International variation in resource utilisation and treatment costs for rheumatoid arthritis
- 70 Downloads
Recent years have witnessed substantial progress in understanding the cost implications of rheumatoid arthritis (RA). To assess the divergent methodologies and their impact on the resulting cost analyses in RA, we conducted a systematic literature review to summarise the scientific evidence of RA-induced costs.
Sixty-five reviews, models or cost analyses on the burden of illness and general costs associated with RA were identified. They covered the US, Canada, Sweden, the UK, The Netherlands, Germany and Finland. Twenty-four cost analyses provided appropriate data about direct and/or indirect costs. Each study was summarised separately. Costs were discounted to 2003 and converted to US dollars.
The costs per RA-year ranged from $US1503 to $US16 514. However, each study has to be interpreted individually, with consideration given to the study population, indication, age of the study, database used, type of therapy, setting, level of cost differentiation and data derivation. Health technology assessment reports offer sufficient space to adequately describe the composite parts and restrictive elements of different methodological approaches and analyses.
Rheumatoid arthritis (RA) is a chronic, systemic, inflammatory disease of unknown aetiology that affects joints, bones and other body systems (e.g. the lungs).
The reported prevalence of RA across Europe varies from 0.5% to 5.2%, depending on the year of reference and the definition of RA. Studies of RA show a 12% prevalence in Scandinavia,[3,4] 1.0–1.5% in the Dutch population aged >40 years, 3% of the German population[2,6] and 1% prevalence in the US adult population.[7, 8, 9, 10] Females are three times more likely to have RA than males. The rates increase with age in both sexes.
Recent years have witnessed substantial progress in understanding the cost implications of RA without therapy or when using different treatment regimens. Taking into account the wide spectrum of methodological approaches of cost evaluation, careful consideration of potential country-specific variations in resource utilisation associated with RA needs to be taken into account.
To assess the divergent methodologies and their impact on the resulting cost analyses in RA, we conducted a systematic literature review to summarise the scientific evidence of RA-induced costs.
1. Search and Selection Strategy
A literature survey of Medline, Medline Alert, Cochrane Database of Systematic Reviews (CDSR93) and Cochrane Database of Abstracts of Reviews of Effectiveness (CDAR94) was performed. Searching terms (FT/TI) were ‘Rheumatoid Arthritis AND Costs’. The search period (year of publication) was unrestricted.
Using the UK NHS Centre for Reviews and Disseminations, the Database of Abstracts of Reviews of Effects (DARE), the NHS Economic Evaluation Database (NHS EED) and the Health Technology Assessment Database (HTA) were searched. The searching terms were ‘Rheumatoid Arthritis AND Costs’.
full text in English
original research (no comments, no letters)
no methodological paper
focus on RA
cost analysis, not cost-effectiveness analysis or cost-benefit analysis.
absolute amount of costs were reported, not only resources or percentages of costs
costs were reported as costs per treatment case and not as costs with respect to national burden of disease
costs per patient-year are reported.
Finally, we used models and reviews for the discussion, if they included relevant empirical data.
Information about identification, measurement and valuation of costs were tracked as a methodology suggested by Raftery. Each eligible study was presented separately.
For our immediate comparison, the study results were adjusted from the year of reference up to 2003 with a discounting rate of 5%. We did not count graduated interest rates.
Additionally, the national currencies were converted into US dollars. The exchange rates of 1 January 2003 were taken using the FX-Converter (http://www.oanda.com/convert/classic?lang = de). One Canadian dollar ($Can), one Swedish krona (SEK), one British pound (£), one Dutch guilder (NLG) and one Finish markka (Fmk) were converted to US dollars using the following respective conversion rates: 0.636, 0.115, 1.611, 0.477 and 0.177.
2. Search Results
The Medline search identified 36 publications, of which 22 articles were selected. The reasons for rejection were as follows: (i) publication not in English (n = 1); (ii) focus not on RA (n = 7); (iii) cost-effectiveness analysis (n = 3); and (iv) undeliverable (n = 3).
The search in the databases DARE, NHS EED and HTA identified 53 records, of which 25 articles were selected. The reasons for rejection were as follows: (i) not original research (n = 2); (ii) methodological paper (n = 1); (iii) focus not on RA (n = 9); (iv) cost-effectiveness analysis (n = 8); (v) not available (n = 4); and (vi) duplication of previously reported data (n = 2). Thirty three remaining articles were identified by hand search.
