Delayed timing of physical therapy initiation increases the risk of future opioid use in individuals with knee osteoarthritis: a real-world cohort study.

OBJECTIVE: We assessed whether late versus early initiation of physical therapy (PT) was related to greater risk of future opioid use in people with knee osteoarthritis (OA) who receive PT. METHODS: We used Commercial and Medicare Advantage claims data from 1999 to 2018 from American adults with incident knee OA referred for PT within 1 year of diagnosis. We categorised people as opioid naïve or opioid experienced based on prior prescriptions. We examined the association of timing of PT initiation with any and chronic opioid use over 1 year. RESULTS: Of the 67 245 individuals with incident knee OA, 35 899 were opioid naïve and 31 346 were opioid experienced. In the opioid naïve group, compared with PT within 1 month, PT 1 to <3, 3 to <6, 6 to <9, 9-12 months from diagnosis was associated with adjusted risk ratio (aRR (95% CIs)) for any opioid use of 1.18 (1.10 to 1.28), 1.49 (1.37 to 1.61), 1.73 (1.58 to 1.89) and 1.93 (1.76 to 2.12), respectively; aRRs (95% CIs) for chronic opioid use were 1.25 (1.01 to 1.54), 1.83 (1.48 to 2.26), 2.29 (1.82 to 2.89) and 2.50 (1.96 to 3.19). Results were similar among opioid experienced; aRRs (95% CIs) for any opioid use were 1.19 (1.14 to 1.24), 1.32 (1.26 to 1.37), 1.39 (1.32 to 1.45) and 1.54 (1.46 to 1.61); aRRs (95% CIs) for chronic opioid use were 1.25 (1.17 to1.34), 1.43 (1.33 to 1.54), 1.53 (1.41 to 1.66) and 1.65 (1.51 to 1.80). CONCLUSION: Compared with PT initiation within 1 month, delayed PT initiation was associated with higher risk of opioid use in people with incident knee OA. The longer the delay in PT initiation, the greater was the risk.

Warfarin use and risk of knee and hip replacements.

BACKGROUND: Identification of modifiable risk factors and treatments for osteoarthritis (OA) are needed. Warfarin, a vitamin K antagonist, causes fetal and animal model skeletal abnormalities. Vitamin K insufficiency has been associated with OA, but whether warfarin is also detrimental to OA is not known. METHODS: We conducted a nested case-control study using a UK general practitioner electronic medical records database. We identified cases of knee or hip replacement (KR or HR) from among adults with atrial fibrillation newly prescribed either warfarin or direct oral anticoagulants (DOACs). Cases were matched with four controls by age and sex. We assessed the relation of warfarin compared with DOAC use to risk of joint replacement using conditional logistic regression. We also evaluated different durations of warfarin use. RESULTS: We identified 857 subjects with KR or HR (cases), of whom 64.6% were warfarin users, and 3428 matched controls, of whom 56.1% were warfarin users (mean age 75, 47% female). Warfarin users had a 1.59 times higher risk of joint replacement than DOAC users (adjusted OR 1.59, 95% CI 1.31 to 1.92). Longer duration of warfarin use was associated with higher risk of joint replacement in comparison with <1 year of warfarin use. CONCLUSION: Warfarin, a vitamin K antagonist, was associated with greater risk of KR and HR (an indicator for end-stage knee OA) than DOAC use, supporting the importance of adequate vitamin K functioning in limiting OA progression.

Effect of bisphosphonates on knee replacement surgery.

