Arthritis & Rheumatism (Arthritis Care & Research) Vol. 61, No. 2, February 15, 2009, pp 158 –165 DOI 10.1002/art.24214 © 2009, American College of Rheumatology
ORIGINAL ARTICLE
Economic Burden of Knee and Hip Osteoarthritis in Spain ´ S MAESE,3 ESTIBALIZ LOZA,1 JUAN MIGUEL LOPEZ-GOMEZ,2 LYDIA ABASOLO,1 JESU 3 2 LORETO CARMONA, ENRIQUE BATLLE-GUALDA, AND THE ARTROCAD STUDY GROUP
Objective. To estimate the direct and indirect osteoarthritis (OA)–attributable costs and predictors of costs of knee and hip OA in Spain. Methods. This study included consecutive patients age >50 years with symptomatic and radiologic knee and/or hip OA who were seen at primary care centers in all provinces of Spain. Information on demographics, health status (Short Form 12 Health Survey), comorbidities (Charlson Index), clinical (Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC]) and radiologic OA severity (Kellgren/Lawrence [K/L] scale), data related to OA health resources utilization (medical and nonmedical), and subjects’ and caregivers’ expenses and time lost in the previous 6 months were collected in 2 separate, structured, and detailed interviews. Costs in euros were assigned using market prices and official sources if available, and were annualized (to 2007). The predictors of costs were assessed in multivariate regression models. Costs were log-transformed before being modeled. Results. A total of 1,071 subjects were analyzed (74% women, mean ⴞ SD age 71 ⴞ 9 years). Average total annual costs were €1,502 per patient. Direct costs accounted for 86% of the total cost. We estimated a national cost of €4,738 million, representing 0.5% of the gross national product. Higher total costs were associated with comorbidity (Charlson Index odds ratio [OR] 1.27, 95% confidence interval [95% CI] 1.03–1.58), poorer health status (P < 0.050), worse WOMAC scores (OR 1.05, 95% CI 1.03–1.08), and grade 4 K/L scores (OR 1.76, 95% CI 1.15–2.69). Conclusion. The economic burden of knee and hip OA is substantial. Costs increased with comorbidity, poorer health status, and clinical and radiologic OA severity.
INTRODUCTION Osteoarthritis (OA) has a major impact on patients’ functioning and quality of life, as well as a high prevalence worldwide (1– 6). In the UK, between 1.3 and 1.75 million people have OA (7), whereas in France, 6 million new diagnoses of OA are made each year (8). In a health survey in 2000, 10.2% and 6.2% of the Spanish population reported symptomatic knee and hand OA, respectively (9). The Artrocad study was supported by Lacer Laboratories. 1 Estibaliz Loza, MD, Lydia Abasolo, MD: Hospital Clinico San Carlos, Madrid, Spain; 2Juan Miguel Lopez-Gomez, MD, Enrique Batlle-Gualda, MD, PhD: Hospital General Universitario de Alicante, Alicante, Spain; 3Jesu´s Maese, MD, PhD, Loreto Carmona, MD, PhD: Spanish Foundation of Rheumatology, Madrid, Spain. Additional members of the Artrocad Study Group are shown in Appendix A. Dr. Batlle-Gualda has received consultant fees, speaking fees, and/or honoraria (more than $10,000 each) from the Spanish Foundation of Rheumatology, the Spanish Society of Rheumatology, Abbott, Schering, and Wyeth. Address correspondence to Estibaliz Loza, MD, C/ Profesor Martin Lagos s/n, Madrid, Spain 28040. E-mail:
[email protected]. Submitted for publication April 1, 2008; accepted in revised form October 27, 2008.
158
Recently, it has been estimated that among US adults, ⬃27 million have clinical OA (3). Furthermore, OA prevalence is expected to increase in the coming years in developed countries, as life expectancy is also increasing and OA more often affects the elderly (3). Nevertheless, the burden of OA relates not only to its prevalence but also to the costs of the disease to individuals and health care systems. In this context, studies of cost of illness (COI) have revealed the great economic impact of OA (10 –13). Data from a Canadian cohort found an average annual cost of $12,200 per patient (2002 US dollars) (14). Moreover, in France, using a macroeconomic approach, OA direct costs in 2002 contributed to 1.7% of expenses of the French health insurance system (15). Even among employed persons, the total costs attributable to OA are substantial, as reported in a Belgian cohort (16). However, most of these studies refer to specific regions or do not include detailed direct and indirect costs. Consequently, updated and detailed COI studies of OA are required to support informed decision making at different levels of health care organizations. In addition, identifying patients at a higher risk of incurring costs is essential in decision making and ranking priorities. Finally, this knowledge will also facilitate clinicians’ daily
OA Costs in Spain practice, because they are active and important participants in health care provision. The objectives of the present national prevalence-based COI study were to 1) estimate the economic burden of OA to individuals and their caregivers, incorporating detailed estimates of both direct and indirect costs, and 2) identify factors that predict OA-related costs and the magnitude of the economic burden.
