Function and employment after total hip replacement in older adults: A narrative review

The burden of osteoarthritis (OA) has increased steadily due to an aging population, increasing life expectancy, obesity and lifestyle factors. Total hip replacement has become one of the most prevalent and successful operations globally and it is projected that demand will continue to grow as the incidence of OA continues to increase. Patients undergoing the operation expect much-improved function and pain relief but also increasingly need to return to work postoperatively, especially given the growing demand for the procedure and the encouragement of older people to continue working by most governments in the developed world. This review provides an overview of function and employment outcomes after hip arthroplasty. Despite the generally good success rate, some patients do not attain good functional outcomes and it is important that we develop ways to identify these patients preoperatively. We describe the effect of demographic, clinical and other factors on functional outcomes, as well as trajectories of physical function and pain recovery beyond the first few weeks after total hip replacement. Regarding employment outcomes, many people in work preoperatively are likely to resume to work after recovery; however, patients feel that they lack guidance from clinicians about returning to work post-operatively. Our review encompasses factors associated with return to work, timing of return to work, and potential temporary or permanent limitations that people might experience at work depending on type of employment.


Introduction
Osteoarthritis (OA) is the most common form of arthritis worldwide [1], causing joint pain, stiffness, and functional disability.The burden of disability caused by OA has been growing steadily over the last three decades [2] due to increasing life expectancy, rising prevalence of obesity and lifestyle factors.According to the Global Burden of Disease study, which described changes in morbidity and mortality levels between 1990 and 2019 for 369 conditions (diseases and injuries) across 204 countries, among people aged 50-74 years, OA went from the 24th most disabling condition in 1990 to the 18th in 2019.Indeed, the burden caused specifically by hip OA specifically increased during this period, mainly among people early in their seventh decade of life (60 to 64 years old) [3].
In the event of joint failure, and after trials of conservative therapies, total hip replacement (THR) may be offered.THR has been one of the most effective orthopaedic operations developed during the 20th century [4].Estimates of the lifetime risk of undergoing THR have been found to vary by country, with rates ranging from 8.7 % to 15.9 % in women and 6.3 % to 8.6 % in men [5,6].Since the earliest THRs, there have been improvements in implant-materials, surgical techniques and biomechanics which have produced very significant reductions in the failure of primary THRs (reducing the need for revision THR).Currently, the rate of revision of THR is estimated to be <1 % at one year follow-up.When failures occur, the three main causes are: aseptic loosening (42.3 %), pain (15.4 %); and dislocation/subluxation (14.7 %) [7].Longer term, a meta-analysis pooling data from 44 case series and Finnish and Australian orthopaedic registries over 25-years of follow-up reported a survival rate of 58 % (95 % CI 57 %-59 %) and if data were limited to those from case series, the survival rate was 77 % (95 % CI 76 %-79 %) [8].National and regional orthopaedic registries were introduced at the beginning of this century in response to concerns about long-term outcomes in relation to cemented arthroplasty survival [9].More recently, many registries have also introduced collection of other types of outcomes than revision, particularly patient-reported outcome measures (PROMs) including pain and function [10].This review focuses on employment and functional outcomes after THR in middle-aged people and does not consider short-term (perioperative) outcomes.These are important given the aging population, the growing requirement for THR among people aged <65 years [11] and the fact that many countries have been introducing financial incentives and/or legislation to encourage workers to work beyond traditional state pension age.

Methods
Our focus was on mid-to longer-term outcomes after THR and therefore we did not consider papers published in the immediate perioperative period.We identified studies published in the English Language from searches of the PubMed database from its inception until March 2022.The keywords used to run the search were combinations of "total hip replacement", "total hip arthroplasty", "return to work", "employment", "function", "pain", "trajectory" "mid-term" and "longterm".Included papers were systematic reviews; meta-analyses; casecontrol studies; observational studies (prospective and/or retrospective); or randomized control trials.Where available, we focused on the findings from systematic reviews or meta-analyses.Relevant studies published since the most recent systematic review were also included.

