Perceived self-efficacy gained from, and health effects of, a rehabilitation program after hip joint replacement
To examine whether a routine multidisciplinary inpatient rehabilitation program can increase patient self-efficacy, and to investigate the effects of high self-efficacy at admission, and increases in self-efficacy, on health changes in patients who undergo such rehabilitation after hip joint replacement.
Participants in this longitudinal study were 1,065 patients who underwent inpatient rehabilitation after hip joint replacement. Questionnaires were administered at admission, discharge, and 6-month followup. The main outcome variables were disability, pain, depressive symptomatology, and self-efficacy to cope with disability and pain.
Significant improvements from admission to discharge from the inpatient rehabilitation program in disability, pain, depressive symptoms, and self-efficacy were found. In addition, higher levels of self-efficacy at admission and larger increases in self-efficacy over the course of the program predicted larger health changes (i.e., greater decreases in disability, pain, and depressive symptoms). Results were generally similar for health changes from discharge to 6-month followup.
A routine multidisciplinary inpatient rehabilitation program after hip joint replacement can result in enhanced self-efficacy.
There is a growing body of empirical evidence demonstrating that a person's perceived self-efficacy positively affects his or her physical health and emotional wellbeing. For example, patients with arthritis have been found to experience less pain and functional disability, as well as better emotional wellbeing, if they perceive high self-efficacy for coping with the negative consequences of the disease (1–6).
The construct of perceived self-efficacy was introduced by Bandura, who described in his social-cognitive theory how cognitive and social factors contribute to health and disease (7–9). Self-efficacy refers to the belief in one's capabilities to mobilize the motivation, cognitive resources, and courses of action required to produce given levels of attainment and is considered to be influenced by information from 4 principal sources.
Of these 4 sources of self-efficacy, mastery experience may be the most powerful influence. Personal successes (i.e., experiences of overcoming obstacles through perseverant effort) create a robust belief in one's personal efficacy. For example, an arthritis patient who starts exercising at an easily mastered intensity and frequency may become more and more confident of being able to improve functional disabilities through regular exercise.
Second, vicarious experience provided by social models conveys knowledge and effective skills/strategies for managing environmental demands. An arthritis patient observing other patients successfully exercising may realize that he or she also possesses the capabilities to master comparable activities for improving functional disabilities.
Third, social persuasion encourages positive appraisals of capabilities to master given tasks and achieve goals. An arthritis patient may receive corrective and reinforcing feedback and in that way acquires confidence in his or her ability to perform physical exercises necessary for improving functional disabilities.
Finally, interpretations of somatic and emotional states provide information about one's self-efficacy. While exercising, an arthritis patient may interpret pain as a sign of personal or physical inefficacy. Another patient, in contrast, may know that pain does not commonly mean worsening disease and may therefore not misinterpret the information with respect to his or her efficacy belief.
It has been proposed that self-efficacy mediates health through 2 processes (7, 8). First, the belief that one is able to successfully cope with stressors should activate biologic systems that promote health and healing processes (7, 8), and self-efficacy affects people's perception of, interpretation of, and ways of coping with pain experiences (10). Second, perception of high self-efficacy increases the likelihood of consideration, adoption, and maintenance of health-promoting behaviors.
Increasing evidence shows that people with arthritis who perceive high self-efficacy have better physical health and emotional wellbeing (i.e., fewer depressive symptoms). It has been found, for example, that arthritis patients with high self-efficacy for coping with pain or disability report less pain, disability, and depressive symptoms (1, 3–5, 11–13). Health-promoting effects of self-efficacy have also been observed in longitudinal studies of patients with arthritis (1–3, 6, 14, 15). In addition, a number of studies of arthritis patients have shown that increases in self-efficacy are associated with short- and long-term improvements in health (3, 14, 16–19). Furthermore, recent research with arthritis patients has demonstrated enhancement of self-efficacy through self-management programs and cognitive-behavioral and other psychological interventions (2, 3, 14, 16–29).