PATIENTS AND METHODS Study design. Artrocad is a national cross-sectional study conducted in 2003 by the Spanish Society of Rheumatology (SER) and the Spanish Society of Rural and General Physicians (SEMERGEN). The aim of the study was to assess the use of health care resources and the socioeconomic impact of symptomatic knee and hip OA in the primary care setting. Patient sample and data collection. The study sample was recruited by primary care physicians (PCPs) and consisted of consecutive patients age ⱖ50 years with radiologic knee or hip OA who reported pain at that location lasting for ⱖ3 months in the last year. Radiologic OA was defined as the presence of osteophytes and/or joint space narrowing. Patients with inflammatory arthritis or a previous fracture/trauma requiring a leg immobilization longer than 3 weeks were excluded. Those with prior joint surgery were allowed to participate. Based on a prevalence ratio of 3:1 for knee versus hip OA in the general population (9), we estimated a study sample of 1,300 patients, 900 with knee OA and 300 with hip OA, expecting to find an additional 100 with both knee and hip OA. The sampling scheme was as follows: Spain was divided into 11 regions, excluding the Canary and Balearic Islands as well as the cities of Ceuta and Melilla for logistic reasons. In each region, PCPs were recruited in a number that was proportional to the population age ⱖ50 years in order to reach the projected study size. One SER member and 1 SEMERGEN member responsible for the project invited the PCPs to participate; those who accepted were then asked to recruit 10 consecutive patients meeting the inclusion criteria, irrespective of the reason for the consultation. Data collection was carried out in 2 interviews using structured questionnaires. Data were then checked against the information available from the patient’s medical records. In the first interview, the PCPs recorded physician and patient sociodemographic characteristics, comorbidity, and OA-related costs, and assessed OA radiographic severity (using the Kellgren/Lawrence scale) (17) with the help of an atlas. Because knee and hip OA could coexist in the same patient, the PCP also determined which joint was generating more demand for health care, as defined by the number of PCP visits (signal joint) in each patient included. An anteroposterior radiograph in the previous 2 years was considered valid. When unavailable, a weightbearing anteroposterior radiograph of the signal joint was requested. Several days later, in the second interview, additional patient socioeconomic features and answers to
159 specific questionnaires were obtained by a different health professional (usually a nurse). The study protocol was approved by the Ethics Committee of the Alicante Hospital, and all patients signed an informed consent form. Identification, assignment, and classification of costs. We recorded OA-related costs in the 6 months before the date of the interview. Costs were classified as 1) direct costs, including medical costs: professional time (all consultations with health professionals in the public and private sector, including type of professional and number of visits), image and laboratory tests and others (type and number), all medications taken including joint injections (both prescribed and over the counter, doses, frequency, and duration), and hospital admissions (considering type of hospital department, length of stay, and costs related to knee and hip replacements); and nonmedical costs: derived from help at work, home, and self-care, adaptive aids, devices, assistive household equipment purchased, and transport (including moving expenses and time spent in the different forms of care); 2) indirect costs: compensation payments for lost labor/productivity (days on sick leave, permanent work disability, and related days off work due to patient’s OA) and costs of housekeeping help if the patient was a homemaker; 3) total costs: direct plus indirect costs; and 4) national costs: knee and hip OA costs in 2007 in the Spanish general population (18). Each cost variable was assigned a financial value by using market prices based on the following sources: 1) direct medical costs: national and regional official publications (Boletı´n Oficial del Estado [19], Boletı´n Oficial de la Junta de Andalucı´a [20], Boletı´n Oficial de la Regio´n de Murcia [21], and Boletı´n Oficial del Principado de Asturias [22], from 2002–2004), 2004 National Pharmacotherapeutical Catalog (Vademecum from 2002–2004) (23), Quiropra´ctica Melville (24), and from the Spanish Association of Biopathology (25); 2) direct nonmedical costs: national and regional official publications and the Madrid Municipal Transport Company (26); and 3) indirect costs: national and regional official publications. When several prices were possible, the cheapest was assigned. All costs were annualized and adjusted for inflation (27), and are expressed in 2007 euros. Other variables. Additional variables obtained include 1) sociodemographic features including age, sex, marital status (married, single, divorced, widowed), and educational level (no studies, elementary school, high school, and post high school); 2) socioeconomic variables such as characteristics of the principal residence, employment status, and social class according to the British Registrar General (28); 3) PCP data, including sex and years of clinical practice in primary care; 4) OA-related data: signal joint (knee, hip, or both), radiologic severity assessed using the Kellgren/Lawrence scale (grade 1 [doubtful] ⫽ minute osteophyte, doubtful significance; grade 2 [minimal] ⫽ definite osteophyte, unimpaired joint space; grade 3 [moderate] ⫽ moderate diminution of joint space; and grade 4 [severe] ⫽ joint space greatly impaired with sclerosis of subchondral bone), and OA duration; 5) comorbidity measured by the Charlson index (29); and 6) specific
