Time of administration of treatment for hypertension in renal patients

Abstract

Wang et al.1 make a valuable contribution to an aspect of blood pressure management often ignored. In 1961, in blood pressure studies over 24 h, we observed blood pressure did not fall at night in patients with malignant (accelerated) hypertension who had renal failure or were at risk of this.2 Wang et al.1 now provide a meta-analysis of subsequent studies and support our suggestion that ‘non-dipping’ hypertension is important in the generation of morbidity, and that evening administration of medication may be optimal. Their conclusion from the analysis is that evening dosing, when contrasted with morning administration, improves two surrogate markers of long-term benefit: restoring ‘dipping’ in ‘non-dipping’ hypertensives and decreasing urinary albumin. Information on renal function was provided in only two of six studies, totalling 138 patients. There was no significant change in kidney function when measured, but there was no information on the rate of progression of renal dysfunction. The studies did not follow patients for long enough to evaluate ‘hard’ endpoints, death and dialysis, and longer follow-up might have demonstrated renal function benefit from evening dosing. Another surrogate marker of prognosis in chronic kidney disease, urinary albumin excretion (UAE), was reported in only one study included in this meta-analysis. A reduction in UAE with evening dosing was highly correlated with the decrease in nocturnal blood pressure and with the increase in diurnal/nocturnal ratio of blood pressure but no rate of change in UAE was calculated. Renal function is usefully considered as the rate of deterioration. Plasma creatinine, expressed as the reciprocal of plasma creatinine, or the estimated glomerular filtration rate, often demonstrates linear progression, allowing calculation of rates of deterioration in individual patients. Future studies could usefully record more frequent measurements of plasma creatinine and UAE in each individual and identify those responding to changes in time of administration. Time series analysis, utilising Bayesian statistics with Kalman Filtering, were developed for my group to monitor progression of renal function and to detect and evaluate the probability of change.3 If this method is utilised in any prospective trial of change in administration time (or retrospectively if enough data points are available) many fewer subjects will be needed to confirm or refute the conclusions of this meta-analysis. My personal clinical experience provides examples of a change in the rate of progression of kidney failure in some patients when there were therapeutic changes to modify nocturnal blood pressure, I have proposed3, 4 that in the clinic records of renal function and of urinary albumin/creatinine ratio should be considered graphically and also analysed for changes in rate of deterioration when changes, including change in time of administration of medication, have been introduced.