Outcomes of total knee replacement: a qualitative study (2023)

Abstract

Total knee replacement (TKR) has an established place in the treatment of knee osteoarthritis and is considered to be an effective intervention [1–9]. TKRs are reported to relieve pain and improve mobility, the best published results reporting a ‘good’ or ‘excellent’ outcome in approximately 90% of patients [9]. Approximately 30 000 TKRs are carried out in England and Wales each year [10], and it is estimated that there is likely to be a 30% increase in the need for all total joint replacements (TJRs) in the next three decades, due to changing demography [11]. The effectiveness of TKR has been dominated by quantitative research and the perspectives of patients have been relatively neglected.

Although TKR is considered an effective intervention, problems have been highlighted in the literature [1]. Systematic reviews of the studies that examined the effectiveness of TKR reported that most of the studies were observational, and stated that the time to prosthesis failure or revision surgery were the main or only outcome measures, rather than patient-centred outcomes [1, 12]. The wide variations in the types of prosthesis and in the outcome measures used also made it difficult to come to a clear conclusion about the effectiveness of TKR [1]. In addition, there was often an overemphasis on physician-defined pain relief and measures of technical success [13], which may not necessarily accord with symptomatic improvement.

The use of physician-based rating scales assumes that physicians and patients concur with regard to the degree of success of TKR. However, it has been shown that there is a marked disparity between the patient's and the physician's evaluation of outcome [14], especially when the patient is not completely satisfied with the result. For example, physicians and patients may have a different definition of what constitutes an excellent outcome. Patients also may not state their problems clearly for fear of disappointing the physician, and even if they state their problems clearly, the physician still may not comprehend the true nature of the pain and the patient's level of dissatisfaction [15]. Furthermore, these scoring systems summarize the clinical examination, symptoms and functional limitations into one score and such summary scores obscure opposing trends between dimensions (for example, a patient may still experience pain but have improved function, range of movement, muscle strength and stability). Thus, data gained by summary or global ratings of outcome may miss important aspects of patients’ views or fail to capture individual meanings.

To overcome the limitations of physician-based scoring systems, a number of health status questionnaires have been developed to be completed by patients undergoing TKR [5, 16–18]. Studies have reported that the Western Ontario and MacMaster Universities Osteoarthritis Index (WOMAC) and the short form 36-item (SF-36) are the instruments of choice for evaluating the outcome of knee replacement surgery in osteoarthritis [19, 20]. In the domains of function and pain, studies have shown that patients consistently showed significant benefit after knee replacement surgery [16, 17, 21]. A recent study reported that WOMAC measures improved significantly after 1 yr for osteoarthritis of the knee: there was a reduction of pain of 53%, reduction of stiffness of 43% and improvement in physical function of 43% [22]. The SF-36 also showed significant improvement for pain (175%), physical function (197%), physical role functioning (275%), vitality (125%) and social functioning (119%) [22].

Even with the use of patient-based measures, the assessment of outcome may not be fully explained as the context and the experiences of the intervention are not addressed. In general, structured questionnaires reflect the issues of major importance to physicians, combined with more general aspects of health. All the physician-based and structured measures suggest that TKR is very effective operation for most, but state that some do not benefit. However, what is missing from these assessments is an understanding of the patient's perspective of undergoing and recovering from a TKR. This study was undertaken to explore the experience of outcome from a TKR of a range of patients, and to investigate whether existing measures capture such experiences sensitively and effectively.

Methods

Local ethics committee approval was obtained from the South West Local Research Ethics Committee. In-depth interviews were conducted with 25 patients on the waiting list for a primary TKR, of whom 10 were interviewed again 6 months after TKR. Patients were sampled from three orthopaedic surgeons’ waiting lists to include individuals with a range of demographic characteristics, including older and younger ages and both sexes. Forty patients were approached and 25 gave written consent to be interviewed. Ten did not reply to the invitation and five refused to participate. One of the authors (G.M.W.) conducted semistructured interviews using a checklist of topics to ensure that the same basic issues were covered with each participant [23]. At the pre-operation interview, patients’ views on the onset, aetiology and experience of joint problems, referral and listing for TKR, and expectations of the TKR were explored. At the postoperation interviews, patients discussed the experience of the hospital stay, operation and recovery, and their perception of the TKR outcome.

