Factors Associated with Mortality in Adults Admitted with Heart Failure at the University Teaching Hospital in Lusaka, Zambia

Abstract

Background: Heart failure is a major public health problem and has been recognized as an important cause of morbidity and mortality for several years. It is one of the leading non-infectious causes of death among
hospitalized patients at the University Teaching Hospital (UTH) in Lusaka, Zambia. This study aimed to investigate the predictors of 30-day mortality in heart failure patients admitted to the medical wards at the UTH
using routinely obtained clinical data.

Methods: We enrolled 390 heart failure patients and followed them up over a period of 30 days. Data collected included demographic characteristics (age, sex), medication use and co-morbidities (hypertension, diabetes mellitus, Human Immunodeficiency Virus (HIV) infection). Clinical data included vital signs, blood urea, serum sodium, serum potassium, serum creatinine, and haemoglobin level. Trans-thoracic echocardiographs
and electrocardiographs were also done to determine left ventricular ejection fraction (LVEF) and to check for the presence of arrhythmias. Patients were dichotomized into those with preserved (LVEF>=40 percent) and reduced (LVEF< 40 percent) systolic function. Recruited patients
were then prospectively followed up to determine outcome by day 30 (i.e. dead or alive). Cox proportion Hazard regression analysis (on Epi Info software version 3.5.3) was used to analyse the effect of each of these
parameters on outcome.

Results: Of the recruited patients, 59% were female (95% CI 54-64). The median age was 50 years (IQR 33-68). 138 patients (35%, 95% CI 31-40) died within 30 days of admission. 94 (68%) of these deaths occurred inhospital. The factors shown to be independent predictors of death on multivariate logistic regression analysis were LVEF<40 percent (OR=2.86, 95%CI 1.68- 4.87), NYHA class IV (OR=2.15, 95%CI 1.27- 3.64), serum urea above 15mmol/L (OR=2.48, 95%CI 1.07-5.70), and haemoglobin level below 12g/dL (OR=1.79, 95%CI 1.11-2.89). The additional factor associated with increased risk of mortality on univariate analysis was systolic blood pressure below 115mmHg (OR=1.63, 95%CI 1.05- 2.51).
However, serum creatinine (OR=1.49, 95%CI 0.49-4.48) and HIV seropositivity (OR=0.96, 95% CI 0.53-1.72) had no bearing on the risk of death in this patient population.

Conclusions: Left ventricular ejection fraction <40 percent, New York Heart Association class IV, serum urea
above 15mmol/L, haemoglobin level below 12g/dLand systolic blood pressure below 115mmHg are predictors of poor 30-day outcome in hospitalised heart failure patients.