Perceived satisfaction with care, service atmosphere and mood fluctuations among patients with schizophrenia spectrum disorder in Italian mental health supported accommodations: a multicenter study

Design setting and recruitment

This research is a component of the national initiative titled “DAily time use, Physical Activity, quality of care and interpersonal relationships in patients with Schizophrenia spectrum disorders (DiAPAson)” [29]. The DiAPAson project involved 20 Departments of Mental Health (each encompassing a different number of SA) and 17 private SA situated in various regions across Italy.

The inclusion criteria comprised individuals aged between 20 and 55 years, a DSM-5 diagnosis of SSD [26], and proficiency in the Italian language for effective participation in research interviews. Exclusion criteria encompassed severe cognitive impairments (e.g., Mini-Mental State Examination corrected score < 24), inability to provide informed consent, a recent (last six months) DSM-5 diagnosis of substance use disorder [30], a cerebrovascular/neurological disease, and a history of clinically significant head injury.

In the participating SA, facility leaders generated an alphabetical list of SSD patients on a specific day, and residents were consecutively invited to join the study until the recruitment goal was met. Between October 2020 and October 2021, 340 residents with SSD were enrolled from 98 SA (81 belonging to the 20 participating Departments of Mental Health, and 17 from the private sector) across Italy (12.8±5.7 residents per SA). Six patients (1.8%) were excluded due to severe cognitive deficits, and 21 (6.2%) withdrew after providing informed consent. Consequently, 313 patients were included in the original DiAPAson project, with each SA contributing an average of 3.5 ± 2.6 residents. Thus, we recruited approximately 27.0% of patients from each participating SA based on the total number of occupied beds.

Because of logistical and financial constraints, the ecological ESM study could only be conducted in a subset of SA (N = 30), which accepted to carry out this part of the project having more human resources to manage this specific research activity, and which were included in the current study. In these facilities 68 residents (21.7%) were assessed using the ESM (Fig. 1).

Assessments

Sociodemographic data (age, gender, years of schooling, marital and working status) and clinical data on the patient’s condition (length of illness, overall duration of hospitalizations, total number of hospitalizations in the last year, and time in the SA) were obtained from clinical records. Questionnaires were filled in by the treating physicians and/or by the residents with the help of research assistants, as appropriate, before starting the ESM. The monitoring ESM phase was preceded by a 30-minute briefing session delivered by research assistants, who provided detailed instructions on the procedures.

The severity of psychopathology was assessed using the 24-item Brief Psychiatric Rating Scale (BPRS) [31, 32]. Ratings for BPRS items were conducted on a 7-point Likert scale, ranging from 1 (no symptoms) to 7 (extremely severe symptoms), in response to the question “Please enter the score for the term which best describes the patient’s condition” and were categorized into five groups (depression/anxiety (e.g.,”Worry, fear, or over-concern for present or future”), excitement (e.g., “Heightened emotional tone, agitation, increased reactivity”), positive symptoms (e.g., “Perceptions without normal external stimulus correspondence”), negative symptoms (e.g., “Reduced emotional tone, apparent lack of normal feeling or involvement”), and cognitive symptoms (e.g., “Confusion or lack of proper association for person, place or time”).

To evaluate the severity of negative symptoms specifically, the Brief Negative Symptom Scale (BNSS) [33, 34] was employed. BNSS items (e.g., “intensity of pleasure during activities”) were rated on a 6-point Likert scale, spanning from 0 (no symptoms) to 6 (severe symptoms), and encompassed assessments of blunted affect, alogia, asociality, anhedonia, and avolition. In both the BPRS and BNSS, higher mean scores indicated a greater severity of symptomatology.

