Version of Record online: 27 MAR 2008
Copyright © 2008 by the American College of Rheumatology
Arthritis Care & Research
Volume 59, Issue 4, page 600, 15 April 2008
How to Cite
Côté, P., Cassidy, J. D. and Carroll, L. (2008), Reply. Arthritis & Rheumatism, 59: 600. doi: 10.1002/art.23522
- Issue online: 27 MAR 2008
- Version of Record online: 27 MAR 2008
To the Editors:
We would like to thank the editors for the opportunity to respond. Dr. Croft states that we assumed that claim closure is a proxy for recovery; however, we have answered this assumption in 2 previously published letters to the editors in other journals (1, 2). As clearly indicated in our article in Arthritis Care & Research, we studied the validity of claim closure as an indicator of health recovery and reported a strong and independent relationship between the rate of recovery, as measured by self reports of neck pain intensity, physical functioning, depressive symptomatology, and the rate of claim closure (3, 4).
Dr. Croft speculates that using claim closure as a proxy for recovery led to outcome misclassification. However, he fails to provide any evidence of the presence of misclassification bias. Moreover, he fails to explore what impact this bias would have had on our results. If misclassification was present, it would have likely been nondifferential and led to an underestimation of the measured effect sizes. In other words, the true association between patterns of care and delayed recovery would be even stronger than what we reported in our study.
We are baffled that Dr. Croft labels our study as retrospective. We designed a cohort study where the exposure (patterns of care) clearly preceded the outcome (time to recovery). Cohort studies that are conducted in a properly defined population at risk are by design prospective. Our cohort design allowed us to extensively control for confounders including injury severity and other known prognostic factors. A description of the potential confounders included in our analysis (including 26 markers of injury severity) is demonstrated in the Patients and Methods section of our article.
We understand that our results are challenging to clinicians primarily invested in the treatment of whiplash injuries. However, we strongly believe that advancement in knowledge is only achieved through the conduct of methodologically rigorous investigations. These investigations occasionally overturn expectations based only on anecdotal clinical observations. We encourage Dr. Croft to conduct well-designed large population-based cohort studies or large randomized controlled trials that will support or refute our findings.
Pierre Côté DC, PhD*, J. David Cassidy DC, PhD, DrMedSc*, Linda Carroll PhD, * University of Toronto and Toronto Western Research Institute and Rehabilitations Solutions, University Health Network, Toronto, Ontario, Canada, University of Alberta, Edmonton, Alberta, Canada.