Prevalence, factors associated with hypoglycaemia and two days outcome among patients admitted on the medical emergency unit at Mulago hospital.

Abstract

BACKGROUND Hypoglycaemia is a clinical syndrome with diverse causes in which low blood glucose concentrations lead to symptoms and signs in affected persons. Hypoglycaemia occurs in patients with and without diabetes and has been associated with increased in-hospital mortality. There is paucity of data in Mulago hospital about hypoglycaemia and its associated factors among medical hospitalized patients. Objectives: To determine the prevalence, factors associated with hypoglycaemia and two days outcomes among patients admitted to the medical emergency unit at Mulago hospital. Methods: A cross-sectional study was done among patients presenting to the medical emergency unit at Mulago hospital from October 2012 to December 2012. Systematic sampling of participants was done and all patients with hypoglycaemia diagnosed by random blood sugar test were recruited until the sample size was attained. The recruited participants were interviewed using a pre-tested standardised data collecting tool on selected social-demographics, history and physical examination. Participants were followed up for two days for an outcome of death or alive (still on ward, discharged or run away). Data was entered into the standardised questionnaire which we checked for completeness on daily basis. It was double entered into EPIDATA and then later was exported to STATA. Analysis was done using Stata and data was described as follows: the prevalence of hypoglycaemia was determined as a proportion of those that were found to have RBS<3.9mmol/L of the total number of participants that were enrolled during the study period. Using bivariate and multivariate analysis, factors associated with hypoglycaemia were determined. The outcome was obtained as odds of those that died with hypoglycaemia and those without hypoglycaemia. Results: The prevalence of hypoglycaemia was found to be 11 %( 47/427). After adjusting for confounders; time since the last meal ≥24 hours (p-value=0.006, CI 3.94[1.37-11.38]), reduced level of conscious with Glasgow coma scale13-14(p-value 0.005, CI 3.13[1.42-6.87]), Glasgow coma scale9-12 (p-value=<0.001, CI 6.94[2.94- 16.38]), Glasgow coma scale ≤8 (p-value=<0.001, CI 9.57[3.34-27.13]) were significant. A total of 22/47(47%) patients who had hypoglycaemia were dead by day two. Among those without hypoglycaemia, 72/380 (19%) were also dead by day two. Patients with hypoglycaemia were more likely to die (OR 3.76, CI=1.97-7.16) compared to those without hypoglycaemia.

Conclusion The prevalence of hypoglycaemia among patients admitted to the medical emergency unit at Mulago hospital was 11% Level of consciousness with low Glasgow coma scales below 15, and lack of a meal for more than twenty four hours were more associated with hypoglycaemia. Patients with hypoglycaemia were more than three times at risk of death in the first 48 hour period than those without hypoglycaemia. Recommendations • Routine screening of hypoglycemia should be adopted as part of basic medical routine care service for all patients that are admitted for emergency care despite their diabetic status. • Patients with Glasgow coma scale below 15, and lack of a meal for more than twenty four hours should be prioritized for blood glucose assessments on the medical emergency unit. • Patients with hypoglycemia should be adequately and closely monitored after admission to prevent death. • A larger study would be very important to establish the causes of death among hypoglycemic adults admitted to the medical emergency unit.