Which disease-related factors influence patients’ and physicians’ willingness to consider joint replacement in hip and knee OA? Results of a questionnaire survey linked to claims data

Osteoarthritis (OA) is the most prevalent chronic joint disease in the world. Half of the world’s population aged 65 and older suffers from some form of OA [1]. It is one of the most common sources of pain and disability in the elderly [1–3], and international reports indicate that the prevalence of OA is on the rise [4]. Stepped-care strategies include recommendations for non-surgical treatment of hip and knee OA [5–8]. Failed conservative therapy, pain, loss of function and radiological changes are generally considered in the decision to perform total hip replacement (THR) or total knee replacement (TKR) [9, 10]. Nevertheless, the decision when to perform total joint replacement (TJR) is not well defined. As a result, there is great variation both in indications for surgery among orthopaedic surgeons [11, 12] and in the utilization of arthroplasty in general [13].

In addition to surgeon recommendations, patient preferences also play an important role in the decision process. In advanced OA, as with other disorders for which multiple treatments are available, shared decision-making helps patients and physicians to choose the treatment that best fits a patient’s preference [14]. Several studies investigating the interrelationship between surgeons’ recommendations and patients’ willingness to undergo TJR report a wide gap between the views of both partners in this decision-making process [15–18]. Patients’ perceptions of the appropriateness of and their desire for TJR are of substantial importance in the discussion of treatment alternatives [12] and have to be appropriately included in shared decision-making [10]. Although several international studies have investigated factors potentially influencing patients’ decisions to undergo surgery [12, 19–21], it is still unclear why patients’ perspectives often differ from surgeons’ recommendations. The aim of our study, therefore, is to investigate the influence of disease-related cofactors on the willingness of patients and the advice of surgeons to consider TJR in a large population of insured individuals with hip or knee OA.

The characteristics of the study population are summarized in Fig. 1 and Table 1.

A total of 3775 patients completed and returned the questionnaires, and 478 patients with symptomatic hip OA and 932 with symptomatic knee OA were analyzed. Mean age was 64 years (95% confidence interval (CI) 63–65) in the group with hip OA and 65 years (95% CI 64–65) in the group with knee OA.

Mean symptom duration was 14 years (95% CI 12–15) for hip OA and 15 years (14–16) for knee OA. The total WOMAC score was 42 (95% CI 39–44) in hip OA and 41 (95% CI 39–42) in knee OA.

An orthopaedic specialist) was reported as the mainly treating physician by 63% (hip OA) and 71% (knee OA) of the patients. Physical therapy was prescribed in 46% (hip OA) and 49% (knee OA), whereas any kind of analgesic was prescribed in 58 and 66%, respectively.

Individuals who returned the completed questionnaire were slightly older (67.2 vs. 65.5 years), and more of them were seeing an orthopaedic specialist (54% vs. 44%) or had a prescription of NSAIDs (48% vs. 41%) than those who did not answer the questionnaire. The percentage of women, Elixhauser comorbidity index (median 2.5 vs. 2.7), and prescriptions of opioids (13.8% vs. 13.5%) did not differ clinically meaningful between responders and non-responders.

Overall, 17% of the patients with hip OA were willing to undergo TJR while 83% stated they were not. The distribution was similar for knee OA (Table 2). Most of the patients who had discussed TJR had done so with an orthopaedic specialist (83% hip OA, 86% knee OA). In the group willing to undergo TJR, patients had higher WOMAC scores, more patients reported deterioration of their symptoms over the last 2 years before the survey or an impact on personal life, and there were more patients with previous arthroplasty of other joints and treatment by an orthopaedist compared with the group of patients who were unwilling (Table 3). We performed logistic regression adjusted for age, sex and symptom duration to determine factors associated with the willingness to undergo TJR (Table 4). For both hip and knee OA, it was associated with reduced joint function, as reflected by a high WOMAC score (OR hip 1.47, 95% CI 1.24–1.75; OR knee 1.42, 95% CI 1.26–1.61), a greater impact on personal life (OR hip 10.3, 95% CI 1.04–103; OR knee 6.81, 95% CI 2.23–20.79), worsening of symptoms over the last 2 years (OR hip 2.73, 95% CI 1.14–6.55; OR knee 3.66, 95% CI 2.05–6.53) and already having one artificial joint (OR hip 6.01, 95% CI 3–12.06; OR knee 3.69, 95% CI 2.2–6.18). For knee OA, decreased well-being was associated with a greater willingness to undergo TJR while for hip OA alone, only treatment by an orthopaedic specialist had an effect on the patient’s wish to undergo TJR.

