- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
Rheumatoid arthritis (RA) presents an enormous economic burden on society in terms of the direct medical costs, the indirect costs that include lost wages and a caregiver's time, and the intangible costs of pain, fatigue, lowered self-esteem, or other psychological problems (1–7). Using data from the 2003 Medical Expenditure Panel Survey (MEPS), the Centers for Disease Control and Prevention estimated that arthritis and other rheumatic conditions collectively cost the US economy nearly $128 billion ($80.8 billion in direct and $47.0 billion in indirect costs, excluding pain and suffering) (8). This spending equaled approximately 1.2% of the US gross domestic product (GDP) in 2003 (8). Furthermore, in this cohort, the average per-person total direct expenditure in 2003 was $1,752, of which ambulatory care ($914) was the highest subcategory of expenditure, followed by emergency department and inpatient services ($352), prescriptions ($338), and other costs ($146) (8). Information about incremental economic burden of chronic diseases can inform health policies aimed at reducing disparities and unnecessary medical expenditure and quantify the excess economic burden of the disease borne by society. It can also help define and improve treatment and disease management guidelines.
Not much is explicitly known about the economic burden distinctively associated with RA, specific to the US population. In chronic conditions such as RA, primary cost drivers tend to be in demand for new health care technology, pharmacologic treatment, and surgical procedures. Prior to the introduction of biologic response- modifying agents, medications constituted the second largest component of RA-related costs, accounting for approximately 8–24% of total medical expenditure (9, 10). Among the medications, disease-modifying antirheumatic drugs (DMARDs) were responsible for approximately two-thirds of the total drug cost and nonsteroidal antiinflammatory drugs for most of the remainder (9). Recently, the introduction of the expensive biologic DMARDs in RA has more than doubled prescription spending, and the future expectation is that this expenditure will continue to increase (11–13). However, since current therapeutic options are not curative, one-third of RA patients may require surgery, although this rate may have decreased considerably in recent years for patients below age 60 years (1, 14). Escalation in pharmacy expenditures has increased the interest in the comparative analyses of biologic and traditional DMARDs (15–17).
Given these remarkable changes in the magnitude and distribution of spending in RA, it is important to understand the impact of direct medical expenditures attributable to RA (17). There are very few studies reporting total and subgroups of expenditures specific to RA based on data from recent years, and none of these studies have used samples that are nationally representative of the US population (18–24). There is no literature that can quantify the incremental expenditure of RA as compared to the US general population to inform the incremental economic burden presented by RA. Results from non-US studies have limited use to enlighten the economic impact of RA in the US due to significant differences in structure and financing of health care markets (20, 21).
The objective of this study was to evaluate the incremental total annual expenditure associated with RA based on a nationally representative sample from the MEPS. Secondary objectives were to assess the expenditures associated with pharmacy, office-based visits, emergency department visits, hospital inpatient stays, and residual expenditure, including hospital outpatient visits, ambulatory hospital visits, home health care, dental care, vision care, and other medical supplies and devices.
Significance & Innovations
This study quantifies the incremental expenditure of rheumatoid arthritis (RA) as compared to the US general population to inform the incremental economic burden presented by RA, using postbiologic era contemporary data from samples that are nationally representative of the US population.
Key findings from our study indicate that RA exerts a significant incremental economic burden of $22.3 billion (in 2008 US dollars) annually on US health care. Furthermore, in less than a decade, the primary driver of this incremental expenditure in RA has shifted from hospital expenditure to pharmacy expenditure.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
The mean age of the RA patients was 61.0 years (95% CI 59.2–62.5 years), while that of the control group was 47.4 years (95% CI 46.9–48.0). Based on Table 1, the majority of the RA patients were female (61%), white (81%), married (55%), unemployed (62%), and residing in a metropolitan area (77%). As compared to the RA patients, the control group was more likely to be younger, employed, privately insured, physically active, and less likely to be obese. While 26.6% of the RA patients had only 1 comorbid condition and 24.0% had 2 or more comorbid conditions, in the control group, approximately 71.5% had no comorbid conditions, 22.1% had exactly 1 comorbidity, and 6.5% had 2 or more comorbid conditions.