The above-mentioned intervention studies involved programs in which self-efficacy–enhancing strategies were specifically incorporated (i.e., mastery experience, role modeling, persuasion, reinterpretation of physical states). In the present study, in contrast, we investigated changes in self-efficacy from a routine multidisciplinary inpatient rehabilitation program that offers exercise therapy, occupational therapy, and patient education but does not explicitly target self-efficacy as one of its components. Considering the sources of self-efficacy described above, we hypothesized that a routine program that does not specifically target self-efficacy would nevertheless lead to increased self-efficacy because it indirectly provides various self-efficacy–enhancing experiences. During exercise therapy, patients may learn to perform and monitor their exercises for improving functional ability and to interpret accompanying physical states (e.g., pain). In occupational therapy in which joint-protection strategies and the use of assistive devices (e.g., canes) are taught, patients may learn how to compensate for limitations. Furthermore, patients receive health knowledge and information about health-promoting behaviors during patient education as part of a routine program.
Routine rehabilitation programs should therefore indirectly provide relevant learning experiences through direct experience, feedback and encouragement, and the opportunity for patients to observe other patients. We thus propose that routine programs that do not specifically target self-efficacy nevertheless lead to enhanced self-efficacy and, consequently, improved health outcomes. Building on research that demonstrates the benefits of self-efficacy in patients with mild or moderate arthritis, we examined benefits in patients who underwent hip joint replacement (i.e., whose arthritis was in an advanced stage). Although joint replacement provides marked pain relief and functional improvement in most arthritis patients (30–32), patients must apply continued self-management behaviors and coping strategies to improve their health outcomes (33). As suggested in previous studies (34, 35), self-efficacy plays a role in rehabilitation after hip joint replacement by contributing to improved health outcomes. Therefore, we investigated whether a routine multidisciplinary inpatient rehabilitation program after hip joint replacement can increase patients' physical health, emotional wellbeing, and self-efficacy, and whether higher and/or increased levels of self-efficacy result in better health outcomes across the course of the treatment and after discharge.
PATIENTS AND METHODS
Study participants were 1,065 patients from 13 rehabilitation facilities in Germany who were admitted for inpatient rehabilitation after hip joint replacement. The mean ± SD age of the participants was 64.58 ± 10.54 years. Sixty percent were women, and the majority (90%) had joint replacement for osteoarthritis of the hip.
Assessments were performed at admission (time 1), discharge (time 2), and 6-month followup (time 3). Measures included perceived self-efficacy, indicators of physical health (i.e., pain and disability), and emotional wellbeing (i.e., depressive symptomatology). Mean values for these measures are presented in Table 1. Additionally, age and sex were assessed as control variables.
Table 1. Mean ± SD values for the main study variables at admission (time 1), discharge (time 2), and 6-month followup (time 3)
|Self-efficacy for pain||3.19 ± 0.60||3.27 ± 0.60||–|
|Self-efficacy for disability||3.24 ± 0.53||3.44 ± 0.53||–|
|Pain||2.33 ± 1.85||1.39 ± 1.47||1.28 ± 1.70|
|Disability||5.89 ± 1.71||4.75 ± 2.02||1.27 ± 2.15|
|Depressive symptoms||11.42 ± 7.36||8.62 ± 6.35||8.96 ± 6.99|
We adapted the Arthritis Self-Efficacy Scale (3, 36) to a more specific measure of self-efficacy for coping with different consequences of joint replacement, and accordingly, we measured patients' confidence in their ability to successfully improve pain and disability. The scale's introduction referred to active participation in the rehabilitation program that exemplifies how patients may improve their health, and to barriers (e.g., pain, fatigue) that could impede active participation. Patients were then asked to indicate their perceived confidence in their ability to reduce their pain and achieve improvement in each of 8 daily activities, using 4-point Likert-type scales (1 = very uncertain; 4 = very certain). Self-efficacy for pain was assessed by a single item (“reducing pain”), whereas the scale of self-efficacy for disability comprised 8 items (e.g., “improve your walking distance,” “improve stair climbing”), with high levels of internal consistency reliability at both time 1 (α = 0.92) and time 2 (α = 0.93).