After excluding irrelevant articles, 65 publications on the burden of illness and general costs associated with RA were identified (table I). Twenty publications were models/simulations/regressions. Nine publications were reviews, of which only one review was published before 1997 (see table I for further details).
Fifteen cost studies covered RA in the US,[23, 24, 25,33,34,44,45,52,53,63,65,67, 68, 69,90] seven studies in Canada,[24,35,41,42,70,71,93] seven studies in Sweden,[4,55, 56, 57, 58, 59,61] three studies in the UK,[11,30,62] two studies in The Netherlands,[5,50] one study in Germany and one study in Finland. Additionally, four publications concerning the US[28,91,92,96] and one from The Netherlands focused only on resource consumption and these were excluded from the analysis.
3. Selected Studies
The focus and approach of the cost analyses reviewed were highly varied. While some publications investigated consequential costs of RA patients like length of inpatient care,[30,35,67] liver biopsy, hip and knee arthroplasty or hip and knee replacement, some studies focused rather on specific settings or perspectives such as from that of the employer, outpatient versus inpatient care, a managed-care setting or health maintenance organisation (HMO). Interestingly, most of the models[21,60,75] as well as some of the empirical cost analyses[4,11,44,55,70,71] focused on specific treatment regimens.
Twenty-four cost analyses from seven countries provided detailed data about direct and/or indirect cost data. Focus, year of costing, publication year, currency and costs per cost domains are summarised in table II. Each study is described below.
Twelve suitable studies performed in the US were selected for review.
Birnbaum et al. estimated the total annual RA costs for a workplace using employer administrative files that covered the years 1995–97 and included employees, spouses, dependents and retirees <65 years of age. 1415 individuals with RA were identified and matched to 469 control employees with similar characteristics but no RA expenditure.
The study revealed three main topics of value for the discussion of the cost impact of RA. First, the yearly costs per patient associated with RA therapy are one-third ($US2337) of the total treatment costs of an RA patient ($US7193), with the other two-thirds arising because of comorbid conditions. Second, treatment costs of RA patients are twice as much as the costs of a non-RA patient ($US3405). Third, the costs for absenteeism and disability per patient-year are one-third ($US1407) of the costs related to employees ($US4245). This indicates that the family disposition of the employees has a high impact on the overall average costs per person.
Interestingly, the costs for RA medication amounted to only $US460 per patient-year, of which $US151 was for NSAIDs, $US137 for disease-modifying antirheumatic drugs (DMARDs), $US102 for gastrointestinal drugs and $US67 for others. The most influential cost component was the medical costs associated with inpatient care ($US4423).
Bloom used the Medicaid Management Information System to identify 527 patients with various forms of arthritis, who were followed up between December 1981 and November 1983. The findings showed that $US211 per patient was being spent each quarter, with $US31 going for inpatient hospitalisation, $US36 for physician/clinic visits, $US78 for outpatient pharmaceuticals and $US66 for the treatment of gastrointestinal adverse effects.
Gabriel et al. accumulated data from two US practice groups and three affiliated hospitals in Olmsted County, MN in 1987. Two key findings that are interesting for the present discussion were identified. First, the total medical care charges per patient were higher for patients with RA ($US3802) compared with patients with osteoarthritis (OA) ($US2655) and nonarthritic patients ($US1388). Second, it was evident that the RA prevalence cohort (n = 397) incurred more costs on average than the nonarthritic patients (n = 25 904) for respiratory, cardiovascular, renal and mental health disorders, although these conditions are not primarily associated with RA.
In 1992, in a separate study, Gabriel et al. analysed the responses of 123 RA patients and 94 non-RA patients recruited from two 200-patient random samples of population-based data resources. The average indirect and nonmedical expenditures by RA patients surpassed the equivalent expenditures for non-RA patients one and a half times ($US2269 vs $US816).
Lanes et al. obtained data from 365 RA patients among 160 000 HMO members that covered a period from July 1993 to July 1994. The total average cost of care per person-year was $US2162, resulting from medications ($US1342), ambulatory care ($US464) and hospital care ($US356). Interestingly, when the RA population was clustered into 14 groups, according to annual cost in dollars per person per year, there was an additional cost of $US400 per ascending group highlighting that the associated costs per patient per year are highly scattered.
Liang et al. evaluated the costs of 148 inpatients with RA or OA recruited in one hospital to be $US683 per patient-year in 1979. The costs for medication, aids and devices were relatively low ($US147) compared with the costs for outpatient visits ($US207), non-hospital costs ($US354) or hospital costs ($US329).