Functional outcomes
Functional improvement post-THR can be assessed objectively using performance-based tests, or subjectively using disease-specific patient reported outcome measures (PROMs).These are validated and selfadministered scoring tools developed to assess the impact of the operation from the patient's perspective by measuring changes in function and/or pain pre-and post-surgery.A meta-analysis exploring mid-term outcomes after THR found that whether measured using the total Harris Hip Score (HHS) [12], combined pain and physical function from the Western Ontario McMaster Universities Osteoarthritis Index (WOMAC) [13] and HHS, or other measures of health-related quality of life, patients were significantly better off in terms of pain and physical function, 3.6 to 7 years post-operatively than they were pre-operatively [14].Despite this, a literature review [15] estimated that the proportion of patients who continued to experience long-term pain post-THR (6 months to a median of 43 months follow-up) ranged between 7 % and 23 %.Consistent with this were findings from another study which assessed whether patients had achieved normal function postoperatively using the Outcome Measures in Rheumatology -Osteoarthritis Research Society International (OMERACT-OARSI) criteria [16].They studied 908 THR recipients and assessed change using WOMAC [17], and found that, although there was overall pain relief and function disability improvement, between 14 % -36 % of recipients experienced either little or no improvement at follow-up after one year [17].In a subsequent study following patients 5 years post-operatively, Judge et al. [18] reported that 2.3 % of participants had worse function by one year and 1.2 % did not show improvement by 5 years.Importantly, patients with worse functional pre-operative scores (Oxford Hip Score (OHS) < 5) showed the greatest improvement with a mean change score of OHS score of 28 points.Given that it appears a small proportion of people do not attain good outcomes after THR, it is important to be able to identify this group pre-operatively if possible.

Trajectories of function recovery
Many, but not all, THR recipients experience a steady recovery postoperatively followed by either a plateau or decline in function over time [19].However, there is some variation in the time taken to reach best function post-THR (Table 1).Two studies with 12-months of follow-up [20,21], reported that patients experienced a steep functional improvement within the first three months, at which point they reached the best functional scores.The first of these studies, by Lenguerrand et al. [20] investigated recovery of physical function separately from symptoms of pain.Participants were categorised according to their preoperative symptoms (more severe symptoms at baseline versus less severe symptoms at baseline) and their trajectories compared postsurgery.At 3 months post-surgery, both groups reported a similar level of pain relief and level of physical function, suggesting that arthroplasty had been more beneficial for people who were more limited pre-operatively.Furthermore, no further changes in pain or function were observed between 3 and 12 months post-operatively [20].The second study [21] identified three different trajectories of recovery using the OHS in a sample of 6030 THA recipients.In the first and most common trajectory (87.7 % of participants) overall function improved quickly by 3 months post-THR, not showing any further improvement afterwards.In a second trajectory, (7.7 % of participants) there was a more gradual recovery in the first 3 months followed by a slight decline in function around 12-months post-operatively.Factors associated with this trajectory were older age (>75 years) or being less fit (ASA score III-IV) at the time of operation, obesity, smoking, direct lateral surgical approach, hybrid fixation, and poorer pre-operative perceived health.A third trajectory (4.6 % of participants) had no functional improvement at 3 months however function improved at 1-year post-THR.Factors associated with this third trajectory were female sex, BMI ≥ 30 kg/m 2 and cigarette smoking, however operation through an anterior approach incision had a protective effect.
Two further studies [22,23], followed patients for 5 years postoperatively but captured functional limitation and degree of pain at different follow-up times than the two previous studies which had only 1-year follow-up data.The authors reported that recovery of physical function and pain relief was experienced by most patients within the first 6 months post-surgery [22,23] but that the best function was achieved at one year and sustained over 5 years [23].Analysis from the UK's National Joint Registry and including knee and hip patients identified two different trajectories over the first 5 years post-operatively.The first trajectory (70 % of patients) experienced marked improvement in pain/ function which was sustained over 5 years.The other trajectory saw sustained improvement but at a more modest level.People with most pre-operative pain and worst pre-operative function were more likely to experience the second trajectory [23].Additionally, it was possible to predict which trajectory an individual would follow based on their scores for pain and function at 6 months post-operatively.