To assess pain, we asked patients to indicate their joint-specific pain intensity in 3 different situations (at rest, with first steps, and when walking), using 11-point visual numerical rating scales (0 = no pain; 10 = pain at its worst). The levels of reliability were high at admission (α = 0.90), discharge (α = 0.90), and followup (α =0.91).
Disability was measured by asking whether patients had difficulty performing each of 8 activities of daily living. The list of daily activities was derived from an internationally used clinical hip score (37) and included walking distance, stair climbing, and putting on shoes and socks. For use in the present analyses, a simple count of the number of daily activities for which the patient reported difficulty was obtained.
Depressive symptomatology was assessed with the Allgemeine Depressionsskala (38), the German 15-item version of the Center for Epidemiologic Studies Depression Scale (CES-D) (39). The CES-D was developed to measure depressive symptoms in the general population and is frequently used as an indicator of emotional well-being in arthritis populations (11, 16, 40). Patients are asked to indicate how often each statement (e.g., “I felt depressed,” “I felt lonely”) applied to them during the past week, using 4-point Likert-type scales (0 = rarely or none of the time; 3 = most or almost all of the time). Possible scores range from 0 to 45, with higher scores indicating more depressive symptoms. The scale showed a Cronbach's alpha of 0.83 at admission, 0.82 at discharge, and 0.86 at followup.
Inpatient rehabilitation treatment
In Germany, most patients undergoing joint replacement are referred, within 4 weeks postsurgery, from acute care to an inpatient orthopedic rehabilitation facility that provides multidisciplinary rehabilitation programs (41). Complete coverage of the costs for 21 days is available through health or pension insurance. Programs for patients who have undergone hip replacement offer exercise therapy, occupational therapy, and patient education. The goal of exercise therapy is to reduce disability and pain by improving muscle strength, stability of joints, range of motion, and aerobic fitness. Supervised by a physical therapist, it is performed in single or group sessions. Additionally, patients are instructed in a specific home exercise program. Occupational therapy aims at improving performance of activities of daily living and compensating for functional limitations by educating about joint protection strategies and use of assistive devices (e.g., canes, crutches, walkers). Patient education is conducted in 1-hour group sessions by an interdisciplinary team (physician, physical therapist, occupational therapist) and is based on a curriculum consisting of modules. In the first module, a physician conveys disease-specific information (e.g., impact of undue stress, importance of musculature, special concerns regarding infection risk). In the second module, a physiotherapist or occupational therapist provides information about health-promoting behaviors (e.g., use of assistive devices, everyday behaviors, and exercise behaviors). Moreover, patients receive physical therapy (e.g., hydrotherapy, balneotherapy, electrotherapy, massage).
In the first set of analyses we addressed changes in health status and self-efficacy throughout the treatment period, and the influence of self-efficacy on health changes from admission to discharge. A second set of analyses was performed to examine health changes after discharge and the predictive value of self-efficacy on health changes from discharge to 6-month followup.
Both sets of data analyses were carried out in 2 steps. First, paired t-tests were computed to test changes in health status and self-efficacy. Second, hierarchical regression analyses were performed to test the predictive value of self-efficacy for changes in health outcomes. To operationalize changes in health outcomes, we used the “residualized change” approach (i.e., a conservative approach to longitudinal analysis). With this method, the health outcome of interest (i.e., assessed at time 2 or time 3) is regressed on its previous assessment (at time 1 or time 2) to compute residualized scores that are algebraically identical to change (42). With this residualized change score entered as the dependent variable, regression analyses can test predictions of health changes independent of previous levels (i.e., assessed at time 1 or time 2).
Two hundred ninety-six participants were no longer available for study at the 6-month followup. To identify factors that may have contributed to study withdrawal, differences in the main study variables between the 769 participants who completed the followup and the 296 who did not were tested. There were significant differences between the 2 groups only for pain, depressive symptoms (both at time 2), and age (Table 2). Specifically, participants who completed the followup were a mean of 2 years older than those who dropped out (t[1,026] = 2.62, P < 0.01, d = 0.16) and reported lower pain levels (t[457.85] = −2.15, P < 0.05, d = −0.20) and lower levels of depressive symptoms (t[448.27] = −3.62, P < 0.01, d = −0.34) than patients who did not return the followup questionnaire. We took these differences into account by controlling in all followup analyses for pain and depressive symptoms at time 2, and for age. No differences were found with respect to sex distribution or the other main study variables.
