Mohammad Adil was a PGY1 Pharmacy Practice Resident, Austin De La Cruz is a Clinical Pharmacy Specialist in Mental Health, and Matthew Wanat is a Clinical Pharmacy Specialist in Critical Care, all in the Department of Pharmacy at the Michael E. DeBakey VA Medical Center in Houston, Texas. J. Douglas Thornton is an Assistant Professor and Policy and Director of the Prescription Drug Misuse Educationand Research (PREMIER) Center; Matthew Wanat is a Clinical Associate Professor and Assistant Director of the PREMIER Center; and Austin De La Cruz is a Clinical Assistant Professor; all at the University of Houston College of Pharmacy. Correspondence: Matthew Wanat (mawanat@uh.edu)
Author disclosures The authors report no actual or potential conflicts of interest with regard to this article.
Disclaimer The opinions expressed herein are those of the authors and do not necessarily reflect those of Federal Practitioner, Frontline Medical Communications Inc., the US Government, or any of its agencies. This article may discuss unlabeled or investigational use of certain drugs. Please review the complete prescribing information for specific drugs or drug combinations—including indications, contraindications, warnings, and adverse effects—before administering pharmacologic therapy to patients.
The incidence of opioid-naïve patients receiving opioids after discharge was 76.3% (n = 90) at 3 months, 19.5% (n = 23) at 6 months and 7.6% (n = 9) at 12 months (Figure). The daily morphine milligram equivalent (MME) of patients prescribed opioids at 3, 6, and 12 months was similar (3 months, 22.7; 6 months, 19.7; 12 months, 20.9). In the univariate regression analysis, several variables were found to be associated with converting to chronic opioid use. Prior history of alcohol use disorder (OR, 0.3; 95% CI, 0.10-0.88; P = .03), ICU type (OR, 3.9; 95% CI, 1.73-8.75; P = .001) and ICU LOS (OR, 0.88; 95% CI, 0.81-0.95; P = .01) had a statistically significant association on opioid use at 3 months. (Table 2). No variables evaluated had a statistically significant effect on chronic opioid use at 6 months, and only age (OR 0.93; 95% CI. 0.87-0.99; P = .02) was statistically significant at 12 months. In the multivariate logistic regression analysis, history of alcohol abuse, admission for surgery, and hospital LOS were significant at 3 months (Table 3).
Discussion
In this single-center analysis conducted at a VA academic hospital of opioid-naïve patients who were administered opioids in the ICU, the incidence of patients subsequently receiving outpatient opioid prescriptions at 12 months after discharge was 7.6%. There also was a decrease in the amount of opioids received by patients (daily MME) after discharge at 3, 6, and 12 months. This trend did not demonstrate the propensity for inpatient opioid use to convert opioid-naïve patients to chronic opioid users.
The most common outpatient opioids prescribed were hydrocodone/acetaminophen, morphine, and tramadol. Logistic regression showed few factors that correlated significantly with opioid use in the long-term (12 month) period. This finding is a deviation from the findings of Yaffe and colleagues who found hospital LOS to be one of the only predictors of long-term opioid use in their population (defined as use at 48 months).16 One important difference between our study and the Yaffe and colleagues study was that they evaluated all patients who were admitted to the ICU, regardless of the exposure to opioids during their inpatient stay. Consequently, this difference may have resulted in the differences in findings.
Strengths and Limitations
Location was a strength of our study, as this analysis was conducted at an integrated health care system that provides comprehensive inpatient and outpatient care. The VA uses a closed electronic health record, which allowed patients to be followed both in the inpatient and outpatient settings to determine which medications were prescribed at each time. In other health care systems this information would have been difficult to follow as patients often fill outpatient prescriptions at community pharmacies not affiliated with the treating hospital. However, any patient not using a VA prescriber for subsequent opioid prescriptions or patients who received opioids through other sources would not have had their continued opioid use captured. These data may be available in the states prescription monitoring program, but it was not available to query for research at this time.
