Blood pressure trend in hospitalized adult dengue patients

Abstract

Background: Monitoring of blood pressure is an important part of management of dengue illness. Large scale studies of temporal trend of blood pressure in adult dengue are lacking. In this study, we examined the differences in time trend of systolic (SBP) and diastolic blood pressure (DBP) in patients with and without severe dengue (SD), dengue hemorrhagic fever (DHF) and pre-existing hypertension, and elderly versus non-elderly patients.

Methods: We studied a retrospective cohort from 2005 to 2008 of 6,070 hospitalized adult dengue patients confirmed by polymerase chain reaction or clinical criteria plus positive dengue serology. Dengue severity was defined according to World Health Organization 1997 and 2009 guidelines. We used Bayesian hierarchical Markov models to compare the daily mean SBP and DBP between different subgroups. Analysis was conducted by day of defervescence (denoted as day 0), and day of illness onset (denoted as day 1) respectively.

Results: SBP decreased to a nadir during the critical phase before defervescence and was significantly lower for patients with SD or DHF, compared with patients without SD or DHF. DBP increased marginally more for patients with SD or DHF in the critical phase before defervescence. By day of defervescence, comparison of patients with and without SD showed significant difference in SBP from day -6 to day +6, except days +1, +3 and +5, and similarly in DBP except days 0, and +4 to +6. Comparison of patients with and without DHF showed significant difference in SBP from day -6 to day -1, but for DBP, significant difference was noted from day -6 to day +6, except day -2 to day 0. By day of illness, SBP differed significantly between patients with and without SD from illness days 1 to 10, and DBP from illness days 7 to 12. Between patients with and without DHF, SBP differed significantly on illness days 1, 2, 4 to 7, while DBP from days 7 to 12. On analysis by days of defervescence or by days of illness, elderly patients and those with hypertension showed consistently higher SBP and DBP throughout their hospitalization, as compared with their younger and non-hypertensive counterparts.

Conclusion: In SD or DHF, SBP decreased to a nadir around the day of defervescence, and recovered to a level exceeding that in febrile phase by days 2 or 3 post-defervescence. Elderly patients and patients with pre-existing hypertension maintained higher SBP and DBP throughout the duration of dengue infection.

Fig 1. Temporal trend of SBP and DBP between patients with Severe Dengue and non-Severe Dengue.

Overall means are indicated as solid lines with 95% credible intervals as dashed lines. The black bar on X-axis indicated days with a ‘‘significant” difference between the groups. The day of defervescence was denoted as day 0, day before defervescence as day -1, and day after defervescence as day 1 respectively, for SBP (A) and DBP (B). Day of onset of fever was denoted as illness day 0, for SBP (C) and DBP (D).

Fig 2. Temporal trend of SBP and DBP between patients with Dengue Hemorrhagic Fever and Dengue Fever.

Overall means are indicated as solid lines with 95% credible intervals as dashed lines. The black bar on X-axis indicated days with a ‘‘significant” difference between the groups. The day of defervescence was denoted as day 0, day before defervescence as day -1, and day after defervescence as day 1 respectively, for SBP (A) and DBP (B). Day of onset of fever was denoted as illness day 0, for SBP (C) and DBP (D).

Fig 3. Temporal trend of SBP and DBP between the elderly and non-elderly.

Overall means are indicated as solid lines with 95% credible intervals as dashed lines. The black bar on X-axis indicated days with a ‘‘significant” difference between the groups. The day of defervescence was denoted as day 0, day before defervescence as day -1, and day after defervescence as day 1 respectively, for SBP (A) and DBP (B). Day of onset of fever was denoted as illness day 0, for SBP (C) and DBP (D).

Fig 4. Temporal trend of SBP and DBP between patients with and without hypertension.

Overall means are indicated as solid lines with 95% credible intervals as dashed lines. The black bar on X-axis indicated days with a ‘‘significant” difference between the groups. The day of defervescence was denoted as day 0, day before defervescence as day -1, and day after defervescence as day 1 respectively, for SBP (A) and DBP (B). Day of onset of fever was denoted as illness day 0, for SBP (C) and DBP (D).