Table 1. Descriptive statistics on the sociodemographic, health status, and expenditure variables*
|Variables||RA (n = 5.8 million)||Non-RA (n = 190 million)|
|Age, years|| || |
|Race/ethnicity|| || |
|Family size|| || |
| 3 or 4||23.4||34.1|
|Education|| || |
| No degree||23.9||14.6|
| GED/high school diploma||50.8||49.6|
| Bachelor's degree||12.7||18.0|
| Other degree||7.8||8.2|
|Family income†|| || |
| Near poor||8.4||4.0|
| Low income||15.3||12.9|
| Middle income||30.8||30.8|
| High income||29.9||41.6|
|Reside in metropolitan statistical area||76.8||84.1|
|Region|| || |
|Insurance coverage|| || |
|Health behavior|| || |
| Physically active§||39.2||57.1|
|Chronic conditions|| || |
| Myocardial infarction||8.2||2.9|
| Congestive heart failure||3.5||1.0|
| Peripheral vascular disease||3.9||0.6|
| Cerebrovascular disease||5.9||2.3|
| Chronic pulmonary disease||23.7||12.7|
| Peptic ulcer disease||1.7||0.5|
| Liver diseases||1.7||0.7|
| Diabetes mellitus||22.5||10.5|
| Renal disease||0.8||0.4|
|Annual expenditure outcomes, median (IQR) dollars|| || |
| Total||4,677 (1,628–12,165)||1,229 (268–4,199)|
| Pharmacy||1,418 (357–3,750)||126 (0–933)|
| Office visits#||720 (222–2,212)||252 (20–919)|
| Emergency department**||0 (0–0)||0 (0–0)|
| Inpatient††||0 (0–0)||0 (0–0)|
| Residual‡‡||170 (0–1,155)||84 (0–494)|
Based on unadjusted analyses (Table 1), average expenditure of RA patients was nearly 1.5- to 3-fold higher than that of the non-RA control cohort in the various subgroups of expenditures. On average, as a proportion of total expenditure, pharmacy expenditure of RA patients was approximately 39.9% (29.1% in non-RA), office-based visits expenditure was 30.4% (35.3% in non-RA), emergency department expenditure was 2.3% (4.5% in non-RA), hospital inpatient expenditure was 9.4% (5.8% in non-RA), and residual expenditure was 18.0% (25.3% in non-RA patients).
After adjusting for sociodemographics, employment, insurance, health behavior, and 13 chronic conditions in the GLM, the incremental total expenditure was significantly higher in the RA patients as compared to the control group. Additionally, the incremental expenditure in all the subgroups of expenditure was also significantly higher in the RA patients as compared to the control group. The average total annual expenditure in the RA patients was $13,012 (95% CI $1,737–$47,081), while that in the control group was $4,950 (95% CI $567–$17,425) (Table 2). The additional/incremental annual total expenditure of the RA patients as compared to non-RA controls was $2,085 (95% CI $250–$7,822). RA patients also had significantly higher average pharmacy expenditure of $5,825 (95% CI $446–$30,998) that was on average approximately $1,380 (95% CI $94–$7,492) higher as compared to the controls. While the incremental office-based visits expenditure was higher by $587 (95% CI $83–$1,939) in RA patients, the incremental differences between inpatient and emergency department visits were $258 (95% CI $6–$1,208) and $13 (95% CI $5–$35), respectively. Based on the survey bootstrapped CIs, the incremental differences were statistically significant in all expenditure categories at an alpha level of 0.05.
Table 2. Adjusted average and incremental annual expenditures (in 2008 US dollars) in rheumatoid arthritis (RA) and non-RA control group*
|Annual expenditure outcomes||RA (n = 5.8 million)||Non-RA (n = 190 million)||Incremental difference|
|Total||13,012 (1,737–47,081)||4,950 (567–17,425)||2,085 (250–7,822)|
|Pharmacy||5,825 (446–30,998)||1,264 (83–6,358)||1,380 (94–7,492)|
|Office visits†||2,838 (464–9,510)||1,221 (156–3,607)||587 (83–1,939)|
|Emergency department‡||279 (84–1,004)||183 (70–492)||13 (5–35)|
|Inpatient§||5,021 (141–30,519)||1,765 (47–8,461)||258 (6–1,208)|
|Residual¶||2,098 (226–7,276)||1,036 (100–3,337)||261 (27–934)|
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
The objective of this study was to evaluate the incremental direct annual medical expenditure associated with RA based on a nationally representative sample of the US population from the contemporary postbiologic DMARD period. A key finding from our study indicates that in the postbiologic era, the primary driver of the incremental total expenditure associated with RA has shifted from hospital stays to pharmacy expenditure. These findings are noteworthy since about a decade ago, the most important driver of direct costs in the US and internationally was hospitalization, especially in moderate and severe RA, while costs of medication represented only a minute proportion of these direct costs (37). This shift in the primary cost driver is very likely related to early aggressive DMARD treatment, as well as the introduction of expensive biologic DMARDs and their ever-increasing use in recent years. These findings further reinforce the need for evidence-based judicious prescribing of expensive DMARDs as proposed in the recent American College of Rheumatology and European League Against Rheumatism recommendations (15, 17, 38). Our results also quantify the enormous incremental economic burden posed by RA and can inform providers, decision makers, and payers how current dollars are being spent in the treatment of RA patients.
