To assess the prevalence of unmet health care demands among rheumatoid arthritis (RA) patients, and to determine if these unmet demands indicate underuse.
To assess the prevalence of unmet health care demands among rheumatoid arthritis (RA) patients, and to determine if these unmet demands indicate underuse.
A total of 679 patients with RA participated in a questionnaire survey and clinical examination. Unmet health care demands and health care use were assessed for orthopedic care, allied health care, home care, and psychosocial care. Indications for underuse were determined by comparing health outcomes of patients with unmet health care demands and of health care users.
Of the 679 patients, 28.7% had an unmet demand for 1 of the 4 services: 13.4% for allied health care, 9.7% for orthopedic care, 9.4% for home care, and 6.2% for psychosocial care. Underuse of allied health care, home care and psychosocial care was observed.
Unmet demands for health care are frequent among RA patients. Most unmet demands indicate underuse. Health care professionals should therefore be more responsive to the demands of patients.
Underuse of health care is a widely acknowledged problem (1–4) because it provokes that treatment and care are less effective than possible. In assessing underuse of health care, 2 perspectives may be distinguished: the perspective of professionals and of patients (5). The professional perspective is usually reflected in guidelines. The perspective of patients is related to personal views of patients, their health care expectations and experiences (5).
Underuse of health care has been studied frequently among patients with chronic diseases, and underuse of health care services has been identified from the professional perspective in asthma (6, 7), cardiovascular diseases (8–10), depression (11), diabetes (12), gastrointestinal disease (13), and stroke (14), and from the patient perspective in cancer (15), dementia (16), human immunodeficiency virus (HIV) infection (17), stroke (18), and rheumatoid arthritis (RA) (19, 20). The results of these studies show that underuse of health care is a widespread problem among patients with chronic diseases.
The previous studies regarding underuse of health care in RA focused on single disciplines, namely on allied health care (19), and on home care (20). However, patients with RA rely on multiple health care disciplines simultaneously (21).
Because patients with RA are higher users of health care when compared with the general population (22, 23), and with patients with other chronic conditions (24, 25), it is important to investigate underuse of multiple health care services concurrently.
Studying underuse of health care among RA patients from the professional perspective is hampered because the professional guidelines that have been developed for RA (26–28), mainly focus on pharmacologic therapy rather than on referral criteria for the use of health care services. Furthermore, several researchers have stressed the importance of assessing the patients' perspective in health care by stating that professionals are unable to recognize aspects of the disease that are important to patients (29, 30). Moreover, the focus on the perspective of patients is necessary, enabling treatment and interventions to comply with patient-centered care.
In this study we therefore focused on the perspective of patients, aiming to assess underuse of health care among patients with RA by studying unmet demands for several health care services. First, we investigated the prevalence of unmet demands for health care among patients with RA. Second, we assessed if the unmet demands of patients indicate underuse of health care by comparing the health outcomes of non-users of health care who expressed an unmet demand with the health outcomes of health care users and of non-users of health care without unmet demands.
For this study, patients were asked to fill in a postal questionnaire, and were invited to have a clinical examination. Information on disease duration was abstracted from the patients' files.
In 1997, 882 patients diagnosed with RA took part in a longitudinal study on health and health care. These patients originated from a group of 1,200 patients, who were selected from the database of a rheumatology outpatient clinic in Amsterdam, which also included patients from affiliated outpatient clinics. Inclusion criteria for the study were being ≥16 years old, meeting the 1987 revised American College of Rheumatology (formerly, the American Rheumatism Association) criteria for RA (31), having received rheumatology care in the 2 years previous to patient selection, and being able to read and understand the Dutch language. Eligible patients were asked to participate and to sign a letter of informed consent. The medical ethics committee approved the study design. Baseline respondents were younger than those who declined participation (mean ± SD 58.3 ± 15.8 years versus 63.6 ± 15.6 years; P = 0.002), but no differences were found in sex, educational level, disease duration, or disability (32).
The present study was performed in 1999 among the respondents of the baseline study. Of these respondents, 41 patients were lost to follow-up, (30 died and 11 relocated to an unknown address). These drop-outs were significantly older (P = 0.001) than the other respondents (n = 841), but no differences were found regarding sex or educational level.