Table 2. Values for the main study variables at admission (time 1) and discharge (time 2), in patients grouped according to those who remained in the study at 6-month followup (participants) (n = 769) and those who did not (dropouts) (n = 296)*
|Age, years||65.11 ± 10.45||63.18 ± 10.69†||–||–|
|Pain||2.28 ± 1.83||2.45 ± 1.92||1.33 ± 1.41||1.56 ± 1.61‡|
|Disability||5.93 ± 1.69||5.77 ± 1.78||4.71 ± 2.01||4.87 ± 2.04|
|Self-efficacy for pain||3.21 ± 0.60||3.14 ± 0.59||3.31 ± 0.60||3.16 ± 0.59|
|Self-efficacy for disability||3.26 ± 0.51||3.20 ± 0.57||3.45 ± 0.53||3.40 ± 0.54|
|Depressive symptoms||11.16 ± 7.17||12.18 ± 7.86||8.15 ± 6.12||9.85 ± 6.78†|
|Changes in self-efficacy for pain||–||–||0.27 ± 0.48||0.24 ± 0.60|
|Changes in self-efficacy for disability||–||–||0.16 ± 0.69||0.10 ± 0.62|
Analysis of health changes from admission to discharge
The inpatient rehabilitation began ∼3 weeks after surgery (mean ± SD 21.56 ± 8.18 days), with a mean ± SD length of stay of 22.64 ± 4.13 days. We expected that patients' health status would be better at discharge from the rehabilitation program than at admission. Paired t-tests with pain, disability, and depressive symptomatology as dependent variables showed that pain (t = −19.45, P < 0.001, d = −0.95), disability (t = −20.24, P < 0.001, d = −0.99), and depressive symptomatology (t = −13.25, P < 0.001, d = −0.65) decreased significantly during the treatment period.
We also examined whether patients' self-efficacy changed during the treatment period. Paired t-tests with self-efficacy for pain and for disability as dependent variables revealed significant increases from admission to discharge in self-efficacy with regard to pain (t = 4.70, P < 0.001, d = 0.31) and disability (t = 9.84, P < 0.001, d = 0.65).
Effects of self-efficacy on health changes
We carried out separate hierarchical regression analyses, with changes in disability and pain as dependent variables to test whether high self-efficacy at admission to the rehabilitation program and its increase throughout the program result in better health outcomes. In the first step, we entered patient age and sex as control variables. We then tested, in the second step, the significance of the effects of self-efficacy at admission and changes in self-efficacy (i.e., whether self-efficacy accounts for unique variance in health changes over and above health status at time 1, age, and sex). To operationalize changes in self-efficacy, difference scores were computed by subtracting self-efficacy levels at time 2 from levels at time 1, with a positive score indicating an increase in self-efficacy throughout the program.
Table 3 presents the results of the regression analyses. The control variables significantly accounted for variance only in disability, i.e., in men, difficulties in daily activities decreased more than in women.
Table 3. Hierarchical regression analyses predicting change in disability and pain from admission (time 1) to discharge (time 2), by admission levels of perceived self-efficacy and change in perceived self-efficacy during the treatment period
|Control variables||0.02¶|| || || ||0.00|| || || |
| Age|| ||0.01||0.00||0.05|| ||0.01||0.00||0.06|
| Sex#|| ||0.41||0.10||0.12¶|| ||−0.05||0.07||−0.02|
|Main effects||0.10¶|| || || ||0.04¶|| || || |
| Self-efficacy at admission|| ||−1.11||0.11||−0.33¶|| ||−0.46||0.08||−0.20¶|
| Change in self-efficacy|| ||−0.85||0.12||−0.22¶|| ||−0.43||0.08||−0.18¶|
In support of our hypotheses, adding admission levels and change scores of self-efficacy in the second step of the analyses contributed significantly to the prediction of changes in disability in pain (Table 3). Admission levels of self-efficacy significantly predicted disability and pain. The more confidence patients had in their ability to improve disability and pain at admission, the more disability and pain decreased across the course of the treatment. Changes in self-efficacy also contributed significantly to the prediction: the larger the experienced increase in self-efficacy for disability or pain, the greater the observed decrease in disability or pain.