160 questionnaires, including the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) (30) and the Short Form 12-item questionnaire (SF-12) (31). Statistical analysis. We described the subjects’ sociodemographic and clinical characteristics by the measures of central tendency appropriate for the distribution of each variable. Because the cost variables were not normally distributed, we described them with several statistics (mean ⫾ SD, median, range, and 25th, 75th percentiles). In addition, to assess the national economic impact of knee and hip OA, we first calculated the number of subjects with knee and hip OA in the country. For this purpose, we used data from the Spanish population of 2007 (18) and the reported prevalence of knee and hip OA published in a Spanish health survey (9). This health survey (1999 – 2000) comprised a probabilistic and representative sample of 2,192 Spanish inhabitants age ⱖ20 years. The diagnosis of OA was established by a rheumatologist according to American College of Rheumatology criteria (32,33). After that, we estimated the disease-related costs by multiplying the calculated number of patients with knee and hip OA with the mean costs in our study results. Finally, the total impact was compared with the 2007 gross national product (GNP) (27). The robustness and consistency of the results obtained were tested with a one-way sensitivity analysis. Taking into account the number of cost variables included for the sensitivity analyses, 95.5% confidence intervals (95.5% CIs) were used. We assumed 2 different scenarios: the best scenario (least cost) using the lower limit of the 95.5% CI, and the worst scenario (most cost) using the upper limit of the 95.5% CI. Finally, we assessed the potential predictors of costs. For this purpose, we analyzed total costs (direct plus indirect costs) and several demographic and clinical factors in multivariate regression models. Total costs were skewed and therefore log-transformed before being modeled, as this was the function that achieved the normalization of the dependent variable. The parameters obtained from least squares regression were then interpreted as multipliers (e.g., the parameter estimate of 0.087 for sex indicated that women, on average, reported costs that were 100.084 ⫽ 1.22 times higher than men [odds ratio (OR)], after adjusting for the other variables in the model, although this difference was not statistically significant). The same methodology was used to calculate the 95% CIs.
RESULTS Study sample characteristics. A total of 1,071 patients were included (82% of the 1,300 initially projected) by 113 PCPs (87% of those invited to participate). As shown in Table 1, the majority of patients included were women in their 70s, either retired or housewives. The knee was the signal joint in two-thirds of patients, and according to the Kellgren/Lawrence scale, ⬃80% and 67% of patients had moderate or severe knee and hip OA, respectively (Table 2). The mean ⫾ SD duration was 7 ⫾ 8 years for knee OA and 3 ⫾ 6 years for hip OA. The mean ⫾ SD WOMAC score was 42.5 ⫾ 17.5, and the mean ⫾ SD scores on the SF-12
Loza et al
Table 1. Sociodemographic and occupational characteristics of the study sample and primary care physicians* Value Study sample Sex (women) Age, mean ⫾ SD years Marital status (married) Educational level (none or elementary studies) Employment status Retired Housewife Working Other Low social class† Principal residence (flat without an elevator) Primary care physicians Sex (women) Age, mean ⫾ SD years Years of medical practice, mean ⫾ SD
74 71 ⫾ 9 62 40 39 31 9 21 56 62 71 45 ⫾ 6 17 ⫾ 7
* Values are the percentage unless otherwise indicated. † Social class was obtained according to the British Registrar General (28).
were 35.1 ⫾ 8.7 for the physical component and 44.8 ⫾ 11 for the mental component. More than 95% of patients used at least 1 medication for OA. A total of 69% had used analgesics, 70% nonsteroidal antiinflammatory drugs, 10.5% a symptomatic slow-acting drug for OA, 9% cyclooxygenase 2 selective nonsteroidal antiinflammatory drugs, and 1% other drugs. Costs incurred by subjects with knee or hip OA and sensitivity analysis. Average total annual costs were €1,502 per patient (Table 3). Direct costs accounted for 86% of the total cost, mostly due to home, work, and self-care help, professional time, and hospital admissions (medical and surgical admissions). Indirect costs were clearly lower (14% of total costs), with the largest component related to help for housewives at home. Table 4 shows costs according to Kellgren/Lawrence scale grades. Given that the size of the Spanish population age ⱖ20 years in 2007 was 36,389,028 inhabitants, and that the prevalence of knee and hip OA was 10.2% and 4%, respectively, we estimated (using median total costs) a national cost of €4,738 million for knee and hip OA, which represents 0.5% of the 2007 GNP. We then carried out a one-way sensitivity analysis of the measured costs (Figure 1). We found that total annual costs might range from €1,400 per patient in the best scenario (represented by the lower limit of the 95.5% CI) to €1,603 per patient in the worst scenario (represented by the upper limit of the 95.5% CI). As shown by the narrow 95.5% CIs, we found that the different cost estimations were generally quite precise. Determinants of costs. Finally, we looked for factors associated with incurring high costs in several regression models by log-transforming cost variables. As shown in Table 5, total costs were lower in patients age 60 –79 years
OA Costs in Spain
161
Table 2. Clinical features related to osteoarthritis (OA) and patient health-related quality of life in the study sample* Value Signal joint† Knee Hip Knee and hip Radiographic OA severity‡ Grade 1 Knee Hip Grade 2 Knee Hip Grade 3 Knee Hip Grade 4 Knee Hip Not evaluable (prosthesis) Knee Hip Charlson Index, mean ⫾ SD SF-12, mean ⫾ SD Physical component Mental component WOMAC, mean ⫾ SD Total score Pain score Stiffness score Physical function score
66.3 23.5 10.2
1.6 1.5 11.8 16.2 45.2 39.9 32.5 26.9 8.9 13.5 1.61 ⫾ 0.68 35.1 ⫾ 8.7 44.8 ⫾ 11 42.5 ⫾ 17.5 8.1 ⫾ 3.7 3.2 ⫾ 1.8 31.3 ⫾ 13.3
* Values are the percentage unless otherwise indicated. SF-12 ⫽ Short Form 12-item questionnaire; WOMAC ⫽ Western Ontario and McMaster Universities Osteoarthritis Index. † Defined by the patient’s primary care physician (PCP) as the joint meeting the inclusion criterion that was generating more OA-related visits to the PCP. ‡ Assessed by the PCP using the Kellgren/Lawrence scale.