Data were collected on actual health care use, unmet health care demands, health outcomes, and sociodemographic characteristics.
Actual health care use in the preceding 12 months was determined for the following services: orthopedic care, physiotherapy, occupational therapy, chiropody, home nurse care, home help, social work, and mental care by asking the following question for each health care professional (“Did you receive care/help from a ___ because of your rheumatoid arthritis during the last 12 months?” [yes/no]). Because of small numbers, the health care services were aggregated according to care function. Physiotherapy, occupational therapy, and chiropody were aggregated into allied health care; home nurse care and home help were aggregated into home care; and social work and mental care were aggregated into psychosocial care.
Unmet health care demands were assessed for the aforementioned services by asking the following question for each health care professional: “Do you want to receive (more) care/help from a ___ because of your rheumatoid arthritis?” [yes/no]).
Health outcomes assessed were disease activity, pain, disability, and mental health. The 28-joint count (33) and erythrocyte sedimentation rate, as assessed in the clinical examination, were used to compute the modified Disease Activity Score without the visual analog scale on general health (DAS28) (34). Pain was assessed with a visual analog scale (VAS-pain), ranging from 0 mm (no pain) to 100 mm (worst possible pain). Disability was measured by the Dutch version of the Health Assessment Questionnaire (HAQ) (35), ranging from 0 to 3 with higher scores indicating more disability. In this version of the HAQ the use of aids does not increase the scores. Mental health was assessed with the Dutch version of the Center for Epidemiological Studies-Depression scale (CES-D) (36). The CES-D has a range of 0 to 60 with higher scores indicating more depressive symptomatology. The cut-off point for identifying possible depression is 16, according to the manual (36).
The sociodemographic characteristics investigated were age, sex, marital status, and educational level. Marital status was rated as “single,” or “married or cohabiting.” Educational level was categorized into 2 groups, “'low” (none or primary school), and “high” (secondary education, college, or university).
Patients were divided into 3 health care service groups: users of a health care service (health care users); non-users of the health care service with an unmet demand for this service (unmet demanders); and non-users of the health care service without a demand for this service (non-demanders).
For the aggregated health care services (i.e., allied health care, psychosocial care, and home care), patients were assigned to the group of “health care users” if they received care from at least 1 of the included single services. Patients were assigned to the group of “unmet demanders,” if they did not receive care from any of the included single services, and perceived an unmet demand for at least one of these services. Patients were assigned to the group of “non-demanders” if they did not receive care from any of the included single services, and did not perceive an unmet demand for any of these services.
To determine whether patients' unmet demands indicate underuse of health care, patients' health outcomes (i.e., disease activity, pain, disability, and mental health) were compared between the three groups of patients for each health care service. Unmet demands were assumed to reflect underuse of a health care service if unmet demanders had comparable health outcomes as health care users, and also if they had worse health outcomes than non-demanders.
All analyses were performed using SPSS 10.0 for Windows (SPSS, Chicago, IL). To identify differences between the 3 patient groups on sociodemographic characteristics, chi-square and analysis of variance statistics were performed. Results were considered statistically significant if P values were less than 0.05.
Health characteristics of the 3 groups of patients were compared for each health care service with linear regression analysis. Patients who received care from a health care service and also expressed an unmet demand for that service were excluded from the analysis because of the small numbers. These patients, 11 (1.6%) for orthopedic care, 73 (10.8%) for allied health care, 23 (3.4%) for home care, and 11 (1.6%) for psychosocial care, had worse health outcomes than the health care users without a health care demand.
Regression models were built with each of the health outcomes as the dependent variable, and being health care user, unmet demander or non-demander of a health care service as the independent variable, with unmet demanders as reference category. Health outcomes were found to differ significantly between health care users, unmet demanders and non-demanders, if the P value of the linear regression was < 0.05. The effect sizes were expressed as the coefficient (B) with 95% confidence intervals (95% CIs). Because we conducted multiple analyses, we also applied the Bonferroni correction to our results for comparison with the uncorrected results.
Of the 841 eligible patients, 683 (81%) took part in the study; however, 4 respondents were excluded because of missing data on health care demands. No statistically significant differences were found on age and sex between non-respondents (n = 158) and respondents (n = 683).