Predictors of physical function
Given that some patients obtain better results from surgery than others, it is important to understand what factors might affect functional improvement or contribute towards function decline after THR.Predictors of function post-arthroplasty have been examined, with some studies focusing exclusively on pain or physical function domains as an outcome, whilst other studies have focused on overall function (Table 2).There have been three systematic reviews [24][25][26] which have examined predictors of functional outcomes following THR.However, although published around the same time, they differed in: inclusion criteria; the length of follow-up required in included studies; the proportion of missing data which was deemed acceptable; and whether or not the indication for THR was confined to primary OA only.

Demographic variables
Based on the results of the three systematic reviews, age may or may not predict outcomes after THR.In the review by Buirs et al. [26] 10 out of the 15 included studies found that undergoing THR at a younger age was associated with better functional outcomes post-operatively.

E. Zaballa et al.
Consistent with this, another review [25] reported that in 4 out of the 11 studies retrieved it was older THR recipients who had poorer functional outcomes or pain.However, a third review [24] found no conclusive evidence that age affected function post-arthroplasty.Regarding gender, one systematic review [26] found that in some studies male gender was a predictor of better function, whereas in other studies it was female gender.A second systematic review concluded that gender did not affect functional outcomes after THR [25].

Body mass index (BMI)
There is not a clear and consistent message from the literature about the effect of BMI on functional outcomes.Two systematic reviews [25,26] reported an association between higher BMI and worse postoperative functional or pain outcomes.Specifically, Vincent and colleagues [27] reported that obese patients (BMI ≥ 30 kg/m 2 ) at the time of THR had worse functional performance post-operatively, suggesting that although these patients benefitted from the surgery their physical function improvement might be poorer than that achieved by non-obese patients.However, a more recent meta-analysis revealed no differences in changes in functional scores when the authors compared non-obese (BMI < 25 kg/m 2 ) people with: severely obese (BMI > 35 kg/m 2 ) patients (0.04: 95 % CI 20.02-0.10),morbidly obese (BMI > 40 kg/m 2 ) patients (0.19 95 % CI, 20.08-0.46),and severely obese (BMI > 50 kg/ m 2 ) patients (20.12 95 % CI 20.57-0.33)[28].Likewise, in a cohort of 191 THR recipients, obesity was not related with hip function or pain at 10 years post-operatively [29].

Comorbidities
Evidence was consistent showing that preoperative comorbid conditions measured as number of comorbidities or specific medical conditions were associated with poorer outcomes post-operatively including overall function [25,26], pain and physical function [25].

Pre-operative function (physical function and pain)
Level of function deterioration before THR is a moderate to strong predictor of functional outcomes post-arthroplasty.Patients with more severe pain/function before surgery are more likely to achieve worse pain/function outcomes post-operatively [18,25], similarly people with better preoperative physical function are more likely to reach better physical functional performance after replacement [26].Additionally, studies reporting changes in pre-and post-surgical function scores have consistently suggested that patients with better pre-operative function, and less severe radiological OA benefit from less changes in pre-and post-operative function, however these patients obtained better functional scores post-arthroplasty [24,25].

General health and mental health
The review by Lungu and colleagues [25] found that poor THR outcomes in terms of function, pain or a combination of function and pain were associated with worse reported general health in 4 out of 4 studies.In one study, people with good mental health reached better physical function post-THR after follow-up over 72 months [26].However, Geeske et al. [30] collected mental health status in a cohort of middle-aged women in the 13 years prior to surgery to assess the impact of mental health on physical function and pain post-arthroplasty.The authors concluded that both women with poor and good mental health obtained similar benefit from the surgery up to 10 years post-THR.

