Health status in patients awaiting hip replacement for osteoarthritis

Background . Hip osteoarthritis is a major cause of pain and disability, especially in the elderly. As part of a study investigating factors that could be associated with advanced osteoarthritis of the hip, we compared the health status of patients awaiting arthroplasty for hip osteoarthritis with controls. We further investigated the interaction of hip osteoarthritis with other variables (age, gender, social class and concurrent pain) in relation to health status. Methods . A case–control study was performed in two English health districts (Portsmouth and North Staffordshire) during 1993–1995. A total of 611 patients (210 men and 401 women) listed for hip replacement because of osteoarthritis over an 18-month period formed the case group and were compared with an equal number of controls selected from the general population and individually matched for age, gender and general practice. Cases and controls completed a structured interviewer-administered questionnaire, which included queries about their medical condition, occupation (from which a measure of social class was derived), and general health status using the SF36. Results . Physical function ( t = 32.1, P < 0.001), social function ( t = 16.8, P < 0.001) and perceived general health ( t = 4.1, P < 0.001) were worse in the case group, but energy u vitality and mental health showed little difference between cases and controls. Cases were more likely to report knee pain than controls, but case–control status was not associated with pain in the ﬁngers or shoulders, or with social class. However, differences in physical and social function between cases and controls did vary with socio-demographic factors and concurrent knee pain status. Conclusion . Patients awaiting hip-replacement because of osteoarthritis were more likely to be restricted in their physical and social life than adults in the general population, but mental state and vitality appear unimpaired in this group. This contrasts with ﬁndings from other chronic pain disorders. Manual social class is not linked to being on a waiting list for osteoarthritic hip replacement but does add to the burden on health status, particularly social functioning in those with osteoarthritis of the hip.

Osteoarthritis is a clinical syndrome of joint pain and stiffness, accompanied by radiological changes of cartilage loss and changes in bone w1x. Its prevalence rises with age and its impact is determined by the extent of the disability that it causes, as well as by the severity of associated pain. Osteoarthritis of the hip joints is a major cause of restricted locomotor activity and functional disability w2x. It also has associated psychosocial implications, elderly subjects with chronic hip pain having a significantly lower quality of life w3, 4x. In addition older people are more likely to have multiple health problems, and psychological distress and physical dysfunctioning is greater in the context of such comorbidity w5x.
Several studies of patients who have had arthroplasties have focused on the subsequent improvement in quality-of-life, particularly reduction in pain and improvement in physical and social functioning, but mental health has shown little change w6-8x. Furthermore, a recent UK study found that mental health, perceived general health and vitality were not significantly associated with hip pain among primary care consulters w9x.
People who live in socially deprived areas have more musculoskeletal symptoms, and experience a greater severity of hip and knee disease w10-12x. Although osteoarthritis has been studied in relation to different occupational groups, little is known about whether the impact of advanced lower limb osteoarthritis varies with social class.
As part of a case-control investigation of factors associated with advanced osteoarthritis of the hip, we examined the health status of patients awaiting arthroplasty for osteoarthritis of the hip, and compared this with an age-and sex-matched sample of population controls. Interactions with age, gender, social class and other musculoskeletal pain were explored.

Methods
The study area comprised two health districts in England (Portsmouth and North Staffordshire), with a combined population of around one million residents. These districts are each served by one secondary care orthopaedic centre and have a broad social and economic catchment population. All men and women, aged 45 yr and over, who were placed on the waiting list for primary total hip arthroplasty over an 18-month period were identified. After review of their medical records, patients were excluded if they had sustained a hip fracture within the past year; fulfilled ACR criteria for rheumatoid arthritis w13x or modified New York criteria for ankylosing spondylitis w14x or had a history of Perthes' disease, congenital hip dislocation, slipped capital epiphysis or other established causes of secondary osteoarthritis.
For each case, a control of the same sex and age (to within 4 yr) was selected from the list of the same general practice held by the local health authority. In the UK, most people are registered with a general practitioner, so such lists provide a convenient sampling frame of a local population. Controls who had undergone previous hip surgery for osteoarthritis were excluded, and controls who declined to participate were replaced.
After giving informed consent, cases and controls completed a structured interviewer-administered questionnaire, enquiring about their medical history, lifestyle and leisure time physical activities. In addition, and relevant to the purpose of this paper, they completed a reduced version of the Short Form 36 (SF36) as a measure of health status w15x. This included five of the original eight dimensions: physical functioning, social functioning, mental health, energyuvitality and general health perception. Questions on physical functioning asked about limitations to activities during a typical day: walking, running, sports, lifting, climbing stairs, bending, bathing and getting dressed. There were three additional items concerned with using the toilet, getting into and out of a car and sex life. Mental health was based on the five questions included in the SF36, addressing both anxiety and depression. Energyuvitality was based on the four items in the SF36, and enquired about energy levels and tiredness. Our question on social functioning was: 'Has your health limited your social activities (like visiting friends or close relatives)?'. The general health perception question was: 'In general, would you say your health is: excellent, very good, good, fair or poor?'. Rating scales were the same as those employed in the SF36, and our methods were the same as those detailed by Jenkinson et al. w16x. Each measurement scale ranges between 0 and 100, with 0 representing worst health or perception score for that dimension and 100 indicating the best score, i.e. the higher the score, the better the health status.
Participants were asked to give details of their most recent occupation (prior to retirement in the case of those who had already retired) from which a measure of social class was derived, using the classification procedure of the Office of National Statistics w17x. Other musculoskeletal pain complaints were represented by self-reported knee, hand and shoulder pain in the previous month.
In total, 726 cases and 1060 controls were approached. The analysis presented here focuses on 611 case-control pairs with complete questionnaire information. Mean scores of the different SF36 dimensions were compared between cases and controls, categorized separately by age group and gender and then by social class and concurrent pain complaints. Multivariate analysis of the association with concurrent pain was carried out using conditional logistic regression to compare cases and controls through estimates of odds ratios (OR) with 95% confidence intervals (CI) adjusting for BMI (weight measured by scales, height by portable stadiometer), occupational lifting and physical activity (from structured interviewer-administered questions) to take account of potential confounders. Data analysis was carried out using SPSS version 10.0 w18x, and STATA version 6.0 for multivariate analysis w19x.