Analysis was conducted according to the method of constant comparison [24]. Thus, the processes of sampling, data collection and analysis were continuous and iterative. Initially, a small number of interviews were carried out and coded. Analyses of subsequent interviews were then compared with analysis of the first interviews so that similarities and differences between cases could be examined. This in turn shaped further sampling, data collection and analysis. Interviews were recorded on audiotape and fully transcribed. These transcripts, together with field notes and reflective comments made at the time of the interviews, formed the raw data for further analysis. Data were analysed by detailed scrutiny of the transcripts to identify common themes, which were then coded, using the computer software package Atlas.ti [25]. Several transcripts were coded independently by G.M.W., P.A.D. and J.L.D., the resultant codings were compared, and any differences were discussed and resolved. Once coding had been completed, data were examined for similarities and differences within themes, retaining the context of the discussion and the characteristics of the participants. The themes were refined as new transcripts were analysed. Sampling continued with the aim of achieving data saturation: when new themes no longer emerged from the data. Descriptive accounts and case studies of the pre- and postoperation interview data were produced and discussed by the authors to check the credibility (internal validity), plausibility (reliability) and clinical relevance of the findings.

This paper focuses on data relating to the perception of outcome from the patient's point of view. The themes that emerged from the data are presented below together with illustrative quotations. All names have been replaced by letters to protect confidentiality.

Results

Twenty-five patients (14 women and 11 men) were purposively sampled and interviewed. Participants were Caucasian and reflected the profile of those operated in the UK [26, 27]: aged between 40 and 84 yr (mean 65 yr), more older females (55–84 yr; median = 68 yr) and more younger males (40–80 yr; median 66 yr). All of the 25 pre-operation participants agreed to be interviewed 6 months after the operation. However, a sample of these participants were purposively chosen for a postoperation interview to obtain a reasonable distribution of gender and age, in order to represent the main cohort. For example, the postoperation informants included the younger male and older female range of those undergoing TKR in the UK. Of the 25 pre-operation informants, 10 were invited to be interviewed again, and all accepted. Table 1 presents the characteristics of these participants. Six were female and four were male, and their ages ranged from 40 to 81 yr. At the time of the postoperation interview, seven of the participants were retired, three were on disability allowance, five were married and all lived in their own home. Seven had previous operations on either knee or hip, three operations being a total joint replacement. The remaining 15 patients were informed that they would not be interviewed again but were encouraged to contact the researcher if they had any comments. The following results relate only to those 10 interviewed 6 months after the operation.

Perception of TKR outcome

On direct questioning, the majority (9/10) stated that their TKR operation was ‘excellent’ (2), ‘very good’ (3) or ‘good’ (4). Questions such as ‘How satisfied are you with your knee replacement outcome?’ often resulted in what appeared to be a socially and personally desired response:

Miss D: I have had a very good outcome, oh yeah … I’m happy with the result, yeah very worthwhile doing. (64, single female, retired, no previous TKR.)

However, despite these positive responses, almost all (8) admitted they still experienced continued pain and immobility. Thus, it was only on further questioning about the outcome that many of the participants qualified their original assertion:

Miss D: I do get a lot of pain sometimes still. I was kneeling on the bed to get myself across to the window and I couldn’t, that was very painful with both knees … and it's still not very happy about shopping… I think that's the only thing now and getting up and down stairs but that's a bit more difficult because there is the problem of falling on stairs so I have got to be a bit more circumspect on that. (64, single female, retired, no previous TKR.)

There were a number of reasons why individuals reported a successful outcome despite the continued experience of pain and immobility. Mrs L, who admitted to being in as much pain after the operation as before it, was grateful for the operation, believing that the ‘bad’ osteoarthritis had been removed. She stated that the knee was ‘bound to be better’ and therefore regarded her outcome as good. Others expressed gratitude for having had the operation and the care given in the hospital, for which they had waited many years:

Mr O: I am still restricted but, you know, don’t get me wrong I am grateful for the attention I got, they looked after us and saw that things were done properly … and this [TKR] was the last resort … so all in all, it was pretty good. (40, married male with family, on disability allowance, previous knee operations.)