Moreover, we hypothesized that high levels of self-efficacy may contribute to better emotional wellbeing. To test this hypothesis, we conducted a hierarchical regression analysis with changes in depressive symptoms from admission to discharge as the dependent variable. In the first step of the analysis, we entered data on health status variables at time 1 (i.e., disability and pain) in addition to patients' age and sex as control variables. By controlling for disability and pain, we attempted to exclude the possibility that decreases in depressive symptoms were merely due to lower levels of disability or pain (at time 1), and not to high/increased levels of self-efficacy. In the second step, we tested the significance of the effects of self-efficacy at admission and changes in self-efficacy for disability and for pain.
As seen in Table 4, control variables accounted for variance in changes in depressive symptoms. In particular, pain had a significant effect: the lower the reported level of pain at admission, the greater the decrease in depressive symptoms across the course of the treatment. Findings of the analysis after addition of the self-efficacy variables in the second step supported our hypotheses (Table 4). Levels of self-efficacy for disability and for pain at admission predicted changes in depressive symptoms: the more confident patients were of their ability to improve their disability and pain at admission, the larger the observed decrease in depressive symptoms over the course of the treatment. Moreover, changes in self-efficacy for disability and for pain also contributed significantly to the prediction: the larger the experienced increase in self-efficacy for disability and pain during the course of the program, the larger the observed decrease in depressive symptoms.
Table 4. Hierarchical regression analyses predicting change in depressive symptoms from admission (time 1) to discharge (time 2) and from discharge to 6-month followup (time 3), by admission levels of perceived self-efficacy and change in perceived self-efficacy during the treatment period
|Control variables||0.02§|| || || ||0.04§|| || || |
| Age|| ||0.02||0.01||0.05|| ||−0.03||0.02||−0.05|
| Sex¶|| ||0.15||0.32||0.01|| ||−0.09||0.46||−0.01|
| Disability (time 1/time 2)|| ||−0.03||0.10||−0.01|| ||0.38||0.12||0.12§|
| Pain (time 1/time 2)|| ||0.36||0.09||0.13§|| ||0.52||0.17||0.12§|
|Main effects||0.03§|| || || ||0.01§|| || || |
| Self-efficacy for disability at time 1|| ||−1.04||0.43||−0.10#|| ||−1.35||0.62||−0.11#|
| Change in self-efficacy for disability|| ||−1.03||0.45||−0.08#|| ||0.26||0.68||0.02|
| Self-efficacy for pain at time 1|| ||−1.05||0.41||−0.11§|| ||0.44||0.57||0.04|
| Change in self-efficacy for pain|| ||−0.82||0.40||−0.08#|| ||−0.82||0.56||−0.07|
Analysis of changes in health from discharge to 6-month followup
We expected that patients would experience further health improvements after discharge. Paired t-tests with pain, disability and depressive symptomatology as dependent variables revealed significant decreases in disability (t = −28.43, P < 0.001, d = −1.67) and increases in depressive symptomatology (t = 2.92, P < 0.01, d = 0.17) after discharge. Pain did not change significantly after discharge (P > 0.10). However, levels of depressive symptoms (t = −7.13, P < 0.001, d = 0.39) and pain (t = −14.92, P < 0.001, d = 0.79) were on average still lower at 6-month followup than at admission.
Effects of self-efficacy on health changes after discharge
To statistically test whether self-efficacy at admission and its change throughout the program are also associated with health changes after discharge, we carried out separate hierarchical regression analyses with changes in disability and pain from discharge to 6-month followup as dependent variables. We maintained the structure of the previously reported analyses but added pain and depressive symptoms at time 2 (the factors that showed significant differences between followup participants and dropouts) as control variables in the first step.