Social support
Pooled data from two studies comprising 2022 lower limb arthroplasty recipients suggested that patients who had social support at oneyear post-THR benefitted in terms of their function measured as WOMAC score [31].

Other predictors
Other factors have been less well studied.Buirs et al. [26] concluded that there was weak evidence for a role of level of education and pre-  operative quadriceps strength and very limited evidence regarding any effect of alcohol intake, vitamin D insufficiency and allergies on THR functional outcomes.On the other hand, 3 studies from the review performed by Lungu et al. found that lower level of education was associated with worse pain and or function at 3 to 24 months postoperatively [25].One study by Vergara et al. [27] found that patients waiting longer than 6 months for their THR experienced less improvement in function 1 year post-operatively, when compared with the benefits experienced by those who had waited for <3 months.A systematic review [32] reported a relationship between frailty and poorer functional outcomes at 3 year follow-up post-THR (1 out of 2 studies).

Work outcomes
Approximately 20 % of THRs are performed on people below of 60 years of age [33], therefore for these younger patients it is important to have information about the effect of arthroplasty on employment.Workrelated outcomes are not routinely collected by clinicians, however some surgeons as part of the decision to perform the arthroplasty may ask patients about their occupation [34].Most patients who are working pre-operatively will intend to return to work (RTW) post-operatively [35] and will need advice about when they can expect to RTW.Patients may also need advice regarding specific demands of their job e.g., physically demanding activities and/or loading or driving.Unfortunately, guidelines for clinicians are lacking.Qualitative research with patients has suggested that when occupational advice is offered to patients this generally relates to time needed off work rather than strategies to facilitate RTW effectively [36] and sustain employment postoperatively.In the absence of current guidance this section describes relevant epidemiological evidence focused on employment outcomes after hip replacement.The main characteristics of the studies discussed hereafter are shown in Table 3.

Return to work and time off paid work
Hoorntje and colleagues [37] systematically reviewed studies that examined rates of RTW until October 2017 and time taken off sick by THR recipients.Pooled data from 23 THR studies showed that the mean rate of RTW for people working pre-operatively was 69 %, whereas when studies rated as high quality were considered the mean rate was 83 %.Even a subsequent study based on 408 employees from the Finnish Public sector reported that 94 % THR recipients returned to work at 3 months post-operatively [38].Furthermore, THR can be beneficial for some people who are unable to work pre-operatively but who are fit for work post-operatively [39][40][41].Most people go to back to the same type of occupation pre-operatively, however people in more physicallydemanding jobs might need to change to less strenuous jobs.In the study by Cowie et al. 13/169 (7.7 %) of those who worked after arthroplasty changed to a less physical job [42].Similarly 6 out of 62 THR recipients from another study moved to jobs with lower physical demands [41].Nunley et al. reported that 42/672 people did not RTW after their THR however in 12 out of these 42 people the operated hip was mainly the reason not to RTW [43].
In terms of time taken off work post-arthroplasty, there is wide variation reported with a mean duration ranging between 1 and 17 weeks (7 studies) [37].Pooled data from these studies estimated that the mean time off work is 8.9 weeks, or the mean is 11.4 weeks if only the results of high-quality studies are considered.