PURPOSE: Bone remodelling as a therapeutic target in knee osteoarthritis (OA) has gained much interest, but the effects of antiresorptive agents on knee OA have been conflicting, with no studies to date examining the effects of bisphosphonate use on the clinically relevant endpoint of knee replacement (KR) surgery. METHODS: We used data from The Health Improvement Network (THIN), a general practitioner electronic medical records representative of the general UK population. We identified older women who had initiated bisphosphonate use after their incident knee OA diagnosis. Each bisphosphonate initiator was propensity score-matched with a non-initiator within each 1-year cohort accrual block. The effect of bisphosphonates on the risk of KR was assessed using Cox proportional hazard regression. Sensitivity analyses to address residual confounding were also conducted. RESULTS: We identified 2006 bisphosphonate initiators, who were matched to 2006 non-initiators(mean age 76, mean body mass index 27), with mean follow-up time of 3 years. The crude incidence rate of KR was 22.0 per 1000 person-years among the initiators, and 29.1 among the non-initiators. Bisphosphonate initiators had 26% lower risk of KR than non-initiators(HR 0.74, 95% CI 0.59 to 0.93); these results were similar when additionally adjusted for potential confounders in the propensity score (HR 0.76, 95% CI 0.60 to 0.95). Results of sensitivity analyses supported this protective effect. CONCLUSIONS: In this population-based cohort of older women with incident knee OA, those with incident bisphosphonate users had lower risk of KR than non-users of bisphosphonates, suggesting a potential beneficial effect of bisphosphonates on knee OA.

Risk of myocardial infarction with use of selected non-steroidal anti-inflammatory drugs in patients with spondyloarthritis and osteoarthritis.

OBJECTIVES: Spondyloarthritis (SpA) is associated with an increased risk of myocardial infarction (MI) due to underlying inflammation and possibly due to medications such as certain non-steroidal anti-inflammatory drugs (NSAIDs). We sought to describe MI risk among patients with SpA who were prescribed NSAIDs, and to compare the pattern of risk in SpA with that in osteoarthritis (OA). METHODS: Nested case-control studies were performed using The Health Improvement Network (THIN). Underlying cohorts included adults with incident SpA or OA who had >1 NSAID prescription and no history of MI. Within each cohort, we matched each MI case to four controls without MI. NSAID use was categorised as: (a) current (prescription date 0-180 days prior to index date), (b) recent (181-365 days) or (c) remote (>365 days). We performed conditional logistic regression to compare the odds of current or recent NSAID use relative to remote use of any NSAID, considering diclofenac and naproxen specifically. RESULTS: Within the SpA cohort of 8140 and the OA cohort of 244 339, there were 115 and 6287 MI cases, respectively. After adjustment, current diclofenac use in SpA was associated with an OR of 3.32 (95% CI 1.57 to 7.03) for MI. Naproxen was not associated with any increase (adjusted OR 1.19, 95% CI 0.53 to 2.68). A ratio of ORs for SpA/diclofenac relative to OA/diclofenac was 2.64 (95% CI 1.24 to 5.58). CONCLUSIONS: MI risk in SpA is increased among current users of diclofenac, but not naproxen. The MI risk with diclofenac in SpA appears to differ from that in OA.

Improved survival in rheumatoid arthritis: a general population-based cohort study.

OBJECTIVE: Mortality trends of rheumatoid arthritis (RA) are largely unknown over the past decade when new drugs and management strategies have been adopted to effectively treat RA. METHODS: Using The Health Improvement Network, an electronic medical record database representative of the UK general population, we identified patients with incident RA and up to five individuals without RA matched for age, sex and year of diagnosis between 1999 and 2014. The RA cohort was divided in two sub-cohorts based on the year of RA diagnosis: the early cohort (1999-2006) and the late cohort (2007-2014). We compared mortality rates, HRs (using a Cox proportional hazard model) and rate differences (using an additive hazard model) between RA and non-RA cohorts adjusting for potential confounders. RESULTS: Patients with RA diagnosed between 1999 and 2006 had a considerably higher mortality rate than their comparison cohort (ie, 29.1 vs 18.0 deaths/1000 person-years), as compared with a moderate difference in patients with RA diagnosed between 2007 and 2014 and their comparison cohort (17.0 vs 12.9 deaths/1000 years). The corresponding absolute mortality rate differences were 9.5 deaths/1000 person-years (95% CIs 7.5 to 11.6) and 3.1 deaths/1000 person-years (95% CI 1.5 to 4.6) and the mortality HRs were 1.56 (95% CI 1.44 to 1.69) and 1.29 (95% CI 1.17 to 1.42), respectively (both p values for interaction <0.01). CONCLUSION: This general population-based cohort study indicates that the survival of patients with RA has improved over the past decade to a greater degree than in the general population. Improved management of RA and its associated comorbidities over recent years may be providing a survival benefit.

