Abstract

Objective. To explore the relationship between measures of self-efficacy, health locus of control, health status and direct medical expenditure among community-dwelling subjects with rheumatoid arthritis (RA) and osteoarthritis (OA).

Methods. This analysis is part of a larger ongoing study of the costs and outcomes of arthritis and its treatments. Community-dwelling RA and OA respondents completed questionnaires concerning arthritis-related expenditure, health status, arthritis related self-efficacy and health locus of control.

Results. Data were obtained from 70 RA respondents and 223 OA respondents. The majority of respondents were female with a mean age of 63 yr for RA respondents and 68 yr for OA respondents. Among the RA respondents, those with higher self-efficacy reported better health status and lower overall costs. Health locus of control was not consistently correlated with health status. OA respondents with higher self-efficacy reported better health status and lower costs. Health locus of control had more influence. OA respondents with higher external locus of control reported worse pain and function. A higher belief in chance as a determinant of health was correlated with more visits to general practitioners and a higher cost to both the respondent and the health system.

Conclusion. Higher self-efficacy, which is amenable to change through education programmes, was associated with better health status and lower costs to the respondent and the health system in this cross-sectional study. Locus of control had less of an influence; however, the tendency was for those with higher external locus of control to have higher costs and worse health status. As the measurement of these constructs is simple and the outcome potentially affects health status, these results have implications for future intervention studies to improve quality of life and reduce the financial impact of arthritis on both the health-care system and patients.

It is known that psychological factors such as self-efficacy and locus of control play a role in adaptation to chronic illness, such as arthritis. Self-efficacy was first described by Bandura as a person's belief that they can perform a specific behaviour or task in the future [1]. These beliefs determine whether a person attempts behaviours or tasks, how long they will persist in the face of obstacles and the level of success eventually achieved. Self-efficacy is a behaviour-specific characteristic which can be altered through education programmes and enhanced self-efficacy is associated with improved health and subsequently reduced health care costs in areas affected by those specific behaviours [2].

The concept of locus of control developed from Rotter's social learning theory [3] and has been extended by Wallston et al. to cover the multidimensional aspect of health-related behaviour [4]. Health locus of control is the extent of an individual's perceived control over health outcomes. The scale of Wallston et al. measures locus of control as ‘internal’, whereby people believe they are personally responsible for their own health; ‘external–powerful others’, which is the belief that others, such as health professionals, are responsible for one's health; and ‘external–chance’, whereby people believe that health depends on luck, fate or chance. Those with high internal locus of control are more likely to take control of their own health, seek health information and maintain physical well-being [5]. Relationships exist between locus of control and reporting of pain intensity, and it has been suggested that patients with chronic pain and external locus of control do not believe in recovery, and thus avoid increasing their activity level and report poor ability to reduce and control their pain [6]. Those with internal locus of control may develop strategies to deal with pain, so report lower pain intensity [7].

Both arthritis-specific self-efficacy and locus of control have an impact on health behaviour. While they have been investigated separately, the combination of these characteristics has not been widely studied, and not in the arthritis population. Those with high self-efficacy and an internal locus of control are likely to feel in control of their own health and feel capable of performing the necessary behaviours to maintain health. This high self-efficacy in combination with internal locus of control may have a different impact on health status than low self-efficacy and external locus of control. The aim of the present study was to explore the relationship between measures of self-efficacy, health locus of control, health status and direct medical expenditure among subjects with rheumatoid arthritis (RA) and osteoarthritis (OA).

Methods

This study is a subset of a larger ongoing study, the Arthritis Cost and Outcome Project, which has been described previously [8, 9]. RA and OA participants, as diagnosed by rheumatologists, were recruited through public and private out-patient clinics in northern and eastern Sydney. People who responded to a random population survey of residents in the northern Sydney area indicating symptoms of OA were also asked to join the study if, after examination by a rheumatologist, a diagnosis of OA was shown to be correct. No RA patients were recruited for the present study through this community survey. Patients who had not had joint replacement as well as those who had undergone hip or knee joint replacement more than 5 yr previously were included. All patients signed written consent forms to participate and the study was approved by the relevant institutional ethics committees.