In addition, the majority of participants reported that they had coped better than expected with the TKR operation, and this may be a reason for a positively reported outcome. The TKR was indeed beneficial for some aspects of the participants’ life. In most cases, movement, pain or both were improved to varying degrees, and for some this made a significant difference to their life. For these participants, the remaining pain was mild enough for them to be able to live and cope with it, and therefore they perceived the TKR outcome as being a success. Comparisons were also made with their physical state before the operation and the problems they would have experienced had they not had the operation. In addition, comparisons of outcome were made with other people's TKR outcome. For example, the TKR outcome was viewed as positive when compared with other people worse off than themselves. Overall, most wanted to believe that they had a good outcome and gathered evidence to support this and present it to others, including this interviewer.

Participants struggled to make sense of their continued pain/immobility. More than half believed the recovery process was not complete 6 months after the operation, and therefore they could not give a final verdict on their TKR outcome because they needed more time to improve:

Mrs L: I mean I have got to give it another 2 or 3 months anyway, you know I thought when I had it done I would be running and walking around but I am not, I still can’t do a lot yet so it's got to take time hasn’t it … 6 months time I will probably tell you different … but really I expected to be running around but it doesn’t work like that. (80, widowed female, retired, previous THR.)

Others acknowledged that because the TKR involved such major surgery, it was only natural to experience pain. Mrs J, who stated she had an ‘excellent’ outcome and had a pain-free joint and improved mobility believed, that her ‘good healing skin’ facilitated her recovery and outcome. Four, who were still experiencing pain and immobility, accepted that they had slower healing times. Other reasons for a less good outcome than anticipated included flu, being overweight, having a weak knee, feeling depressed or worried about the knee and therefore not exercising it, or that the knee had been under too much stress leading up to the operation. Furthermore, the impact of other health problems, such as sciatica, back pain or their other knee, were thought to affect the outcome. Lay beliefs, such as TKRs being less successful than THRs, were also raised by three participants in an attempt to make sense of their outcome:

Mrs E: Well, they say that the hip is a bad one but apparently the knee is the worse because you are putting all your weight on that joint, you know. (64, married female, no previous TKR.)

A major factor that emerged from the struggle to make sense of outcome was the tendency for patients to try and take personal responsibility for the continued pain/immobility. Most of the participants believed that as the knee joint had been replaced by something new, any problems experienced must be their own fault. The participants did not criticise the surgeon or the surgery for their outcome:

Mr P: It's nothing to do with the joint now, it's more down to me, like I say it can’t be the actual joint, there shouldn’t be anything there now so … I think I might have done too much too soon. (48, single male, on disability allowance, no previous TKR.)

Six participants blamed themselves for ‘overdoing’ certain activities too soon after the operation:

Mrs B: I was getting pain down the back of my leg and I wondered whether … I mean you hear a lot of people say ‘you won’t get any pain at all afterwards’ and I thought is there something wrong … have I been stupid and done something silly, I didn’t know whether I had done some damage … cos I did go mad when I came home … my husband went out on the Saturday, and I was just pottering around just doing different things and really I should have rested it. (62, married female, retired, no previous TKR.)

Mrs E was so impatient with the slow recovery process that she forced her knee to do activities too soon after the operation:

Mrs E: I have heard so much about different people, like I told you before I have had no one in the family that I can compare with, like I said I have heard different people and they have said ‘oh so and so is like a two year old since they have had their hip or their knee’, course I have been forcing myself to go upstairs since 4 weeks on, and that was when I got told off because they said the stairs is the last thing really I should accomplish and, er, that was when I was told more or less that I am too impatient. (64, married female, no previous TKR.)