Meloxicam and risk of myocardial infarction: a population-based nested case-control study.

Certain non-steroidal anti-inflammatory drugs (NSAIDs) have been associated with an increased risk of myocardial infarction (MI), a risk linked to cyclo-oxygenase-2 inhibition. There are limited studies assessing the risk of MI associated with meloxicam, an increasingly popular drug with COX-2 inhibiting properties. A nested matched case-control study using The Health Improvement Network, a UK population-based database was conducted. NSAID users between 35 and 89 years of age with at least 1 year enrollment in the cohort were included. Incident MI cases were matched on age, sex, practice and event date with up to 4 controls. NSAID exposure was categorized as remote (between 60 days and 1 year), recent (between 1 and 60 days) or current relative to the event date. Current users were further classified as naproxen (negative control), diclofenac (positive control), meloxicam or other NSAID users. Multivariable conditional logistic regression was conducted to determine the risk of MI for each NSAID use categories compared with that of remote users. 9291 MI cases were matched with 30,676 controls. The cases had a higher prevalence of traditional cardiac risk factors, chronic kidney disease and inflammatory arthritis and cardioprotective drug utilization. The adjusted odds ratio of MI for current user compared to remote users were: meloxicam 1.38 (1.17-1.63), naproxen 1.12 (0.96-1.30) and diclofenac 1.37 (1.25-1.50). In this large population-based study, meloxicam increased the risk of MI by 38%. This study warrants cautious use of this increasingly popular drug.

Independent impact of gout on the risk of diabetes mellitus among women and men: a population-based, BMI-matched cohort study.

OBJECTIVE: Evidence on the potential independent impact of gout on the risk of diabetes is limited to a single study of men with a high cardiovascular risk profile. Our objective was to examine this relation in the general population, particularly among women. METHODS: We conducted a sex-stratified matched cohort study using data from The Health Improvement Network (THIN), an electronic medical records database representative of the UK general population. Up to five non-gout individuals were matched to each case of incident gout by year of birth, year of enrolment and body mass index (BMI). Multivariate HRs for incident diabetes were calculated after additionally adjusting for smoking, alcohol consumption, physician visits, comorbidities and medication use. RESULTS: Among 35 339 gout patients (72.4% men, mean age of 62.7 years), the incidence rates of diabetes in women and men were 10.1 and 9.5 cases per 1000 person-years, respectively, whereas the corresponding rates were 5.6 and 7.2 cases per 1000 person-years among 137 056 non-gout subjects. The BMI-matched univariate and multivariate HRs of diabetes were higher among women compared with those among men (1.71; 95% CI 1.51 to 1.93 vs 1.22; 95% CI 1.13 to 1.31) and (1.48; 95% CI 1.29 to 1.68 vs 1.15; 95% CI 1.06 to 1.24), respectively (p values for interaction <0.001). This sex difference persisted across age-specific subgroups. CONCLUSIONS: This general population-based study suggests that gout may be independently associated with an increased risk of diabetes and that the magnitude of association is significantly larger in women than in men.

Validity of ankylosing spondylitis diagnoses in The Health Improvement Network.