Questionnaires related to arthritis-related expenditure, health status, arthritis self-efficacy and health locus of control were completed in 2003 (Table 1). Respondents recorded in the Cost Questionnaire their arthritis-related expenditure over the previous 3 months. Where respondents reported using medications, both the cost to the patient themselves, in terms of out-of-pocket cost, and the cost to the government-funded Pharmaceutical Benefits Scheme (PBS) were calculated. Under this scheme, the Australian Federal Government subsidizes the cost of medications, so patients pay a reduced amount for pharmaceuticals included in the scheme. In a similar manner, the cost to the government-funded Medicare Benefits Scheme (MBS) was calculated where respondents recorded visits to health professionals, such as a general practitioner, rheumatologist or surgeon, or medical tests. Patients are able to claim from the MBS 85% of the schedule fee for visits and tests, although some practitioners charge in excess of the schedule fee so patients may incur some additional out-of-pocket expense. Both the cost to the patient for visits and tests and to the MBS were calculated.

Table 1.

Questionnaires used in the study

Questionnaire Author Areas covered
Cost Questionnaire Arthritis Cost Study [8, 9Three-month retrospective arthritis-related expenditure for prescription and non-prescription medications, visits to health professionals, tests, assistive equipment, use of community and private services, hospitalization
Short Form-36 (SF-36) Ware [10Eight domains: Physical Function, Role Physical, General Health, Bodily Pain, Vitality, Social Function, Role Emotional, Mental Health; combine to give two scores Physical Component Score (PCS) and Mental Component Score (MCS)
WOMAC Osteoarthritis Index Bellamy [11Pain, Stiffness, Physical Function
Health Assessment Questionnaire Fries [12Overall disability score
Arthritis Self-Efficacy Scale Lorig [2Self-efficacy for Pain, Function and Other symptoms
Multidimensional Health Locus of Control Wallston [4Internal, External-Powerful others, External-Chance
Questionnaire Author Areas covered
Cost Questionnaire Arthritis Cost Study [8, 9Three-month retrospective arthritis-related expenditure for prescription and non-prescription medications, visits to health professionals, tests, assistive equipment, use of community and private services, hospitalization
Short Form-36 (SF-36) Ware [10Eight domains: Physical Function, Role Physical, General Health, Bodily Pain, Vitality, Social Function, Role Emotional, Mental Health; combine to give two scores Physical Component Score (PCS) and Mental Component Score (MCS)
WOMAC Osteoarthritis Index Bellamy [11Pain, Stiffness, Physical Function
Health Assessment Questionnaire Fries [12Overall disability score
Arthritis Self-Efficacy Scale Lorig [2Self-efficacy for Pain, Function and Other symptoms
Multidimensional Health Locus of Control Wallston [4Internal, External-Powerful others, External-Chance

The health status of the respondents was determined through the completion of the Short Form 36 (SF-36) questionnaire [10], the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) [11] and the Health Assessment Questionnaire (HAQ) [12]. For both the WOMAC and HAQ, higher scores indicate a worse health status. The WOMAC gives three scores: pain (minimum 0 to maximum 20), stiffness (0–8) and function (0–68). The overall HAQ disability score varies from 0 to 3. The SF-36 gives scores on a 0–100 scale, higher scores indicating better health. The eight scales of the SF-36 are combined to create a Physical Component Summary (PCS) score and a Mental Component Summary (MCS) score, which are standardized to create scores with a mean of 50 and a standard deviation of 10. The PCS and MCS are derived and scored using a factor analytical method. The PCS comprises the SF-36 scales Physical Functioning, Role–Physical, Bodily Pain and General Health. The scales of Vitality, Social Functioning, Role–Emotional and Mental Health combine to form the MCS. A higher score on the SF-36 indicates a better health state.

To assess arthritis-related self-efficacy, the Arthritis Self-Efficacy Scale (ASES) was used. Respondents are required to indicate how certain they are of performing specific behaviours with regard to pain, function and other symptoms, such as fatigue, depression and frustration [2]. Health locus of control was assessed using the Multidimensional Health Locus of Control questionnaire, which asks respondents to indicate on a six-point scale their level of agreement with a number of statements regarding the control of their health [4]. Respondents are given a score for each of the three areas: internal, powerful others and chance. Each subscale varies from 6 to 36 with a median score of 21. Internality and externality are not opposite ends of the one spectrum, and it is possible to have both internal and external beliefs about health status at the same time. To assess the combined effect of locus of control and self-efficacy on function, those scoring above the median on both the internal locus of control scale and the function self-efficacy scale were compared with those scoring below the median on both these scales. It is assumed that those scoring above the median on both these scales believe they are personally responsible for their health and are able to perform the behaviours to maintain their health.

