Keywords

FormalPara Take-Home Messages

  • On average one in three patients do not return to work after knee arthroplasty

  • Patients return to work around 12 weeks post-surgery, although large differences exist between patients and full return to work may take more than 6 months.

  • The cause for not returning to work is multifactorial, but known prognostic factors are preoperative sick leave of more than 2 weeks, female sex, high body mass index (BMI), patient-reported work-relatedness of knee symptoms, and physically demanding jobs. Age and Knee Injury and Osteoarthritis Outcome Scores (KOOS) were not associated with no return to work.

  • At present, no studies are available that evaluated the effect of exercise-based rehabilitation, active referral to an occupational physician or therapist, or other forms of multidisciplinary care for knee arthroplasty on return to work.

  • Promising interventions for return to work are better expectation management by setting preoperative patient-centered realistic work-related activity goals, preoperative referral to an occupational physician or therapist to actively address prognostic factors hindering return to work, and the use of personalized e/mHealth including activity trackers to support KA patients on a daily basis in return to work.

11.1 Introduction

If knee arthroplasty (KA) surgery has been successful and patients’ pain is reduced and mobility returns, it becomes vital for both patients’ health and society that patients return to normal daily life activities. For many patients, returning to work will require them to accept the fact that their “new” KA knee will not function as their healthy knee. The largest increase in primary KA demands is namely not among the classic knee arthroplasty population of patients aged 70 years and older but among patients of working age [1]. For instance, the number of patients aged 45–65 years who undergo total knee arthroplasty (TKA) has tripled (Swedish knee arthroplasty register) in the past 30 years. Germany – one of the leading countries in the prevalence of knee arthroplasty – foresees the highest increase in patients aged 50–65 years until 2050, and in a similar study using the same database, even among patients aged 40–49 years until 2040 [2, 3]. In several countries, the current proportion of knee arthroplasty patients under 65 years is already substantial at 30–40%. It is expected in 2030 that the USA will be the first country where the majority of these patients will be younger than 65 years, followed by the UK in 2035 [4, 5]. In addition, it was found that the combined loss of productivity plus medical costs for conservatively treated symptomatic knee osteoarthritis for those in paid employment in the Netherlands amounts to €871 per patient per month, with loss of productivity accounting for 83% and medical costs for 17% [6].

Previously little was known about return to work in either employed or self-employed patients undergoing TKA. Because the numbers of working patients undergoing TKA are increasing, it is important to find out which factors will help or hinder patients in returning to work following surgery in a swift and also effective manner. What is the impact of surrounding medical as well as social support, the type of work a patient performs, and the general health of the patient? How do these factors interact with one another?

There is sparsity although increasing data about the variety of outcomes regarding this working population. It seems that patients have varying expectations about returning to work after TKA surgery. Remarkably, it was found that only 72% of the patients expected that TKA would improve their ability to work prior to surgery. Six months after TKA, this was even further reduced to 28%. With respect to knee-demanding activities, only 34% expected severe difficulty in kneeling, 30% in crouching, and 17% in clambering at 6 months after TKA [7].

Rehabilitation with Goal Attainment Scaling (GAS) could be a useful tool to manage expectations of functional postoperative outcome. When preoperative goals are set as studied in unicompartmental knee arthroplasty (UKA) patients for postoperative daily life activity, work, and leisure time, it was found that 100% met these goals, compared to 82% of TKA patients [8]. When realistic goals are set and expectations are adjusted, this might improve perceived outcome.

More detailed knowledge about the impact of KA on ability to return to work can help in making better informed decisions about whether KA is the appropriate treatment for the patient’s problem. Furthermore, no randomized or appropriately adjusted comparison has yet been made to find out whether UKA patients return to work sooner or perform better than patients with TKA. UKA surgery is less invasive, and patients seem to function better and be more active and are even able to return to sport sooner despite reported higher revision rates, but the role of bias is unclear [9, 10].

There is increasing interest in the development of health care toward more outcome-oriented care in a broad sense, in which the choice of treatment looks at what best fits the specific situation of the patient instead of population-based objective group outcomes.

Outcome-oriented care can be defined as the outcome that really matters for the health and well-being of a specific patient. The goal is to focus care better on what matters to the patient which in turn can lead to better decision-making choices and more timely work-directed care. This is of importance given that the first prospective cohort study among working age TKA patients showed that even after 1 year, only 71% of workers had fully returned to work [11].