Consistent with the literature, the majority of the RA sample was comprised of women of an older age. Large majorities of these patients were unemployed, which may be due to the high prevalence of disability and functional limitations common in RA patients (39). Given the functional limitations posed by the disease, the RA patients also had a lower likelihood of being physically active, which in turn may have resulted in a higher proportion with obesity. Modifiable health risk factors such as smoking and obesity are responsible for a majority of the health care utilization and expenditure related to chronic diseases. It has been reported that the average US health care expenditure for people who were obese was $5,148 compared to $3,636 for those who were overweight and $3,315 for people who were normal weight (40).
While a majority of RA patients in our study had at least 1 comorbid condition, many of them had multiple severe comorbid conditions, which is similar to results reported by Sullivan et al (39). In contrast, only a small number of the control cohort had multiple comorbidities. The most prevalent chronic conditions in RA patients were chronic pulmonary disease, diabetes mellitus, and cancer, which seem consistent since all these conditions are associated with underlying inflammation. Evidence suggests that treating patients with multiple comorbid chronic conditions costs up to 7 times as much as treating patients with a single chronic condition (41).
Our results on total expenditure (Table 1) are comparable to those reported recently by Wolfe and Michaud (2009), where they estimated the out-of-pocket expenses and their burden on patients with RA from the National Data Bank for Rheumatic Diseases (NDB) (18). Depending on the level of out-of-pocket expense burden, the average total medical expenditure reported in that study ranged between $9,470 and $11,194 (2007 USD). However, the results on subgroups of expenditure reported in their study, especially for pharmacy and inpatient expenditure, were very different. Inpatient costs ranged between $900 and $1,599, while prescription spending was between $7,522 and $8,790 depending on the out-of-pocket expenses level. In our study, the average unadjusted inpatient expenditure was $2,791, while average prescription spending was $3,099 (both amounts deflated to 2007 USD). There are differences between the duration with disease, age, sex, ethnic, and educational characteristics of the US population as compared to the NDB population, and they may not be comparable (22).
The primary driver for the incremental expenditure in the RA patients was the additional pharmacy expenditure, which accounted for approximately 66% of the incremental total expenditure of $2,085 (95% CI $250–$7,822). Using data from MEPS over multiple years, Cisternas et al (42) and Yelin et al (8) have shown that the average prescription expenditure for patients with arthritis and other rheumatic conditions (including ICD-9 codes 274, 354, 390, 391, 443, 446, 710–716, 719–721, and 725–729) had nearly doubled since 1997; it is likely that RA in particular would be treated much more aggressively as compared to other forms of arthritis (8, 11, 12, 42). Moreover, comparing to trends reported by Cisternas et al (42) in contrast to arthritis and other rheumatic conditions, the average prescription expenditure of RA patients in our study was even higher than the average inpatient expenditure ($3,214 versus $2,894). These results highlight the need for additional data on comparative and cost effectiveness of DMARDs in RA that can reduce suboptimal or inefficient spending and reduce disparities. Currently, the majority of RA patients are treated with nonbiologic DMARDs, although the rate of use of biologic DMARDs has been increasing (39, 43–45). Since the cost of biologic therapy can range between $15,000–$25,000 USD per year depending on dose and other factors, it is important to identify subgroups of patients who will most likely benefit from these expensive treatments.
