Location not quantity of blood pressure measurements predicts mortality in hemodialysis patients

Abstract

Background: Blood pressure (BP) measurements obtained outside the dialysis unit are prognostically superior. Whether it is the greater number of measurements made outside the dialysis unit that correlates with prognosis or whether BPs outside dialysis units are ecologically more valid is unknown.

Methods and results: A prospective cohort study was conducted in 133 patients on chronic hemodialysis. BP was measured by the patients at home for 1 week, over an interdialytic interval by ambulatory recording, and by 'routine' and standardized methods in the dialysis unit for 2 weeks. Up to 6 BPs were randomly selected from a 44-hour recording of ambulatory or 1-week recording of home BPs, such that the dialysis unit BPs were exactly matched to the number of ambulatory or home BPs. The relationship with left ventricular hypertrophy and all-cause mortality was analyzed using receiver-operating characteristic curves and Cox proportional hazards analysis, respectively. Over a median follow-up of 24 months, 46 patients (31%) died. A BP change of 10/5 mm Hg increased the risk of all-cause mortality by 1.22 (95% CI 1.07-1.38)/1.18 (95% CI 1.05-1.31) with the average of the 44-hour recording and 1.20 (95% CI 1.07-1.34)/1.15 (95% CI 1.03-1.27) when up to 6 random BPs from the same ambulatory recording were drawn and averaged. With home BPs the hazard ratios were 1.17/1.15 per 10/5 mm Hg increase in BP with the average of 1-week recording and 1.18/1.13 when up to 6 random BPs were drawn and averaged. Limited duration ambulatory BP monitoring of any 6-hour interval during the first 24 h or 4-day home BP recorded after the midweek dialysis was similarly predictive of all-cause mortality.

Conclusions: In patients on hemodialysis, the location, not the quantity, of the BP recordings obtained outside the dialysis unit is associated with target organ damage and mortality.

Fig. 1.

Relationship of mortality with home blood pressure monitoring obtained on various non-dialysis days. The circles represent the non-dialysis days. Non-dialysis days 3 and 4 represent the weekend. A combination of days (non-dialysis days 2 and over weekend) was statistically significant and is shaded.