Results
The median age for the case group was 70.9 yr winterquartile range (IQR) 63.8-76.4x and 71.1 yr (IQR 64.1-76.8) for the control group. Of the 611 pairs, 401 (65.6%) were women and 210 (34.4%) were men. Of those with a categorized social class, in the case group 298 (49.7%) were in the manual group and 302 (50.3%) in the non-manual group, compared with 305 (50.7%) manual and 296 (49.3%) non-manual classes in the control group.
The scores for the five dimensions of the SF36 are shown in Table 1, for cases and controls separately. The physical function score was substantially lower in cases compared with controls, as was the measure of social functioning. Perception of general health was also lower in cases, but the difference was less marked than for the function measures. Energyuvitality and mental health scores showed little difference between cases and controls.
Average scores for the five dimensions of the SF36 are shown stratified by age and gender in Table 2. As expected, physical function, social function and general health perception decreased with age in both cases and controls. Women had worse physical function, social function and general health perception than men, except in controls aged under 65 yr. Relative differences between cases and controls were similar regardless of age in men, but decreased with rising age in women. Mental health and energyuvitality scores were similar across ageugender bands and showed little variation with case or control status.
Overall, there was no evidence of a link between social class and case-control status: the OR for the association between hip osteoarthritis and manual as compared with non-manual work was 0.99 (95% CI=0.77, 1.27). Within both case and control groups, there were lower SF36 scores in the manual compared with non-manual groups for physical functioning, social functioning and general health perception, but not for energyuvitality or mental health (Table 3). The differences in SF36 scores between case and control groups were of similar magnitude in the manual and non-manual groups except with respect to social functioning where there was a wider gap between cases and controls in the manual group than in the non-manual group.
We also investigated the association of symptoms at sites other than the hip with case-control status ( Table 4). Pain in the shoulders and hands was not associated with hip osteoarthritis. In contrast, cases were significantly more likely than controls to report knee pain, even after adjustment for BMI, occupational lifting and physical activity. A similar association was also observed for stiffness of the knee, and to a lesser extent a recalled history of previous knee injury.
In Table 5 the mean scores for the different dimensions of the SF36 in cases and controls are compared separately in those who reported knee pain and those who stated they had no knee pain. In general, the presence of knee pain was associated with lower physical function and social functioning scores in both cases and controls. There was evidence of some association with general health perception also, but not with scores for mental health and energyuvitality. Significant differences between case and control groups for physical function and social function persisted after stratification by knee pain. There was no association between hip osteoarthritis and general health perception after adjustment for self-reported knee pain status. Thus, concurrent pain in the knee appeared to influence the relationship between osteoarthritis of the hip and general health perception.   Scores for each dimension are based on a scale of 0-100: 0 indicates worst perceived health and 100 indicates best perceived health.

