(A-187) Using a Random Slope Model to Evaluate the Association Between Initial High/Low Opioid Dosage and Kidney Function Change Over One Year Post-Surgical Discharge

Background: Observational studies have shown that opioid use is associated with kidney dysfunction. As opioids are frequently prescribed at post-surgical discharge, it is critical to evaluate the association between opioid use and kidney function change after surgical discharge.

Objectives: To evaluate the effect of initial high/low opioid dosage prescribed at post-surgical discharge on estimated glomerular filtration rate (eGFR) decline within one year of post-surgical discharge.

Methods: We identified a retrospective longitudinal cohort of opioid naive adult patients who had at least one eGFR value during the first-year post-surgery discharge and had been discharged with opioids. The data source was electronic health records from an integrated academic health system in the US, 2012-2019. We used random slope model to evaluate the effect of discharge opioid dosage (exposure) on quarterly eGFR change over time in first year after post-surgical discharge (outcome). The interaction term of time and high versus low total morphine milligram equivalent (> 200 MME vs. ≤ 200 MME) was included to evaluate the time-varying effect of high/low total MME on eGFR decline. Age, sex, surgery types, comorbidities, baseline eGFR and co-prescribing medications at post-surgical discharge were adjusted to control confounding. Surgical types were included as a random effect at cluster level to control the heterogeneous effects of surgery on the change in eGFR. We used an unstructured covariate matrix to address the eGFR correlation due to repeated measurements. Sensitivity analysis included different thresholds of high/low total MME to evaluate the robustness of results and multiple imputation by chained equation technique was used to address eGFR missingness.

Results: 25,230 patients met our eligibility, and 12,487 patients discharged with high MME (49.5%). On average, patients’ eGFR significantly declined by 0.41 ml/min/1.73m2 per quarter (95% CI, -0.54, -0.29) for the first year after surgical discharge. Compared to patients discharged with low MME, high MME was not significantly associated with additional eGFR decline (-0.06 ml/min/1.73m2, [-0.25, 0.12], p=0.52). Sensitivity analysis showed similar results; high total MME was not significantly associated with additional eGFR decline.

Conclusions: High total MME prescribed at post-surgical discharge was not associated with eGFR decline in first year after surgical discharge. Further studies are needed to evaluate the impact of opioids prescribed at post-surgical discharge on long-term change in kidney function.