The Effect of Prescribing Antibiotics with Opioids on the Development of Opioid Use Disorder: A National Database Study

Introduction

Evidence suggests that the majority of patients who have an opioid use disorder (OUD) reported that their first opioid was a prescription drug¹. With the opioid epidemic steadily on the rise in the United States since 2000, in part due to the advent of synthetic opioid products², it is important to increase research efforts that identify risk factors that are correlated with the development of OUD as well as protective factors capable of reducing the development of OUD in vulnerable individuals.

The addictive properties of prescription opioids are often attributed to the agonistic effects on μ-opioid receptors in the central nervous system and their ability to cross the blood-brain barrier via active transport, which allows for rapid increases in brain concentration and fast onset of action³. However, growing clinical and preclinical evidence implicates the gut microbiome in the persistent misuse of illicit substances, including opioids^{4, 5}. These studies highlight that exposure to opioids elicits gut dysbiosis (a microbial imbalance or maladaptation), as well as alterations in the production and expression of microbial metabolites. Given that the gut microbiome is causally associated with anxiety-like behaviors ⁶ and directly modulates drug-withdrawal-induced anxiety⁷, these opioid-induced alterations in the gut microbiome may contribute to critical physiological responses observed in OUD that perpetuate continued drug use, including negative affective states⁸.

In addition to opioids, there is evidence that antibiotics modify the gut microbiome, eliciting either a negative (pathogenic) or positive impact (i.e., eubiotic effect)⁹. This modification is dependent upon the class (bacterial targets), pharmacokinetic profile, pharmacodynamics, dose, duration, and route of administration⁹. Preclinical studies of substance use disorders, employing the conditioned place preference paradigm, a model of drug reward and drug-context associations, have shown that antibiotic treatment blocks cocaine (10 mg/kg)-induced conditioned place preference^{10, 11}. Additionally, preclinical research has shown that the β-lactam antibiotic, ceftriaxone, is effective at attenuating drug-seeking behaviors^{12, 13}. Based on the observed protective effects of antibiotics on addiction like behaviors in preclinical models of substance use disorders, we sought to investigate whether antibiotic co-treatment with opioids in a hospital or emergency department setting influenced the risk of developing OUD 12 months following discharge from the hospital. For this investigation, we employed TriNetX electronic medical records. TriNetX is a third-party software product that provides access to a clinical data network consisting of 170 healthcare organizations across 30 countries¹⁴. Here, we employed TriNetX to test the hypothesis that antibiotics prescribed with opioids in the hospital setting were associated with a protective effect against developing OUD following discharge from the hospital.

Methods

Results

Discussion

Using the TriNetX electronic medical record database, we found that the relative risk of developing OUD was reduced in subjects who were prescribed opioids in combination with antibiotics when compared to subjects that were prescribed opioids in the absence of antibiotic treatment. Additionally, comparing these two cohorts, we found that the risk ratio for developing OUD steadily decreased each year with the lowest risk ratio calculated during the year of 2018. Lastly, in subjects prescribed an opioid in the hospital setting, our results showed that the risk of being diagnosed with OUD twelve months following discharge from the hospital was, overall, less than 3%.

The finding that antibiotics in combination with opioids reduce the risk of developing OUD coincides with preclinical evidence that shows that antibiotics are protective against the expression of addiction-like behaviors. A well-studied antibiotic, ceftriaxone, has proven effective at blocking drug-seeking behaviors in rodent models of psychostimulant, alcohol, and opioid use disorders^{21, 22}. Evidence suggests that the mechanism mediating ceftriaxone’s effects on drug seeking is mediated by its ability to increase the expression of the glutamate transporter, GLT-1, on glia, which in turn enables efficient removal of glutamate from the synapse within key brain regions involved in reward and motivation²³. However, ceftriaxone is also capable of altering the gut microbiota²⁴ and it may be through its effects on the gut microbiota that allows ceftriaxone to regulate drug-seeking behaviors in rodent models of substance use disorders. Additionally, antibiotics may have an indirect effect on OUD through their effects on microbiome depletion²⁰. In patients diagnosed with OUD compared to healthy controls, there is evidence that there are differences in alpha (the diversity within a particular area) and beta diversity (the diversity between species) within the gut microbiome^25-27. These changes, as shown in rodent models, may compromise intestinal barrier function^{28, 29}, resulting in the opening of tight junctions in the large intestine. This would enable bacteria and their toxins to penetrate the bloodstream and lead to systemic inflammation. Given that systemic inflammation is observed in OUD patients as well as in preclinical models of OUD^30-32, antibiotic-induced depletion of the gut microbiota may remove pathogenic taxa that inhabit the gut following repeated exposure to opioids. Despite this aforementioned evidence that antibiotic co-treatment with opioids protects against OUD, it is also possible that antibiotic treatment may interfere with a healthy microbiota enabling pathogenic taxa to colonize and contribute to the development of OUD. The results obtained from this investigation suggest that antibiotics are more protective than facilitative when it comes to the development of OUD. However, our study is limited by our inability to identify how many opioid prescriptions the subjects received after the index event was measured. Additionally, the comparisons between cohorts may include different subject populations based on the requirement of an antibiotic prescription³³.