Based on our model, summing the total expenditure over the entirety of RA patients identified in the study, the total direct medical expenditure of RA on the US civilian population was $73.4 billion USD in 2008. To put into context, this expenditure was equivalent to 0.5% of the GDP of the US in 2008 and 6.5% of the approximately $1.14 trillion USD spent on health care in the US in 2008 by the civilian noninstitutionalized population. Additionally, the incremental or additional direct total expenditure of RA to the US was $22.3 billion USD annually or 0.16% of the GDP in 2008. These results indicate the enormous impact of RA on US society, where approximately 0.5–1.5% of the US population that has RA is consuming the equivalent of 0.5% of the GDP in just direct medical expenditure. The true societal burden of RA is much larger since we have not incorporated the impact of RA on out-of-pocket expenses, productivity, disability, limitations, health-related quality of life, and mortality.
Comparing the observed expenditures in RA patients to the overall arthritis and other rheumatic conditions reported by Yelin et al (8), it is evident that the per person average and incremental expenditure in RA is much higher, and as its subgroup, RA exerts a significant proportion of the net economic burden of arthritis and other rheumatic conditions on US health care. Yelin et al estimated that the aggregate incremental expenditures attributable to arthritis and other rheumatic conditions increased from $64.8 billion USD in 1997 to $80.8 billion USD in 2003 (8, 46). By inflation adjusting these estimates (8, 46) to 2008 USD using the medical component of the consumer price index, we get approximately $99 billion USD as the aggregate incremental expenditures attributable to arthritis and other rheumatic conditions. Ignoring the gains due to advances in health technology and the availability of new therapeutic treatments between 2003 and 2008, the incremental expenditure attributable to RA ($22.3 billion USD) seems to contribute approximately 22.5% of the entire incremental medical expenditure burden ($99 billion USD) attributable to arthritis and other rheumatic conditions. Given that the economic impact exerted by arthritis and other rheumatic conditions is similar to that of a moderate recession, the individual contribution of RA in this context is enormous (46).
These results should be interpreted with caution as they may be plagued similar to a prevalence-based cost of illness analysis. From an economic perspective, marginal costs (or cost effectiveness) of DMARD treatment are much more informative as compared to the incremental expenditure associated with RA. For example, even if a cure for RA were discovered, it does not necessarily imply that we would save the $22.3 billion USD additionally spent on RA patients. Essentially, the costs associated with the cure and remission would remain. However, our estimates of incremental expenditures associated with RA do inform rheumatologists and other providers, policy or decision makers, and payers how these dollars are currently spent in the treatment of RA patients.
There are potential limitations related to the retrospective observational study design. Since MEPS is based on self-report, prevalence estimates of RA and other chronic conditions may be misreported and this could vary systematically by cohort status, race, and ethnicity. In RA, disease progression and disease severity are important factors that can lead to heterogeneity of estimated costs, although our study could not control for these effects. We also study patients at a single point in time, whereas a longitudinal study could reduce the effects of unobserved confounding and provide insights to the long-term burden posed by this debilitating disease. We have only estimated the expenditure models using GLM with log link and gamma distribution; hence, changing the econometric models using different link functions could change the results. For example, although misspecification of the gamma distribution would only lead to efficiency losses, misspecification of the log link used in the study could produce biased results.
Finally, our study only evaluated the direct medical expenditure associated with RA, which only provides a lower bound on the true societal burden imposed by this disease. Due to its chronic nature, coupled with long-term disability occurring during a patient's most productive period, the indirect costs associated with RA are significant (47). Indirect costs such as productivity losses and time spent by family members and caregivers to support the patient can be 3 to 4 times higher than direct medical expenditure, and is a greater burden on society (47). Additionally, the intangible costs of pain, fatigue, lowered self-esteem, or other psychological problems imposed by RA are also significant economic burdens on society. Future studies should additionally incorporate productivity losses and health-related quality of life decrements experienced by RA patients in longitudinal nationally representative cohorts.
Despite these limitations, our results show that RA exerts substantial incremental economic burden on US health care. Furthermore, in less than a decade, the primary cost driver for RA has shifted from hospital expenditure to pharmacy expenditure. To reduce the impact of the incremental pharmacy expenditure in RA, future health policies should be designed to promote judicious use of the expensive biologic agents in the patients most likely to benefit from these agents and limit treatment only during the phases of the disease when the biologic agents can be most effective and are necessary.