The mean age of the patients was 61.5 years (range 23.4–92.1 years). Of the 679 patients, 71% were women, 66.8% were married or cohabiting, and 24.3% had a low educational level (Table 1). Patients had a mean disease duration of 10.7 years (range 1.5–57.7 years). On average, patients had moderate disease activity (mean DAS28 3.53; range 0.46–7.40), a moderate level of pain (mean VAS-pain 38.8; range 0–100), minor disabilities (mean HAQ 0.73; range 0.00–2.90), and good mental health (mean CES-D 12.0; range 0–49).
|Age, mean ± SD years||61.5 ± 13.8|
|≤60 years||286 (42.1)|
|>60 years||393 (57.9)|
|Sex, no. (%) female||485 (71.4)|
|Single, no. (%)||224 (33.1)|
|Together, no. (%)||452 (66.9)|
|Disease duration, mean ± SD, years||10.7 ± 9.2|
|≤5, no. (%)||284 (41.8)|
|>5, no. (%)||395 (58.2)|
|Disease activity, mean ± SD||3.53 ± 1.31|
|≤3.2, no. (%)||217 (43.5)|
|>3.2, no. (%)||282 (56.5)|
|Pain score, mean ± SD||38.8 ± 27.3|
|VAS ≤50 mm, no. (%)||430 (63.3)|
|VAS >50 mm, no. (%)||249 (36.7)|
|Disability score, mean ± SD||0.73 ± 0.67|
|≤0.5, no. (%)||326 (48.1)|
|>0.5, no. (%)||352 (51.9)|
|Mental health score, mean ± SD||12.0 ± 8.9|
|<16, no. (%)||488 (73.5)|
|≥16, no. (%)||176 (26.5)|
|Health care use|
|No services, no. (%)||319 (47.0)|
|≥1 service, no. (%)||360 (53.0)|
|Unmet health care demand|
|No services, no. (%)||484 (71.3)|
|≥1 service, no. (%)||195 (28.7)|
In the study population of 679 patients, 53% received care from at least 1 of the 4 health care services under study, and 28.7% expressed an unmet demand for at least 1 of these 4 services (Table 1). The highest percentage of health care use and unmet demands were found for allied health care, 33.3% received care and 13.4% had an unmet demand (Table 2). The lowest percentages were found for psychosocial care, 6% received care and 6.2% had an unmet demand. For orthopedic care and home care the utilization percentages were 23% and 14.9%, respectively, and 9.7% and 9.4% had an unmet demand for these services, respectively.
|Health outcome||Unmet demanders||Health care users||Non-demanders|
|Orthopedic care, (n = 668)||66 (9.7)||156 (23.0)||446 (65.7)|
|Allied health care, (n = 606)||91 (13.4)||226 (33.3)||289 (42.6)|
|Home care, (n = 656)||64 (9.4)||101 (14.9)||491 (72.3)|
|Psychosocial care, (n = 668)||42 (6.2)||41 (6.0)||585 (86.2)|
Overall, patients who reported an unmet demand for at least 1 of the health care services had similar sociodemographic characteristics and worse health outcomes (i.e., more pain, more disabilities, and more depressive symptomatology) compared with patients who reported no unmet health care demands (data not shown).
Analyzing each health care service separately, differences regarding sociodemographic characteristics were observed between unmet demanders, health care users, and non-demanders. For allied health care, unmet demanders were more often men than health care users and they were significantly younger than non-demanders. For home care, unmet demanders were significantly younger and more often married or cohabiting than health care users and they were more often women than non-demanders. For psychosocial care, unmet demanders were more often men and more often married or cohabiting than health care users (data not shown).
Table 3 shows the results of the linear regression analyses in which the health outcomes of unmet demanders were compared with the health outcomes of health care users and non-demanders of the selected health care services. For orthopedic care, unmet demanders had comparable levels of disease activity and pain as health care users; however they had lower levels of disability than health care users, and lower levels of disease activity, pain and disability than non-demanders. For allied health care, unmet demanders had comparable levels of disease activity, pain, and disability as health care users and they reported lower levels of disease activity, pain, and disability than non-demanders. For home care, unmet demanders had comparable levels of disease activity, pain, and disability as health care users and they had lower levels of these health outcomes compared with non-demanders. For psychosocial care, unmet demanders reported comparable levels of mental health as health care users and they suffered from lower levels of mental health compared with non-demanders. Moreover, the group of unmet demanders, and health care users of psychosocial care both had on average exceeded the cut-off point for possible depression (16 points). The observed differences between unmet demanders and non-demanders remain statistically significant after we applied the very conservative Bonferroni correction method for multiple comparisons (37) to our results.