as compared with the 50 –59 years age group. Furthermore, higher total costs were associated with comorbidity (Charlson Index OR 1.27, 95% CI 1.03–1.58), poorer health-
related quality of life (P ⬍ 0.050 for both physical and mental components of the SF-12), and worse WOMAC score (OR 1.05, 95% CI 1.03–1.08). Finally, patients with Kellgren/Lawrence scale grade 4 were more likely to incur costs than patients with grades 1 or 2 (OR 1.76, 95% CI 1.15–2.69). However, no differences were found between women and men (P ⫽ 0.250). Moreover, when analyzing subcomponents of total costs, we found that the probability of higher direct medical costs was lower in the age 50 –59 years group compared with patients age ⱖ60 years (P ⬍ 0.001). Higher direct medical costs were associated with comorbidity (Charlson Index OR 1.30, 95% CI 1.03–1.64), poorer score on the physical component of the SF-12 (P ⫽ 0.006), worse WOMAC score (OR 1.03, 95% CI 1.01–1.06), and Kellgren/ Lawrence scale grades 3 (OR 1.58, 95% CI 1.06 –2.36) or 4 (OR 2.60, 95% CI 1.63– 4.16). On the other hand, higher direct nonmedical costs were more likely to be incurred by women (OR 1.89, 95% CI 1.13–3.16), patients ages 50 –59 years compared with those ages 80 – 89 years (P ⫽ 0.003), and those with worse WOMAC scores (OR 1.05, 95% CI 1.01–1.09). Interestingly, higher indirect costs were associated only with male sex (P ⫽ 0.026) and Kellgren/Lawrence scale grades 3 (OR 3.61, 95% CI 1.34 –9.75) or 4 (OR 6.74, 95% CI 2.08 –21.8).
DISCUSSION We performed a detailed prevalence-based COI study of symptomatic knee and hip OA in Spain. As shown, the total annual costs amounted to €1,502 per patient, mostly due to direct costs, whereas the national cost was estimated at approximately €4,738 million, representing 0.5% of the GNP (for 2007). In addition, incurring higher costs was associated with comorbidity, poorer health status, and clinical and radiologic OA severity. The present COI study included a large sample of patients with knee and hip OA attending primary care in all provinces of Spain. We considered this setting appropriate to assure the representation of the study population given that subjects from specialized care usually have more severe conditions and therefore may bias the COI results
Table 3. Estimated annual costs per patient, in euros
Direct costs Medical costs Professional time Image and laboratory tests Drugs Hospital admissions Nonmedical costs House/work/self-care help Aid devices Transport Indirect costs Lost labor/productivity Help for housewives at home Total costs
Mean ⴞ SD
Median
Range
25th, 75th percentiles
% of total cost
1,297 ⫾ 1,544 724 ⫾ 1,210 311 ⫾ 422 145 ⫾ 247 77 ⫾ 167 192 ⫾ 1,000 574 ⫾ 854 430 ⫾ 820 140 ⫾ 322 4 ⫾ 10 205 ⫾ 590 88 ⫾ 470 117 ⫾ 377 1,502 ⫾ 1,699
783 424 194 104 20 0 196 0 64 0 0 0 0 917
40–18,155 19–17,862 19–4,512 0–3,134 0–2,428 0–16,463 0–5,143 0–4,984 0–5,135 0–71 0–3,538 0–3,360 0–2,492 40–18,155
437–1,627 211–776 19–4,025 0–1,988 3–83 0 73–674 0–533 43–128 0–68 0 0 0 463–1,943
86 47 22 7 5 13 39 29 9 1 14 6 8 100
162
Loza et al
Table 4. Estimated annual costs per patient according to Kellgren/Lawrence scale grades, in euros* Kellgren/Lawrence scale grade
Direct costs Medical costs Professional time Image and laboratory tests Drugs Hospital admissions Nonmedical costs House/work/self-care help Aid devices Transport Indirect costs Lost labor/productivity Help for housewives at home Total costs
1
2
3
4
693 (300–997) 381 (189–514) 189 (77–365) 54 (27–103) 14 (3–71) 0 (0)† 151 (55–429) 0 (0–667)‡ 108 (49–257) 0 (0–13) 0 (0–922)‡ 0 (0–737)‡ 0 (0–222)‡ 693 (362–997)
657 (366–1,511) 256 (130–521) 118 (58–262) 55 (27–108) 19 (3–69) 0 (0–542)‡ 218 (93–520) 0 (0–1,867)‡ 128 (64–256) 0 (0–22) 0 (0–711)‡ 0 (0–242)‡ 0 (0–356)‡ 704 (399–1,280)
690 (394–1,271) 373 (132–630) 175 (97–345) 81 (27–130) 17 (3–65) 0 (0–256)‡ 165 (73–562) 0 (0–1,868)‡ 73 (51–128) 0 (0–23) 0 (0–1,335)‡ 0 (0–2,240)§ 0 (0–889)‡ 802 (438–1,583)
1,008 (470–2,242) 649 (346–1,122) 233 (116–386) 103 (27–152) 22 (2–75) 0 (0–577)‡ 257 (68–1,034) 0 (0–2,760)‡ 51 (34–103) 0 (0–27) 0 (0–1,870)‡ (0–3,360)§ 0 (0–1,068)‡ 1,208 (533–2,539)
* Values are the median (25th, 75th percentiles) unless otherwise indicated. † There were no osteoarthritis hospitalizations in this subgroup of patients. ‡ Median (5th, 95th percentiles). § Median (1st, 99th percentiles).