11.2 Patient-Reported Outcome Measures in Working KA Patients

To study patients’ physical difficulty experienced in work before or following KA, the Work, Osteoarthritis or joint-Replacement Questionnaire (WORQ) was developed [6]. The WORQ (range 0–100, with a minimal clinically important difference of 13) assesses the experienced difficulty for 13 work-related activities, like kneeling, working with the hands below knee height, and walking on rough terrain. This 13-item questionnaire was tested for internal consistency by factor analysis, internal reliability (Cronbach’s α), and construct validity. A test-retest reproducibility was performed for analyzing standard error of measurement (SEM agreement), reliability (ICC), and smallest detectable change (SDC) in individuals and groups. Lastly, responsiveness (standardized response means [SRM]), floor and ceiling effects, and interpretability (minimal important change [MIC]) were analyzed. It was shown that the WORQ is a reliable, valid, and responsive questionnaire following TKA that can be used to evaluate the impact of knee complaints on patients’ ability to work [12].

Other patient-reported outcome measures (PROMs) commonly applied to TKA patients are the KOOS, Oxford, and the new Knee Society Scoring System questionnaires. These mainly assess home-life activities and do not look at specific activities that are necessary to return to the work. Gagnier et al. performed a review on PROMs for TKA to critically appraise, compare, and summarize their psychometric properties using accepted methods. Although not all psychometric properties were studied, they concluded that the WORQ had the highest overall ratings and thus could be a useful PROM for evaluating patients undergoing TKA [13, 14].

In an early cross-sectional survey, it was found that approximately one-third of TKA patients worked within 2 years prior to surgery [15]. When looking at these working patients, activities that most improved were operating foot pedals, operating vehicles, and standing and walking on level terrain. Activities that least improved were kneeling, crouching, and clambering (Fig. 11.1).

Fig. 11.1
figure 1

WORQ score improvements in % performing work-related activities following knee arthroplasty between T0 (before the knee problems arose) and T2 (at 2 years after TKA) [15]

Full size image

Fifty patients scored 5 or less on the Work Ability Index (WAI), an index from 0 to 10 in which a patient can report how well they are able to perform their work with a TKA. TKA significantly, but unequally, reduces difficulties in carrying out knee-burdening work activities [15]. When UKA patients (median 60 years, 51% male) were compared to TKA patients (median 60 years, 49% male) (n.s.), it was found that WORQ scores improved similarly in both groups. The WAI score was also comparable between the groups. Dissatisfaction with work ability was comparable (UKA 15% versus TKA 18%) (n.s.). TKA and UKA patients have similar WORQ, WAI, and satisfaction scores [16].

11.3 Return to Work Timing Following TKA and UKA

Return to work between TKA and UKA patients has been reported to be around 70–80% (Table 11.1 [16]). In the same multi-center retrospective cohort study as mentioned above, the time period between stopping work and returning to work was assessed [10]. UKA patients (n = 157, median 60 years, 51% male) were compared to TKA patients (n = 167, median 60 years, 49% male) (n.s.). Of the 157 UKA patients, 115 (73%) returned to work within 2 years compared to 121 (72%) of TKA patients (n.s.). More UKA patients returned to work within 3 months (73% versus 48%) (p < 0.01) (Fig. 11.2) [16]. UKA patients return to work significantly sooner after surgery than TKA patients, which might improve their quality of life and allow them to re-participate more actively in society at an earlier time.

Table 11.1 Summary of timing of return to work reported in a KSSTA study in 2020 [16]
Full size table
Fig. 11.2
figure 2

Time when patients resumed work after unicompartmental (UKA, N = 117) and total knee arthroplasty (TKA, N = 117) in absolute numbers, with distribution line [16]

Full size image

11.4 Prognostic Factors for Not Returning to Work

In the Netherlands, it has been studied what patient characteristics are associated with no return to work (RTW) [38]. Backward stepwise logistic regression analyses were performed to predict no RTW. One hundred and sixty-seven patients met the inclusion criteria, and 46 did not RTW. Preoperative sick leave of more than 2 weeks (OR, 12.5; 90% CI 5.0–31.5) was most strongly associated with no RTW. Other associations found were female sex (OR, 3.2; 90% CI 1.3–8.2), BMI 30 (OR, 2.8; 90% CI 1.1–7.1), patient-reported work-relatedness of knee symptoms (OR, 5.3; 90% CI 2.0–14.1), and physically knee-demanding job (OR, 3.3; 90% CI 1.2–8.9). Age and KOOS scores were not associated with no RTW (Fig. 11.3). Especially obese female workers, with a preoperative sick leave duration >2 weeks who performed knee-demanding work and indicated that their knee symptoms were work related, had a high chance for no RTW after TKA. These results stress the importance of a timelier referral for work-directed care of patients at risk for no RTW after TKA.