Data were analysed using Statistical Packages for the Social Sciences v. 12.0.1 for Windows 2003, (Chicago, illinois, SPSS Inc). Product moment correlations were used to assess relationships between health status and locus of control and self-efficacy. Where some items were missing from individual scales on questionnaires, the missing value was replaced with the mean of that scale for that individual, a method recommended for use in the SF-36 and WOMAC questionnaires [10, 11]. Differences of 5% were considered significant, and correlation coefficients are denoted at both the 5 and the 1% level.

Results

Data were collected from 70 RA and 223 OA respondents. Details of respondents are shown in Table 2. The majority of RA respondents were female and the mean age of respondents was 63 yr, while over half of the OA respondents were female. The mean age of OA respondents was 68 yr. The disease duration was in excess of 20 yr for both groups, indicating that these were groups who had been living with the disease for some time. While this study was not directly comparing OA and RA patients, it is interesting to note that the pattern of locus of control scores was very similar between the two groups, with higher internal and powerful others scores and low chance locus of control.

Table 2.

Details of study respondents

RA OA
n 70 223
Females (%) 84.3 55.2
Mean age (yr) 62.7 68.2
Mean duration of disease (yr) 25.9 23.0
Mean (s.d.) SE Pain score 6.2 (1.9) 6.8 (2.1)
Median SE Pain score 6.2 6.8
Mean (s.d.) SE Function score 6.4 (2.3) 7.8 (2.3)
Median SE Function score 6.5 8.3
Mean (s.d.) SE Other Symptoms score 7.2 (2.0) 7.8 (2.0)
Median SE Other Symptoms score 7.5 8.3
Mean (s.d.) Internal MHLC score 22.3 (6.0) 25.4 (5.4)
Median Internal MHLC score 23.0 26.0
Mean (s.d.) Powerful Others MHLC score 23.2 (5.2) 22.5 (6.9)
Median Powerful Others MHLC score 23.0 23.0
Mean (s.d.) Chance MHLC score 16.5 (6.0) 17.2 (6.1)
Median Chance MHLC score 16.0 17.0
Mean (s.d.) HAQ score 1.3 (0.7) 0.8 (0.6)
Mean (s.d.) WOMAC Pain score 5.6 (3.9) 4.8 (4.0)
Mean (s.d.) WOMAC Stiffness score 2.8 (1.7) 2.6 (1.8)
Mean (s.d.) WOMAC Function score 23.1 (13.6) 19.5 (14.3)
Mean (s.d.) SF-36 PCS 32.8 (10.9) 37.6 (11.0)
Mean (s.d.) SF-36 MCS 51.9 (10.7) 52.3 (10.0)
RA OA
n 70 223
Females (%) 84.3 55.2
Mean age (yr) 62.7 68.2
Mean duration of disease (yr) 25.9 23.0
Mean (s.d.) SE Pain score 6.2 (1.9) 6.8 (2.1)
Median SE Pain score 6.2 6.8
Mean (s.d.) SE Function score 6.4 (2.3) 7.8 (2.3)
Median SE Function score 6.5 8.3
Mean (s.d.) SE Other Symptoms score 7.2 (2.0) 7.8 (2.0)
Median SE Other Symptoms score 7.5 8.3
Mean (s.d.) Internal MHLC score 22.3 (6.0) 25.4 (5.4)
Median Internal MHLC score 23.0 26.0
Mean (s.d.) Powerful Others MHLC score 23.2 (5.2) 22.5 (6.9)
Median Powerful Others MHLC score 23.0 23.0
Mean (s.d.) Chance MHLC score 16.5 (6.0) 17.2 (6.1)
Median Chance MHLC score 16.0 17.0
Mean (s.d.) HAQ score 1.3 (0.7) 0.8 (0.6)
Mean (s.d.) WOMAC Pain score 5.6 (3.9) 4.8 (4.0)
Mean (s.d.) WOMAC Stiffness score 2.8 (1.7) 2.6 (1.8)
Mean (s.d.) WOMAC Function score 23.1 (13.6) 19.5 (14.3)
Mean (s.d.) SF-36 PCS 32.8 (10.9) 37.6 (11.0)
Mean (s.d.) SF-36 MCS 51.9 (10.7) 52.3 (10.0)

Self-efficacy (SE) scores vary from 1 to 10; Multidimensional Health Locus of Control (MHLC) scores vary from 6 to 36.

