What makes patients aware of their artificial knee joint?

Patients were consecutively approached for the study if they had had a TKA at the Medical University of Innsbruck and were at least 12 months after surgery. Patients were either recruited at a follow-up visit at the department of orthopaedics or invited via mail. Inclusion criteria were:

Patients recruited at the hospital first completed the FJS-12, the WOMAC and the importance ratings on a selection of outcome parameters measured by orthopaedic PRO instruments. Afterwards patients were asked to describe what makes them aware of their artificial joint within a semi-structured interview.

Patients invited via mail received a letter providing study information and an informed consent form. In addition, they received the –FJS-12, the WOMAC and the questionnaire on the importance ratings. Patients were asked to send back the questionnaires, written informed consent and their phone number and times, when they were available for a phone interview. Patients returning the questionnaires and signed informed consent were then contacted and interviewed on the phone.

For sample size determination we relied on recommendations by Morse [25] suggesting – at least 30 interviews for grounded theory approaches and on the concept of issue saturation from Glaser and Strauss [26], i.e. continuation of patient inclusion until no new issues are reported by five subsequent patients. Therefore, we planned to conduct at least 30 interviews and then continue until issue saturation was reached.

The project was approved by the local ethics committee (AN2015–0086).

All interviews were conducted by a psychologist not involved in the treatment of the patient and who is experienced with interviewing patients for patient-reported outcome studies. The interviews were either conducted face-to-face or by phone following a standardised interview guideline.

To familiarise patients with orthopaedic patient-reported outcome measures we asked them to complete two questionnaires prior to providing the importance ratings. These two questionnaires were the Western Ontario and McMaster Universities (WOMAC) Osteoarthritis Index [17], a 24-item questionnaire assessing pain, stiffness and function, and the Forgotten Joint Score – 12 (FJS-12; [20]), a 12-item questionnaire measuring joint awareness. For the WOMAC (score range 0–100) high scores indicate poor outcome, whereas for the FJS-12 (score range 0–100) high scores indicate good outcome. For aim 1 we asked the patients to rate the importance of frequently measured outcome parameters. Based on commonly used patient-reported outcome measures (e.g. Oxford Knee and Hip Score, WOMAC, FJS-12) an interdisciplinary group of orthopaedic surgeons and psychologists selected a list of 9 parameters to be rated. These parameters were:

Importance ratings for the various outcome parameters were obtained using a rating scale with ‘0’ indicating no importance at all and ‘10’ indicating high importance.

For aim 2 patients were asked to think of situations in everyday life, when they had been aware of their artificial knee (“Please explain in your own words in which occasions or situations you get aware of your artificial joint”). Then they were asked in what way they were made aware of their knee (“Why do you get aware of your artificial knee joint?”). To ensure detailed information on occasions, situations and reasons of joint awareness, the interviewer (FLL) used clarifying questions, encouraging the patients to explain or expand their statements, or give some examples of e.g. performed activities. The interviewer took detailed notes from all patients for later analysis.

Sample characteristics for sociodemographic and clinical characteristics are given as means, standard deviations and absolute and relative frequencies. For sample description we also provide means and standard deviations for the WOMAC and FJS-12 scales. For patients’ ratings for the importance of commonly assessed outcome parameters we calculated medians, means and standard deviations.

The qualitative interview data were independently reviewed by two authors (FLL and JMG). The analysis was based on the Grounded Theory approach described by Glaser [27], performing inductive coding with the help of the software NVivo 11 Pro, a software for organizing, analysing and structuring qualitative data. We first defined occasions, situations and reasons of joint awareness as theme for the analysis. Following this criterion we worked line by line though the material inductively constructing categories for each aspect fitting to the pre-defined theme. Categories could be terms (e.g. meteoropathy) or short sentences (e.g. foreign body feeling because of the implant) which characterized the material as nearly as possible. If passages fitted to the already constructed categories they were subsumed, otherwise new categories were formulated. After a first round, the category system was revised regarding the meaningfulness of the categories. In addition, we assessed if the level of abstraction was adequate to identify important occasions, situations and reasons of joint awareness. After a final work through the interviews, categories were grouped to obtain main categories (e.g. daily tasks and activities). Afterwards, categories and allocated text passages were compared and harmonized within several feedback loops between the two raters. The final category system was again applied to all interview data and codings were harmonised between the two raters. Besides examples for each category, the number of patients and the number of comments are reported as frequencies.