(34). Furthermore, in Spain, as in other countries, PCPs are the main gateway to health care, although patients may also visit other specialists at the same time (35,36). One of the most important strengths of this study is the diagnostic accuracy of OA cases. OA was defined clinically and radiologically, and not by self-report, an important weakness that hampers many COI studies. But perhaps the main strength of this study is the detailed and exhaustive cost domains applied to cover the possible OA-attributable costs. Direct and indirect costs, including out-of-pocket and informal costs in public and private practice, as well as their frequency, were collected, increasing the validity of the cost estimations. Finally, for the cost incurrence analyses, we incorporated not only the classical variables
Figure 1. Sensitivity analysis. Circles in the middle represent the baseline case (arithmetic mean cost) and circles at both sides of the line represent the 95.5% confidence interval (95.5% CI). The best scenario (least cost) is represented by the lower limit of the 95.5% CI, and the worst scenario (most cost) by the upper limit of the 95.5% CI for each measured cost. Results are in 2007 euros.
(demographic, clinical, and health status), but also radiologic variables, which provided new insights into these analyses. However, this study has some limitations. We excluded patients age ⬍50 years, who have a higher probability of being actively employed. However, as has been documented, the OA prevalence in this population subgroup is low (3); therefore, the costs incurred by this age group would probably not significantly modify those reported in the present study. All COI studies have common limitations not easily overlooked. It could be argued that those recruited had more symptomatic or severe OA, since these patients are more likely to contact their PCP and thus are more likely to be included. But PCPs were allowed to include patients with OA regardless of the reason for consultation (not necessarily their OA) in order to include a wide variety of subjects with symptomatic knee and hip OA. We used subjective (patient-derived) cost data (incurred in the previous 6 months) that were subsequently cross-checked and completed, if necessary, by reviewing patients’ medical records so as to collect the most complete and precise information possible. Common to other cost studies, cost assignment is very complicated due to different sources, lack of standardization, etc. In Spain, it is difficult to obtain information on public health costs. Therefore, we used several sources for the cost assignment, which may limit our results. But taking all these considerations into account, we are confident that our COI study provides excellent data on OA costs in Spain that are suitable for further comparisons. Since OA-related costs are enormous, it is worthwhile to identify subjects at a higher risk of incurring higher costs in order to provide targets for public health efforts, for example, to implement specific programs to increase knowledge and self-care of OA to reduce the disease socioeconomic impact, contributing in order to efficiently allocate health care resources in a health area, city, or even
OA Costs in Spain
163
Table 5. Total (log-transformed) hip and knee osteoarthritis costs: multivariate regression results*
Sex (women) Age, years 50–59 (reference group) 60–69 70–79 80–89 ⬎90 Charlson Index SF-12 Physical component Mental component WOMAC (total score) Kellgren/Lawrence scale grade 1 and 2 (reference group) 3 4
Parameter (SE)
OR†
95% CI for multiplier
P
0.087 (0.076)
1.223
0.868–1.724
0.250
0 ⫺0.320 (0.111) ⫺0.329 (0.108) ⫺0.239 (0.123) 0.080 (0.235) 0.104 (0.235)
1 0.478 0.469 0.577 1.203 1.272
– 0.290–0.789 0.288–0.764 0.330–1.006 0.412–3.483 1.026–1.576
– 0.004 0.002 0.053 0.732 0.028
⫺0.018 (0.001) ⫺0.008 (0.003) 0.022 (0.006)
0.958 0.983 1.051
0.939–0.978 0.969–0.996 1.025–1.078
⬍ 0.001 0.013 ⬍ 0.001
0 0.103 (0.081) 0.244 (0.094)
1 1.269 1.755
– 0.881–1.828 1.145–2.692
– 0.201 ⬍ 0.001
* OR ⫽ odds ratio; 95% CI ⫽ 95% confidence interval; SF-12 ⫽ Short Form 12-item questionnaire; WOMAC ⫽ Western Ontario and McMaster Universities Osteoarthritis Index. † Multiplier ⫽ 10parameter.