Rheumatoid arthritis: correlations between health status, self-efficacy, locus of control and arthritis-related expenditure

Table 3 shows the correlations between health status, self-efficacy and locus of control for the RA group. Higher SE was significantly associated with all health outcome measures, particularly for Function self-efficacy and level of disability, indicating that those with higher self-efficacy had a better health status. Health locus of control was not as strongly associated with the health status measures, but a higher internal locus of control was associated with less disability on HAQ, SF-36 PCS, SF-36 MCS and WOMAC Pain and Function.

Table 3.

RA health status, self-efficacy and locus of control: correlations

Health status measures

WOMAC

SF-36

Pain Stiffness Function HAQ PCS MCS
Self-efficacy (ASES)
Pain −0.45b −0.39b −0.39b −0.15 0.38b 0.42b
Function −0.54b −0.37b −0.73b −0.82b 0.61b 0.35b
Other Symptoms −0.51b −0.47b −0.52b −0.43b 0.51b 0.56b
Locus of control (MHLC)
Internal −0.28a −0.11 −0.33b −0.36a 0.47b 0.31a
Powerful Others 0.10 0.20 0.07 −0.18 0.07 −0.12
Chance −0.06 0.11 0.00 −0.23 0.08 −0.17
Health status measures

WOMAC

SF-36

Pain Stiffness Function HAQ PCS MCS
Self-efficacy (ASES)
Pain −0.45b −0.39b −0.39b −0.15 0.38b 0.42b
Function −0.54b −0.37b −0.73b −0.82b 0.61b 0.35b
Other Symptoms −0.51b −0.47b −0.52b −0.43b 0.51b 0.56b
Locus of control (MHLC)
Internal −0.28a −0.11 −0.33b −0.36a 0.47b 0.31a
Powerful Others 0.10 0.20 0.07 −0.18 0.07 −0.12
Chance −0.06 0.11 0.00 −0.23 0.08 −0.17

aSignificant at the 0.05 level; bsignificant at the 0.01 level. WOMAC and HAQ: higher score indicates worse health status; SF-36: higher score indicates better health status. ASES, Arthritis Self-Efficacy Scale MHLC, Multidimensional Health Locus of Control.

Twenty-four RA respondents scored above the median on both Internal LOC and Function self-efficacy, indicating they have both a belief in their ability to perform activities and a belief that they themselves are responsible for their health. Fifteen respondents scored below the median on both these scales. Comparing the health status of these two groups showed that those who scored above the median on both function self-efficacy and internal locus of control reported significantly better PCS [mean 38.7 (s.d. 10.6) vs 24.6 (9.2), P<0.001], MCS [55.2 (7.7) vs 43.7 (11.9), P = 0.001], WOMAC Function [14.9 (8.7) vs 32.7 (12.8), P<0.001], WOMAC Pain [4.1 (3.1) vs 7.6 (3.4), P = 0.003], WOMAC Stiffness [2.2 (1.4) vs 3.6 (1.3), P = 0.007] and less HAQ disability [0.8 (0.4) vs 1.7 (0.8), P = 0.001] than those who scored below the median on both scales.

Cost data were available for 61 (87%) RA respondents. There were no significant differences in age, gender, health status, locus of control or self-efficacy between those who provided cost data and those who did not (data not shown). The mean number of visits to a general practitioner over the 3-month period was 2.1, varying from 0 to 10. The total cost to both the patient and the health-care system varied from AU$10.00 to AU$8450, with a mean of AU$1056. Mean patient out-of-pocket cost over the 3-month period was AU$560. A significant correlation was seen between function self-efficacy and arthritis-related costs, indicating that those with higher self-efficacy reported lower combined cost to both themselves and the health-care system (Table 4). Health locus of control was not significantly correlated with arthritis-related expenditure in this group. There was no significant difference in out-of-pocket cost to respondents or in overall cost to the respondent plus the health system between those above and below the median for internal locus of control and function self-efficacy as described above (mean total cost was AU$842 for those above the median and AU$826 for those below the median).