Lubeck et al. performed a cost analysis using a survey of 940 participants with RA in 1981. Costs per patient for medications ($US406), physician visits ($US206) and laboratory tests ($US217) contributed to the lion’s share of outpatient costs ($US1092). Additive inpatient costs ($US913) and costs for transportation, domestic assistance and others ($US527) added up to $US2532 total costs.
Meenan et al. studied 50 RA patients from five different practices in 1976. Interestingly, the total medical costs per patient varied within the five practice settings from a minimum average of $US1349 in a solo practice to a maximum average of $US3584 in a fee-for-service group practice. The average outpatient costs per patient ($US651) for all five groups could be further broken down into costs for medical visits ($US189), other visits ($US31), laboratory tests ($US104), x-rays ($US91), drugs ($US169) and devices ($US68). Inpatient cost was $US1544. Costs for transportation, expenses and other therapies were $US129.
Ward used a top down approach that examined several national surveys like the national ambulatory medical care survey, the national hospital ambulatory medical care survey, the national hospital discharge survey and the national nursing home survey. The reference year was 1994. The authors determined the annual per-person direct costs attributable to RA for: hospitalisation to be $US210; outpatient physician visits $US139; outpatient non-physician visits $US77; transportation to appointments $US42; medication $US918; medical devices $US10; and nursing-home care $US306. When taken as a whole the resulting annual total per-person costs for RA amounted to $US1702.
Wolfe et al. performed two structured interviews with 816 RA patients in five centres in reference to their RA-related physician and hospital costs that were incurred during the preceding 6-month period (reference year 1981). The highest mean hospital and physician cost for patients with RA per centre was $US7864 and the lowest was $US3234.
Yelin used the University of California San Francisco (UCSF) panel to include 1025 patients with RA in 1994. The total average medical care costs among patients with RA was $US4328 per patient. Hospital admissions made up the main part ($US2188) of the total costs, followed by physician visits ($US911), DMARDs ($US323), NSAIDs ($US365), other healthcare providers ($US261) and outpatient surgery ($US280).
In 1996 Yelin and Wanke based another study on the UCSF panel and included 1156 persons with RA. In this study two key findings are worthy of mention. First, the authors disclosed annual medical care per-patient costs for RA per se to be $US5919 compared with $US2582 non-RA medical costs incurred by persons with RA. Second, it was evident that medical care utilisation was not normally distributed, with cost data being highly skewed. The strength of the publication was that it differentiated between the average costs per cost domain as well as declaring the individual unit costs that were used.
Three suitable studies performed in Canada were chosen for review.
Brooks et al. evaluated the costs of a rheumatic disease unit, differentiating between in inpatient costs and costs within an affiliated outpatient department in 1978. They used a top down approach, dividing the full-accumulated costs of the disease unit by the number of inpatients (n = 194) and outpatients (n = 433), respectively. The economic cost per-patient year of inpatient service ($Can2365) was four times that of outpatient service ($Can262). The cost analysis did not examine personal direct and indirect costs related to the medical service, such as travel costs or ambulatory medications.
Interestingly, 60% of inpatient costs resulted from medical staff costs, 10% for accommodation, 8% for tests and procedures, 6% for surgery, 7% for overhead costs and 7% for others. The aim of the study was to provide information for a business assessment and management analysis rather than to determine the cost impact of RA.
Clarke et al. used the Arthritis, Rheumatism and Aging Medical Information System (ARAMIS) database of a rheumatic disease unit in Saskatoon, as well as data from patients with RA in Montreal who had been tracked through to 1994; patients were followed for up to 12 years in the ARAMIS sample and for up to 4 years in the Montreal sample.
The 1063 selected patients received a self-administered questionnaire asking them about their resource utilisation during the previous 6 months. Resources were then multiplied by the corresponding costs. Interestingly, the length of inpatient stay per patient-year varied considerably over time. The authors reported direct costs ($Can3788) and indirect costs ($Can2165) per-patient year within the 1983–89 study interval. In the study interval 1990–94 direct costs were $Can4656 and indirect costs were $Can1597.
The authors worked out that patients with a higher disability index had higher direct and indirect costs. Patients aged 65 years or older had higher direct and lower indirect costs compared with patients younger than 65 years of age.
Maetzel et al. analysed cost data of 482 US patients with RA who were asked about treatment or disease-specific costs in 1999. The authors reported mean yearly per-patient costs for patients treated with leflunomide ($Can1761), methotrexate ($Can1281) and placebo ($Can1324). Annual drug treatment and routine monitoring costs were estimated to be $Can3853 and $Can483 for leflunomide, $Can258 and $Can599 for methotrexate, and $Can0 and $Can55 for placebo.
