Preoperative opioid use is detrimental to outcomes after hip and knee arthroplasty. This study aims to identify the prevalence of preoperative opioid prescriptions and the specialty and practice setting of the prescriber, as well as the percentage of patients who do not report their opioid prescriptions and any variables associated with preoperative opioid prescriptions.
A total of 461 consecutive new patients evaluated for an arthritic hip or knee were retrospectively studied using institutional data from a tertiary-care, urban center at a university-affiliated private-practice and the state Prescription Monitoring Program to identify opioid prescriptions (including medication, number of pills and dosage, refills, prescriber specialty, and practice setting) within 6 months before their first appointment. Demographic data included age, sex, ethnicity, body mass index, joint, laterality, diagnosis, Charlson Comorbidity Index, duration of symptoms, decision to have surgery, number of days from the first visit to surgery, smoking status, alcohol use, mental health diagnoses, preoperative outcome scores, nonopioid medications, and opioid medications. Patients were separated into opioid and nonopioid cohorts (opioid receivers were further subdivided into those who reported their opioid prescription and those who did not) for statistical analysis to analyze demographic differences using t
-tests and Mann-Whitney U
tests for continuous variables, the Fisher exact test for categorical variables, and multivariate logistic regression.
One hundred five patients (22.8%) received an opioid before the appointment. Fifty-two (11.3%) received schedule II or III opioids, 43 (9.3%) received tramadol, and 10 (2.2%) received both. Primary care physicians were the most common prescriber (59.5%, P
< 0.001) followed by pain medicine specialists (11.3%) and orthopaedic surgeons (11.3%). More prescribers practiced in the community than academic setting (63.8% versus 36.2%, P
< 0.001). Seventy-eight patients (74.3%) self-reported their opioid prescriptions, with the remaining 27 patients (25.7%; 14 schedule II or III opioids and 13 tramadol) identified only after query of the Prescription Monitoring Program. In regression analysis, higher body mass index, diagnosis other than osteoarthritis, and benzodiazepine use were associated with receiving opioids (P
< 0.05), while antidepressant use decreased the likelihood of self-reporting opioid prescriptions (P
A striking number of patients are being treated with opioids for hip and knee arthritis. Furthermore, many patients who have received opioids within 6 months do not report their prescriptions. Although primary care physicians prescribed most opioids for nonsurgical treatment of arthritis, a substantial percentage came from orthopaedic surgeons. Further education of physicians and patients on the ill effects of opioids when used for the nonsurgical treatment of hip and knee arthritis is warranted.
Level of Evidence:
Level III, retrospective cohort study