Comparison of Different Modeling Approaches for Prescription Opioid Use and Its Association With Adverse Events

Abstract Previous research linking opioid prescribing to adverse drug events has failed to properly account for the time-varying nature of opioid exposure. This study aimed to explore how the risk of opioid-related emergency department visits, readmissions, or deaths (composite outcome) varies with opioid dose and duration, comparing different novel modeling techniques. A prospective cohort of 1,511 hospitalized patients discharged from 2 McGill-affiliated hospitals in Montreal, 2014–2016, was followed from the first postdischarge opioid dispensation until 1 year after discharge. Marginal structural Cox proportional hazards models and their flexible extensions were used to explore the association between time-varying opioid use and the composite outcome. Weighted cumulative exposure models assessed cumulative effects of past use and explored how its impact depends on the recency of exposure. The patient mean age was 69.6 (standard deviation = 14.9) years; 57.7% were male. In marginal structural model analyses, current opioid use was associated with a 71% increase in the hazard of opioid-related adverse events (adjusted hazard ratio = 1.71, 95% confidence interval: 1.21, 2.43). The weighted cumulative exposure results suggested that the risk cumulates over the previous 50 days of opioid consumption. Flexible modeling techniques helped assess how the risk of opioid-related adverse events may be associated with time-varying opioid exposures while accounting for nonlinear relationships and the recency of past use.


Comparison of Different Modeling Approaches for Prescription Opioid Use and Its Association With Adverse Events
Siyana Kurteva, Michal Abrahamowicz, Marie-Eve Beauchamp, and Robyn Tamblyn

Table of Contents
Web Figure 1 Web Table 1 Web Appendix 1 Web Table 2 Web Figure 2 Web Table 3 Web Figure 3 Web Figure 4 References ATC codes used to identify opioids: N02A (opioids), R05DA (opium alkaloids and derivatives) Exclusions: Not all drug forms were included in the analyses.Only patches and tablets of these medications were kept.Injectable, liquid and rectal forms were excluded.Methadone and buprenorphine/naloxone combinations were kept to define subclinical patient populations but were excluded from all dosing/duration calculations as these medications are used to treat addiction and we want to focus on the association of duration/dose of opioids used for pain relief.
The daily dose of each opioid was calculated by first dividing the quantity of units dispensed by the prescription duration to determine the number of units per day, and then multiplying the number of units by the strength.To account for concurrent prescriptions, a subsequent dispensation was considered as an early refill if days of overlap were ≤30% of the previous dispensation duration.Otherwise, the opioids were considered to be taken simultaneously.Daily dose of each dispensation was converted to MME doses using the Center for Disease Control Opioid Morphine Equivalent Conversion Factor and the opioid doses determined to be concurrently dispensed were added together.
Cumulative duration of past use assessed the long-term impact of opioids, where the effect on the outcome persisted upon discontinuation: defined as the total number of days exposed, calculated by summing the durations of all dispensations between cohort entry (first opioid dispensation) and a given day during the follow-up.Cumulative users represented patients who used opioids only when needed, thus accumulating use over time.On the other hand, we assessed continuous duration where the effect of opioids accumulated by dispensation supply but the risk returned to baseline after discontinuation.Continuous duration was defined similarly but was allowed to increase only during the periods of uninterrupted use and was reset to zero if there was a gap of >5 days between subsequent dispensations.

Drug Name Conversion Factor
a Centers for Disease Control and Prevention, Atlanta, GA, May 2014 (1).b The MME conversion factor for buprenorphine patches is based on the assumption that one milligram of parenteral buprenorphine is equivalent to 75 milligrams of oral morphine and that one patch delivers the dispensed micrograms per hour over a 24-hour day.Example: 5 ug/hr buprenorphine patch * 24 hrs = 120 ug/day buprenorphine = 0.12 mg/day buprenorphine = 9 mg/day oral morphine milligram equivalent.In other words, the conversion factor not accounting for days of use would be 9/5 or 1.8.However, since the buprenorphine patch remains in place for 7 days, we have multiplied the conversion factor by 7 (1.8X 7 = 12.6).
Codes Used for Drug Classification and Rationale for Opioid Dose and Opioid Duration of Use Calculations.

Table 3 .
the dispensed micrograms per hour over a 24 hour day.Example: 25 ug/hr fentanyl patch * 24 hrs= 600 ug/day fentanyl = 60 mg/day oral morphine milligram equivalent.In other words, the conversion factor not accounting for days of use would be 60/25 or 2.4.However, sincethe fentanyl patch remains in place for 3 days, we have multiplied the conversion factor by 3 (2.4X 3 = 7.2).In this example, MME/day for ten 25 μg/hr fentanyl patches dispensed for use over 30 days would work out as follows: Example: 25 ug/hr fentanyl patch * (10 patches/30 days)* 7.2 = 60 MME/day.Estimated weights function showing the association between current daily opioid use during the entire one-year follow-up period and the current hazard of adverse events.Comparison of goodness of fit in sensitivity analyses for flexible non-linear (NL) MSM additionally adjusting for the non-linear effect of MME current daily dose (logtransformed), with alternative time-varying opioid exposure metrics.
eThe MME conversion factor for fentanyl nasal spray is 0.16, which reflects a 20% greater bioavailability for sprays compared to lozenges/tablets.fTheMME conversion factor for fentanyl patches is based on the assumption that one milligram of parenteral fentanyl is equivalent to 100 milligrams of oral morphine and that one patch delivers Web Figure2.