Two suitable studies performed in Sweden were reviewed.
Jonsson and Larsson estimated mean hospital costs per patient for hip arthroplasty to be SEK34 902 and SEK56 220 for knee arthroplasty. The 1988 study was based on 23 patients with hip replacement and 31 patients with knee replacement.
Jonsson et al. studied the annual costs in a population-based cohort of 82 patients with RA in 1987. Costs included direct and indirect costs. The total average cost per RA patient was about SEK60 000. Patients under age 65 years had higher costs (SEK 79 000) compared with patients aged 65 years and over (SEK 46 000), primarily due to SEK62 000 in indirect costs for the younger patient group. The authors observed that the costs per patient ranged from SEK0 to SEK421 000, with 22 patients having costs of over SEK100 000.
3.4 United Kingdom
Three suitable studies from the UK were reviewed.
Bedi et al. calculated the costs of rheumatology service per inpatient bed-day or per outpatient visit in 1985. Study results were calculated on the basis of 8047 outpatient visits and 12 180 inpatient days in two hospitals. The costs for one outpatient visit were £25.59 in one hospital and £36.90 in the other hospital. Costs for one inpatient bed-day were £49.00 in one hospital compared with £70.71 in the other. Unfortunately, neither the number of hospital days per patient nor the number of visits by patient were reported. Therefore, the results can be used as unit costs only.
Lambert et al. compared the costs for the treatment of day-patients with RA and inpatients with RA in 1992. The costs for hospital, transport, community and indirect costs were lower for day care than for inpatient care. Unfortunately, the costs were calculated on the basis of the total cases being treated and not according to patient-years. Costs per treatment case during 6 months of observation were £1027 for a day patient and £1453 for an inpatient.
McIntosh used the Office of Population Censuses and Surveys (OPCS) Surveys of Disability in Great Britain to include 225 077 RA patients living in private households and 7748 patients living in communal establishments in 1992. They calculated the mean annual medical per-patient costs for RA patients living in communal establishments to be £17 576 compared with £5580 for adults in private households.
3.5 The Netherlands
Two suitable studies from The Netherlands were reviewed.
Van Jaarsveld et al. estimated the annual direct RA-related cost per patient in 1997, using self-reporting questionnaires, based on a 6-year study population of 424 RA patients who had the disease for at least 1 year. Costs included costs of healthcare workers, days admitted to care facilities, medication, monitoring for adverse effects, alternative medicine, home alterations, devices and other direct costs such as travelling expense. The mean annual cost per patient due to RA was estimated to be NLG11550.
Van Jaarsveld and colleagues also calculated the costs per disease duration. Yearly costs per patient were NLG14 455 if disease was present for 1 year, followed by NLG13 800 (2 years), NLG9457 (3 years), NLG6233 (4 years), NLG13 005 (5 years) and NLG11 158 (6 years). The authors similarly determined that the yearly RA-associated costs for inpatients (NLG18 418) were much higher than those for non-hospitalised patients (NLG7561). The mean and median costs differed widely. This skewed distribution is evident in that 96 patients had costs of more than NLG10 000, whereas 267 patients had costs lower than NLG10 000.
Verhoeven et al. compared the RA-related costs of a combination therapy (prednisolone, methotrexate and sulfasalazine) with the costs of sulfasalazine alone. The cost data were based on a multicentre, randomised, double-blind health economic companion trial with a follow-up of 56 weeks. Analysis was performed on an intention-to-treat basis. The direct mean costs per patient in the first year were $US5519 for patients under combination therapy and $US6511 for the sulfasalazine group. The costs in the first 6 months were higher ($US3356 for combination therapy vs $US4218 for sulfasalazine) than the costs in the second 6 months of observation ($US2163 vs $US2293).
Merkesdal et al. investigated the indirect costs per person-year in 133 gainfully employed outpatients with early RA in 1995. The objective was to investigate the different components of indirect costs in the early stages of RA. The authors used the human-capital approach. The total indirect costs for work disability and sick leave were $US11 750.
Rissanen et al. evaluated the costs for joint replacements based on 276 patients with hip replacement and 176 patients with knee replacement in seven orthopedic departments in 1992. Total costs per episode were Fmk45 090 for hip replacement and Fmk49 640 for knee replacement.