Table 5 presents the results of the regression analyses for predicting changes in disability and pain. First, the control variables contributed significantly to the predictions. Age significantly predicted disability, indicating that the younger patients were, the greater the decrease in disability after discharge. Sex significantly contributed to the prediction of pain, i.e., women experienced larger decreases in pain. Depressive symptoms at time 2 accounted for variance in both dependent variables, indicating that the lower the level of depressive symptoms at discharge, the greater the decreases in disability and pain after discharge. Finally, pain at discharge had a significant effect on disability: the lower the level of pain at discharge, the greater the decreases in disability after discharge.
Table 5. Hierarchical regression analyses predicting change in disability and pain from discharge (time 2) to 6-month followup (time 3), by admission (time 1) levels of perceived self-efficacy and change in perceived self-efficacy during the treatment period
|Control variables||0.12¶|| || || ||0.03¶|| || || |
| Age|| ||0.04||0.01||0.20¶|| ||−0.01||0.00||−0.04|
| Sex#|| ||−0.02||0.13||0.00|| ||−0.25||0.11||−0.08**|
| Depressive symptoms at time 2|| ||0.03||0.01||0.09¶|| ||0.04||0.01||0.16¶|
| Pain at time 2|| ||0.29||0.05||0.22¶|| ||–||–||–|
|Main effects||0.02§|| || || ||0.00|| || || |
| Self-efficacy at time 1|| ||−0.60||0.15||−0.16¶|| ||−0.13||0.12||−0.04|
| Change in self-efficacy|| ||−0.54||0.18||−0.11¶|| ||−0.12||0.13||−0.04|
In support of our hypotheses, the addition of admission levels and change scores of self-efficacy in the second step of the analyses contributed significantly to the predictions, but only for disability (Table 5). The higher the level of self-efficacy for disability at admission, the greater the decrease in disability from discharge to followup. Moreover, changes in self-efficacy significantly affected disability, indicating that the larger the increase in self-efficacy for disability experienced throughout the program, the larger the decrease in disability after discharge.
The results of the hierarchical regression analysis with changes in depressive symptoms as the dependent variable are shown in Table 4. First, the control variables contributed to the prediction: the lower the reported levels of disability and pain at discharge, the larger the decrease in depressive symptoms after discharge. The addition of the self-efficacy variables in the second step of the analysis supported our hypothesis (Table 4). In particular, admission levels of self-efficacy for disability had a significant effect: the more confident patients were of their capability to improve their disability at admission, the larger the decrease in depressive symptoms after discharge. Changes in self-efficacy for disability and both variables of self-efficacy for pain did not significantly contribute to the prediction of changes in depressive symptoms after discharge.
The present longitudinal study examined whether a routine multidisciplinary inpatient rehabilitation program that was not specifically targeted to enhance self-efficacy can nevertheless increase self-efficacy, and whether self-efficacy promotes better health in patients who participated in the rehabilitation program after hip joint replacement. On the basis of reported theories and empirical findings, we hypothesized that the program contains elements of self-efficacy–enhancing strategies. In addition, we postulated that the more highly patients perceived their self-efficacy at admission and the larger the experienced increase in self-efficacy throughout the program, the more health outcomes would improve from admission to 6-month followup. The presented findings support these hypotheses.
We found that patients' physical health and emotional wellbeing improved throughout the period of inpatient rehabilitation. Patients reported lower levels of disability, pain, and depressive symptoms at discharge than at admission. Decreases in disability continued during the first 6 months after discharge, while pain levels remained stable and depressive symptomatology increased after discharge. Despite this increase, we note that the level of depressive symptomatology was on average still lower 6 months after discharge than at admission. This pattern of results is in accordance with previous research demonstrating greater and faster improvements after joint replacement with respect to pain, while functional improvements seemed to take place within a protracted recuperation period of 3–6 months postsurgery (30–32, 43). Depressive symptomatology, in contrast, increased from 1 month to 6 months postsurgery after a decrease during the first month postsurgery (43). To explain these different recovery trajectories, one may have to consider that depressive symptomatology can be a function of experiences across different areas of life. In this regard, it may be that the positive surgery-related experiences play a less and less important role the longer the time passed since the surgery.
