Risk of Overdose Associated With Co-prescription of Antipsychotics and Opioids

A Population-Based Cohort Study

Alejandro G. Szmulewicz; Brian T. Bateman; Raisa Levin; Krista F. Huybrechts


Schizophr Bull. 2022;48(2):405-413. 

In This Article

Abstract and Introduction


The US FDA issued a black-box warning against co-prescription of antipsychotic (AP) agents and opioids due to the risk of respiratory depression, but evidence on the comparative safety of sedating vs nonsedating APs is lacking. We classified APs as sedating (eg, quetiapine, olanzapine, and chlorpromazine) and nonsedating (eg, aripiprazole, haloperidol, and risperidone) based on their affinity to the histamine-1 neuroreceptor (Ki < or ≥20, respectively) and sought to compare the rate of overdose between patients using sedating vs nonsedating APs plus opioids. We constructed a population-based cohort nested in the IBM MarketScan database (2004–2017). Patients with concomitant use of sedating APs and prescription opioids ("exposed") were 1:1 matched to patients with concomitant use of nonsedating APs and prescription opioids ("referent") based on the propensity score (PS). The primary outcome was any hospitalization or emergency department visit due to an overdose within 30 days. The final cohort comprised 62 604 exposed and an equal number of PS-matched reference patients. Characteristics of matched exposed and reference patients were similar. There were 178 overdose events among the exposed (35.3 events per 1000 person-years [PY]) vs 133 among the reference group (26.4 events per 1000 PY), for an adjusted hazard ratio of 1.34 (95% CI: 1.07–1.68). This finding was consistent across sensitivity and subgroup analyses. Among patients receiving prescription opioids, concomitant use of sedating APs was associated with an increased risk of overdose compared with nonsedating APs. Caution is required when co-prescribing opioids and APs. If co-prescription is needed, choosing a nonsedating agent should be preferred whenever possible given the clinical context.


Chronic pain and mental health conditions commonly co-occur.[1,2] In the United States, more than half of all opioid prescriptions each year are filled by patients with coexisting mental health conditions,[3] and it has been reported that about 10%–15% of opioid-treated patients use antipsychotic (AP) agents concomitantly.[4,5]

In 2016, the US FDA released a "black-box" warning about the potential risks for respiratory depression after combined use of opioids and benzodiazepines or other central nervous system depressants including APs.[6] Although the supporting evidence linking the risk of opioid overdose and benzodiazepine use was substantial,[4,7–9] the inclusion of APs in the warning was based on the sedative pharmacological properties of some AP agents, even though benzodiazepines and AP agents exert their sedative actions through preferential binding of different receptors (ie, GABA and histamine-1 H1 receptors, respectively).[10] Given that there exists considerable heterogeneity between AP agents in their sedative properties and the potential for respiratory depression (eg, dystonia of the muscles of the larynx),[11] studies exploring whether the risk of unintended overdose varies depending on the specific AP agents are needed to inform clinical practice.

We, therefore, conducted a large cohort study to evaluate the risk of overdose in patients co-prescribed opioids and AP agents with sedating vs nonsedating properties using real-world data from commercially insured beneficiaries in the United States.