The costs per patient differ, to a certain extent, in direct relation to the epidemiological case mix of the study population e.g. with respect to the disease duration, the onset of disease[5,50] or age.
Some studies used observational population-based surveys. Since these surveys are often restricted to specific healthcare settings, like HMOs,[63,91] fee-for-service practices, workplaces, or Medicaid, the subsequent results may yield a strong database bias.
Some studies focused on restricted therapies (e.g. hip or knee replacement),[56,80] or specific pharmacological treatments.[50,70] Therefore, these costs are related to a limited clinical study setting and do not reflect that of a typical RA patient.
Additionally, there are also intra-provider differences, e.g. between hospitals or between inpatient and outpatient facilities[30,35,62] and between communal establishments or private households.
It should also be remembered that some of the studies were performed before 1990.[23, 24, 25,30,34,35,52,56,67] The age of these studies implicates a strong sensitivity of study results with respect to the applied discounting rate and the currency exchange rate. On the one hand, the choice of the discount rate (e.g. 3% vs 5%) is more decisive in studies conducted long ago. On the other hand, the currency rate could be adjusted from the year of study performance or from the year of analysis or submission.
It is also of interest to note which kind of musculoskeletal disorders were reflected in the studies. While some authors distinguished their presentation of results between RA and OA,[23,52,53,63] others did not. The most differentiated approach could be seen in the study of Brooks and colleagues. They reported costs separated for RA, ankylosing spondylitis, psoriatic arthritis, inflammatory polyarthritis, OA, septic arthritis, polymyalgia rheumatica, juvenile rheumatoid arthritis, systemic lupus eryth-ematosus, Reiter’s syndrome, gout, scleroderma and poly- or dermatomyositis. Other diagnoses included Raynaud’s disease, disc and lower back pain, monoarticular arthritis, NYD gene-associated arthritis, temporal arteritis, hypersensitivity, angiitis and fibrositis. All of these diseases were treated in one rheumatic disease unit and had different costs per inpatient admission or outpatient visit associated with them.
Some study results clearly showed that the resource consumption for each RA patient and the associated costs are extremely variable.[5,69] For this reason, it cannot be excluded that the results are randomly influenced by patient selection. This is evident when costs for a single patient with a very long period of hospitalisation are able to noticeably increase the mean value.
Merkesdal et al. reported that most of the 32 reviewed studies (including cost-effectiveness studies and cost-utility studies) reported costs of outpatient medication (n = 31), acute hospital facilities (n = 31) and outpatient visits to physicians (n = 26). All other 35 cost items were reported in a maximum of 17 studies.
This above example serves to illustrate that each study likely contains elements that do not apply to other settings. Therefore, results of each study have to be interpreted individually, with consideration given to the study population, indication, age of the study, database used, type of therapy, setting, level of cost differentiation and data derivation. Furthermore, one should always keep in mind that such reviews, or even meta-analyses of several health economic studies, have to be used and interpreted with due diligence.
It is evident that the costs of RA differ widely with respect to the approach that is used (table II). In the US, the per-patient costs per RA-year ranged from $US1503 to $US16 514. In Canada the per-patient costs per RA-year ranged from $US1272 for a placebo patient to $US5767 for an RA patient receiving treatment. In Sweden the per-patient cost per RA-year was calculated as $US12 420. In Great Britain the per-patient cost per year for an RA-patient in a private household was $US13 934. In The Netherlands the per-patient costs per RA-year were between $US7162 and $US9915 for a specific treatment regimen. In Germany indirect costs for a patient with RA were calculated as $US16 450.
Van Jaarsveld et al. performed one of the most differentiated cost-of-illness studies. Interested individuals who want a comprehensive introduction to the field of health economics in rheumatology should begin with such an article. This review demonstrates the advantage of conducting multinational, well defined health economic studies about the resource consumption of patients with RA, facilitating a high-quality, country-to-country comparison. Unfortunately, the results of such trials are often not as frequently published as the clinical components.
Reviews of various health economic or pharmacoeconomic guidelines reveal differences between these recommendations with respect to study purpose, type of study, perspective, time horizon, method of data capture, costing, discounting and sensitivity analysis.[98, 99, 100] This follows on from the fact that there is no one concerted pharmacoeconomic guideline in effect. Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) has addressed this problem.[101, 102, 103]
On the other hand, there are also problems with respect to the transferability of international health-economic results to national study questions, as Greiner et al. pointed out. Here the authors discussed some of the problems with respect to study design, availability of prices and comparability of resource consumption.