Interestingly, we found that the risk ratio between the subjects prescribed opioids versus opioids + antibiotics decreased from 2012 to 2018. This was due to increases in the risk of developing OUD from 2012 to 2018 in subjects prescribed opioids in combination with antibiotics (~30%), whereas the risk in the opioid cohort increased relatively less (~12%) twelve months following discharge from the hospital (Table II). The observed increased risk in the opioid + antibiotic cohort is potentially due to the increased risk evoked by the antibiotic as the risk for the antibiotic cohort increased the most relative to the other groups (~45%) during this same time frame. Although it is unknown why the risk of developing OUD within the antibiotic and opioid + antibiotic cohorts increased relatively greater from 2012 to 2018 compared to the opioid cohort, potential contributing factors may be the type of antibiotic prescribed as well as the route of antibiotic administration as it has been shown that these factors dictate the pathogenic or eubiotic effects of antibiotics⁹.

Finally, we acknowledge that data gathered in this study which was used to make statistical comparisons comes with limitations. For example, we are limited by the OUD diagnoses, which may not always be accurately diagnosed and/or may be incomplete. This includes geographical discrepancies in OUD diagnosis as well as discrepancies introduced over the time course of this study. For example, physicians/providers may not have been as well trained to identify an opioid-related disorder during the first half of this study timeframe as compared to more recent years. Additionally, we are unable to differentiate opioid or antibiotic treatment based on dose, duration of use, or route of administration. Moreover, if the results found within this data set are dependent upon the microbiome, our analyses did not take into account confounding factors that influence the gut microbiota, including geographic regions, socioeconomic status, pre/probiotic use, exercise, comorbidities, weight, and drug use prior to treatment. Along these same lines, the inability to identify race using the TriNetX software reduces our ability to investigate how microbiomes and race, which constitutes social, environment, and biological factors, are intertwined. As the TriNetX software evolves, the inclusion of race-related data will be essential. We also acknowledge that the TriNetX platform does not represent the general U.S. population, but rather people who received clinical care within the network. Lastly, it is possible that some subjects were included in two or more cohorts (e.g., antibiotic cohort of 2012 and opioid + antibiotic cohort of 2016) in this study. While we are not sure of the quantity of antibiotic and opioid prescriptions before or after the index event of subjects in each cohort, our results represent a snapshot in time that is still representative of relative risk for development of OUD. Despite these limitations, TriNetX does provide large sample sizes from 92 health systems throughout the United States and rigorous analysis based on propensity-score matching. Because of this, we were able to efficiently and cost-effectively identify a large population of subjects within each cohort to address our question, and in this case, support our hypothesis. Although we are unsure what a risk ratio of 1.03 between the opioid vs. opioid+antibiotic cohort (year 2018) means in terms of clinical relevance, it is clear, based on the large sample size gathered from this study, that a 3% change may impact thousands of lives.

Conclusion

Our overall findings, using a broad-based investigative approach, suggest that opioids prescribed in combination with antibiotics in the hospital setting are protective against the development of OUD at later time points following hospital discharge. Importantly, our findings are not meant to suggest that antibiotics should be prescribed along with opioids to reduce the risk of OUD as other potential life threatening complications may arise, including antibiotic resistance. Currently, in the dental setting, there are prevention strategies being implemented to reduce antibiotic and opioid over-prescription to minimize antibiotic resistance, adverse drug events, and illicit opioid use^{33, 34}. Therefore, the intention of this study is facilitate more carefully crafted inquiries that can be pursued to further understand the relationship between antibiotic treatment, the gut microbiome, and the risk of developing OUD, including clinical trials to further explore this area. Continued work on this topic has the potential to lead to scientifically-based treatment options for individuals vulnerable to developing OUD.