We approached 44 eligible patients for study participation, 12 at a hospital visit and 32 via mail. All patients had a total knee replacement between January 2001 and June 2015. Of the 44 eligible patients four did not participate in the interview. The remaining 40 patients had an average age of 69.1 (8.8) years and 65.0% were female. The mean FJS-12 score was 46.7 (SD 28.9) and the mean WOMAC total score was 40.4 (SD 21.5). Assessment duration ranged from 20 to 45 min. Further details are given in Table 1.

All outcome parameters received a mean importance rating of at least 7.4 on the 0–10 scale (Table 2). Lowest importance was given to having no abnormal feelings (7.4 points), and having no foreign body feeling (7.9 points). Highest ratings were found for being able to trust in the joint (8.7 points), not having pain and not having restrictions in doing the household (both 8.3 points).

The analysis of interview data from 40 patients resulted in six main categories: activities, specific movements, pain, bodily sensations, psychological factors, and meteoropathy. Details are given in Tables 3, 4 and 5.

Psychological factors of joint awareness mainly comprised insecurity about the stability of the joint (N = 6; e.g. male, 76 years: ‘You live differently, you are more cautious, you don’t jump up - you have the impression that you have to take care of the new knee’). Patients were also afraid of having revision surgery, an infection or inflammation of the knee and of being in pain again (N = 3; e.g. male, 77 years: ‘I am afraid that something will happen again, that something has to be fixed again’). Two patients reported that they are aware of their artificial joint as they simply know that they do have one (N = 2; e.g. female, 61 years: ‘I necessarily know that I have an artificial knee’) or because it is different from the past (N = 1).

Our design for evaluating the relative importance of the various outcome parameters may not have been optimal as patients had difficulties discriminating levels of importance. This resulted in rather equal importance levels across the parameters. While this may be a true finding it may also indicate that our method of assessing importance with ratings scales did not sufficiently emphasise the aspect of prioritisation. More conclusive results may have been reached by asking patients to rank the various parameters e.g. in a card sort exercise. Another limitation of our study is the sample size of 40 patients that was rather small for the quantitative analysis on importance ratings of outcome parameters. However, in the qualitative analysis we could reach issue saturation with our sample. In addition, our heterogeneous sample allowed the comprehensive identification of factors contributing to joint awareness, but the prevalence of the individual factors may not be generalizable to all TKA patients as our sample is not representative in terms of e.g. revision rates or frequency of bilateral surgery. We would like to note the high response rate in our study that was obtained through reminder calls via phone.

In a recent study [24] on the development of a PRO instrument to assess performance after TKA, patient focus groups were set up to qualitatively investigate different aspects of TKA outcome. In these focus groups joint awareness was identified as an important domain when measuring outcome after TKA. Similar to our findings the authors could demonstrate that beside the positive effects of TKA, patients noted negative changes including stiffness, numbness or tingling below the knee, meteoropathy issues, inability to kneel on the affected knee, problems going downhill or descend stairs. Furthermore the patients felt that more confidence (both physically and mentally) was needed in the knee or leg. These findings are well in line with the results from our interviews. The same study also found that patients also have distinct concepts of confidence, stability, and satisfaction when performing specific activities. This indicates that further concepts in addition to functioning or pain may have to be assessed after TKA. Joint awareness may work as an overarching parameter including several of these distinct concepts and serve as a measure of the above mentioned unselfconsciousness. The ability to forget about a joint in everyday life reflects a natural joint experience and indicates the unity of the self and the (artificial) joint. As a matter of fact this unity is not present while a patient is suffering from acute symptoms, making joint awareness a construct that gains importance at a later stage of joint recovery, i.e. after the initial rehabilitation phase and in mid- and long-term follow-up when the patient might be able to be distracted from his joint as a result of low or no symptom burden. Such a possible longitudinal shift in the relative importance of specific outcome parameters across the recovery trajectory could not be investigated within this study but, may be a worthwhile aim of future research.