country. In this context, politicians and health decision makers should be strongly aware of the economic impact of knee and hip OA, because in many countries and health systems, the resources are not unlimited. Therefore, taking into account our results and the prevalence of the disease, we consider that there are enough reasons for giving more relevance to knee and hip OA. Moreover, we have identified the subgroup of patients who are more likely to incur higher costs. Therefore, these data could also be a good opportunity for clinicians to improve not only patients’ clinical outcomes, but also societal or economic outcomes. Despite methodologic differences that limit the comparability of the study results, some general remarks can be made with regard to other reports. Our results show that the costs incurred due to symptomatic knee and hip OA impose an enormous economic burden on patients and health systems. Similar findings have been published in other developed countries (8,14 –16,37), confirming the great magnitude and economic impact of OA worldwide. In our sample, direct costs clearly accounted for the majority of total costs. Similar results have been published in a Canadian study (37), but they differ from other studies that showed greater indirect costs (14,16). Regarding direct costs, results from previous studies are not directly comparable to ours for several reasons. Differences in demographics, clinical characteristics, and health care systems limit the comparability of the results. In our study, subjects were age ⱖ50 years and very few were working; therefore, most of the costs incurred were direct costs. Moreover, the estimation of indirect costs is controversial. Some authors suggest that the current methodology used gives insufficient weight to diseases that, like OA, mainly affect women, thus underestimating indirect costs (38). On the other hand, economists often consider that indirect costs are overestimated; therefore, they recommend that these costs be estimated conservatively (39,40). Similar to other reports (14,16), we found that total costs were associated with age, except for patients age ⬎90
years. This could be due to the small number of patients in this age subgroup, and may also be because they are less likely to be considered candidates for surgery (due to age and its related complications [41,42]), thus decreasing the costs incurred with regard to medical admissions, professional time, etc. On the other hand, patients ages 50 –59 years compared with older patients were more likely to incur costs. This could be because they clearly reported more indirect costs due to loss of productivity and other work-related compensations (retirement age in Spain is ⬃65 years) and medical costs (data not shown). As expected, patients with poorer health status, worse WOMAC score, and more severe radiologic OA incurred more costs, as other studies have shown (14,16,37). Interestingly, no sex differences were reported except in the case of male patients, who incurred more indirect costs, whereas other authors have reported that women are more likely to incur indirect costs (14). One explanation for these differences could be that, in traditional societies such as Spain, women are less likely than men to seek help, especially for housework. However, we also consider that in the future, as women are progressively incorporating into the work place, their OA-related cost will be higher. From a social perspective, this study highlights the great economic burden of symptomatic knee and hip OA to individuals and to the compensation benefits scheme of the Spanish health system. Additionally, we have identified patients who will eventually gain from specific programs and policies for prevention and treatment that will help minimize total OA costs. As the number of persons affected with OA is expected to increase rapidly, and given that public health resources are finite, all participants involved in the health care system should consider these figures seriously. Further studies based on agreed, standardized methods for calculating COI are clearly warranted to produce better estimates of
164
Loza et al
the lifetime costs of OA so that reliable, valid, and useful data can be provided to health care providers.
12.
ACKNOWLEDGMENTS The authors appreciate the invaluable help of the PCPs of the SEMERGEN and other health professionals who participated in the study, as well as the health centers that provided the time and the space for the study.
13. 14.
AUTHOR CONTRIBUTIONS Dr. Loza had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study design. Batlle-Gualda. Acquisition of data. Maese, Batlle-Gualda. Analysis and interpretation of data. Loza, Abasolo, Maese, Carmona, Batlle-Gualda. Manuscript preparation. Loza, Lopez-Gomez, Carmona, BatlleGualda. Statistical analysis. Loza, Lopez-Gomez.
ROLE OF THE STUDY SPONSOR Lacer Laboratories supported the study financially, but did not have any influence on the study design, collection of data, monitoring, analysis, or interpretation of the results, nor did they have any influence on the decision to submit the manuscript.
15. 16. 17. 18. 19. 20. 21. 22. 23.
REFERENCES
24.
