In this retrospective study of nearly 2 million US patients treated with prescription medications for OA pain, opioids were more commonly used than NSAIDs and intra-articular corticosteroid injections. We found an overall TJR incidence rate of 3.21 per 100 person-years, and the incidence rate of TJR was highest in patients receiving intra-articular corticosteroids and lowest in patients receiving NSAIDs.
Our analysis found a high prevalence of comorbid conditions, including cardiovascular disease, diabetes, obesity, and osteoporosis, in patients with OA receiving pain medications. Similarly, a 2011 claims database cohort analysis by Gore et al. showed that patients with OA had a significantly higher prevalence of comorbidities and greater use of pain medications compared with a control cohort [19]. Postler et al. conducted a cross-sectional study of German claims data and found comorbidities were common among 595 754 patients with hip or knee OA (e.g., arterial hypertension in 78.7%, diabetes in 29.0%, and depression in 27.9%) [20]. Their study showed that 63.4% of patients with OA were prescribed analgesics and 44.1% were prescribed NSAIDs [20].
Currently, NSAIDs are first-line pharmacologic agents for management of pain associated with OA; this is despite their associated risks [9, 21,22,23]. Growing recognition of the risks associated with non-tramadol opioids has led to conditional recommendations against their use [9, 21,22,23]. One recent, large observational study by Nalamachu et al. reported that 710 of 841 (84%) commercially insured patients visiting physicians for treatment of OA had been prescribed a pain medication [5]. Their analysis found that patients with severe OA pain were more likely than those with mild or moderate pain to have been prescribed opioids and multiple medications, and > 80% of treatment switches were because of lack of efficacy [5].
In contrast to our current analysis, which found opioids were the most commonly prescribed treatment at index date in 62.6% of the “any pain medication” cohort, a 2021 retrospective US electronic health records analysis of adults with a new diagnosis of knee OA found that the most frequent first treatment was intra-articular corticosteroids, in 25.5% of patients, and the most frequent second treatment was opioids, in 15.8% [24]. However, our analysis was not limited to patients with newly diagnosed OA, which may explain the treatment difference. Analyses of data from patients with knee OA in the South Korean nationwide claims database showed that 12.2% received opioids (most commonly tramadol) as their first treatment [25] and 82.5% received an NSAID at any point [26]. Notably, a retrospective longitudinal study of US insurance claims data found that among patients with hip or knee OA who received opioids, 34.9% had evidence of opioid regimen failure, as suggested by an increase in regimen intensity, additional non-opioid pain medication, opioid abuse events, or joint surgery [10]. Our current analysis also found that a substantial proportion (20.6%) of patients who received an opioid at index date subsequently received an NSAID.
An analysis by Berger et al. of US health insurance data for patients with OA who had undergone a TJR showed that in the 2 years prior to hip or knee TJR, 55.2% had received prescription NSAIDs, 58.4% had received opioids, and 50.3% had received intra-articular corticosteroid injections [27]. Similarly, an observational cohort study by Jin et al. of Medicare patients who underwent hip or knee TJR showed that 60.2% had received opioids in the 12 months before surgery [28].
Our current analysis of data from patients being treated for OA-related pain between 2013 and 2019 found an overall TJR incidence rate per 100 person-years of 3.21. A 2011 review of trends in knee and hip TJR procedures performed for OA found that TJR rates varied by country, have increased over the preceding 2–3 decades, and were predicted to increase further due to improved longevity and an aging population [29]. The cross-sectional study of German claims data by Postler et al. showed that 5.3% of patients with hip or knee OA underwent a TJR in 2014 [20]. Among Danish postmenopausal women, the incidence rates of knee and hip TJR increased with age up to 80–84 years of age, when the yearly average incidence rates reached 64 per 10 000 population for knee TJR and 115 per 10 000 population for hip TJR [30].
A possible limitation of our findings is the use of claims-based algorithms to define exposures, outcomes, and baseline characteristics, which may have caused some patients to be misclassified. In addition, there was a lack of data available for use of non-prescription pain medication (including non-prescription NSAIDs), complementary and alternative medicines, and homeopathic remedies. Further, chronic pain may be a driver for increased health care utilization, and we would expect to see increased utilization among those with chronic pain [31]. In addition, we described pain medication use among primarily commercially insured patients, and our results may not be comparable to patients with publicly insured health plans [32].