Osteoarthritis: quality of life, comorbidities, medication and health service utilization assessed in a large sample of primary care patients

Osteoarthritis is one of the most prevalent chronic diseases worldwide and is associated with substantial impact on patients' individual quality of life as well as on healthcare costs. Its prevalence is expected to rise significantly in the upcoming decades. Increasing life expectancy and decreasing physical activity, leading to a constant increase in body weight, are regarded as underlying determinants of this development. Facing this situation, the WHO and the United Nations have declared the years 2000 to 2010 to be the "Bone and Joint Decade" [1]. Since in the year 2050 more than 50% of the population will be over 50 years of age, the German health care system will be hit tremendously by chronic illnesses like osteoarthritis [2]. Most of these individuals will receive medical treatment in primary care settings, accounting for the growing number of studies dealing with OA in primary care [3‐5]. However, to date, relatively little is known about osteoarthritis symptoms and their medical treatment in various subgroups of patients in primary care.

Previous studies have focused on the prevalence and prognosis of OA [4, 6]. Regarding prevalence, it is a frequently replicated result that women have a higher probability for developing OA, especially OA of the knee [7, 8]. Several studies have suggested that not only prevalence but also the disease process is related to gender: women were found to have more severe structural progression and a higher need to undergo surgical interventions than men [6]. Other studies suggested that women with OA suffer from pain and disability to a greater extent compared to men and also that these dimensions of QoL are strongly associated with the social situation [9‐11]. However, it remains unclear how these findings can be explained.

The present study was performed to get a comprehensive overview of the health status and the healthcare received by primary care patients with OA in Germany. We particularly focused on differences related to gender because we hypothesized that men and women differ regarding health status and health service utilization (HSU). Furthermore, since it is known that quality of life (QoL) of OA patients is mainly determined by pain and disability our aim was to assess factors that are associated with these two dimensions of QoL [8, 12].

The data used for this study are retrieved from the baseline assessment of the PraxArt project, which is financed by the German Ministry for Education and Research over a period of 6 years, starting in 2003. The aim is to assess the status of OA care and to search for possibilities to improve care as well as patients' quality of life by tailored interventions. A randomly created sample of 75 general practitioners in the area of Baden-Wuerttemberg and Bavaria has been enrolled and recruited the patients for this survey.

In total, 1311 patients were addressed by the GPs. 1250 of them agreed to complete the questionnaire. 1021 of the 1250 (81.7%) patients returned the questionnaires, corresponding to at least 11 questionnaires in each practice. Regarding available data, including sociodemographic variables, comorbidities and medication, no statistically significant differences could be revealed between the non-respondents and the respondents. The main reason given for not participating was time effort. Among the enrolled patients, 347 (34.0%) were male and 674 (66.0%) were female. If missing data occurred, they mainly occurred within the same questionnaire, in total in 271 of the 1021 questionnaires. In 123 cases the data could be completed from the patient file.

278 (80.1%) men were married or lived with a partner. 376 (55.8%) women were engaged. This difference was significant (p < 0.01). Completely retired from work were 233 (67.1%) men and 482 (71.5%) women. T-test for group comparison revealed a significant difference in the (formal) educational level between men (mean 2.61, SD 1.1) and women (mean 2.38, SD 0.83). BMI, age, number of comorbidities or disease duration did not differ significantly. Table 1 displays the characteristics of the study sample separated by localization of OA.

Based on previous findings indicating biological differences (e.g. regarding the destruction of the cartilage) and psychological differences (e.g. perception of pain) we hypothesized that men and women differ regarding many aspects of QoL and received care. This hypothesis could be confirmed: OA has higher impact on women in important aspects of QoL such as pain, disability and mood. Similar gender differences have been found e.g. by Woo et al. among Chinese people [23]. They received more NSAIDs and visited their GP but not their specialist more frequently than men and tended to have less intraarticular injections. Interestingly, minor or major depressive episodes were not more frequent among women, even though the affect scale of the AIMS2-SF indicated lower mood among women.