BACKGROUND/PURPOSE: Because ankylosing spondylitis (AS) is uncommon, large medical record databases offer important opportunities for pharmacoepidemiologic research. However, the validity of AS diagnoses recorded by a general practitioner (GP) is unknown. We assessed the validity of algorithms for identifying AS in The Health Improvement Network (THIN). METHODS: THIN is a database of GP records for over 10 million persons in the UK. In 2014, we administered a questionnaire to GPs of 100 adults for whom an AS diagnosis had been recorded. As high positive predictive value (PPV) is critically important in AS research, we sought to determine the PPV of an AS diagnostic code relative to the GP's clinical impression as the gold standard. Other AS algorithms included: more than one AS diagnostic code, prescription of a nonsteroidal anti-inflammatory drug (NSAID), disease modifying anti-rheumatic drug (DMARD) or biologic. RESULTS: In 61 of 85 returned questionnaires, the GP's clinical impression confirmed AS yielding an overall PPV of 72%. PPV was 89% for two AS codes >7 days apart, and was 86% for an AS code plus a DMARD/biologic. Sensitivity was reduced with algorithms requiring two AS codes (64%) and a DMARD/biologic prescription (30%). Algorithms requiring prescription of an NSAID, or the absence of OA or RA had lower PPV (71-75%) and higher sensitivity (95-98%). CONCLUSION: An AS identification algorithm of two AS diagnoses separated by >7 days provided the highest PPV. This algorithm should be used for pharmacoepidemiologic studies in THIN.

Risk for cardiovascular disease early and late after a diagnosis of giant-cell arteritis: a cohort study.

BACKGROUND: Involvement of large arteries is well-documented in giant-cell arteritis (GCA), but the risk for cardiovascular events is not well-understood. OBJECTIVE: To evaluate the risks for incident myocardial infarction (MI), cerebrovascular accident (CVA), and peripheral vascular disease (PVD) in individuals with incident GCA in a general population context. DESIGN: Observational cohort study. SETTING: U.K. primary care database. PATIENTS: 3408 patients with incident GCA and 17 027 age- and sex-matched reference participants without baseline cardiovascular disease (MI, CVA, or PVD). MEASUREMENTS: Diagnoses of GCA, outcomes, and cardiovascular risk factors were identified from electronic medical records. One combined and 3 separate cohort analyses were conducted for the outcomes of MI, CVA, and PVD. The association of GCA with study outcomes is expressed with hazard ratios (HRs) with 95% CIs after adjustment for potential cardiovascular risk factors. RESULTS: Among 3408 patients with GCA (73% female; mean age, 73 years), the incidence rates of MI, CVA, and PVD were 10.0, 8.0, and 4.2 events per 1000 person-years, respectively, versus 4.9, 6.3, and 2.0 events per 1000 person-years, respectively, among reference participants. The HRs were 1.70 (95% CI, 1.51 to 1.91) for the combined outcome, 2.06 (CI, 1.72 to 2.46) for MI, 1.28 (CI, 1.06 to 1.54) for CVA, and 2.13 (CI, 1.61 to 2.81) for PVD. The HRs were more pronounced in the first month after GCA diagnosis (combined HR, 4.92 [CI, 2.59 to 9.34]; HR for MI, 11.89 [CI, 2.40 to 59.00]; HR for CVA, 3.93 [CI, 1.76 to 8.79]; HR for PVD, 3.86 [CI, 0.78 to 19.17]). LIMITATION: Information on temporal arterial biopsies was not available, and there was a substantial amount of missing data on cardiovascular risk factors. CONCLUSION: Giant-cell arteritis is associated with increased risks for MI, CVA, and PVD. PRIMARY FUNDING SOURCE: National Institute of Arthritis and Musculoskeletal and Skin Diseases.

The risk of cardiovascular disease in systemic sclerosis: a population-based cohort study.