1. Reginster JY. The prevalence and burden of arthritis [review]. Rheumatology (Oxford) 2002;41 Suppl 1:3– 6. 2. Dillon CF, Rasch EK, Gu Q, Hirsch R. Prevalence of knee osteoarthritis in the United States: arthritis data from the Third National Health and Nutrition Examination Survey 1991–94. J Rheumatol 2006;33:2271–9. 3. Lawrence RC, Felson DT, Helmick CG, Arnold LM, Choi H, Deyo RA, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States: part II. Arthritis Rheum 2008;58:26 –35. 4. Loza E, Abasolo L, Jover JA, Carmona L. Burden of disease across chronic diseases: a health survey that measured prevalence, function, and quality of life. J Rheumatol 2008;35: 159 – 65. 5. Boutron I, Rannou F, Jardinaud-Lopez M, Meric G, Revel M, Poiraudeau S. Disability and quality of life of patients with knee or hip osteoarthritis in the primary care setting and factors associated with general practitioners’ indication for prosthetic replacement within 1 year. Osteoarthritis Cartilage 2008;16:1024 –31. 6. McHugh GA, Luker KA, Campbell M, Kay PR, Silman AJ. Pain, physical functioning and quality of life of individuals awaiting total joint replacement: a longitudinal study. J Eval Clin Pract 2008;14:19 –26. 7. Watson M. Management of patients with osteoarthritis. Pharm J 1997;259:296 –7. 8. Levy E, Ferme A, Perocheau D, Bono I. Socioeconomic costs of osteoarthritis in France. Rev Rhum Ed Fr 1993;60:63S–7S. In French. 9. Carmona L, Ballina J, Gabriel R, Laffon A. The burden of musculoskeletal diseases in the general population of Spain: results from a national survey. Ann Rheum Dis 2001;60: 1040 –5. 10. Gabriel SE, Crowson CS, Campion ME, O’Fallon WM. Indirect and nonmedical costs among people with rheumatoid arthritis and osteoarthritis compared with nonarthritic controls. J Rheumatol 1997;24:43– 8. 11. Gabriel SE, Crowson CS, O’Fallon WM. Costs of osteoarthritis:
25. 26. 27. 28.
29.
30.
31.
32.
33.
34. 35.
estimates from a geographically defined population. J Rheumatol Suppl 1995;43:23–5. Lanes SF, Lanza LL, Radensky PW, Yood RA, Meenan RF, Walker AM, et al. Resource utilization and cost of care for rheumatoid arthritis and osteoarthritis in a managed care setting: the importance of drug and surgery costs. Arthritis Rheum 1997;40:1475– 81. MacLean CH, Knight K, Paulus H, Brook RH, Shekelle PG. Costs attributable to osteoarthritis. J Rheumatol 1998;25: 2213– 8. Gupta S, Hawker GA, Laporte A, Croxford R, Coyte PC. The economic burden of disabling hip and knee osteoarthritis (OA) from the perspective of individuals living with this condition. Rheumatology (Oxford) 2005;44:1531–7. Le Pen C, Reygrobellet C, Gerentes I. Financial cost of osteoarthritis in France: the “COART” France study. Joint Bone Spine 2005;72:567–70. Rabenda V, Manette C, Lemmens R, Mariani AM, Struvay N, Reginster JY. Direct and indirect costs attributable to osteoarthritis in active subjects. J Rheumatol 2006;33:1152– 8. Kellgren JH, Lawrence JS. Radiological assessment of osteoarthrosis. Ann Rheum Dis 1957;16:494 –502. Instituto Nacional de Estadı´stica. URL: www.ine.es/jaxi/tabla. do?path⫽/t20/e245/p04/a2007/l0/&file⫽00000002.px&type⫽ pcaxis&L⫽0.htlm. Boletı´n Oficial del Estado (BOE). URL: http://www.boe.es/g/es/. Boletı´n Oficial de la Junta de Andalucı´a. URL: http://www. juntadeandalucia.es/boja/boletines/2008/226/index.html. Boletı´n Oficial de la Regio´n de Murcia. URL: http://www.carm. es/borm/vista/principal/inicio.jsf. Boletı´n Oficial del Principado de Asturias. URL: www.asturias. es/portal/site/Asturias/menuitem. Vademecum Internacional. URL: http://www.vademecum. es/. Centro de Quiropra´ctica Melville. URL: http://bilbao.popsalud. com/post/128n/centro-de-quiroprctica-melville. Asociacin Espan´ola de Biopatologı´a Me´dica. URL: www.aebm. org. Empresa municipal de transportes de Madrid. URL: www. emtmadrid.es/interest/index/mensual.html. Ministerio de Economia y Hacienda. URL: www.igae.pap.meh. es. Alonso J, Perez P, Saez M, Murillo C. Validity of the occupation as an indicator of social class, according to the British Registrar General classification. Gac Sanit 1997;11:205–13. In Spanish. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40: 373– 83. Batlle-Gualda E, Esteves-Vives J, Piera MC, Hargreaves R, Cutts J. Adaptacio´n transcultural del cuestuonario WOMAC especı´fico para la artrosis de rodilla y cadera. Rev Esp Reumatol 1990;26:38 – 45. Alonso J, Prieto L, Anto JM. The Spanish version of the SF-36 health survey (the SF-36 health questionnaire): an instrument for measuring clinical results. Med Clin (Barc) 1995;104: 771– 6. In Spanish. Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D, Brandt K, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hip. Arthritis Rheum 1991;34:505–14. Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis: classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039 – 49. Rosemann T, Joos S, Szecsenyi J, Laux G, Wensing M. Health service utilization patterns of primary care patients with osteoarthritis. BMC Health Serv Res 2007;7:169. Ballina Garcia FJ, Hernandez Mejia R, Martin Lascuevas P, Fernandez Santana J, Cueto Espinar A. Epidemiology of musculoskeletal complaints and use of health services in Asturias, Spain. Scand J Rheumatol 1994;23:137– 41.
