Electronic medical records vs insurance claims: Comparing the magnitude of opioid use prior, during, and following surgery Public Deposited

Downloadable Content

Download PDF
Last Modified
  • September 28, 2022
Creator
  • Young, Jessica
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
  • Funk, Michele
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
  • Chidgey, Brooke
    • Affiliation: School of Medicine, Department of Anesthesiology
  • Dasgupta, Nabarun
    • Affiliation: Injury Prevention Research Center
  • Hudgens, Michael
    • Affiliation: Gillings School of Global Public Health, Department of Biostatistics
  • Stürmer, Til
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
  • Pate, Virginia
    • Affiliation: Gillings School of Global Public Health, Department of Epidemiology
Abstract
  • Background: Pharmacoepidemiology studies often use insurance claims and electronic medical record (EMR) data. However, the implications of data source choice on key study design elements are not well understood. Objectives: Use linked claims-EMR data (separately and together) to characterize opioid use as it relates to study eligibility criteria, and exposure and outcome assessment. Methods: EMR data from a large healthcare system were linked to Medicare insurance claims for patients undergoing invasive surgery in an ongoing study on opioid use. Drug utilization based on order and fill dates came from 3 sources: EMR inpatient orders, EMR outpatient orders, and Medicare Part D claims. We evaluated 3 study design elements: a) Study selection - opioid use 182 days before surgery to identify opioid naïve patients; b) Exposure ascertainment- perioperative opioid use; c) Outcome assessment - prolonged opioid use 90 days post-op. Results: We identified 12,023 surgery patients in the linked claims-EMR data. Eligibility: For baseline opioid exposure, 30% (outpatient EMR) vs 28% (claims) had an opioid order (EMR) or fill (claims). Using both claims and outpatient EMR, 44% of patients had evidence of exposure in at least one data source. Perioperative Exposure: For use during surgical admission, combined inpatient and outpatient EMR orders documented much higher use of opioids (81%) compared to claims-only (32%). Combining the 3 data sources, 87% had evidence of opioids ordered or filled during the surgical admission. On the date of surgical discharge 56% (outpatient EMR), 31% (claims), and 65% (claims or outpatient EMR) had evidence of opioids ordered or filled. In the 7 days immediately after surgery, outpatient EMR found 3.0% with orders, claims found 9.2% with opioid fills, and combining the 2 data sources suggested 11% could have received opioids. Prolonged Use Outcome: Outpatient EMR found 4.7% with 90-day postsurgical opioid orders, claims found 6.0%, and combining claims and outpatient EMR found 9.4%. Conclusions: When characterizing opioid exposure, we found substantial non-overlap between EMR and claims depending on time window relative to surgery and care setting, driven by inpatient opioids only captured in EMR and primary nonadherence. In studies that straddle hospital and outpatient settings, the potential for misclassification of drug utilization must be evaluated carefully, and choice of data source may have large impacts on key study design elements. Both EMR and claims data are needed to provide a more complete picture of opioid exposure prior to, during and after surgery.
Date of publication
DOI
Resource type
Rights statement
  • In Copyright
Conference name
  • ICPE 2020
Language
Parents:

This work has no parents.

In Collection:

Items