Psychosocial predictors for outcome after total joint arthroplasty: a prospective comparison of hip and knee arthroplasty

Besides objective medical parameters like mortality, prosthesis survival or range of motion, health-related quality of life (HRQoL) is an important outcome of surgical interventions. A large number of studies investigated HRQoL after total knee (TKA) or total hip (THA) arthroplasty and reported impressing results with highly significant increases in general HRQoL as well as in disease-specific outcome [1‐6]. Nevertheless, a small proportion of patients do only reach poor or no improvement, raising the question which determinants could have an impact on joint replacement outcome. For this reason, sociodemographic and psychosocial variables have been taken into account, investigating their potential contribution to HRQoL after TKA or THA. For example, educational status, social support, depression [7] and mental distress [2] have been found to influence follow-up pain and physical function scores. As another widely used psychological parameter, sense of coherence (SOC) was shown to play an important role in perceived health and HRQoL of individuals with and without physical or mental illness, independent of age, sex, nationality and ethnicity [8, 9]. SOC, proclaimed by Antonovsky [10], describes “a global orientation that expresses the extent to which one has a pervasive, enduring though dynamic feeling of confidence” (p. 19), and consists of the three components comprehensibility, manageability and meaningfulness. In patients with osteoarthritis, a high baseline SOC was associated with a better mental as well as physical HRQoL after THA [11], yet SOC is generally stronger related to mental aspects of health [8, 9, 12].

Results of studies which evaluated THA and TKA patients separately suggest that patients´ benefit after THA is even greater with respect to pain, HRQoL, function and satisfaction than after TKA [3, 6, 13‐15], but there are also different findings [16]. A review investigating psychological factors affecting the outcome of total joint arthroplasty [17] summarized that evidence is strong regarding TKA whereas results for THA are limited and conflicting.

Against this background, the purpose of this study is to examine the influence of the psychosocial resources SOC and social support as well as of mental distress on outcome separately in patients after THA and TKA to define different predictors in these two patient groups.

This study compared in a prospective design outcomes of patients after THA with those after TKA and investigated which psychosocial parameters had an influence on outcomes in each group. To our knowledge, it is the first study to define predictors for each procedure separately and to contrast them.

As expected, THA as well as TKA patients showed high improvements in disease-specific variables and in general HRQoL after surgery, consistent with the existing literature. Because most variables did not differ between patient groups before surgery, comparability of changes was warranted. Overall, in our study the improvements seen in the WOMAC scores were even stronger for the THA than for the TKA group, which replicates the results of other studies regarding joint-specific instruments [3, 6, 14, 15]. Beyond that, in both patient groups we saw a reduction of psychological distress as measured by the three BSI scales and an increase in psychosocial resources, namely sense of coherence and social support. Whereas the decrease in anxiety and depression is in line with the literature [7, 11, 16, 26, 27] a change in SOC or social support has rarely been investigated [7]. The strong relationship of SOC to perceived health has widely been proven [8] and social support has shown to be an essential protective factor against psychological strain for example after traumatic experiences [28]. So it was our goal to examine the role of these two important constructs in the context of total joint arthroplasty. We know only one study which investigated SOC as predictor after surgery and this was restricted to THA, not including TKA patients [11]. Social support has been studied in different investigations but with conflicting results [4, 7, 29, 30].

Our main finding, which adds to the existing literature, was the difference between THA and TKA patients regarding the relationship of WOMAC with psychosocial variables. With one exception (stiffness baseline score predicted stiffness at 12 weeks), in THA, the regression analyses with all WOMAC scales as dependent variables solely revealed psychosocial parameters as significant predictors and WOMAC baseline scores seemed to be negligible. In contrast, in TKA patients, almost all scales were predicted by the baseline scores of the corresponding scales and with one exception (anxiety as predictor for stiffness) psychosocial variables did not play a role in predicting disease-specific outcome. This suggests that the relationship between physical and psychological aspects is much greater in THA patients. Maybe, the different results for the stiffness scale are due to its lower reliability compared to the pain and function scales because it only consists of two items.

The literature regarding psychosocial variables as predictors for physical outcome after THA or TKA is highly inconsistent. For this reason, we aimed to provide more evidence concerning this topic. In TKA patients, the results of other studies are heterogeneous. Desmeules et al. [31] found psychological distress (anxiety and depression) to influence SF-36 physical functioning but not WOMAC pain and function. Social support did not predict either of the scales. A review by Lewis et al. [32] identified depression, anxiety and social support as predictors for pain. In a study by Lopez-Olivo et al. [7] depression and one form of social support influenced WOMAC function but not WOMAC pain. Anxiety did not reveal as predictor for function or pain in this study. Depression predicted neither WOMAC function nor WOMAC pain in a study by Sullivan et al. [33]. Beyond anxiety, depression and social support, which were also investigated in our study, different coping styles, locus of control [7], (pain) catastrophizing [32, 33] and mental health [32] revealed as predictors for some physical variables but not for others in different investigations. However, despite all these conflicting findings, the baseline score of a particular physical variable nearly always predicted the outcome score after TKA [16, 31‐33], which confirms the result of our study. With respect to THA, we found only few studies investigating the influence of psychosocial variables on physical outcome after surgery. Quintana et al. [30] found baseline scores and mental health but not social support to influence WOMAC and SF-36 scores. Sense of coherence, anxiety as a trait and neuroticism but not anxiety as a state, depression or extraversion predicted SF-36 PCS and MCS in a study by Badura-Brzoza et al. [11]. Only partially in accordance with that, we identified anxiety, social support and sense of coherence as predictors for the different WOMAC scales in our study. But as all these investigations show significant methodical differences they are difficult to compare.