Another participant also forced his knee in the recovery process. As a result of reading the information booklet provided by the hospital, Mr T compared his own recovery to the reported recovery times in the booklet. Thus, at 6 weeks, when he did not have the reported range of movement in his knee, he forced it to achieve this range, potentially causing damage to the knee:

Mr T: You sort of get this book and it tells you what exercises to do, and I done all them and it says after 3 weeks you must come off your sticks and you can bear weight and after 6 weeks you should be able to walk up and down the stairs normally … well I can’t walk up and down the stairs normally after 6 months. When I went to see the doctor, you could almost feel the heat there … he said ‘what have you been doing to this knee’ and I said I have been trying to bend it by pulling it back and I force that up … hour after hour I used to do that … I was thinking the more I get it on the way the better, you don’t get any gain without pain and I don’t want any more operations. (66, divorced male, retired, previous operations.)

A key factor in allowing an understanding of the accounts of TKR outcome was the participant's life context. Individual case studies showed that the participants’ perception of the outcome made sense in the context of the preoperation data and the changes that had occurred since the operation. For example, Mrs L, although in as much pain after as before the operation, stated that she had a ‘good’ outcome because her recovery process coincided with her move from a lonely neighbourhood to a new, more community-spirited residential home, which positively affected her view of her TKR outcome. On the other hand, Mr S made sense of his remaining pain and immobility in the context of his religious belief that ‘you only receive what you deserve from God’. Although he was still experiencing pain and immobility 6 months after the TKR, he thought that the outcome was better than he had anticipated. He believed that this was God's way of trying to improve him as a person. These environmental and personal factors demonstrate the importance of exploring the life context of the individual when attempting to understand outcomes.

Discussion

This study has shown that patients had a strong desire to state that their TKR outcome was successful despite the continued experience of pain and immobility. Different reasons and rationalizations were made by the participants in an attempt to diminish any disappointment with their remaining pain and disability. As a result of these explanations, and despite the fact that they had considerable pain and disability, they continued to consider the TKR with high regard. Although it would be impossible to state that the small sample in this study was representative of all those who undergo a TKR, their demography corresponded well with the general profile of those operated in the UK [26, 27].

The results highlight a number of important issues in relation to the effectiveness of TKR. Previous quantitative work suggests that TKRs relieve pain and improve mobility, with a ‘good’ or ‘excellent’ outcome in approximately 90% of patients [9]. However, this qualitative study found that while outcome was reported initially in positive terms (‘good’, ‘very good’ or ‘excellent’), the majority of participants continued to experience pain and restricted mobility 6 months after the operation. The participants gave a socially desirable and positive summary of their TKR outcome when asked direct global questions about their outcome. This may be seen as their public expression of their outcome. It was only after further in-depth questioning that the participants admitted to the limitations of their TKR outcome. For example, 8 of the 10 participants admitted to the continued experience of pain and/or immobility, after previously reporting a ‘good’ or ‘excellent’ outcome. This may be seen as their private expression of their outcome.

The public expression of the outcome may be as a result of responding to a formal question, perhaps in the same way an individual might respond to medical staff or a questionnaire. Thus, individuals responded in simple terms and provided a summary of their experience. In the private view, participants were more willing and able to describe in further detail the overall summary of their outcome. They were given time to focus on what they could and could not do which revealed their concern and discomfort with the continuing pain and mobility difficulties.

Other social researchers have attempted to account for conflicting views presented in interviews [28–30]. For instance, Cornwell claimed that private accounts were indicative of people's real views, but they often felt obliged to present a publicly acceptable point of view [28]. Thus, the public accounts may represent the pat response: the right thing to say and to do, reflecting what they felt the researcher wanted to hear. Private accounts tend to represent the true feelings, and incorporate a range of psychological, emotional, social and contextual influences. It may be that in questionnaires completed in the quantitative studies, which reported TKR to be a highly effective intervention [5, 16, 18–20], the participants only expressed their public view of the outcome. The private expression that there is still the remaining pain and immobility may only be captured using a qualitative approach.