Table 4.

RA correlations between self-efficacy, locus of control and 3-month arthritis-related expenditure

No. of visits to GP Out-of-pocket cost Total cost to patient and health system
Self-efficacy (ASES)
Pain 0.04 −0.00 −0.01
Function −0.22 −0.25 −0.30a
Other Symptoms −0.11 −0.07 −0.04
Locus of control (MHLC)
Internal −0.18 0.01 0.01
Powerful Others −0.10 −0.04 −0.04
Chance −0.22 −0.16 −0.16
No. of visits to GP Out-of-pocket cost Total cost to patient and health system
Self-efficacy (ASES)
Pain 0.04 −0.00 −0.01
Function −0.22 −0.25 −0.30a
Other Symptoms −0.11 −0.07 −0.04
Locus of control (MHLC)
Internal −0.18 0.01 0.01
Powerful Others −0.10 −0.04 −0.04
Chance −0.22 −0.16 −0.16

aSignificant at the 0.05 level. ASES, Arthritis Self-Efficacy Scale MHLC, Multidimensional Health Locus of Control.

Osteoarthritis: correlations between health status, self-efficacy, locus of control and arthritis-related expenditure

Table 5 shows the correlations between health status, self-efficacy and locus of control for the OA group. As seen in the RA group, higher self-efficacy was significantly correlated with all health outcome measures. Health locus of control had more of an influence amongst the OA group, suggesting that those with higher external locus of control reported more pain and worse function, as shown by the negative correlations with the SF-36 scores and positive correlations with the WOMAC scores (Table 5). Similarly, the tendency was for those with higher internal locus of control to report less pain and better function, as shown by positive correlations with SF-36 scores and negative correlations with the WOMAC scores.

Table 5.

OA health status, self-efficacy and locus of control: correlations

Health status measures

WOMAC

SF-36

Pain Stiffness Function HAQ PCS MCS
ASES Self-efficacy
Pain −0.66b −0.58b −0.66b −0.66b 0.61b 0.49b
Function −0.59b −0.52b −0.68b −0.82b 0.66b 0.29b
Other Symptoms −0.69b −0.54b −0.71b −0.66b 0.60b 0.59b
MHLC (locus of control)
Internal −0.15a −0.08 −0.15a −0.20 0.31b 0.18b
Powerful Others 0.11 0.08 0.14a 0.30 −0.23b −0.09
Chance 0.21b 0.13 0.20b 0.30 −0.16a −0.21b
Health status measures

WOMAC

SF-36

Pain Stiffness Function HAQ PCS MCS
ASES Self-efficacy
Pain −0.66b −0.58b −0.66b −0.66b 0.61b 0.49b
Function −0.59b −0.52b −0.68b −0.82b 0.66b 0.29b
Other Symptoms −0.69b −0.54b −0.71b −0.66b 0.60b 0.59b
MHLC (locus of control)
Internal −0.15a −0.08 −0.15a −0.20 0.31b 0.18b
Powerful Others 0.11 0.08 0.14a 0.30 −0.23b −0.09
Chance 0.21b 0.13 0.20b 0.30 −0.16a −0.21b

aSignificant at the 0.05 level; bsignificant at the 0.01 level. WOMAC and HAQ: higher score indicates worse health status; SF-36: higher score indicates better health status. ASES, Arthritis Self-Efficacy Scale MHLC, Multidimensional Health Locus of Control.

Ninety-seven OA respondents scored above the median on both internal locus of control and function self-efficacy, indicating that they have both a belief in their ability to perform activities and a belief that they themselves are responsible for their health. Twenty-six respondents scored below the median on both these scales. The comparison of health status of these two groups showed that those who scored above the median on both function self-efficacy and internal locus of control reported significantly better PCS [mean 44.8 (s.d. 9.1) vs 27.8 (7.9), P<0.001], MCS [55.1 (8.2) vs 49.3 (12.2), P = 0.005], WOMAC Function [10.4 (8.9) vs 29.6 (14.1), P<0.001] and WOMAC Pain [2.6 (2.5) vs 7.0 (4.7), P<0.001], WOMAC Stiffness [1.8 (1.4) vs 3.5 (1.8), P<0.001] and less HAQ disability [0.2 (0.2) vs 1.8 (0.8), P<0.001] than those who scored below the median on both scales.