Fig. 11.3
figure 3

Prognostic factors for not returning to work – stepwise logistic regression analyses [38]

Full size image

In a qualitative study performed in 50 TKA patients by Bardgett et al., three key factors were identified that influenced RTW from the patients’ perspective [39]. These patients reported an improved physical and psychological performance at work after surgery in comparison to preoperative functioning. The three factors reported were that (1) patients did not receive specific advice to facilitate their RTW following surgery, (2) patients perceived that the current provision of information for joint replacement patients is focused on the needs of elderly patients and reported that more clarity and consistency are required regarding RTW advice, and (3) these patients reported a lack of support and adaptation in the workplace and described a negative influence on their experience of RTW although this was not reflected in increased duration of sickness absence [39].

Furthermore, patients who had a slower return to work often reported that comorbidities, especially musculoskeletal like low back pain or OA affecting other joints, prevented their RTW even when the surgical outcome was positive [40]. However, the most recent review on prognostic factors for return to work concluded that based on 14 studies and 3073 patients, the most important prognostic factors associated with a slower or no RTW were a more physically demanding job and preoperative absence from work [41].

11.5 Interventions Aimed at Improving Return to Work After KA

Remarkably, little to no evidence is available for effective return-to-work interventions for KA patients. Although the provision of exercise-based rehabilitation after KA is almost universal, a systematic literature review performed in Ovid Medline and EMBASE concluded that “no studies were found evaluating the effect of rehabilitation programmes for knee arthroplasty on return to work” [42]. To come to this conclusion, a detailed search was performed with the support of a clinical librarian specialized in the outcome work participation, and despite that, 3788 studies were independently assessed by two reviewers. If the search was broadened and also included integrated multidisciplinary care, like active referral to an occupational physician or an occupational therapist or including e/mHealth interventions, again no studies were found for KA patients and RTW [43]. Therefore, to develop an occupational advice intervention to support early recovery to usual activities including work that is tailored to the requirements of KA patients, Baker and colleagues performed an intervention mapping approach, including 110 stakeholder interviews and a survey of 152 practices [44]. The intervention included information resources, a personalized return-to-work plan, and coordination from the health-care team. To support delivery, a range of tools (e.g., occupational checklists, patient workbooks, and employer information), roles (e.g., return-to-work coordinator), and training resources were created. The intervention was assessed in 26 patients and staff and showed high rates of adherence to the defined performance objectives. The overall results demonstrated that the occupational advice intervention developed for KA patients is deliverable. However, the intervention warrants a randomized controlled trial to assess its clinical effectiveness and cost-effectiveness to improve rates and timing of return to work. Two other promising return-to-work interventions for KA patients and using limited health-care resources are Goal Attainment Scaling (GAS) and the use of a personalized e-Health application, iRecover (in Dutch: ikHerstel) [45, 46]. GAS personalizes exercise-based rehabilitation by setting patient-specific, activity-oriented rehabilitation goals in close collaboration between the patient and the physical therapist, thereby setting realistic patient expectations and securing close monitoring of these goals during the rehabilitation period. A randomized controlled trial (RCT) among 120 working-age KA patients showed that GAS resulted in higher patient satisfaction with work activities compared to care as usual in the control group: an increase of 11 points on a scale from 0 to 100 with a 98% confidence interval of 2–19 points [45]. For the iRecover application, multidisciplinary consensus on recommendations regarding the resumption of 27 activities of daily life, including work, has been reached among a multidisciplinary expert panel of six orthopedic surgeons, three physical therapists, five occupational physicians, and one physician assistant for fast, average, and slow recovery [46]. These consensus recommendations are integrated into the algorithm of the iRecover application (Fig. 11.4) [47]. In combination with the use of an activity tracker and GAS for work-related activities, this intervention is currently evaluated in the so-called Active RCT among 368 patients (https://www.trialregister.nl/trial/8525).