Regarding QoL, we found lower scores than Sany et al. did in a sample of rheumatoid patients regarding physical limitation. However, we observed nearly the same mean scores regarding the symptom scale. This finding may indicate that patients suffering from OA are less limited in their mobility but appear to suffer from equivalent pain intensity than patients with rheumatoid arthritis (RA). With regard to comorbidities which have an important impact on the QoL of patients suffering from osteoarthritis as well as on the outcome of surgical interventions, gender differences occurred only regarding high blood pressure [24‐27]. Unfortunately, reliable data regarding comorbidities in OA patients are difficult to compare since different comorbid conditions have been assessed with different methods (e.g. self reports) in previous studies. Groessl et al. who enrolled 363 OA patients in a primary care setting in a health management organization (HMO) in the United States reported on somewhat lower rates of HBP (28.8 %), which was the commonest comorbidity in their sample. Similar numbers were found by Nilsdotter et al. [26]. Compared to national data, the prevalence of HBP in Germany in this age group is expected to be over 55%, as was found in a large international comparison [28]. However, a limitation of our findings is that no control group was available.

Regarding pain medication, Paracetamol, which is the first choice treatment according to most guidelines, was only marginally prescribed. The main pillar in pharmacological treatment are NSAIDs such as Diclofenac [29‐31]. This is in accordance with the fact that NSAIDs are known to be increasingly used worldwide [32]. Interestingly, COX-2-inhibitors played no important role in prescriptions. Our data also confirmed previous findings showing that the use of NSAIDs is more frequent among women than men [33]. In the study of Linsell et al. 45.9% of OA patients stated to take pain killers frequently, which is comparable to our results[3].

Regarding HSU, our data indicated a high HSU by OA patients. However, it has to be noted that the German health care system is characterized by a high physician contact-rate. The number of mean contacts per year and person in Germany, including all contacts to GPs and specialists, is 6.6 [34]. In Germany patients have free access to secondary care, a referral is not required [35]. Thus, the revealed high amount of x-rays for example may also be due to the unlimited accessibility of health care in Germany [36]. The reason why women visited their GP more often than men could be related to the higher pain scores of women, since it is known that pain is a strong predictor for HSU among OA patients [27]. Interestingly, gender differences could only be revealed regarding contacts to GPs but not to specialists. An important weakness of the presented data is that, even though the analyses were adjusted for important covariates, HSU may have been related to other reasons except arthritis, even though we asked patients to mention only contacts which were related to OA. It should also be mentioned that we did not control our data for patients' insurance. About 10% of patients are "privately" insured, resulting in higher reimbursement for physicians. This may have influence on treatments, prescriptions as well as on referral rates. Our data regarding HSU reflect a finding that may be ignored by many physicians: the important role of complementary alternative medicine (CAM) for patients with OA. As Rao et al. could show the use of CAM is very common among patients with RA. Our data regarding visits to healers and received acupuncture are lower in comparison to the findings of Rao, who reported a frequent use of up to 90% among the RA patients, but at least more than a quarter of our patients reported on current use of CAM [37]. According to Rao, only half of the patients discuss the use of CAM with their physician, so they should be aware of this issue and address it in order to avoid treatment conflicts or side effects. Interestingly, comparable findings regarding CAM have been reported by Linsell et al. in a sample of OA patients in the UK [3]. Many studies have assessed depression in patients with rheumatoid arthritis, some of which indicate a higher risk among patients with RA than OA patients [38, 39]. None of them enrolled as much OA patients as we did. The importance of depression for OA patients is related to the fact that it is an important predictor for functional disability and an independent risk factor for mortality in RA [40]. Previous findings regarding the prevalence among OA patients indicated no increased prevalence [41, 42]. Our data showed that 19.7% of women and 18.9% of men fulfilled the criteria for a major or minor depressive episode. Data regarding the point prevalence among the German population vary between 5–10 % in the general population [19]. In contrast to the general population, no gender differences could be revealed in our study sample. Our findings indicate a significant increase in the point prevalence but the numbers of about 30% reported for RA patients were not met [43].

Pain and disability have often been shown as the major burden of OA. Similar as in various other studies, women achieved higher scores regarding both symptoms of OA [23]. But interestingly, no gender differences could be revealed regarding their predictors.

Despite the fact that our study has certain limitations and acknowledging the characteristics of the German health care system with e.g. a large number of non-surgical orthopedics, the study gives a comprehensive overview. However, because of the wide range of aspects addressed in this paper, it is not possible to describe the findings in detail e.g. in the sense of revealing predictors for each variable. The study represents the largest assessment of OA patients in a primary care setting in Germany.

Our findings regarding QoL and the burden of the disease suggest that OA patients differ from patients suffering from other forms of arthritis, especially RA. Our findings suggest that the impact of OA on men and women differs. Even we could not prove causality we assume that this may be have lead to the revealed differences in the pharmacological treatment and the use of the health care system. Further research is needed to confirm our results and assess causality.