Cost data were available for 163 (73%) OA respondents. There were no significant differences in age, gender, health status, locus of control or self-efficacy between those who provided costs data and those who did not (data not shown). While no significant difference was detected between these two groups, it should be noted that the power to detect differences may be diminished due to 27% of those eligible not providing cost data. The mean number of visits to a general practitioner over the 3-month period was 0.72, varying from 0 to 12 visits. The total cost to both the patient and the health-care system varied from AU$0 to AU$3493, with a mean of AU$317. Mean patient out-of-pocket cost over the 3-month period was AU$190. As shown in Table 6, a significant correlation was seen between self-efficacy for pain, function and other symptoms and number of visits to a general practitioner in the 3-month period, indicating that those with higher self-efficacy were less likely to visit their general practitioner. Self-efficacy for pain and function were both correlated with the combined arthritis-related cost to the patient, MBS and PBS: those with lower self-efficacy experienced higher cost. There was no detectable difference in out-of-pocket cost to respondents or to overall cost to the respondent plus the health system between those above and below the median for internal locus of control and function self-efficacy, as described above (mean total cost was AU$265 for those above the median and AU$356 for those below the median).

Table 6.

OA correlations between self-efficacy, locus of control and 3-month arthritis-related expenditure

No. of visits to GP Out-of-pocket cost Total cost to patient and health system
Self-efficacy (ASES)
Pain −0.32b −0.21a −0.24b
Function −0.29b −0.11 −0.16a
Other Symptoms −0.26b −0.12 −0.14
Locus of control (MHLC)
Internal −0.08 0.00 −0.02
Powerful Others −0.01 −0.01 −0.02
Chance 0.22b 0.15 0.11
No. of visits to GP Out-of-pocket cost Total cost to patient and health system
Self-efficacy (ASES)
Pain −0.32b −0.21a −0.24b
Function −0.29b −0.11 −0.16a
Other Symptoms −0.26b −0.12 −0.14
Locus of control (MHLC)
Internal −0.08 0.00 −0.02
Powerful Others −0.01 −0.01 −0.02
Chance 0.22b 0.15 0.11

aSignificant at the 0.05 level; bsignificant at the 0.01 level. MHLC, Multidimensional Health Locus of Control.

Discussion

This study investigated self-efficacy and health locus of control amongst a group of community-dwelling people with OA and RA who were not undergoing any particular treatment or programme to enhance self-efficacy. Locus of control showed less of an influence on health status than self-efficacy in this community-dwelling sample. While self-efficacy is the dominant construct, when it is used in combination with health locus of control a greater understanding of the influence of psychological factors on health status may be ascertained.

For the OA respondents, lower self-efficacy was associated with higher overall expenditure and a greater number of visits to general practitioners, which would result in a greater cost to the Medicare system. Locus of control had less of an effect on expenditure. It has been reported previously that psychological factors such as self-efficacy and coping style can affect health status, health-care utilization and health-care costs [13]. For both RA and OA, respondents with a combination of high self-efficacy for function and high internal locus of control, i.e. those who believe they are both responsible for their own health and have the capacity to perform behaviours to maintain health, did not have significantly different overall expenditure than those with low self-efficacy and external locus of control despite better health status. This may be explained by the fact that, in general, respondents reported high scores on both the internal and powerful others scales, with much lower scores on the chance locus of control scale. As Wallston has suggested, it may not be the locus per se that is important, but the belief in some form of control, whether it comes from within or externally, may be important for health outcomes [14]. It may be that a belief in powerful others combined with internal control may be better for adhering to treatment regimens and result in better health.

Self-efficacy for function was strongly associated with the functional status measures, suggesting that those with higher self-efficacy also have less disability. However, as the correlations were high it may be postulated that the questions in the ASES are in fact measuring the same concepts as the WOMAC Index, HAQ or SF-36. The ASES specifically asks how certain respondents are that they can perform specific behaviours, but whether the respondents are reporting on their belief or actual ability we are not certain.

