Exercise therapy in severe osteoarthritis and total knee replacement: Patient Perspectives, Research Practice, and Trial Feasibility

Knee osteoarthritis (OA) is a degenerative joint disease characterised by progressive pain and loss of function. When symptoms become severe, patients may be offered knee replacement surgery. Before surgery, patients are often advised to undertake “prehabilitation” exercise therapy to manage symptoms and potentially improve surgical outcomes. However, challenges persist as patients often struggle to adhere to exercise programmes, their preferences are poorly understood, and there is uncertainty about how best to deliver exercise after surgery. Additionally, the scientific literature evaluating postoperative exercise therapy is methodologically inconsistent, often lacking prospective trial registration, which raises concerns about the reliability of the reported effects. Compounding this, the widespread implementation of exercise therapy as usual care makes it ethically and practically difficult to conduct trials comparing it to no intervention.

Aim: The aim of this PhD thesis was to explore these challenges, through three distinct studies investigating patient perceptions, research practice, and trial feasibility. Study I aimed to explore the patients’ perceptions and experiences of engaging in prehabilitation exercise therapy. Study II
aimed to explore the prevalence of prospective trial registration and how registration influences reported effects. Study III aimed to evaluate whether it is feasible to conduct a study where patients are discharged to usual-care exercise therapy or no exercise therapy, after knee replacement surgery.

Findings: In Study I, semi-structured interviews were conducted with 10 patients eligible for knee replacement surgery, before and after they participated in a three-month prehabilitation exercise therapy programme. Three overarching categories were developed through content analysis, reflecting key determinants for their experience and exercise adherence. 1) Simplicity was beneficial to getting used to exercising as it was easy to initiate and to learn but for those used to exercising a simplified exercise program appeared insufficient. 2) Motivation was predominantly driven by patients desire to achieve pain relief and to avoid surgery, but some succeeded in developing exercise habits using physical reminders. To some social support was important to succeed in exercising, for others it was not. 3) Having a comprehensive and supportive plan with the healthcare institution was important to all patients. In Study II, 94 randomised controlled trials comparing exercise therapy interventions were included in a systematic review. Of the 94 trials 14% were registered before the first participant was enrolled,
34% were registered after the first participant was enrolled, and 52% were not registered at all. The outcomes reported in the trials were at low risk of bias, for 23% of the prospectively registered trials, 3.2% of the retrospectively registered trials, and none in the un-registered trials. Prospectively registered trials tended to report smaller effect sizes than the retrospectively registered (difference in SMD 0.552 [95%CI:-0.03;1.13]) and un-registered trials (difference in SMD 0.359 [95%CI:-0.19;0.91]). In Study III, 25 persons were randomised between usual-care exercise therapy, and no exercise therapy after total knee replacement surgery to test the feasibility of conducting a large-scale trial. A large proportion (45.8%) of otherwise eligible patients refused to participate as they wanted the usual care exercise therapy. Clinicians and participants found the trial procedurally feasible and acceptable, and 3 of 4 predetermined feasibility criteria were met. Although the trial was procedurally feasible, the high refusal rate indicated that a full-scale trial would lack generalisability and was therefore not recommended.

Conclusion: In relation to the overarching aim of this thesis, Study I provided insights in patient perspectives, Study II examined current research practices, and Study III assessed the feasibility of conducting non-intervention trials - in perioperative exercise therapy for knee replacement surgery. The key conclusions were as follows: Study I demonstrated that the patients perceived key determinants differently meaning that some components facilitated their adherence to the exercise therapy, where for others it was a barrier. This indicates that greater emphasis could be placed on accommodating individual preferences to improve adherence to exercise therapy like the one tested in the trial. Further, having a comprehensive care-pathway that supports patients can provide them a sense-of-security which may enhance their engagement. Study II showed that prospective registration in the trials that make up the evidence base for exercise therapy after knee replacement is lacking and trials without prospective registration tended to overestimate their reported effects. This effectively means that the registration status of a trial
should be considered when results are being referenced in treatment guidelines and systematic reviews as they may skew the outcomes if not accounted for. Study III found that the basic premise of randomising between exercise and no exercise introduced significant risk of selection and attrition bias. This would severely limit the generalisability and thereby usefulness of a full-scale trial if conducted with the same design. Therefore, study III concluded that the randomised controlled trial design may not be suitable with a pragmatic approach like the one tested in its design.