Persistent opioid use after surgery and trauma

Problem

Opioids are prescribed after surgery and after trauma for short-term pain relief. Some patients keep filling those prescriptions long after the original injury has healed. That pattern is called persistent opioid use. In Aotearoa New Zealand, no one had counted how often it happened, who it happened to, or whether it caused real harm. Without that picture, prescribers, pharmacists, and policy-makers were working from overseas estimates and clinical intuition.

What I did

I led a series of linked-data studies covering every public-hospital surgical or trauma admission in the country from 2007 to 2019. The result was a national, decade-long picture of how opioid prescribing plays out after people leave hospital.

A descriptive baseline first. Of 1.78 million surgical patients, 20.9% were dispensed opioids within a week of discharge. Among Māori, Pacific, and Asian patients the rate was lower, pointing to an equity gap in access to prescribing.

A surgical cohort next, then a trauma cohort, and finally a cohort linking persistent use to longer-term outcomes. Same data, same definitions, each study answering the next question. All four were published in international peer-reviewed journals, two of them as featured articles.

Evidence

• After surgery, 1 in 11 opioid-naïve patients developed persistent use. The strongest modifiable signals were opioid switching mid-prescription, multiple opioids on discharge, and a high total dose. Co-prescribing simple non-opioid pain relief was protective.
• After trauma, 1 in 7 opioid-naïve patients developed persistent use, higher than the surgical rate, with slow-release formulations standing out as a risk.
• Persistent use was not a benign label. We followed more than 500,000 patients for up to a year. Those who developed persistent use had 6.6 times the all-cause mortality after surgery and 2.8 times the mortality after trauma, with fewer days alive and out of hospital.
• The flagship surgical paper received the School of Pharmacy Best Publication Prize 2025. The trauma and outcomes papers were both featured by their journals.

Impact

The series closed a causal chain. Persistent opioid use is not a tidy statistical endpoint; it predicts real mortality and hospitalisation. That has shifted the policy conversation from whether the problem is real to which prescribing patterns can be modified at the point of discharge, and it has given clinicians, pharmacists, and regulators the local numbers needed to act.

The work is observational. Causation in any single patient remains a clinical judgement rather than a statistical claim, and the next round of studies tightens the analysis by linking dispensing patterns to opioid use after gout, opioid beliefs in primary care, and opioid use after admissions for inflammatory bowel disease. Funded by a Health Research Council fellowship.