: 22079683] [DOI] [PubMed] [Google Scholar]
WHO 1998
- Alberti KM, Zimmet PZ. Definition, diagnosis and classification of diabetes mellitus and its complications. Part I: diagnosis and classification of diabetes mellitus. Provisional report of a WHO consultation. Diabetic Medicine 1998;15:539‐53. [DOI] [PubMed] [Google Scholar]
Wild 2004
- Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27(5):1047‐53. [PUBMED: 15111519] [DOI] [PubMed] [Google Scholar]
Wong 2006
- Wong SSL, Wilczynski NL, Haynes RB. Developing optimal search strategies for detecting clinically sound treatment studies in EMBASE. Journal of the Medical Library Association 2006;94(1):41‐7. [PMC free article] [PubMed] [Google Scholar]
Wood 2008
- Wood L, Egger M, Gluud LL, Schulz KF, Juni P, Altman DG, et al. Empirical evidence of bias in treatment effect estimates in controlled trials with different interventions and outcomes: meta‐epidemiological study. BMJ 2008;336(7644):601‐5. [DOI] [PMC free article] [PubMed] [Google Scholar]