Predicting benzodiazepine prescriptions: A proof-of-concept machine learning approach.

Introduction: Benzodiazepines are the most commonly prescribed psychotropic medications, but they may place users at risk of serious adverse effects. Developing a method to predict benzodiazepine prescriptions could assist in prevention efforts. Methods: The present study applies machine learning methods to de-identified electronic health record data, in order to develop algorithms for predicting benzodiazepine prescription receipt (yes/no) and number of benzodiazepine prescriptions (0, 1, 2+) at a given encounter. Support-vector machine (SVM) and random forest (RF) approaches were applied to outpatient psychiatry, family medicine, and geriatric medicine data from a large academic medical center. The training sample comprised encounters taking place between January 2020 and December 2021 (N = 204,723 encounters); the testing sample comprised data from encounters taking place between January and March 2022 (N = 28,631 encounters). The following empirically-supported features were evaluated: anxiety and sleep disorders (primary anxiety diagnosis, any anxiety diagnosis, primary sleep diagnosis, any sleep diagnosis), demographic characteristics (age, gender, race), medications (opioid prescription, number of opioid prescriptions, antidepressant prescription, antipsychotic prescription), other clinical variables (mood disorder, psychotic disorder, neurocognitive disorder, prescriber specialty), and insurance status (any insurance, type of insurance). We took a step-wise approach to developing a prediction model, wherein Model 1 included only anxiety and sleep diagnoses, and each subsequent model included an additional group of features. Results: For predicting benzodiazepine prescription receipt (yes/no), all models showed good to excellent overall accuracy and area under the receiver operating characteristic curve (AUC) for both SVM (Accuracy = 0.868-0.883; AUC = 0.864-0.924) and RF (Accuracy = 0.860-0.887; AUC = 0.877-0.953). Overall accuracy was also high for predicting number of benzodiazepine prescriptions (0, 1, 2+) for both SVM (Accuracy = 0.861-0.877) and RF (Accuracy = 0.846-0.878). Discussion: Results suggest SVM and RF algorithms can accurately classify individuals who receive a benzodiazepine prescription and can separate patients by the number of benzodiazepine prescriptions received at a given encounter. If replicated, these predictive models could inform system-level interventions to reduce the public health burden of benzodiazepines.

Case of buprenorphine-associated central sleep apnea resolving with dose reduction.

Aside from respiratory suppression in overdose, full opioid agonist agents are known to cause sleep-disordered breathing (SDB). The increasing rates of opioid overdose in the United States have led to increasing use of medication-assisted treatments for opioid use disorders. Dose-dependent increase in SDB has been documented with methadone. There is emerging literature in the form of case reports providing evidence of buprenorphine and buprenorphine-naloxone contributing to sleep apnea. We report an additional case of a female patient developing central sleep apnea during initiation of buprenorphine-naloxone treatment. The condition resolved with dose reduction.

Opioid overdose history and awareness of naloxone in patients seeking outpatient detoxification.

OBJECTIVES: The goal of this study was to investigate whether patients were knowledgeable about naloxone, and whether they would accept overdose (OD) education and naloxone distribution (OEND), if available. DESIGN: This is a cross-sectional study to ascertain participants knowledge about OEND. A questionnaire was designed requesting information regarding age, gender, ethnicity, previous OD experience, type of substance(s) used, availability and use of naloxone for reversal, other treatment interventions utilized if witnessed OD, ways of seeking help for OD, knowledge about OEND, and willingness to accept OEND. SETTING: Study was conducted at the ambulatory detoxification clinic of a tertiary healthcare institution. PARTICIPANTS: Consecutive 131 patients, who presented for primary treatment of opioid detoxification during their visits to the ambulatory detoxification clinic between a predefined timeline from October 2014 through February 2015, were invited to participate, and all 131 agreed to be included in the study. A total of 124 participants returned completed questionnaires (95 percent response rate.) RESULTS: Overall, 68 (52 percent) of the participants indicated that they would accept OEND. A logistic regression analysis showed that younger participants (95% CI: 0.9-1, p = 0.02) and those who identified as non-white (95% CI: 0-0.8, p = 0.01) had higher odds for accepting OEND. Furthermore, prior administration of naloxone was significantly associated with OEND acceptance (95% CI: 1.6-68.6, p = 0.01). CONCLUSIONS: Results indicate more than half of participants presenting for outpatient detoxification from opioids have had an OD or witnessed an OD. More than half of the participants were willing to accept OEND. This study provides evidence that patients starting their recovery are willing to accept naloxone.