This study has some limitations in that it was cross-sectional, measuring self-efficacy, health locus of control and health status at one time point. No assumptions about causality can be made and no assessment of the impact of changes in self-efficacy or locus of control can be made. Patients were not asked whether they had participated in any programme designed to enhance self-efficacy. We did not measure the value of health and its ability to act as a reinforcer of behaviour, and this may account for locus of control not showing statistical significance in many of the analyses. Cost data were not available from all study participants. While the proportion of RA patients providing cost data was reasonably high at 87%, only 73% of OA patients provided details of costs. While there was no difference in several factors between these responders and non-responders, it may be that there was inadequate power to detect clinically important differences, and as a result bias may have been created in the cost estimates.

Conclusions

Higher self-efficacy, which is the more amenable to change, was associated with better health status in this cross-sectional study. While there was no difference in health expenditure, respondents with high internal health locus of control in combination with high self-efficacy reported better health status than those with low internal locus of control and low self-efficacy. As the measurement of these constructs is simple and the outcome potentially affects health status, these results have implications for future intervention studies aimed at improving the quality of life and reducing the financial impact of arthritis on both the health-care system and to patients themselves.

Acknowledgements are due to the members of the Arthritis COST Study project group: David Champion, Brett Courtenay, Myles Coolican, Mervyn Cross, Andrew Ellis, Michael Neil, Michael O'Sullivan, David Parker, Leo Pinczewski, J. Stephen Quain, Frank Robertson, Stephen Ruff, William Walter and Bernie Zicat.

The authors have declared no conflicts of interest.

References

1
Bandura A. Self-efficacy: toward a unifying theory of behaviour change.
Psychol Rev

1977
;
84
:
191
–215.
2
Lorig K, Chastain RL, Ung E, Shoor S, Holman HR. Development and evaluation of a scale to measure perceived self-efficacy in people with arthritis.
Arthritis Rheum

1989
;
32
:
37
–44.
3
Rotter JB. Generalized expectancies for internal versus external control of reinforcement.
Psychol Monogr

1966
;
80
:
1
–28.
4
Wallston KA, Wallston BS, de Vellis R. Development of the Multidimensional Health Locus of Control (MHLC) Scales.
Health Educ Monogr

1978
;
6
:
160
–70.
5
Waller KV, Bates RC. Health locus of control and self-efficacy beliefs in a healthy elderly sample.
Am J Health Promot

1992
;
6
:
302
–9.
6
Gustafsson M, Gaston-Johansson F. Pain intensity and health locus of control: a comparison of patients with fibromyalgia and rheumatoid arthritis.
Patient Educ Couns

1996
;
29
:
179
–88.
7
Crisson JE, Keefe FJ. The relationship of locus of control to pain coping strategies and psychological distress in chronic pain patients.
Pain

1988
;
35
:
147
–54.
8
Lapsley HM, March LM, Tribe KL, Cross MJ, Brooks PM. Living with osteoarthritis: patient expenditures, health status and social impact.
Arthritis Rheum

2001
;
45
:
301
–6.
9
Lapsley HM, March LM, Tribe KL, Cross MJ, Courtenay BG, Brooks PM. Arthritis Cost and Outcome Project Group. Living with rheumatoid arthritis: expenditures, health status, and social impact on patients.
Ann Rheum Dis

2002
;
61
:
818
–21.
10
Ware JE Jr, Snow KK, Kosinski M, Gandek B.
SF-36 Health survey manual and interpretation guide
. Boston, MA: Health Institute, New England Center,
1993
.
11
Bellamy N.
WOMAC Osteoarthritis Index: A user's guide
. London, Ont., Ontario, Canada, University of Western Ontario,
1995
.
12
Fries JF, Spitz PW, Kraines RG, Holman HR. Measurement of patient outcome in arthritis.
Arthritis Rheum

1980
;
23
:
137
–45.
13
Miller C, Cronan T. The effects of coping style and self-efficacy on health status and health care costs.
Anxiety, Stress and Coping

1998
;
11
:
311
–25.
14
Wallston KA. Hocus-pocus, the focus isn't strictly on locus: Rotter's social learning theory modified for health.
Cogn Ther Res

1992
;
16
:
183
–99.

Author notes

Institute of Bone and Joint Research, University of Sydney, Sydney, NSW, 1Centre of National Research on Disability and Rehabilitation Medicine and 2Health Sciences, University of Queensland, Brisbane Qld, Australia.