Our findings also show that patients' self-efficacy increased throughout the inpatient rehabilitation period. This supports the assumption that self-efficacy can be enhanced through a routine multidisciplinary inpatient rehabilitation program that comprises exercise therapy, occupational therapy, and patient education. Such routine programs may provide experiences and information about task demands and capabilities for mastering certain tasks and achieving goals, and thereby indirectly strengthen patients' self-efficacy.
On the basis of previous research on arthritis patients, we had predicted that individual differences in self-efficacy at admission and in its increase across the inpatient rehabilitation period would predict physical and emotional health changes. The reported results strongly support this hypothesis by demonstrating that higher self-efficacy at admission and larger increases in self-efficacy throughout the program result in larger decreases in disability, pain, and depressive symptoms from admission to discharge, over and above patients' baseline health status. Thus, these findings provide evidence in support of a central assumption of social-cognitive theory, i.e., that confidence in one's capabilities to mobilize the motivation, cognitive resources, and courses of action required to produce given levels of attainment facilitates the adoption of behaviors that promote the attainment of desired outcomes (8).
Taken together, the results suggest that patients acquired skills and resources that facilitated their adjustment to the hip joint replacement and its consequences. This finding thus supports the following conclusions: 1) The higher the level of perceived self-efficacy for disability, the more effort patients may invest in exercise therapy. If this is the case, then they may acquire more muscle strength, stability of joints, range of motion, and aerobic fitness, which are assumed to reduce pain and disability (44). 2) Patients who perceive high self-efficacy to cope with pain may have lower pain levels because they use more adaptive pain-coping strategies, as demonstrated in previous studies (6, 11, 14, 45, 46). 3) The belief in one's ability to cope with disability and pain can have positive consequences regarding emotional wellbeing (e.g., depressive symptomatology) (3, 4, 11, 14, 16, 34, 47). However, these interpretations are speculative, and more research is needed to directly test the postulated mediating processes.
Moreover, our results showed that both self-efficacy measures for disability contributed to health changes after inpatient rehabilitation following hip joint replacement. Both the level of self-efficacy for disability at admission and its increases accounted for differences in changes in disability, and admission levels of self-efficacy for disability accounted for differences in changes in depressive symptomatology from discharge to 6-month followup. This is consistent with findings from self-efficacy research suggesting that when patients experience high and/or increased self-efficacy, they experience substantial and prolonged health benefits (1, 15, 48), better long-term health outcomes (18), and reduced utilization of health care services in the future (49). Thus, benefits of self-efficacy seem to be sustained, and self-efficacy–enhancing interventions (whether direct or indirect) provided early in the rehabilitation process may promote positive long-term health.
Given the predictive value of (enhanced) self-efficacy for health improvements, more efforts should be made to improve self-efficacy in the context of hip joint replacement surgery and the subsequent rehabilitation process. Our results underscore the fact that a routine multidisciplinary inpatient rehabilitation program provides opportunities to enhance self-efficacy. It does not seem necessary to implement an intervention specifically designed to increase self-efficacy. In rehabilitation settings, simple techniques such as guiding mastery experiences in the context of exercise therapy, facilitating vicarious experiences, providing specific feedback, and/or providing support in interpretation of somatic states can simply be implemented by therapists. Furthermore, implementation of similar techniques early in the treatment process may promote self-efficacy even before admission to a rehabilitation facility.
Although the results of this study supported our hypotheses and general expectations, some limitations should be addressed in future research. First, participants remaining in the study through the followup period were older and less affected by their pain and emotional condition at discharge than patients who dropped out. Second, our study did not examine the variables that may explain the health-promoting effects of self-efficacy. Future research may assess possible mediating variables such as investment of effort in exercise therapy or adaptive pain-coping strategies. Finally, although our findings confirmed the self-efficacy–enhancing effects of a routine rehabilitation program, it is unclear which program components or experiences (indirectly) enhanced self-efficacy. Future research is needed to study directly the self-efficacy–enhancing factors in routine rehabilitation programs, based on the 4 described sources of self-efficacy. In light of the apparent health-promoting effects of high and/or enhanced self-efficacy, understanding the contributing factors would be most relevant to improving routine care early in the rehabilitation process after hip joint replacement.