Predictors of resuming work
Four reviews have summarised the evidence about factors which determine RTW after THR [37,[44][45][46].Kuijer et al. highlighted the scarcity of literature published between 1998 and 2008 on predictors of RTW after lower limb arthroplasty.The authors found that restricting movement during the early post-operative period could limit RTW   whereas following guidelines to discharge THR patients had no effect on timing to resume to work [44].Systematic reviews published subsequently have reported that age was not found associated with RTW in some studies, but that some studies found that younger patients were more likely to RTW [37,45,46].Similarly, most studies have not found an association between sex and likelihood of RTW, as was reported in the review by Hoorntje and colleagues (10 out of the 12 studies) [37].However, THR recipients were less likely to RTW if they were on sick leave either at the time of surgery [37,45] or >30 days in the year prior to arthroplasty [38].There is some evidence that having attained higher levels of educational qualifications is associated with both a higher rate of RTW and less time taken to RTW [37] but this may relate to these individuals having availability of less physically demanding and more flexible types of work.Certainly, Laasik et al. [38] reported that people in non-physically demanding occupations (e.g.physician or teacher) were twice as likely to RTW compared with manual workers.It not surprising therefore that it takes longer to RTW for people with occupations that entail higher physical demands as compared with those in jobs with no physical demands [37].Other predictors that have been examined to a small extent in the literature are self-employed versus employed; availability of a disability pension; job satisfaction; and surgical approach however the existing evidence is scarce, and no conclusions can be made.

Workability post-operatively
Currently it is unclear whether people can sustain physicallydemanding jobs or whether certain activities should be avoided postoperatively.Return to work can be part of successful rehabilitation but, in some cases, a phased RTW with modification of tasks or hours may be required [43].Sankar et al. found that some occupational activities which are difficult or impossible to perform pre-operatively can become possible post-operatively but that some THR recipients continued to experience difficulties with lifting, carrying and moving objects, crouching and kneeling [47].In a more recent study [40], 7 % of the 411 THR patients who worked post-operatively reported that they had stopped working partly or mainly due to problems with their operated hip.Standing ≥4 h, kneeling/squatting and carrying/ lifting≥10 kg on an average day at work were strongly associated with having to stop work post-operatively.

Conclusion
In summary some studies have described different trajectories of physical function and pain recovery for up to 5 years post-arthroplasty, showing that most people achieve good results from surgery.Preoperative function and severity of radiological OA may be important determinants of the risk of a poor outcome.However, the evidence about other predictors (e.g.age, gender) is conflicting and no reliable predictive model is as yet available.Among people working pre-operatively, most RTW post-THR, however being on sick leave pre-operatively can be associated with reduced odds of return to work.Conversely, it seems that attainment of a higher level of education, which obviously is not unrelated to the nature of the job to which someone needs to return, is a predictor of better chances of RTW.

Clinical implications and need for research
Although THR is a very successful procedure some recipients fail to achieve physical function improvement and pain relief.More research that might enable prediction of those likely to have poor outcomes preoperatively is indicated.Now that PROMs are collected more regularly through arthroplasty registries [10] (i.e.The American Joint Replacement Registry, The Australian Orthopaedic Association National Joint Replacement Registry and Swedish National Registry), more data will be available to better understand the functional outcomes after THR in the mid-to long-term (beyond 10-year follow-up) as well as which factors contribute towards a decline in function.
Arthroplasty recipients who need to RTW post-operatively should receive occupational advice and support covering how long it might take to RTW and taking account of the nature of their job, and also the types of occupational activities involved.Future research should address whether more physically-demanding jobs may affect work sustainability, or function in the mid-to long-term post-operation.

Table 1
Characteristics of studies that described trajectories of functional outcomes after total hip replacement.
ASA; American Society of Anesthesiologists, BMI; body mass index, EQ-5D; EuroQol-5 dimension, FFbH; Hannover Functionality Status Questionnaire, NJR; The National Joint Registry from England, Wales, Northern Island and Isle of Wright, OA; osteoarthritis, OHS; Oxford Hip Score, THR; total hip replacement, VAS; Visual Analogue Scale, WOMAC; Western Ontario and McMaster Universities Osteoarthritis Index.

Table 2
Characteristics of systematic reviews that reported functional outcomes after total hip replacement.

Table 3
Characteristics of studies that evaluated return to work after total hip replacement.