OBJECTIVES: To evaluate the risk of incident myocardial infarction (MI), stroke and peripheral vascular disease (PVD) in individuals with systemic sclerosis (SSc) in a general population context. METHODS: We conducted a cohort study using a UK primary care database containing records from 1986 to 2011. SSc diagnoses, outcomes and cardiovascular risk factors were identified from electronic medical records. We conducted two cohort analyses: (1) MI and stroke, and (2) PVD, excluding individuals with prevalent disease at baseline for each analysis. We estimated HRs comparing SSc with age-, sex- and entry time-matched comparison cohorts, adjusting for potential cardiovascular risk factors. RESULTS: Among 865 individuals with SSc (85.8% women, mean age 58.7 years), the incidence rates (IRs) of MI and stroke were 4.4 and 4.8 per 1000 person-years (PY), versus 2.5 and 2.5 per 1000 PY in the comparison cohort. The corresponding adjusted HRs were 1.80 (95% CI 1.07 to 3.05) for MI and 2.61 (95% CI 1.54 to 4.44) for stroke. Among 858 individuals with SSc (85.3% female, mean age 58.9 years), the IR of PVD was 7.6 per 1000 PY versus 1.9 per 1000 PY in the comparison cohort, with an adjusted HR of 4.35 (95% CI 2.74 to 6.93). CONCLUSIONS: These findings provide the first general population-based evidence that SSc is associated with an increased risk of developing MI, stroke and PVD. Further insight into disease mechanisms, as well as how disease subtype, organ involvement and medication use may alter these increased risks, is needed.

Validation of knee osteoarthritis case identification algorithms in a large electronic health record database.

Purpose: To facilitate studies of knee osteoarthritis (OA) in large databases, case finding algorithms with high levels of diagnostic performance are needed. Methods: From a UK general practitioner (GP) practice derived database, we selected adults ages 40-90 years meeting algorithms that included various combinations of codes for knee OA or knee pain and imaging. The GP for each patient was mailed a questionnaire to assess the cause of knee pain and provide knee x-ray and/or MRI findings. We considered knee pain with x-ray and/or MRI findings consistent with OA the gold standard. We calculated positive predictive values (PPV) and sensitivity for case identification algorithms. Results: Of 100 questionnaires sent, 93 were returned; we excluded 8 subjects who had other rheumatic disorders or total knee replacements. Among those with one code for OA, the PPV was 64% (95% CI = 49%-79%) and it increased to 92% (95% CI = 76%-100%) when two or more OA codes over six months were required. The increase in PPV was accompanied by a drop in sensitivity from 44% (95% CI = 31%-57%) to 19% (95% CI = 9%-30%). Use of one pain code yielded similar results to use of one OA code. Requiring two or more knee pain codes over six months yielded a PPV of 68% (95% CI = 49%-88%) and sensitivity of 26% (95% CI = 15%-38%). Discussion: A case identification algorithm requiring two or more knee OA codes yielded the highest PPV at the cost of reduced sensitivity. Tradeoffs between PPV and sensitivity will need to be weighed alongside study goals when selecting a case identification algorithm.

Proton-Pump Inhibitors and Risk of Calcium Pyrophosphate Deposition in a Population-Based Study.

OBJECTIVE: There are no proven effective medical treatments to prevent calcium pyrophosphate crystal deposition (CPPD). Hypomagnesemia is a known CPPD risk factor. The present study was undertaken to carry out a real-world epidemiologic study on proton-pump inhibitor (PPI) use, which can cause hypomagnesemia, and CPPD risk. METHODS: We conducted a time-stratified, propensity score (PS)-matched cohort study using the UK-based IQVIA Medical Research Data. We compared risk of incident CPPD among PPI users versus H2 blocker users using Cox proportional hazards models. We used greedy matching of incident PPI users 1:1 to incident histamine receptor 2 (H2 ) blocker users in 1-year cohort accrual blocks. Subjects were censored at time of drug switch. We evaluated incident use of PPI and H2 blockers prior to incident CPPD using a nested case-control study within the same cohort, matched 1:4 by age and sex using risk-set sampling. RESULTS: We identified 81,102 PPI and H2 blocker initiators, with 113 and 63 incident cases of CPPD, respectively. In the case-control study when compared with nonusers, both PPI and H2 B users had higher risk of incident CPPD, with odds ratios (ORs) of 1.79 (95% confidence interval [95% CI] 1.55-2.07) and 1.52 (95% CI 1.14-2.03), respectively. Incident PPI use was nonsignificantly associated with incident CPPD (hazard ratio 1.03 [95% CI 0.75-1.41]) compared with H2 blocker use. CONCLUSION: In this study using real-world data, incident use of PPIs was not associated with a higher risk of CPPD compared with incident H2 blocker use, although use of PPI and H2 blockers had higher risk compared with nonuse.

