Background Identification of patients who benefit from Total Hip or Knee Arthroplasty (THA or TKA) is necessary to indicate the need for surgery. This seems difficult as 20% of all patients report unfavourable outcomes. To provide better treatment options for patients with unfavourable outcomes, they should be identified at an early stage after surgery.
Objectives I To investigate the prevalence of four trajectories in THA and TKA patients for both pain and function separately: (1) Patients with favourable initial and long-term outcomes, (2) patients with favourable initial but unfavourable long-term outcomes, (3) patients with unfavourable initial and favourable long-term outcomes and (4) patients with unfavourable initial and long-term outcomes. II To determine if long-term outcomes in all patients and in patient groups stratified by initial outcome (favourable/unfavourable) can be predicted on the basis of pre- and early postoperative variables.
Methods Prospective, multicentre, observational study including consecutive patients undergoing primary THA and TKA (Longitudinal Leiden Orthopaedics and Outcomes of OsteoArthritis Study (LOAS). Measurements were done preoperatively and at 6 and 12 months. Pain and function were assessed with the Hip Disability/ Knee injury and Osteoarthritis Outcome Scores (HOOS or KOOS) Pain and Activity Daily Living (ADL) subscales (score range 0–100). In addition, preoperative variables (sex, age, comorbidities) and early postoperative complications (reoperations or readmissions) were gathered by questionnaire. Scores of the Pain and ADL subscales above and below 70 were defined as favourable and unfavourable outcomes, whereas initial and long-term outcomes were defined as 6 months and 12 months after surgery, respectively. First the prevalence's of the trajectories were determined. Secondly, multivariable logistic regression analysis were done to examine if preoperative variables, postoperative complications and initial outcomes predicted long-term pain and function outcomes. Identical analyses were performed in the patients with favourable and unfavourable initial outcomes separately.
Results 972 THA and 892 TKA patients were included. With respect to pain, the prevalence of trajectories 1, 2, 3 and 4 were 79% (n=740), 5% (n=47), 7% (n=63) and 9% (n=88) in THA and 75% (n=606), 4% (n=33), 9% (n=71) and 12% (n=93) for TKA. With respect to function, the prevalence's were 74% (n=696), 6 (n=53), 7% (n=63) and 14% (n=131) for THA and 72% (n=624), 5% (n=49), 9% (n=78) and 14% (n=119) for TKA, respectively. Moreover, one-third of the patients with initial unfavourable pain or function outcomes attained favourable long-term outcomes. Overall, only the initial postoperative HOOS/KOOS pain and function predicted long-term outcomes (Table 1).
Conclusions One-third of patients with unfavourable initial outcomes improved further after the first six months after surgery. More research into the recovery early after surgery may help to identify patients at risk for unfavourable outcomes and modify their course by appropriate interventions.
Acknowledgements This study was funded by the Dutch Arthritis Foundation.
Disclosure of Interest None declared