The Journal of Medical Virology has just published a study called “Effect of using a structured pacing protocol on post-exertional symptom exacerbation and health status in a longitudinal cohort with the post-COVID-19 syndrome.” The investigators are from the University of Leeds and the Leeds Community Healthcare NHS Trust.
Post-COVID-19 syndrome (PCS) is the name the investigators are using for what is commonly called long Covid. Post-exertional symptom exacerbation (PESE) is an alternate and arguably more accurate name for what has generally been called post-exertional malaise (PEM).
Here’s the description from the abstract: “This prospective study investigated the effect of a 6-week structured World Health Organization (WHO) Borg CR-10 5-phase pacing protocol on PESE episodes and quality of life in a cohort of individuals with long-standing PCS (average duration of symptoms was 17 months).”
(Personally, I’ve always preferred the Klingon and Romulan CR-10 5-phase pacing protocols over the Borg version. But more on that matter below.)
This a small study–31 patients underwent a six-week rehabilitation program that was essentially a form of graded exercise therapy and included weekly telephone calls with a clinician. The investigators found a reduction in reported PESE episodes and an improvement in a quality-of-life measure. But these outcomes–both subjective–would have been infused with an unknown amount of bias.
Todd Davenport, a professor of physical therapy at University of the Pacific in Stockton, California, slammed the study for statistical and other failings in a stream of tweets that @lucibee helpfully pulled into a thread. For one, he pointed out that 20% of the participants had reported no episodes of PESE at baseline—a fact that would likely skew any findings. @lucibee also added her own pertinent observations on questionable aspects.
The study included no comparison group. This means that it could only yield descriptive findings. For example, the study could legitimately report that 31 patients did X, Y, and/or Z, and it could legitimately report on differences between the data provided by participants before and after the intervention. By virtue of its design, however, the study could provide no insight into the impact or effect of the intervention.
In particular, it is impossible to know if any reported changes might have happened on their own as part of the natural course of illness. Moreover, rather than being helpful, the intervention could be completely neutral, or even harmful. Perhaps, in the absence of the intervention, study participants would have done even better than they did. With no comparison group, no inferences can be drawn in any direction.
Therefore, any language indicating causality, or suggesting that the study is examining the intervention’s impact or effect, is unwarranted. The investigators acknowledge this limitation in the limitations section—but ignore it in the rest of the study...............