Regarding the interaction with their mainly treating physician, 44% of hip OA and 45% of knee OA patients had already discussed a possible TJR. However, the majority of patients who reported a previous discussion with their physician (68% hip OA, 72% in knee OA) were not willing to undergo TJR (Table 2). The unadjusted data (Table 5) show certain differences between hip OA and knee OA patient groups regarding the effect of prior discussion of TJR.

Among patients who had already discussed TJR with their physician, we searched for factors associated with willing to undergo TJR. In this subgroup, both knee and hip OA patients who were willing to under TJR had poorer joint function than those who were not willing (WOMAC score of 52 vs. 44 for hip OA and 50 vs. 39 for knee).

We fitted logistic regression models to identify factors associated with having discussed TJR among all hip and knee OA patients. The models were adjusted for age, sex and symptom duration (Table 6). In both hip and knee OA patients, significant associations with physician discussion were only found for impact on personal life and already having received another TJR. Older age, male sex, longer symptom duration, deterioration of symptoms over the last 2 years before the survey, a higher WOMAC score and reduced well-being were associated with having discussed TJR in knee OA, but not in hip OA.

No significant association was found between the number of prescribed medications as an indicator of comorbidity or prescription of physical therapy and discussion of TJR.

Hip and knee replacement is among the medical and surgical interventions with the highest cost effectiveness and capacity to improve patients’ quality of life [27, 28]. Nevertheless, many studies have shown that even patients with advanced OA who are good candidates for TJR are often unwilling to consider surgery [20, 29]. This is confirmed by our findings in a large unselected cohort of insured individuals with the diagnosis of OA. Patients with both hip and knee OA showed little willingness to consider arthroplasty (17 and 14%, respectively). Although not all participants of our study had end-stage OA, a significant disease burden can be assumed since many patients reported a long history of symptoms, worsening over the last 2 years and a high impact on their personal situation. This apparent contradiction between burden and willingness to undergo TJR is what Hudak et al. termed “unmet need” [30]. They found that just a small proportion of patients who were identified as perfect candidates for surgery were actually willing to undergo TJR.

In view of this situation, the question arises: what may induce patients to contemplate surgery as a suitable treatment option? According to Dieppe et al. [31] the decision about TJR is a “judgment call that has to be made by the physician and patient working together, and which has to take account of a large range of complex psychological, social and other issues, in addition to pain, disability and X-ray changes”. They propose a capacity-to-benefit concept consists of interacting disease-related factors (state of the joint, impact on the individual) and treatment-related factors (risk-benefit of TJR and alternative treatment options). This inspired us to analyze the relevance of these possible influencing factors for patients contemplating the need for TJR following many years of conservative treatment of OA.

In both groups - hip and knee OA - we found that patients with poorer joint function (high WOMAC score), deterioration of symptoms over the last 2 years before the survey and impact on personal life were more readily willing to consider TJR. This is in accordance with previous studies describing disease burden as a major driver for the willingness to undergo surgery [12, 29, 32, 33]. A major factor with an impact on the decision about surgery, for both hip and knee OA patients in our study, is the experience with previous TJR of another joint. Although this result reflects common clinical experience in that patients often proceed with additional TJR of other affected joints more readily once they have overcome the barrier of their first decision, it has not been systematically addressed in studies before. Age and sex had not influence the willingness of patients to consider surgery in our study. This is in contrast to earlier investigations [29, 34, 35] reporting associations between demographic factors and the decision for TJR. Another factor not appearing to have had an impact on patients’ attitudes towards surgery in our study is conservative therapy. This result is difficult to interpret, as the proportion of patients with daily intake of analgesics and even opioids was rather low and only about half of the patients had a prescription of physical therapy.

An interesting finding of our study is that, although most patients were treated by an orthopaedic specialist, their counselling led to heterogeneous effects. While patients’ decisions to consider TJR were associated with orthopaedic treatment in hip OA only, significantly more orthopaedists had discussed surgery with knee OA patients than with hip OA patients. As we had no direct access to physicians’ perspectives or to any information objectively documenting the discussions between physicians and their patients, it is impossible to interpret this finding. Nevertheless, the role of physicians in the decision process is important and several studies have highlighted the expectations and attitudes of surgeons and other physicians when proposing TJR to their patients [10, 11, 16, 17, 32, 36, 37]. The most important predictor for patients’ decisions in the study of Hawker et al. [20] was having previously discussed the surgical procedure with a physician.