As mentioned above, in TKA, the influence of physical baseline scores on the corresponding outcome is the most consistent finding [31]. From our results we conclude, that they even have the strongest effect on physical outcome after surgery. In contrast, the impact of psychosocial variables seems to be much weaker in this group. Even if we had detected additional psychosocial predictors with a larger sample size, physical baseline scores still would seem to play the most important role. It is the strength of our study that we investigated TKA as well as THA patients, so we found out that in THA, the results are contrary. In this group we saw a stronger influence of presurgery psychosocial parameters than of physical baseline scores on the physical outcome.

These findings could possibly be explained by the differences in recovery after both procedures. Several studies pointed out that in both groups the greatest improvements occur in the first 3 to 6 months after surgery followed by slighter improvements thereafter [6, 13, 16, 34]. But even if THA and TKA patients lastly reached a similar level of specific or general HRQoL, THA patients showed faster recovery than TKA patients [3, 6, 34]. Dailiana et al. [16] reported the same trend for WOMAC pain and stiffness but not for WOMAC function and total score, in which the TKA group showed higher improvements (on both scales, TKA patients already exhibited better baseline scores). Furthermore, patients who underwent both, THA and TKA, reported more postoperative pain, lower improvements in range of movement and in quality of life, longer time before they felt physically better than they had before surgery and more required physiotherapy after TKA than after THA. THA was only rated inferior to TKA regarding the longer acute hospital length of stay [15]. So, patients after THA were more overall satisfied than TKA patients [3, 15, 34, 35] and more willing to undergo another total joint arthroplasty [15, 35]. It can therefore be concluded that TKA is a more difficult procedure than THA with a longer and more complicated recovery process. This could be the reason why a better physical status presurgery constituted a predictive factor for a better physical outcome after surgery in this group. Possibly, psychosocial factors also had an influence in TKA, but this was overlapped by the high impact of WOMAC baseline scores. Instead, in THA, results seem to be more independent from baseline physical conditions, at least in the early recovery process, so psychosocial parameters´ effects are more evident. Indeed, TKA patients showed psychological distress as indicated by significantly higher BSI GSI scores compared to the THA group, both before and after surgery. The worse prognosis and concerns of possible complications might have been the reason for higher distress in this group already prior to operation. Maybe patients informed themselves about the planned intervention and TKA patients found more discouraging information than THA patients. After arthroplasty, the higher distress presumably was related to the inferior physical condition.

As the follow-up period in our study took no more than 3 months, patients experienced a critical phase in recovery. If we had investigated our sample 6 or 12 months later, it could have been possible that the observed differences between THA and TKA patients regarding predictors for physical outcome scores no longer would have been found. Suggestive of this possibility are the results of two studies which investigated THA and TKA patients together. Serra-Sutton et al. [4] found that THA as well as TKA patients with low baseline scores in the SF-36 subscale mental health presented worse values in all SF-36 and all WOMAC scales 1 year after surgery. In the study by Dailiana et al. [16], in both patient groups predictive variables for changes in WOMAC scales 12 months after surgery were the corresponding baseline scores (with the exception of stiffness in TKA), whereas none of the investigated sociodemographic variables had an influence.

Overall, in our study, from all psychosocial variables tested, anxiety most often revealed as a significant predictor and showed elevated preoperative levels at least in the TKA group. So in preparing for total joint arthroplasty, distressed patients should be offered interventions for minimizing anxiety (for example psychosomatic contacts) to reveal a better outcome after surgery. Also, to a lesser extent, social support and sense of coherence influenced outcome in the THA group. In our study, the patients showed average or even above average scores in both parameters, but other patients could have a need to improve sense of coherence before surgery or could benefit from contacts to social services.

A limitation of our study is the relatively short follow-up period of 12 weeks. A longer follow-up period would have provided the opportunity to analyze if the identified differences between THA and TKA patients will maintain over time or maybe decrease. However, as mentioned above, the most significant changes occur shortly after surgery, so this was the phase we were primarily interested in. Furthermore, we did not investigate possible confounders like comorbidities or BMI. So other studies with greater sample sizes and controlling for different confounders are needed to further test our hypothesis.

To our knowledge, this is the first study to contrast predictors of physical outcome after surgery in THA vs. TKA patients. Our results suggest that although disease-specific dimensions and general HRQoL as well as psychological distress and psychosocial resources show great improvements after a total joint arthroplasty, the improvements in physical outcome are predicted by different variables in THA and TKA in the early recovery process. Whereas in the first 3 months after surgery physical outcome seems to be stronger influenced by physical baseline scores in TKA patients, in THA patients it seems to be stronger influenced by psychosocial aspects. However, TKA patients also reported mental distress before and after surgery, so patients in both groups, who exhibit a respective need, should be offered interventions to reduce psychological distress.