OA Costs in Spain 36. Rasker JJ. Rheumatology in general practice. Br J Rheumatol 1995;34:494 –7. 37. Maetzel A, Li LC, Pencharz J, Tomlinson G, Bombardier C. The economic burden associated with osteoarthritis, rheumatoid arthritis, and hypertension: a comparative study. Ann Rheum Dis 2004;63:395– 401. 38. Clarke AE, Zowall H, Levinton C, Assimakopoulos H, Sibley JT, Haga M, et al. Direct and indirect medical costs incurred by Canadian patients with rheumatoid arthritis: a 12 year study. J Rheumatol 1997;24:1051– 60. 39. Bloom BS, Bruno DJ, Maman DY, Jayadevappa R. Usefulness of US cost-of-illness studies in healthcare decision making. Pharmacoeconomics 2001;19:207–13. 40. Koopmanschap MA. Cost-of-illness studies: useful for health policy? Pharmacoeconomics 1998;14:143– 8. 41. Parvizi J, Mui A, Purtill JJ, Sharkey PF, Hozack WJ, Rothman RH. Total joint arthroplasty: when do fatal or near-fatal complications occur? J Bone Joint Surg Am 2007;89:27–32. 42. Mantilla CB, Horlocker TT, Schroeder DR, Berry DJ, Brown DL. Frequency of myocardial infarction, pulmonary embolism, deep venous thrombosis, and death following primary hip or knee arthroplasty. Anesthesiology 2002;96:1140 – 6.
APPENDIX A: MEMBERS OF THE ARTROCAD STUDY GROUP In addition to the authors, members of the Artrocad Study Group are Abadı´a F, Acevedo J, Aganzo F, Aganzo F, Aicart M, Alcarzar F, Almenar E, Alonso A, Alonso A, Alonso R, Andreu M, Anto´n N, Aparicio M, Arribas F, Arriola E, Baba Z, Bassa A, Begonte G, Berenguer J, Bernad
165 M, Blanco E, Blanco J, Borreguero V, Brenes F, Buded F, Cabrera M, Calvo J, Caman˜o J, Carpi M, Casado J, Casado P, Cereijo A, Cerijo A, Chamizo E, Chamorro J, Chulvi J, Climent J, Coladas C, Colmenarejo J, Comas J, Corrales J, Costa N, Cristel L, de Miguel E, Del Val M, Delgado J, Diaz N, Domı´nguez O, Fernandez A, Ferna´ndez C, Ferna´ndez E, Ferna´ndez M, Ferna´ndez M, Ficote M, Fle´ I, France´s M, Gallego A, Garavı´s J, Garcı´a E, Garcı´a J, Garcı´a T, Gime´nez S, Gimeno M, Ginel L, Gomariz J, Go´mez A, Go´mez J, Guevara D, Guillard M, Gurt A, Guzma´n J, Henrı´quez A, Herna´ndez J, Hidalgo C, Iturregui E, Jime´nez D, Jime´nez M, Jime´nez S, Lara E, Leal B, Lenzos J, Linares L, Lisbona M, Lo´pez A, Lo´pez L, Lo´pez M, Lo´pez M, Lo´pez M, Lo´pez V, Lo´pez V, Luna S, Mallo´n S, Manero F, Manpel C, Marcos A, Martı´ D, Marti M, Martı´n I, Martı´n L, Martincano J, Martı´nez A, Martı´nez C, Martı´nez P, Medel F, Mediavilla J, Mediavilla L, Me´ndez-Cabeza J, Mene´ndez J, Mila´n A, Miralles M, Montan˜e´s F, Morago V, Morales J, Mosquera J, Mosquera J, Murga N, Nicola´s J, Nieto M, Noya J, Ocan˜a C, Olive X, Ortega A, Ortega A, Otero J, Peral O, Pe´rez M, Pe´rez M, Pizarro G, Puente M, Ramı´rez R, Rı´os R, Rivas A, Rivera I, Rodrı´guez G, Rodrı´guez J, Rodrı´guez J, Rodrı´guez T, Roma´n J, Ruiz A, Ruiz E, Safont M, Sainz N, Sainz-Maza M, Salanova A, Salcedo J, Sa´nchez P, Santiago L, Senan R, Sua´rez F, Tenorio P, Tornero J, Trujillo P, Valero A, Vergara J, Vico F.