The investigation into the perceived ward atmosphere by both patients (WAS-P) and healthcare staff (WAS-T) utilized the “Ward Atmosphere Scale” (WAS) developed by Moos and Houts in 1968 [20]. Only the patients’ perception (WAS-P) was considered in the current study. This instrument consists of 100 dichotomous (Yes/No) items that capture various aspects of the ward. The higher-order dimensions of the scale are relational, personal growth, and system maintenance, each associated with ten corresponding subscales (e.g., “Patients put a lot of energy into what they do around here”, “Doctors have very little time to encourage patients”). These subscales, including involvement, support, spontaneity, autonomy, practical orientation, personal problem orientation, anger and aggression, order and organization, program clarity, and staff control, assess specific aspects of the ward environment. Each item is scored as 0 or 1 (Y/N), contributing to a subscale score ranging from 0 to 10. Subscale averages are then calculated to provide an overall assessment. The WAS scale, widely employed in clinical settings and program evaluations, allows for the comparison of perceptions between patients and healthcare staff [21, 35].

Satisfaction with care was assessed using the Verona Service Satisfaction Scale (VSSS-32) which comprises 32 items (e.g., “How satisfied are you with the ability of staff to listen/understand your problems?”), each rated on a five-point Likert scale (1 = not at all satisfied; 5 = extremely satisfied). The items are grouped into 7 subscales (overall satisfaction, professionals’ skills and behaviour, efficacy, types of intervention, information, access, relatives’ involvement) [36].

Psychosocial functioning was evaluated using the Specific Levels of Functioning Scale (SLOF), which measures the level of support a participant requires to carry out various tasks (e.g., “Engages in activities without prompting”). The scale includes 43 items, each rated on a 5-point Likert scale, yielding a total score between 43 and 215. Higher scores indicate better psychosocial functioning [37].

Functioning and disability were assessed with the WHO Disability Assessment Schedule (WHODAS 2.0), consisting of 12 items (e.g., “In the last 30 days how much difficulty did you have in concentrating on doing something for ten minutes?”) rated on a 5-point Likert scale (0 = no difficulty; 4 = extreme difficulty), where higher scores denote greater levels of disability [38].

Following clinical assessment and questionnaire completion, Experience Sampling Method was employed to assess daily activities, social interactions and mood with a questionnaire administered on a smartphone-based application, developed ad hoc for the project, over a seven-day period. In details, each participant received semi-randomised notifications (i.e., randomly sent within the scheduled time slots) across eight time slots (8–10 a.m., 10 a.m.–12 p.m., 12 p.m.–2 p.m., 2–4 p.m., 4–6 p.m., 6–8 p.m., 8–10 p.m., 10 p.m.–12 a.m.), each followed by a reminder notification after 15 minutes. The participants had a maximum of 30 min to reply. Mood was assessed with the question “Right now, I feel …”, with possible answer options being “happy, sad, tired, relaxed, nervous, quiet, full of energy”.

Statistical analysis

Data are presented as mean, standard deviation, median, minimum and maximum for continuous variables, and as count and percentage for categorical variables. In this analysis, a structural equation model (SEM) was developed to explore the factors influencing mood. This model incorporates observed variables and the individual’s perception of their environment as a latent construct. Univariate normality was assessed using the Shapiro–Wilk test, along with skewness and kurtosis indices, and by visually inspecting histograms and Q–Q plots. Three variables—Mood, Age, and the total score on the BPRS—displayed statistically significant but minimal deviations from normality. In contrast, the other continuous variables exhibited distribution patterns consistent with the assumption of normality. The screening for univariate and multivariate outliers did not reveal any influential outliers. Given the small sample size and the minimal departures from normality, non-parametric bootstrap standard errors were employed in the SEM to ensure robust estimation and inference, despite potential violations of distributional assumptions. The model is based on the hypothesis that a combination of demographic factors, clinical variables, and environmental perception shapes mood. Individual’s environmental perception is a latent variable called Env-Perc (environment perception), indicating a psychological dimension that cannot be directly measured but is inferred from observable indicators. In this case, two self-report instruments were used: the VSSS-32 total score for perceived social support and the WAS-P total score for environmental satisfaction. Env-Perc, derived from the combination of VSSS-32 and WAS-P, along with other variables (age, sex, illness duration, BPRS total score, BNSS total score and SLOF total score), was examined in relation to mood. Statistical analyses were performed using R software, version 4.3.1. A significance level of 0.05 was adopted for all statistical tests.