Rothfuss et al. described presence/absence of eight main elements in eleven socioeconomic studies in RA. They concluded that most of the studies did not meet the basic methodological quality standards with respect to presentation of perspective, resource use, cost identification and discounting.
Ruof et al. assessed 15 cost/utilisation questionnaires by describing major characteristics and the presence/absence of cost domains. The researchers discovered differences in the number of total tracked cost items, the presence/absence of self-administered format, the use of monetary or physical units, the recall period and the predominant response category (item lists, free space, yes-no items).
The selected studies that were used in this review confirm the difficulties one can expect to encounter when attempting to deal with these inherent problems systematically. The currently available studies reflect only a small portion of the influencing factors of RA to society and the healthcare system. Because of methodological gaps and/or incongruity, any conclusions drawn should be made with a degree of caution.
It is anticipated that there will be increasingly more Health Technology Assessment (HTA) reports concerning the treatment costs of RA similar to the two extensive reports prepared on behalf of the National Institute for Clinical Excellence.[105,106] Such HTA reports offer sufficient space to adequately describe the composite parts and restrictive elements of different methodological approaches and analyses.
The pharmaceutical industry will recognise that this effort also serves in their best interest and helps to support the endeavours of international committees like the European Collaboration for Health Technology Assessment (ECHTA).
The authors have no conflicts of interest that are directly relevant to the content of this review. The authors would like to thank Fred Sorenson for his review of the manuscript. The authors have provided no information on sources of funding directly relevant to the content of this review.
- 8.Gabriel SE. Epidemiology of rheumatic diseases. In: Harris Jr ED, editor. Kellys textbook of rheumatology. Philadelphia (PA): WB Saunders, 2000: 321–33Google Scholar
- 19.Raftery J. Costing in economic evaluation. BMJ 2003; 320: 1579Google Scholar
- 20.Khan SK, Kleijnen J et al. editors. Undertaking systematic reviews of research on effectiveness - CRD’s for those carrying out or commissioning reviews. CRD report no. 42nd ed. York Publishing, March 2001Google Scholar
- 37.Callahan LF. The burden of rheumatoid arthritis: facts and figures. J Rheumatol 1998; 25 Suppl. 53: 8–12Google Scholar
- 50.Verhoeven AC, Bibo JC, Boers M, et al. on behalf of Combinatietherapie Bij Reumatoide Artritis (COBRA) Trial Group. Cost-effectiveness and cost-utility of combination therapy in early rheumatoid arthritis: randomized comparison of combined step-down prednisolone, methotrexate and sulphasalazine with sulphasalazine alone. Br J Rheumatol 1998; 37 (10): 1102–9PubMedCrossRefGoogle Scholar
- 68.Yelin E. The costs of rheumatoid arthritis: absolute, incremental and marginal estimates. J Rheumatol 1996; 23 Suppl. 44: 4751Google Scholar
- 71.Marra CA, Esdaile JM, Sun H, et al. The cost of COX inhibitors: how selective should we be? J Rheumatol 2000; 27 (12): 27313Google Scholar
- 77.Magnuson S. Treatment of rheumatoid arthritis: does it affect society’s cost for the disease? J Rheumatol 1996; 23 Suppl. 44: 47–51Google Scholar
- 78.Naujoks C. Die volkswirtschaftlichen kosten von rheumy in der bundersrepublik Deutschland [bulletin]. Deutsche Rheuma I Iga, Merkblatt Nr. 6, 7, 1990Google Scholar
- 79.Tevar LIRE. Datos socioeconómicos de las emfermadades reumatológicas. 1st National Congress of Liga Reumatológica Espanol; 1997 March 3-7Google Scholar
- 94.Canadian Coordinating Office for Health Technology Assessment (CCOHTA). A guidance document for the costing process. Version 1.0. Ottawa (ON): Jean-Francois Baladi, 1996: 1–15Google Scholar
- 105.NICE Appraisal Team. The clinical effectiveness and cost effectiveness of celecoxib, rofecoxib, meloxicam and etodolac (COC-II inhibitors) for rheumatoid arthritis and osteoarthritis. London: The National Institute of Clinical Excellence, 2000: 1-109Google Scholar
- 106.Jobanputra P, Barton P, Bryan S, et al. The clinical effectiveness and cost-effectiveness of new drug treatments for rheumatoid arthritis: etanercept and infliximab. Report commissoned by NHS R&D HTR Programme. West Midlands DES, University of Birmingham. London: The National Institute for Clinical Excellence, 2001: 1–138Google Scholar