Fig. 11.4
figure 4

An example of a dashboard of the iRecover app, providing knee arthroplasty patients tailored guidance on resumption of activities of daily life, including work. Guidance on the resumption of activities and the recovery status of these activities (upper two panels) can be provided, while wearable devices can be used to provide patients with feedback on their physical activities, helping them to work toward self-chosen (work-directed) goals [47]

Full size image

11.6 Discussion

11.6.1 Cost-Effectiveness of KA From a Personal and Societal Perspective

Despite good results with respect to return to work for most of the KA patients, a large proportion of these patients do not return to work. Also within this population, when patients return to work, not all aspects of functional recovery improve, but patients’ overall satisfaction can still be high. Information regarding time to return to work including what work-related activities will improve most after KA is of primary importance for patients and care providers, as meeting preoperative expectations is key for satisfaction [48, 49]. This is especially true for patients approaching their retirement age. As, on average, patients return to work 3 months after KA, it is still unclear if the cost of the 3-month sick leave in addition to the cost of KA weighs up to the potential improvement of productivity from a societal economical perspective. This may be relatively easy in patients who are already on full sick leave and unable to perform their job because of OA-related knee complaints as they can only improve. Such a simple cost-benefit analysis is less straightforward in patients who are still able to perform their work, but in a less productive manner. Will they improve sufficiently from surgery? As shown earlier, activities like kneeling and crouching only improved marginally. Therefore, KA will probably improve general quality of life but not necessary productivity for most plumbers, gardeners, and builders getting close to pension age as the “return of investment” time is too short. On the other hand, if a patient’s work mainly consists of driving a vehicle, such as is the case in taxi drivers or lorry drivers, it might be advantageous to perform arthroplasty surgery earlier on as these activities do seem to improve. It needs to be further investigated what timings are advantageous for specific working groups and whether active referral to an occupational physician or therapist might be advantageous for return to work, as well as for professions in order to find work-related solutions for activities that improve less after KA.

11.6.2 UKA or TKA?

It seems that UKA patients return to work sooner than TKA patients. Despite the fact that prosthetic survival of a UKA is shorter than that of TKA, a well-informed decision can be made in the case of anteromedial osteoarthritis. If it is paramount for a patient to return to work as soon as possible, UKA could be the prosthesis of choice. This can be the case for patients who are self-employed. However, if a patient finds it most important to receive an arthroplasty which will last longer, a TKA can be chosen despite the longer return-to-work interval. Future research will focus on translating research data into optimal decision-making in the workplace. It will be interesting to see if patients will be more satisfied if they are better informed on what to expect from return to work after TKA or UKA surgery. With better insight into what a specific patient needs to be able to return to work, better coaching on the choice and timing of treatment can be provided. Specific physiotherapy could be focused to prioritize the performance of work activities to see if patients can return sooner. Interventions can be tested for effectiveness by assessing WORQ scales prior to surgery as well as post-surgery. Future research will need to focus not only on outcome but also on cost-effectiveness. As the combined loss of productivity plus medical costs for conservatively treated symptomatic knee osteoarthritis for those in paid employment in the Netherlands amounts to €871 per patient per month (with loss of productivity accounting for 83% and medical costs for 17% [6]), better assessment of cost-benefit and cost-effectiveness will become possible. One might expect that arthroplasty surgery may reduce these costs. If arthroplasty surgery would reduce the loss of productivity to zero at the moment of return to work at 3 months, and the total cost of arthroplasty surgery is on average around €10.000 [50], surgery would accrue positive cost-benefit outcome if absence from work could be shortened by 12 months (=10.000/871) or more. However, these rough estimates ignore the fact that three out of ten patients do not return to work and that surgery will produce adverse outcomes in others. To make an accurate assessment of when is the best time to perform surgery for specific patients, new comparative prospective studies should be performed. Challenges for future research are the difference between the intervention and the control arm, not only with respect to the choice and timing of surgery but also with respect to other covariates such as management of expectations, quantification of medical and societal costs (such as loss of productivity), and adequate as well as feasible follow-up. The results of one study demonstrated that the total economic cost to society for treatment of severe knee osteoarthritis in a relatively young working person is markedly lower with TKA than it is with non-operative treatment [51]. As furthermore stated by the authors of this paper:

The results of this model illustrate the need to account for the implications of treatment choices, not only at the individual patient level, but also for society at large. When deciding among available treatment options, patients, p hysicians, payers, and policymakers must consider individual treatment cost and effectiveness but also should account for future potential earnings generated when a treatment may restore a patient’s ability to contribute to society [51].

11.7 Conclusion

Knee arthroplasty is becoming more and more important to keep patients active as members of the workforce. Therefore, not only in clinical practice and in research but also in guideline development, this important outcome should be more often addressed, especially regarding effective multidisciplinary care.