December 22, 2011

Bedtime Dosing of Antihypertensive Medications Reduces Cardiovascular Risk in CKD

.   .   

This open-label single-center trial examines the impact on cardiovascular events in hypertensive patients with CKD treated with ≥1 hypertension medication at bedtime vs taking all anti-hypertensive medications upon waking.
See the accompanying editorial:
Free full text kindly provided by the American Society of Nephrology

Related Articles:

  • Anonymous

    “Bedtime Dosing of Antihypertensive Medications Reduces Cardiovascular Risk in CKD” by Ramón C. Hermida et al catches the eye as one peruses the December issue of JASN. While it seems to suggest an easy and sure-shot remedy, careful reading however, leaves many questions unanswered. While it is stated to be a randomized clinical trial, several essential tenets of such a trial are missing from the article.
    From the description on materials and methods, the reference population studied is not clear, nor how they were recruited. The usual flow diagram describing the recruitment and enrolment into the study is missing. The acceptance rate for the study seems fortuitously high – a rather uncommon scenario compared to other randomized clinical trials. It is unclear which stages of CKD were excluded and why. It is also unclear whether all subjects who tested positive for HIV infection were excluded or only those with AIDS defining criteria at the start of the study.
    Further, the process of randomization has not been described adequately.  While at first, the article  seems to suggest simple randomization, the subsequent description in the methodology section indicates that each class of anti-hypertensive medication was randomized separately into either a bedtime dose or a morning dose.  If then, the bedtime dose group consisted of all subjects on at least one medication at bedtime, and if the distribution of the numbers of antihypertensives was the same at the beginning of the trial as at the end, this group would be about 74% of the total, not half as reported.  A sample size calculation is conspicuous by its absence.
    The actual study intervention is also unclear since the baseline medication of the subjects has not been described. It is unclear whether the subjects’ medication was prescribed to follow a certain protocol or whether their existing regimen was altered to either a bedtime or morning dose of each class of drug.  One wonders why alpha blockers were not similarly randomized..
    There were issues with the description of the follow-up procedures. Variation in follow-up in those who required alteration in their medication has the potential to cause error in reporting because of potential detection bias. Unfortunately, the length of follow-up for each group was not described in the paper. The trial is described as being blinded but, given all of the exceptions to blinding, it is hard to see who was actually in the dark regarding group allocation.
    While the authors suggest that blood pressure parameters by ambulatory blood pressure monitoring was better in the bedtime group, diastolic blood pressure mean – both awake and mean over 48 hours – were actually better in the daytime group, making the inference of “better” blood pressure control in the bedtime dose group perhaps not completely accurate.
    Finally, there are issues with the generalizability of the results. While the study’s title is suggestive of hypertension in CKD, by exclusion of secondary hypertension entirely and exclusion of patients with ‘kidney failure’ from the trial makes the study less generalizable. Is it a given that all the subjects had hypertension predating their albuminuria? Not permitting divided doses of medication also reduces the clinical relevance of the study.
    In conclusion, the take-home message of the trial purports to be that event-free survival for both total cardiovascular events and the composite end-point were better in the bedtime group.  While these are intriguing observations, for the methodological issues mentioned above, they cannot yet be judged as completely valid or generalizable and one awaits more robust data to throw more light on this subject.
    Santosh Varughese
    David Naimark
    for the Sunnybrook NERD (nephrology epidemiological research discussion) group