In this issue of Arthritis Care & Research, Yelin et al examine the University of California, San Francisco Rheumatoid Arthritis Panel for differences in the use of specific antirheumatic treatments, especially anti–tumor necrosis factor (anti–TNF) agents, among patients in health maintenance organizations (HMOs), other managed care health plans (preferred provider organizations or point of service plans), and fee-for-service plans (Yelin EH, Trupin LS, Katz PP. Impact of managed care on the use of biologic agents for rheumatoid arthritis. Arthritis Rheum 2005;53:423–30). The authors have been following these patients since 1982–1983 using annual telephone interviews. Their data now includes an impressive 11,669 person-years of observation, although there has unfortunately been an ∼65% attrition (∼25% died, ∼30% declined participation, and ∼10% were lost to followup). There also appears to be an ∼20% turnover during each year, with 6–10% patients starting and 8–10% stopping anti–TNF agents, for example.
This study has significant strengths, such as a long-term followup and a clear and thorough analysis, which the authors achieved by attempting to account for the above-mentioned turnover. It also has some weaknesses, including the very significant attrition, the fact that all followups are by telephone interview (introducing memory biases), and the fact that one of the major control groups (the fee-for-service group) comprises only 92 patients. Taking into consideration these weaknesses, the authors point out that allocation of certain medications differs between patients in HMOs, other managed care plans, and fee-for-service plans; namely, patients in HMOs use fewer biologic agents or cyclooxygenase 2 inhibitors than patients in fee-for-service plans and are less likely than patients in other managed care plans to start taking biologic agents. This implies that this difference is the result of a bottom-line orientation (i.e., the agents are expensive and therefore are used less frequently in patients in HMOs). This analysis is used in other studies that show no differences in other outcomes such as hospitalizations or joint surgeries. The results of this study are of real interest because they investigate beyond the usual outcomes of joint replacements or hospitalizations.
One of the reasons that hospitalizations and joint replacements are used as outcomes is that they are clear (dichotomous) and easy to record. Perhaps this study provides proof that other outcomes are also important for patients with chronic conditions, such as rheumatoid arthritis (RA). It may well be the tip of an iceberg in that a difference in the cost of biologic agents may represent a surrogate for quality of life or productivity that goes beyond the high-altitude simple measures of cost of care represented by hospitalizations or joint surgeries. Could this difference, for example, be a surrogate for the ability to maintain job productivity? Certainly such a connection is possible; less use of biologic agents could lead to less function, resulting in less ability to maintain employment (either in or out of the home). In fact, employment status has been examined in a previous study regarding biologic agents (Yelin E, Trupin L, Katz P, Lubeck D, Rush S, Wanke L. Association between etanercept use and employment outcomes among patients with rheumatoid arthritis. Arthritis Rheum 2003;48:3046–54).
The study by Yelin et al does not change the focus of the research. It still examines cost of care from the societal, or at least from the managed care organization's point of view rather than the individual's point of view. Although the larger view (societal) is valid, so is the individual's view. By examining biologic use, this study may open the way to examining cost from the individual's view point. Do patients who use biologic agents use fewer antidepressants, anxiolytics, or analgesics? Do they use more antibiotics? How do biologic agents affect depression, home function, or social interactions? Although the Health Assessment Questionnaire, the Arthritis Impact Measurement Scales, and the Short Form 36 approach these issues and some of their domains, more careful examination of these areas is certainly justifiable and worthwhile from the patient's quality-of-life perspective.
It may be necessary to include more complete examinations in these long-term registries and observational cohorts, rather than simply conducting telephone interviews, so that other aspects of individual outcomes can be examined. Databases such as Paulus' Rheumatologist Consortium, the Consortium of Rheumatology Researchers of North America database, or Rheumatoid Arthritis Disease-Modifying Anti-Rheumatic Intervention and Utilization Study include more complete data on physical examinations, depression, and laboratory results, and they may represent an additional route toward understanding the effect of RA and its treatment on society and on patients with RA.
Perhaps this study's greatest importance is not in the findings per se, but in the authors' attempt to begin to reflect on aspects of life that are different from those that are usually examined (such as hospitalizations, mortality, or joint replacements). It is clear that there continues to be a need to search for valid ways to reflect the quality of life of individuals with chronic diseases, such as RA, that do not rapidly curtail the duration of life (although they may do so, as well) but rather change the individual's productivity and “joi de vie.”