The important question concerning the implications of outcome and health status is which of the two accounts (public or private) should be taken as superior? The answer to this is that both should be considered, or at least the possibility of the existence of multiple realities should be taken into account. The public response represented an immediate and summarized formal response, whereas the private expression enabled the participant to qualify their global response. They acknowledged their improvement and desire for a good outcome and indicated that it was good in their public view. However, when allowed time to describe their summarized outcome in more detail, they then qualified this with statements about the remaining pain and immobility. Structured questionnaires fail to capture this qualification. The existence of these two realities does not mean that one is superior to the other, but it represents two different ways of portraying the same thing. Capturing only one of these accounts provides only one portrayal of the experience. However, as multiple realities exist, these should be explored in further detail as well.

A fundamental finding of this study was the process by which the participants struggled to understand and adapt to their outcome. All of the 10 postoperation participants provided rationalizations and attempted to make sense of their disappointment with their remaining pain and disability. The very human desire to say they had made the right decision in going ahead with the operation emerged as an important theme. They were pleased that the operation had been performed and that something had been done to relieve some of the pain and disability they were previously experiencing. Thus, their lives had improved and they were able to do more activities than before the operation. In addition, the fact that the majority of participants felt they had coped with the operation better than expected led to them feeling positive about their outcome. The process of ‘downward social comparison’ also reminded the participants that they had adjusted to their TKR outcome better than other people they knew [31, 32]. Others compared their healing process with other people's and used this to explain why their recovery was slow. Many held the continued hope for improvement as they felt that the recovery process was not complete after 6 months. Thus, they felt they needed more time to improve and achieve the full benefit of the operation. However, the evidence suggests that if an individual has not achieved pain relief or mobility at approximately 6 months, then they are unlikely to improve any more [27, 33].

The overriding rationalization was self-blame, taking personal responsibility for the continued pain and immobility. They had either overworked the knee too soon or had not listened to the information given about the recovery. This self-blame may represent attempts to re-establish some control over the outcome of the TKR. There was never any criticism of the surgeons or the surgery for the remaining symptoms. Other researchers have observed individuals’ reluctance to express criticism for the NHS or health professionals, resulting in a so-called normative effect [34, 35]. One of the reasons given for this reluctance arises from a sense of sympathy or understanding for the constraints that prevail in the NHS; it would be unreasonable or inconsiderate to criticize.

Examination of the case studies demonstrated that these apparently contradictory accounts were consistent in the context of the informants’ lives, and represented adaptation, rationalization or accommodation to their changed health state. Research has shown that changes in health status may induce changes in the meaning of one's self-evaluation, a phenomenon called ‘response shift’ [36]. In medicine, response shift refers to a change—as a result of an event such as a therapy or an operation—in the meaning of one's self-evaluation of quality of life. Thus, when health fails or an outcome of an intervention has not improved the quality of life to the full extent, individuals are said to reappraise the values and activities that had given their lives meaning [37]. Patients change their internal standards, values or reconceptualization of quality of life and therefore assess outcome differently than they would have if they had not adapted to their situation. In this study, it was apparent that the participants struggled to understand their TKR outcome, rather than give a straightforward representation of the symptoms they were still experiencing. This reconceptualization of their outcome became a means of trying to make sense of their outcomes.

Although TKR has been shown to be a highly effective procedure using quantitative methods, these studies may need to be qualified by these qualitative findings. Outcomes of TKR gained by simple questionnaire methods may represent only a limited view of the patients’ perception of TKR outcome and fail to capture individual meanings or adaptations. In contrast, this qualitative work has highlighted the complexity of the process of surgery and recovery, and the struggle to come to terms with their outcome. More sensitive assessments of outcome are needed to capture patients’ experiences, which incorporate the process of reconceptualizing outcome and take into account the context of the individual.

We thank all the study participants who were willing to be interviewed in this study. We also acknowledge Mr J. Newman, Mr C. Ackroyd and Professor I. Learmonth for allowing us access to their waiting list patients. The Department of Social Medicine at the University of Bristol is the lead centre for the Medical Research Council's Health Services Research Collaboration.

No conflict of interest has been declared by the authors.

References

Author notes

Department of Social Medicine and 1Medical Research Council Health Services Research Collaboration, Canynge Hall, University of Bristol, UK.

© The Author 2005. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For Permissions, please email: journals.permissions@oupjournals.org