Additional prescription of benzodiazepines, z-drugs and gabapentoids and mortality risk in people with opioid use disorder receiving opioid substitution therapy in primary care: prospective observational study using linkage between the CPRD and ONS death
To estimate associations of additional prescription of benzodiazepines, z-drugs (zaleplon, zolpidem and zopiclone) and gabapentoids with treatment duration and cause-specific mortality amongst patients receiving Opioid Substitution Treatment (OST)
Cohort study of 12,118 patients aged 15-64 prescribed OST between 1998 and 2014 extracted from the Clinical Practice Research Datalink (CPRD) of which 7,106 patients had specific cause of death information on drug related poisonings (DRP). In these patients we identified episodes of prescription of benzodiazepine, z-drugs and gabapentoids and assessed the associations of these exposures with treatment duration and cause-specific mortality using linear and poisson regression. Prescription of benzodiazepine, z-drugs or gabapentoids at any time during or within a year of the end of an OST treatment episode (co-prescription). Prescription of benzodiazepine, z-drugs or gabapentoids during an OST treatment episode was classified as concurrent prescription.
In 36,126 person years of follow-up there were 657 deaths and 29,540 OST treatment episodes of which 42% involved benzodiazepine co-prescription, 20% z-drug co-prescription and 8% gabapentoid co-prescription. Concurrent prescription was 29%, 11% and 5% for benzodiazepines, z-drugs and gabapentoids respectively. Concurrent prescription of benzodiazepine was associated with an approximate doubling of the duration of methadone treatment (adjusted mean duration of treatment episode 466 days compared to 286 days. Benzodiazepine co-prescription had little association with non-DRP but was associated with substantially increased risk of DRP with evidence of a dose response effect (mortality rates per 100 person years 0.45 off treatment, 0.87 normal dose, 1.77 high dose, p<0.0001, fully adjusted IRR compared to DRP off treatment: 2.48 (95% CI 1.55 to 3.95) normal dose, and 4.47 (2.41 to 8.30) high dose). Adjustment had little effect on these estimates. In contrast z-drug and gabapentoid co-prescription was associated with increased risk of both all cause mortality and DRP with no evidence of a dose response effect and, in the case of gabapentoids, substantial attenuation on adjustment. In analyses considering the increased duration of OST associated with concurrent prescription an overall adverse effect on mortality risk was still apparent (adjusted IRR for DRP with benzodiazepine concurrent prescription 3.32 (2.08 – 5.29)). We found no evidence of an interaction with either OST type or treatment period in relation to this effect.
Co-prescription of benzodiazepine is associated with increased risk of DRP in problem opioid users both on and off OST. Co-prescription of z-drugs and gabapentoids is also associated with increased mortality risk, however evidence that this association is causal was less strong. Concurrent prescription of benzodiazepine was associated with increased treatment duration however still had an adverse association with mortality risk. Clinicians should avoid prescribing benzodiazepines to individuals with opioid use disorder who continue to use opioids.