Editorial: Is Rheumatoid Arthritis a Mortal Disease?
J. Askling,M. Holmqvist,L. Ljung
DOI: https://doi.org/10.1002/art.40145
2017-08-01
Arthritis & Rheumatology
Abstract:Is “best practice” management of rheumatoid arthritis (RA) good? Are we as rheumatologists doing better than previously? Are our patients doing any better? In complex chronic diseases such as RA, only repeated evaluations of the long-term outcomes of clinical practice can help us understand whether our improvements to “best practice” pay off in terms of a reduced long-term disease burden. During the past 20 years, new concepts of “best practice” in RA management have been introduced. We have learned the importance of early diagnosis and treatment, careful evaluation of the response to treatment, and treat-to-target strategies, and disease remission has arisen as a reachable target. New treatment options have made it possible to achieve these increasingly ambitious treatment goals (1,2). During the same time period, societal changes such as the declining prevalence of smoking (3) may have impacted the risk of developing RA. Other trends in public health such as declining morbidity and premature mortality from cardiovascular diseases (4) may have impacted the outcome of patients with RA, at least in high-income countries (5). Thus, instead of uncritically attributing time trends in RA outcomes to advancements in the field of rheumatology, evaluations of trends in RA outcomes over time must factor in both RA-centric time trends as well as societal changes in general. In this issue of Arthritis & Rheumatology, 2 studies (6,7), each representing different ends of this spectrum, from patient to society, provide critical insights into the potential benefits of modern rheumatology. The study by Kiadaliri and colleagues used publicly available World Health Organization death statistics from 1987 through 2011 to compare the mortality from RA across 31 (mainly high-income) countries. Kiadaliri et al convincingly demonstrated an average annual 3% decrease in deaths for which RA was listed as the underlying cause. The study also showed how between-country differences in RA mortality declined considerably during the same time period. These estimates must, however, be interpreted in light of the quite profound secular trends in demography and mortality in the general population. Kiadaliri et al demonstrated how the impact of population growth and aging is more than offset by secular trends in other (unspecified) factors such that the balance is negative, resulting in a net decrease in mortality from RA. “Mortality” is in many ways a simple, robust, and indisputably binary outcome. In contrast, “mortality from RA” can be a challenging concept, for several reasons. First, mortality is a function of both the incidence of disease and its outcome. Therefore, a decrease in mortality attributable to RA may reflect a declining risk of developing RA as much as a declining risk of dying of it. Second, “mortality listing RA as cause of death,” the metric used by Kiadaliri et al, is a concept different from “mortality in individuals with RA,” which is the metric used in the study by Gwinnutt and colleagues on the long-term outcomes of RA, which also appears in this issue of Arthritis & Rheumatology. Whereas the former metric is conditioned on RA being the attributed cause of death, the latter metric reflects mortality irrespective of why it occurred and to what it Johan Askling, MD, PhD, Marie Holmqvist, MD, PhD: Karolinska Institutet, Stockholm, Sweden; Lotta Ljung, MD, PhD: Karolinska Institutet, Stockholm, Sweden, and Umeå University, Umeå, Sweden. Dr. Askling has received research grants from AbbVie, Bristol-Myers Squibb, MSD, Pfizer, Roche, AstraZeneca, Eli Lilly, and UCB. Dr. Ljung has received consulting fees, speaking fees, and/or honoraria from Pfizer and Bristol-Myers Squibb (less than $10,000 each). Address correspondence to Johan Askling, MD, PhD, Clinical Epidemiology Unit T2:01, Karolinska University, Hospital Solna, SE-171 76 Stockholm, Sweden. E-mail: johan.askling@ki.se. Submitted for publication March 23, 2017; accepted in revised form May 2, 2017.