1University of Washington, Seattle, WA, USA. 2Pfizer Inc, New York, NY, USA. 3Eli Lilly and Co, Indianapolis, IN, USA. 4Pfizer Ltd, Surrey, United Kingdom
Purpose Chronic pain is typically defined as pain that lasts for more than 3 to 6 months and can be the result of a wide array of issues, including underlying medical conditions or disease such as osteoarthritis (OA). OA is a degenerative joint disease involving the cartilage and surrounding tissues that impacts approximately 32.5 million adults in the United States (US). Two of the most commonly affected OA joints are the hips and the knees (it can also affect the hands, facet joints, and feet). The current standard of care for OA focuses on symptomatic pain management and improving joint movement. Pain management options are classified into non-pharmacological and pharmacological management, and joint surgery. Pharmacological treatment options for OA pain commence with the use of oral analgesics, followed by topical/oral non-steroidal anti-inflammatory drugs (NSAIDs) and opioids. A number of studies have assessed the relative effectiveness and safety of OA pain medications. A literature review conducted by Costa et al (2014) assessed the effectiveness of opioids (oral and transdermal) for the treatment of hip and knee OA (vs placebo or no intervention) and found a limited effect of opioids on pain and functional outcomes. For safety outcomes, adverse events and substance abuse could potentially limit the use of opioids for OA pain management. Although the American College of Rheumatology’s clinical practice guideline conditionally recommends the use of tramadol and conditionally recommends against the use of non-tramadol opioids for patients with OA, opioids continue to be prescribed. The aim of this study was to characterize the demographics of patients who received treatment for moderate-to-severe OA pain, and to compare them to the non-moderate-to-severe OA pain cohort. Methods This was a retrospective cohort study using data from the 2013–2018 IBM® MarketScan®Commercial and Medicare Supplemental Database (integrated US commercial claims database that contains patient-level demographic, diagnosis, inpatient, outpatient, procedure, prescription and payment information). Patients aged ≥45 years, with at least one diagnosis of hip and/or knee OA (ICD9/10) or unspecified diagnosis of OA plus a diagnosis of pain in the hip or knee within 3 months during the study period were included. The date of the first OA diagnosis was defined as the index date. Patients who presented with at least one of the following criteria: a visit to a specialist within 3 months post-index date, a surgical or non-surgical invasive procedure relating to OA treatment within 12 months post-index period, had ≥2 prescriptions for various NSAIDs within 3 months post-index date, ≥2 prescriptions for any opioids within 3 months post-index date, or an Emergency Room (ER) visit for hip or knee OA within 12 months post-index date, with a subsequent primary care physician visit within 14 days were classified as having moderate-to-severe OA pain. To minimize selection bias, moderate-to-severe OA pain patients were propensity-score matched 1:1 with patients with an OA diagnosis to non-moderate-to-severe pain (controls) using age, sex, Charlson Comorbidity Index (CCI) score, type of health plan, obesity, anxiety, depression, and geographical region. Descriptive analyses were performed to gain a better understanding of the data used in this study and of the characteristics of the study populations. Binary and categorical variables were summarized as counts and proportions of the total study populations, and by subgroups where appropriate. Continuous variables were reported as means (standard deviations) or median and ranges, where appropriate. Bivariate analyses were conducted to identify significant demographic/clinical characteristics between the treatment/outcome groups and to identify potential covariates for inclusion in multivariable models. For dichotomous and categorical measures, Chi-squared tests or Fisher’s exact tests were used to test for differences between groups. For continuous measures, t-tests or Wilcoxon rank-sum tests were utilized to test for differences for two-sample comparisons and one-way analysis of variance in the case of multi-sample comparisons. Results A total of 546,254 patients with OA were eligible for the study; after propensity score matching, the final cohort count was 186,374 in each group. Overall, 65.6% of patients were aged 45-65 years and 34.4% were ≥65 years and there were more females (61.0%) than males in both cohorts. More than half of patients within the moderate-to-severe cohort were on a Preferred Provider Organization (PPO) healthcare plan (51.0%) and most were from the Southern and North-Central regions of the US. Prior to matching, patients in the moderate-to-severe cohort had a lower mean CCI score (0.95) compared with the non-moderate-to-severe cohort (1.11; P<0.0001). Comorbidities of interest were reported for 37.6% of patients (moderate-to-severe) vs 32.6% (non-moderate-to-severe) with the majority being sleep-related (16.8% vs 14.0%, respectively; P<0.0001).After propensity score matching, overall pain medication use was greater in the moderate-to-severe pain cohort 12-months (89.6% vs 76.1%) and 24-months’ post-index (94.0% vs 85.0%) vs the non-moderate-to-severe cohort (P<0.0001).The three most commonly prescribed medications at baseline were NSAIDs (39.5% vs 32.3%), non-tramadol opioids (41.2 vs 28.6%) and antidepressants (28.3% vs 21.7%) in the moderate-to-severe vs non-moderate-to-severe cohorts, respectively (all P<0.0001). In the 12-month follow-up period, prescriptions for non-tramadol opioids, tramadol and intra-articular injections of corticosteroids increased (38.6%, 54.6% and 133.6%, respectively) in the moderate-to-severe cohort, while prescriptions for non-tramadol opioids, tramadol and intra-articular injections of corticosteroids increased by smaller rates (2.6%, 16.4% and 88.5%, respectively) in the non-moderate-to-severe group over the same period. In the 24-month follow-up period, prescriptions for non-tramadol opioids, tramadol and intra-articular injections of corticosteroids also increased (66.4%, 101.8% and 185.3% in the moderate-to-severe cohort, and 44.9%, 51.0%, and 35.8% in the non-moderate-to-severe group) over the same period. Conclusions This study provides valuable information on the characteristics of patients with moderate-to-severe OA pain of the hip and/or knee. In this study, most patients with moderate-to-severe OA pain were female, from the Southern and North-Central regions of the US, aged 45-65 years, and on a PPO healthcare plan.A greater proportion of patients with moderate-to-severe OA pain reported comorbidities of interest and greater use of pain medication that those with non-moderate-to-severe OA pain.This study also provides insights into current treatment paradigms and showed that patients with moderate-to-severe OA pain had a substantial increase in prescriptions for non-tramadol opioids, tramadol and intra-articular injections of corticosteroids 12 months from baseline, while those with non-moderate-to-severe OA pain received smaller increases for the same medications over the same time period. Moreover, greater increases in prescriptions for non-tramadol opioids, tramadol and intra-articular injections of corticosteroids were noted 24 months from baseline in the moderate-to-severe group than the non-moderate-to-severe group. Data are limited as causal relationship between factors and outcomes cannot be inferred; however, these data demonstrate potential areas for further exploration in improving patient outcomes.