Relation of therapies for ankylosing spondylitis and psoriatic arthritis to risk of myocardial infarction: a nested case control study.

BACKGROUND: Risk of myocardial infarction (MI) is elevated in ankylosing spondylitis and psoriatic arthritis (AS/PsA) compared to the general population. We evaluated the risk of MI related to the use of tumor necrosis factor inhibitor (TNFi) and other therapies in AS/PsA. METHODS: We conducted a nested case-control study using 1994-2018 data from OptumLabs® Data Warehouse, which includes de-identified medical and pharmacy claims, laboratory results, and enrollment records for commercial and Medicare Advantage enrollees. The database contains longitudinal health information on enrollees and patients, representing a diverse mixture of ages, ethnicities and geographical regions across the United States. Assessing AS/PsA separately, MI cases were matched to 4 controls by sex, age, diagnosis year and insurance type. We evaluated treatment within 6 months prior to MI including NSAIDs (AS referent), disease-modifying anti-rheumatic drug (DMARDs; PsA referent) and TNFi alone or in combinations. We evaluated the relation of treatment categories to MI risk using conditional logistical regression adjusting for confounders. RESULTS: Among 26,648 AS subjects, there were 237 MI cases and 894 matched controls. Among 43,734 PsA subjects, there were 404 cases and 1596 controls. In AS, relative to NSAID use, the adjusted odds ratio (aOR) for MI among TNFi only users was 0.85 (95% CI 0.39-1.85) and for DMARD only users was 1.04 (95% CI 0.65-1.68). In PsA, relative to DMARD use, the aOR among TNFi only was 1.09 (95% CI 0.74-1.60). Combination therapies also had no effect. CONCLUSIONS: Among AS/PsA, no combination of therapies appeared to be protective or harmful with regards to MI. Future studies should capture more AS and PsA patients and include longer term follow up to further investigate this question.

Relation of NSAIDs, DMARDs, and TNF Inhibitors for Ankylosing Spondylitis and Psoriatic Arthritis to Risk of Total Hip and Knee Arthroplasty.

OBJECTIVE: Ankylosing spondylitis (AS) and psoriatic arthritis (PsA) often affect the hip and/or knee. If effective, treatments might reduce risk of total hip or total knee arthroplasty (THA/TKA). We evaluated risk of THA/TKA related to use of medical therapies in AS/PsA. METHODS: We conducted a nested case-control study using 1994-2018 data from the OptumLabs Data Warehouse, which includes deidentified medical and pharmacy claims, laboratory results, and enrollment records for commercial and Medicare Advantage enrollees. Among those with AS/PsA, THA/TKA cases were matched up to 4 controls by sex, age, AS/PsA diagnosis, diagnosis year, insurance type, obesity, and prior THA/TKA. We assessed AS/PsA treatment 6 months prior to THA/TKA, including disease-modifying antirheumatic drugs (DMARDs) and tumor necrosis factor inhibitors (TNFi), alone or in combination, stratified by nonsteroidal antiinflammatory drug (NSAID) use. We evaluated the relation of treatment to risk of THA/TKA using conditional logistical regression with adjustment for confounders. RESULTS: Among 16,748 adults with AS, there were 444 THA/TKA cases and 1613 matched controls. Among 34,512 adults with PsA, there were 1003 cases and 3793 controls. Adjusted ORs for treatment category and THA/TKA ranged from 0.60 to 1.92; however, none were statistically significant. Results were similarly null in several sensitivity analyses. CONCLUSION: Odds of THA/TKA were not reduced with any combinations of NSAIDs, DMARDs, or TNFi among persons with AS or PsA. Given current utilization patterns in this population of US adults with AS and PsA, these medical therapies did not appear to be associated with less end-stage peripheral joint damage.