Sociodemographic and clinical characteristics of the sample

The sample included 68 participants, comprising 20 females (29.4%) and 48 males (70.6%). Their socio-demographic and clinical characteristics are reported in Table 1. Briefly, their mean age was 42.8 years (SD = 10.5). Most of them were single (80.9%), while 8.8% were married or living together, and 10.3% were divorced or widowed. On average, participants had 11.3 years of education (SD = 3.5). Regarding employment status, 16.2% were employed, 4.4% were studying, and 79.4% were unemployed. The average duration of illness was 19.2 years (SD = 10.2). In terms of lifetime psychiatric hospitalization, 54.4% had spent between 1 and 5 years in psychiatric facilities, while 45.6% had been hospitalized for more than 5 years. Additionally, 35.3% reported being smokers. In terms of clinical assessment, the mean score on BPRS (Cronbach’s alpha = 0.80) was 47.3 (SD = 12.9), and the BNSS (Cronbach’s alpha = 0.91) had a mean score of 23.3 (SD = 14.2). The SLOF (Cronbach’s alpha = 0.92), which evaluated functioning, showed a mean score of 175.7 (SD = 20.9), while the WHODAS 2.0 (Cronbach’s alpha = 0.71), which assessed disability, had a mean score of 33.3 (SD = 25.2). Service satisfaction, measured by VSSS-32 (Cronbach’s alpha = 0.89), yielded a total mean score of 3.8 (SD = 0.5), while WAS-P (Cronbach’s alpha = 0.87), which assessed perceived ward atmosphere by patients, had a mean score of 56.8 (SD = 10.3).

Results of structural equation model

The SEM developed to investigate the determinants of mood revealed that environmental perception, represented as a latent construct (Env-Perc), was the only statistically significant predictor of mood among the variables examined. This latent factor, derived from the total scores of the VSSS-32 and the WAS-P, demonstrated strong and significant factor loadings. Specifically, the WAS-P total score had a high standardized loading of 0.779 (p = 0.032), while the VSSS-32 total was fixed at 1.00 for identification purposes and showed a standardized loading of 0.709, confirming their contributions to the underlying construct of environmental perception (Fig. 2). The VSSS-32 total score was fixed at 1.00 to establish a scale for the latent construct. This approach aligns with standard procedures in SEM, where one indicator is constrained to 1. This constraint helps set the metric for the latent variable without influencing the estimation or interpretation of the other parameters.

The model fit indices indicated an excellent fit to the data: the chi-square test was non-significant (χ² = 8.072, p = 0.779), and all fit indices surpassed conventional thresholds, with a Comparative Fit Index (CFI) of 1.000, a Root Mean Square Error of Approximation (RMSEA) of 0.000 (90% CI: 0.000–0.088), and a Standardized Root Mean Square Residual (SRMR) of 0.060.

Among the predictors considered, only Env-Perc was significantly associated with mood (β = 0.344, 95%CI: [0.084;0954], p = 0.033), suggesting that individuals who reported more positive perceptions of their environment tended to experience better mood states. In contrast, demographic variables and clinical scales showed no significant relationship with mood (Table 2).

Ethics approval and consent to participate

The study has been approved by the ethical committees (ECs) of the three main participating centres: EC of IRCCS Istituto Centro San Giovanni di Dio Fatebenefratelli (31/07/2019; no. 211/2019), EC of Area Vasta Emilia Nord (25/09/2019; no. 0025975/19), and EC of Pavia (02/09/2019, no. 20190075685). All participants provided informed consent before the study began. The authors affirm that all procedures involved in this research adhere to the ethical guidelines set by the pertinent national and institutional committees on human experimentation, as well as the Helsinki Declaration of 1975, updated in 2008.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.