Patients who abuse, or are dependent on, opioids have higher readmission rates after major surgery than their counterparts who do not abuse the agents, according to a national database study. Opioid abuse and dependence, the investigators added, also increase costs, both at initial admission and at readmission after a major procedure.
“We know that given the current opioid crisis, both Medicare and Medicaid have started imposing penalties for excess readmissions,” said Atul Gupta, MBBS, a clinical associate in anesthesia and critical care at the University of Chicago. With that in mind, Dr. Gupta—along with co-investigator Sajid S. Shahul, MBBS—examined the effect of opioid abuse on hospital mortality and readmissions in adult patients.
As part of the retrospective cohort analysis, the researchers analyzed data from the National Readmissions Database for 2013 and 2014. As Dr. Gupta noted, the database accounts for 51.2% of the U.S. population and 49.3% of all U.S. hospitalizations each year. Records for all patients aged at least 18 years who had undergone a major operating room diagnostic or therapeutic procedure were included.
“We used ICD-9 [International Classification of Diseases, Ninth Revision] codes to identify opioid abuse or dependence,” Dr. Gupta explained. “We specifically did not include patients who were nonopioid drug abusers or were going through drug withdrawal.” Controls were defined by the absence of opioid abuse and dependence codes.
Readmission was defined as the first admission occurring within 30 days of discharge from the initial major procedure. Of 70,886,775 weighted discharges, 16,016,842 were patients aged at least 18 years who underwent a major operating room procedure. Of these, 94,903 (0.59%) were opioid abusers.
As Dr. Gupta reported at the 2017 annual meeting of the American Society of Anesthesiologists (abstract JS08), the readmission rate in opioid abusers was significantly higher than in nonopioid users (14.3% vs. 8.8%; P<0.0001; odds ratio [OR], 2.08; 95% CI, 2.01-2.16).
The investigators used clustered logistic regression modeling to predict risk factors for readmissions. They found that the most common reasons for readmission in opioid abusers included infectious causes (14.9%), acute renal failure (1.7%), drug withdrawal (1.2%) and chest pain (1.0%). By comparison, the most common reasons for readmission among nonopioid users were infectious causes, such as endocarditis, sepsis and cellulitis (14.4%); cardiac complications (3.7%); and pulmonary embolism (1.3%). Of note, a significantly greater number of opioid abusers had pain diagnoses, opioid dependence or withdrawal as the primary reason for readmission than nonopioid users (6.7% vs. 3.1%; P<0.0001).
The study also found that the mean hospital length of stay was significantly greater in opioid abusers upon initial admission than in nonopioid users (9.7 vs. 5.4 days; P<0.0001). It was no surprise that mean hospital charges per patient were greater in opioid abusers than in nonabusers, both at initial admission ($27,929 vs. $19,968; P<0.0001) and at readmission ($15,248 vs. $14,564; P=0.01). Nevertheless, mortality rate was no different in opioid abusers than in nonopioid users (1.36% vs. 1.35%; P=0.94).
“There are many other factors that can impact readmission after major operations, including patient demographic factors, hospital practice patterns and the types of procedures,” Dr. Gupta said. “So we also adjusted for a variety of covariates.” After performing this analysis, the readmission rate was found to be 6.9% in opioid abusers and 5.3% in nonabusers.
Duplicating a Prior Analysis
Similar results were found in an analysis by researchers at Massachusetts General Hospital (abstract A3012). Researchers there turned to the Healthcare Cost and Utilization Project’s California Inpatient Database for the years 2010 and 2011. They included 498,055 elective surgical admissions and 47,318 readmissions in their primary analysis, using a multivariable logistic regression model to determine the adjusted OR (aOR) of readmission for patients with opioid abuse or dependence after adjusting for 42 covariates.
Alan D. Kaye, MD, PhD, found the study to be valuable in bringing greater awareness to anesthesia providers, given the prevalence of opioid abuse nationwide.
“Some of the data is not surprising because opioids suppress natural killer cells and propagate infectious processes,” said Dr. Kaye, professor, chairman and program director of anesthesiology at Louisiana State University, in New Orleans. “However, a part of these data may reflect these patients not following directions or continuing antibiotic regimens.
“The mean length of stay in the hospital was significantly higher in opioid abusers,” Dr. Kaye continued. “This provides an opportunity for anesthesiologists to deliver enhanced recovery after surgery protocols, which may significantly improve this important statistic.”