Discussion
We found that both physical and social functioning were substantially restricted in patients awaiting hip replacement compared with similarly aged controls selected from the general population. However, such patients' own perceptions of their health status showed much less contrast between those with and without hip osteoarthritis than did the actual function scores. Mental health, energy and vitality seemed unaffected by the presence of advanced hip osteoarthritis.
Case series of patients awaiting hip and knee replacement have made similar observations. Jones et al. w6x, for example, noted that bodily pain and physical function scores in patients listed for hip or knee replacement were lower than population normal values, whereas scores on other dimensions were similar to the norms for the relevant age group. Furthermore, there is evidence that the general decline in physical health with advancing age is independent of mental health, which by contrast remains relatively stable w20, 21x. These studies all used the SF36, and it may be that the mental health dimension of this instrument is insensitive to psychological problems in older people or in those with chronic physical illness. However, studies of other chronic illnesses suggest that this is not the explanation. Reduction in mental health status scores using the SF36 have been observed in studies of hospital attenders with, for example, type 2 diabetes w22x, coronary heart disease w23x and chronic renal failure w24x. In the coronary heart disease study w23x the SF36 performed in a similar fashion to the GHQ28, an instrument designed more specifically to identify symptoms of anxiety and depression w25x. Chronic illness in the general population is also associated more generally with lower mental health scores on the SF36 w26x.
The apparent well being of those awaiting hip replacement may have a number of explanations. It Mantel-Haenszel estimate of the OR for the association between concurrent pain and hip osteoarthritis, where absence of concurrent pain is the reference category. c OR obtained by multivariate conditional logistic regression, adjusted for the other concurrent pain variables listed in the table (and excluding knee stiffness and knee injury), BMI, occupational lifting and exercise. Statistical analysis of mean paired-differences was based on stratification of the 611 case-control pairs by social class (manual or non-manual). Difference mean scores were based on the 332 pairs that were matched for social class as well as age and gender: 165 non-manual pairs and 167 manual pairs. may be that the more alert, healthier patients are being selected for operation, particularly among the older age groups. However, the controls in our study would have been selected in comparable ways-they had to be well enough and alert enough to agree to interview. The greater concern would be that selection bias would work in the other direction, i.e. that we had a healthier group of controls than the pool of patients from which our waiting list cases came. Our waiting list patients are highly likely to represent all those waiting for total hip replacement-most were recruited; the hospital centres drew most of the patients in the districts studied; private patients were included. A different sort of selection bias might relate to waiting list status: it may be that, once people with hip osteoarthritis are on a waiting list, aspects of their health status improve (such as mental health). There is evidence from elsewhere that optimism in older patients is associated with higher scores on mental health, vitality and general health perception w27x. However, this does not affect the general conclusion from our data that patients with advanced hip osteoarthritis, for which an operation is to be performed, have a clear reduction in physical and social functioning, but that mental health, general energy and vitality, and to some extent their own perception of their general health, appear unaffected by the condition.
This result is in direct contrast to observations on other chronic musculoskeletal pain problems. Studies of patients with chronic back pain, shoulder pain, neck pain and widespread pain, both in the general population and in clinics, all emphasize a major association of these disorders with mental health, fatigue and general health status. A specific difference is that these latter conditions affect people of working age predominantly, and issues of employment and job loss may be added factors that are less important post-retirement. However, there was no evidence in our study of a differential impact on mental health or vitality pre-and post-retirement age. It may be important that the hip osteoarthritis was, by definition, soon to be specifically treated, whereas chronic back pain and fibromyalgia lack specific cures. However, similar findings to those in this study have been reported from a cohort of new consulters with hip pain in primary care w9x. Alternatively, it may be that osteoarthritis of the large weight-bearing joints in older people presents a different 'biopsychosocial' mixture to chronic pain elsewhere.
The association of hip osteoarthritis with knee pain was not unexpected: clinically hip disease can present with knee pain, the presence of hip disease leading to gait alterations and consequent knee pain. Radiographic studies of osteoarthritis have shown an association between osteoarthritis at the hip and osteoarthritis at the knee and the hand w28x. However, in our patient group there were only weak links with hand pain, and a similar lack of association with shoulder pain. Once again this contrasts with the 'regional musculoskeletal pain syndromes' of young and middle-aged adults (back, shoulder, neck), where concurrence of pain at different sites is consistently reported w29x. It seems that with hip osteoarthritis we are dealing with a different type of problem compared to axial, upper-limb and generalized pain syndromes. There have been previous reported associations of BMI, occupational lifting and physical activity with hip osteoarthritis and knee osteoarthritis w30-32x. However, these factors did not explain the link between hip osteoarthritis and knee pain in our study.
There was no evidence of an association between social class and waiting list status for hip osteoarthritis Statistical analysis of mean paired-differences was based on stratification of the 611 case-control pairs by presence or absence of knee pain. Difference mean scores were based on the 238 pairs matched with respect to knee pain status as well as age and gender: 99 no knee pain pairs and 139 knee pain pairs. surgery. However, social class was linked with health status, and hence cases from semi-or unskilled manual social classes awaiting hip replacement had lower mean function scores than those from higher social classes. Indeed, when average scores were compared across age, gender and social class bands, the absolute function scores of two adjacent people on a waiting list could be considerably different. Men aged under 65 yr in social class 1 or 2 awaiting total hip replacement were estimated as having a mean physical function score of 41.0, whereas females aged over 75 yr in social class IV or V had a mean physical function score of 20.6. Given that many of these differences also apply to people without hip disease who are in different age, gender and social class groups, such contrasts are to be expected. However, the differences between cases and controls, with respect to social functioning, were greater than expected on the basis of differences observed in social class. We are currently planning a prospective study to investigate the extent to which social functioning changes after total hip replacement in this cohort.