|Disease activity||Pain||Disability||Mental health|
|Mean ± SD||β (95% CI)||Mean ± SD||β (95% CI)||Mean ± SD||β (95% CI)||Mean ± SD||β (95% CI)|
|Health care users||3.85 ± 1.27||−0.09 (−0.48, 0.31)||46.4 (26.8)||−1.89 (−9.16, 5.37)||1.04 ± 0.75||0.18 (0.00, 0.35)||12.1 ± 9.0||−2.16 (−4.60, 0.29)|
|Unmet demanders||3.82 ± 1.36||0.0||46.5 (27.5)||0.0||0.84 ± 0.61||0.0||14.2 ± 9.8||0.0|
|Non-demanders||3.33 ± 1.27||−0.61 (−0.96, −0.26)†||34.4 (26.3)||−13.93 (−20.36, −7.51)†||0.60 ± 0.60||−0.26 (−0.41, −0.11)†||11.5 ± 8.6||−2.74 (−4.92, −0.57)|
|Allied health care|
|Health care users||3.79 ± 1.26||0.06 (−0.23, 0.35)||43.6 (25.8)||−4.48 (−9.74, 0.77)||0.87 ± 0.68||−0.05 (−0.18, 0.08)||11.8 ± 8.5||−2.56 (−4.35, −0.77)|
|Unmet demanders||3.47 ± 1.34||0.0||44.6 (28.8)||0.0||0.78 ± 0.66||0.0||12.9 ± 9.0||0.0|
|Non-demanders||3.18 ± 1.25||−0.55 (−0.83, −0.27))†||29.5 (25.1)||−18.67 (−23.68, −13.65)†||0.52 ± 0.60||−0.39 (−0.52, −0.27)†||10.8 ± 8.8||−3.59 (−5.31, −1.88)†|
|Health care users||3.98 ± 1.29||−0.07 (−0.49, 0.36)||47.1 (28.6)||−5.64 (−13.24, 1.95)||1.16 ± 0.71||0.03 (−0.15, 0.20)||15.4 ± 8.8||−0.73 (−3.22, 1.76)|
|Unmet demanders||4.03 ± 1.09||0.0||53.1 (27.6)||0.0||1.04 ± 0.66||0.0||15.6 ± 9.5||0.0|
|Non-demanders||3.35 ± 1.30||−0.70 (−1.03, −0.37)†||34.5 (25.8)||−18.16 (−24.17, −12.16)†||0.57 ± 0.57||−0.56 (−0.70, −0.43)†||10.5 ± 8.3||−5.60 (−7.58, −3.63)†|
|Health care users||4.06 ± 1.26||0.21 (−0.40, 0.83)||53.3 (22.7)||5.85 (−5.09, 16.79)||1.20 ± 0.69||0.25 (−0.02, 0.51)||17.6 ± 7.47||0.49 (−3.05, 4.03)|
|Unmet demanders||3.84 ± 1.27||0.0||47.1 (26.3)||0.0||0.90 ± 0.61||0.0||16.2 ± 9.7||0.0|
|Non-demanders||3.47 ± 1.30||−0.38 (−0.80, 0.04)||36.9 (27.3)||−10.55 (−18.07, −3.04)||0.68 ± 0.65||−0.27 (−0.45, −0.09)†||11.1 ± 8.5||−5.97 (−8.41, −3.52)†|
Patients expressed unmet demands for health care services regarding health problems on which the specific health care services are directed. Disabled patients reported unmet demands for allied health care and home care, while depressed patients reported unmet demands for psychosocial care. Therefore, with regard to these services, unmet demands seem to indicate underuse. For orthopedic care, the results were less conclusive. Unmet demanders of this service had worse health outcomes compared with non-demanders, but they were not comparable with health care users on all these health outcomes. Unmet demanders reported fewer disabilities than health care users of orthopedic care.