Association of Physical Therapy Interventions With Long-term Opioid Use After Total Knee Replacement.

Importance: Many individuals who undergo total knee replacement (TKR) become long-term opioid users after TKR. Associations of physical therapy (PT) interventions before or after TKR with long-term use of opioids are not known. Objectives: To evaluate associations of PT interventions before and after TKR with long-term opioid use after TKR. Design, Setting, and Participants: This cohort study used data from the OptumLabs Data Warehouse on 67 322 individuals aged 40 years or older who underwent TKR from January 1, 2001, to December 31, 2016, stratified by history of opioid use. The analyses for the study included data from January 1, 1999, to December 31, 2018. Exposures: Any PT interventions within 90 days before or after TKR, post-TKR PT dose as number of sessions (ie, 1-5, 6-12, and ≥13 sessions), post-TKR PT timing as number of days to initiation of care (ie, <30 days, 31-60 days, or 61-90 days after TKR), and post-TKR PT type (ie, active vs passive). Main Outcomes and Measures: The association of pre- and post-TKR PT with risk of long-term opioid use occurring more than 90 days after TKR was assessed using logistic regression while adjusting for confounders, including age, sex, race and ethnicity (Asian, Black, Hispanic, or White), obesity, type of insurance, geographical location, and physical and mental health comorbidities. Results: A total of 38 408 opioid-naive individuals (21 336 women [55.6%]; mean [SD] age, 66.2 [9.2] years) and 28 914 opioid-experienced individuals (18 426 women [63.7%]; mean [SD] age, 64.4 [9.3] years) were included. Receipt of any PT before TKR was associated with lower odds of long-term opioid use in the opioid-naive (adjusted odds ratio [aOR], 0.75 [95% CI, 0.60-0.95]) and opioid-experienced (aOR, 0.75 [95% CI, 0.70-0.80]) cohorts. Receipt of any post-TKR PT was associated with lower odds of long-term use of opioids in the opioid-experienced cohort (aOR, 0.75 [95% CI, 0.70-0.79]). Compared with 1 to 5 sessions of PT after TKR, 6 to 12 sessions (aOR, 0.82 [95% CI, 0.75-0.90]) and 13 or more sessions (aOR, 0.71 [95% CI, 0.65-0.77) were associated with lower odds in the opioid-experienced cohort. Compared with initiation of PT within 30 days after TKR, initiation 31 to 60 days or 61 to 90 days after TKR were associated with greater odds in the opioid-naive (31-60 days: aOR, 1.45 [95% CI, 1.19-1.77]; 61-90 days: aOR, 2.15 [95% CI, 1.43-3.22]) and opioid-experienced (31-60 days: aOR, 1.10 [95% CI, 1.02-1.18]; 61-90 days: aOR, 1.32 [95% CI, 1.12-1.55]) cohorts. Compared with passive PT, active PT was not associated with long-term opioid use in the opioid-naive (aOR, 1.00 [95% CI, 0.81-1.24]) or opioid-experienced (aOR, 0.99 [95% CI, 0.92-1.07]) cohorts. Conclusions and Relevance: This cohort study suggests that receipt of PT intervention before and after TKR, receipt of 6 or more sessions of PT care after TKR, and initiation of PT care within 30 days after TKR were associated with lower odds of long-term opioid use. These findings suggest that PT may help reduce the risk of long-term opioid use after TKR.