Overdose Risk in Young Children of Women Prescribed Opioids

Setting and Data Sources

We conducted a population-based, nested case control study between April 1, 2002 and March 31, 2015 in Ontario, Canada. We identified prescription claims using the publicly funded Ontario Public Drug Benefit Database. This database contains prescription information for all Ontario residents who receive social assistance. This is a subset of the Ontario population (∼13.8 million): those receiving prescription medications subsidized by the province. All such residents receive all their medications, including over-the-counter medications, such as nonsteroidal antiinflammatory drugs (NSAIDS), free-of-charge, and all dispensed medications, including over-the-counter drugs, are captured in this database. Therefore, we identified all opioid and NSAID prescriptions reimbursed under this drug benefit plan during the study period. We used the Canadian Institute for Health Information Discharge Abstract Database, which contains detailed diagnostic information for all hospital admissions in Ontario, and we identified emergency department visits using the National Ambulatory Care Report System. We used the Ontario Health Insurance Plan database to identify claims for physician services, and obtained basic demographic data and dates of death from the Registered Persons Database, a registry of all Ontario residents eligible for health insurance. The Ontario Registrar General Death Database identified the cause of death reported on individual death certificates up to December 2012, the most recent data available. We identified mother–child pairs using birth records, as done previously.^12,13 These databases were linked in an anonymous fashion using encrypted health card numbers and are routinely used to study drug safety.^14–16

Study Subjects

We defined cases as children aged ≤10 years whose mothers received subsidized medications via Ontario’s publicly funded drug plan and who presented to the hospital for opioid toxicity. Opioid overdose was defined using the International Classification of Diseases and Related Health Problems, Tenth Revision codes T40.0, T40.1, T40.2, T40.3, T40.4, and T40.6. We also included children who died of opioid overdose in the prehospital setting, identified using the International Classification of Diseases and Related Health Problems, Ninth Revision codes X40–X49, X60–X69, X85–X90, Y10–Y19 and the Ontario Registrar General Death database, as validated previously.¹⁷ We excluded children >10 years of age to avoid inclusion of overdoses related to voluntary drug misuse, which is rare before adolescence. Only the first hospital presentation was considered for children with multiple episodes. The date of hospital presentation or death served as the index date for all analyses. For each case, we randomly selected 4 control children without an opioid overdose, matching on age (within 6 months) and sex. Each individual could serve as a control only once. To account for temporal changes in analgesic prescribing over the study period, each control subject was randomly assigned an index date within 6 months of his or her corresponding case.

Case and control children were linked to their mothers using birth records, as done previously.¹⁸ Among mothers of case and control children, we identified those who received either an opioid or an NSAID in the year preceding their child’s opioid overdose. We chose NSAIDs as the reference exposure because, like opioids, they are primarily prescribed for analgesia and because individuals who receive analgesics for chronic pain are frequently inherently different from those who do not. Because those women’s prescription costs are covered by the provincial government, they are unlikely to obtain NSAIDs without a prescription. Finally, to focus on maternal prescriptions, we also excluded children who were prescribed an opioid or NSAID for themselves in the 100 days before the index date.

Statistical Analysis

We compared the baseline characteristics of cases and controls using P values (Table 1). We used conditional logistic regression to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) for the association between pediatric opioid overdose and maternal receipt of an opioid prescription. This is a specialized type of logistic regression used when cases with a specific outcome are matched with multiple controls; each case and his or her corresponding controls represent a stratum in the analysis. Other covariates were not included in the model to preserve model fit. In all analyses, maternal receipt of an NSAID prescription in the year preceding the index date constituted the reference exposure. We used SAS version 9.3 (SAS Institute, Inc., Cary, NC) for all analyses, and a two-tailed type 1 error rate of .05 as the threshold for statistical significance. The study was approved by the Research Ethics Board of Sunnybrook Health Sciences Centre (Toronto, Ontario, Canada).