In this study we investigated the prevalence of unmet demands among patients with RA in 4 health care services (orthopedic care, allied health care, home care, and psychosocial care), and we explored if these unmet demands of patients might indicate underuse of these health care services.
Unmet demands for health care are frequently reported by patients with RA. In our study, 28.7% of the patients had an unmet demand for 1 or more of the health care services being examined. Specifically, 9.7% of the patients had an unmet demand for orthopedic care, 13.4% for allied health care, 9.4% for home care, and 6.2% for psychosocial care, whereas 23.0%, 33.3%, 14.9%, and 6.0%, of the patients received care from these services, respectively. Previous studies that focused on unmet demands among patients with RA (19), and other chronic conditions such as cancer (15), dementia (16), HIV infection (17), and stroke (18), also described high overall percentages of unmet demands. These findings raise concern about the access to health care, because the demands for health care for chronic conditions widely exceed the use of health care. On some points, however, the expectations of patients with respect to health care may be too high, which might be exacerbated by the fact that treatment strategies in health care, by definition, tend to be less effective for chronic conditions, such as RA than for acute conditions (5).
The unmet health care demands as observed for allied health care, home care and psychosocial care correspond with the types of health problems at which these services are directed. These unmet demands, therefore, seem to indicate underuse of these health care services. Certainly for home care, underuse seems a longstanding and persistent problem in the Netherlands as it has been observed previously in RA (20) and in stroke (18). Moreover, the waiting lists for home care in the Netherlands have increased rather than declined over the years (38, 39). In a previous study (21), in which we investigated the equity in the use of health care services among patients with RA according to the epidemiological model by Andersen (40, 41), we also observed problems regarding allied health care and psychosocial care. Our current findings combined with our previous findings stress the relevance of patients' unmet health care demands for identifying deficits in health care.
We have shown statistically significant differences in health outcomes between patients with unmet demands for health care services and patients with no demands for health care. In addition, it seems likely that the observed differences are clinically relevant as well. For instance, low disease activity has been defined as a DAS28 score ≤3.3 (42). We have shown that on a group level, the non-demanders of orthopedic care, allied health care, and home care had a disease activity score ≤3.3, whereas unmet demanders and health care users of these services had much higher levels. Regarding mental health, a CES-D score ≥16 points indicates “possible depression” (36). Both unmet demanders of psychosocial care and psychosocial care users had on average exceeded this cut-off point of 16, whereas non-demanders reported much lower scores.
Our study was undertaken to explore unmet health care demands resulting in underuse of health care services among patients with RA. However, by interpreting our results some remarks could be made. First, more specific insight into patients' demands will be gathered if these demands are explored with a more detailed instrument than we have used. Such an instrument could detect, for example, not only the presence of an unmet demand, but also information on why and for what specific health problem patients report the demand. The detailed information could be used to pinpoint which unmet demands require action by professional health care services. Efforts have been made to develop such a detailed instrument for rehabilitation care (43), but this questionnaire has not been validated for chronic patients or for other health care services. Second, more insight into possible deficits in health care for patients with RA will be gathered if the professional perspective could be studied as well. The perspectives of patients and professionals could be combined to strengthen patient participation in daily clinical practice and to make sure that health care services are more responsive to the demands of patients.
Our results indicate that the use of health care might increase considerably for allied health care, home care, and psychosocial care if the unmet health care demands for these services are taken into account. At the time of this study, 33% of the patients received allied health care and 13% had an unmet demand for this service, which might result in allied health care use by 46% of the patients. From this perspective the use of home care might increase from 15% to 24%, and the use of psychosocial care from 6% to 12% of the patients. Moreover, as the unmet demanders of orthopedic care reported statistically significant worse health outcomes compared to non-demanders of this service, it may be expected that the use of this service also will increase.
In conclusion, unmet demands are reported relatively frequently among patients with rheumatoid arthritis for orthopedic care, allied health care, home care, and psychosocial care. Indications for underuse were found for the latter 3 health care services. The results of this study show that unmet health care demands of patients are relevant for identifying deficits in health care. Our results, therefore, stress the importance for health care professionals to be more responsive to the health care demands of patients. Efforts have to be made to decrease underuse of health care for RA patients by systematically incorporating patient's demands in chronic disease management.
We would like to thank M. Kammeijer for her large contribution to the collection of data.