Our interest was to assess the association of orthopaedist counselling with disease-related cofactors. According to the literature, physicians’ recommendations of TJR should be based on failed conservative therapy, pain, loss of function and radiological changes [9, 10]. As we had no data on radiographic OA grading in our patient population, we focused on prior and current treatment and subjectively perceived disease burden. Neither in patients with hip OA nor in patients with knee OA could we find an association between physicians’ discussion of TJR and prescription of physical therapy or analgesics. This observation confirms the recently described lack of adherence to conservative treatment recommendations in our total cohort of patients with OA in a large German statutory health insurance fund [38, 39]. Regarding the recommendation to base an indication for surgery on pain and functional impairment, we found an association between physicians’ discussion and symptoms (in terms of both duration and deterioration) as well as poorer WOMAC scores and impact on personal life at least in patients with knee OA. In patients with hip OA, however, we only found an association with an impact on patients’ personal life. These results suggest that adherence to treatment guidelines in hip OA is even worse than in knee OA and underline the need for implementing not only already established guidelines for TKR in Germany [10] but also for indication to THR. We also found a high impact of a previous history of TJR at another site on physicians’ discussions of arthroplasty in hip as well as knee OA patients. A positive experience with earlier TJR not only increases patients’ willingness to consider further surgery but also the readiness of physicians to mention this treatment alternative.

Physician support is crucial in the process of shared decision-making. The steps which patients are going through in deciding about arthroplasty include mental preparation, addressing concerns around pain, quality of life and social isolation, weighing risks and benefits, facing fear and uncertainty and the question when “the time is right” for surgery [19]. The consultation is the central part in the preparation of joint replacement surgery. It is where risks and benefits are discussed, trust established, and decisions made [14, 40]. A clear positioning of consulting physicians, whether there is a medical indication, would be helpful in this process. However, published studies suggest that general practitioners, orthopaedists and surgeons differ widely in their attitudes regarding the appropriateness of TJR [32, 36, 37]. Surgeons and referring physicians differ in their opinion about the benefits and timing of surgery [36]. A recent study investigating the influence of patient characteristics on physicians’ and surgeons’ decisions to refer patients to or to perform TKR confirms this variability and also reveals a significant attribution to individual unreliability over time [37]. This inconsistency may enlarge an already existing gap between patients’ and physicians’ perception of the need for surgery and patients’ preferences. This gap can make patients feel anxious and discontent [40], thus contributing to their unwillingness to undergo surgery.

Our results, which are based on a combination of claims data and self-reported patient data, confirm previous reports that the percentage of patients who are willing to undergo TJR is lower than the proportion of patients recommended to have surgery by their physicians. While previous TJR seems to enhance patient’s and surgeons’ willingness, no consistent picture emerges regarding the influence of other cofactors. Especially for patients with hip OA, our results do not reflect internationally established indication criteria for replacement surgery.

We thank the BARMER health insurance, in particular Dr. Ursula Marschall, Joachim Saam and Janine Schwarz, for their support and for the opportunity to use the data for these analyses.

Conception and design: AP, JG, KPG, TL, IR, JS, AZ and JC. Acquisition of data: AP, JG, KPG, TL, AZ and JC. Analysis and interpretation of data: AP, KPG, TL and JC. Drafting the manuscript: AP, KPG, JC. Critical revisions for important intellectual content: all authors. Final approval of the version to be submitted: all authors.

The study was sponsored by the Federal Ministry of Education and Research, Germany (01EC1405). The funding had no role in study design, data collection, analysis, interpretation, writing, or submission of the manuscript.

Open Access Funding was covered by the Publication Fund of TU Dresden (institutional code: IN-1502334).

The datasets generated during and analyzed during the current study are available from the corresponding author on reasonable request.

Ethical approval was obtained from the ethics committee of the Charité University Medicine in March of 2015 (EA1/051/15). This study was conducted in agreement with the Declaration of Helsinki.

Not applicable.

The authors declare no conflicts of interest relevant to this work.

Supplementary information accompanies this paper at 10.1186/s12891-020-03368-1.