Consideration of Multiple Chronic Diseases in Randomized Controlled Trials

To the Editor: Before approval of a new drug, the US Food and Drug Administration (FDA) requires phase 3 clinical trials demonstrating efficacy. Usually trials are conducted in homogeneous populations and rarely include individuals with multiple chronic conditions. In the United States, more than 50% of people with chronic conditions have 2 or more diseases.¹ Twenty-eight percent of the population lives with multiple chronic conditions, including 2 of 3 older individuals.² Multiple chronic conditions account for 66% of the country's overall health expenditures² and more than 95% of Medicare expenditures.³

This study tested the hypothesis that patients with multiple chronic conditions are underrepresented in randomized controlled trials (RCTs) published in high-impact journals and that there are no significant differences by type of journal or changes over time.

We examined all RCTs published from January through March in 1995, 2000, 2005, and 2010 in the 5 highest impact factor general medical journals (BMJ, Canadian Medical Association Journal, Journal of the American Medical Association, Lancet, and New England Journal of Medicine ]] Table 1) and specialized journals that focus on the most prevalent chronic conditions (American Journal of Respiratory and Critical Care Medicine, Annals of General Psychiatry, Circulation, Diabetes, and Journal of Clinical Oncology ]] Table 2).

Trials on the effects of interventions for chronic conditions (any incurable or long-lasting illness) in adults were eligible. Pediatric studies, posttrial follow-up studies, or secondary subgroup analyses were excluded. Data for each trial were extracted and coded by 2 independent raters. When necessary, discrepancies were resolved by consensus and arbitration by a third individual. χ² Tests were used to compare the proportion of trials in general and specialized journals that considered multiple chronic diseases implicitly (used names of specific conditions) or explicitly (mentioned terms such as comorbidities or coexisting diseases), or that used them as a selection criterion. Two-sided tests of significance were performed for all statistical analyses using SPSS version 18.0 (SPSS Inc)]] P values lower than .05 were considered statistically significant.

Of the 284 trial reports included in the analysis, 165 RCTs (58%) were published in specialized journals. Two hundred trial reports (70%) mentioned multiple coexisting diseases]] general medical journals described them more often than specialized journals (72% vs 69%]] P = .02).

Of the 200 trial reports that mentioned multiple coexisting diseases, its presence affected the eligibility of participants in 190 RCTs (95%). Patients with polypathology were excluded in 179 of the trial reports, which represent 63% of the 284 RCTs identified, 90% of the 200 RCTs that mentioned coexisting diseases, and 94% of the 190 RCTs that considered polypathology as part of the selection process. Six RCTs (2.1%) included patients with multiple chronic diseases explicitly. There was no difference across the publication years (65% in 1995, 67% in 2000, 74% in 2005, and 75% in 2010]] P = .49).

Few RCTs published in the last 15 years included patients with multiple chronic conditions. Although the external validity of this finding is limited by the small sample of journals and the short period covered by the study, it invites reflection about the risk of unintended harm from inappropriate generalization of trial results conducted in populations with a single disease.⁴ Given the possible drug-to-drug, drug-to-disease, and disease-to-disease interactions that remain unexamined, most of the evidence gathered to date by RCTs is of limited value to guide decisions about medication use by patients with multiple chronic diseases.⁵

It may be beneficial for the FDA to consider large observational studies as sources of supplementary data on the value of interventions for multiple coexisting diseases]] have manufacturers include subgroups of patients with the most frequent combinations of diseases in their drug development processes]] increase the number of n-of-1 trials to assess the safety of new drugs added to polymedicated patients with conditions that could be temporarily alleviated]] and have postmarketing surveillance studies include risk stratification and standardized outcomes that could allow meta-analyses across populations.

Author Contributions: Dr Jones 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 concept and design: Jadad, To, Jones.

Acquisition of data: To, Emara, Jones.

Analysis and interpretation of data: Jadad, To, Emara, Jones.

Drafting of the manuscript: Jadad.

Critical revision of the manuscript for important intellectual content: Jadad, To, Emara, Jones.

Statistical analysis: To, Emara, Jones.

Administrative, technical or material support: To.

Study supervision: Jadad, To, Jones.

Funding/Support: Dr Jadad was supported by funds from the Canada Research Chair in eHealth Innovation, University of Toronto, and the University Health Network. Mr To received research funding from the University of Toronto.

Role of the Sponsor: The University of Toronto and the University Health Network had no role in the design and conduct of the study]] collection, management, analysis, and interpretation of the data]] and preparation, review, or approval of the manuscript.