An Audit Of The Management Of Diabetic Ketoacidosis At The Kenyatta National Hospital

Abstract

BACKGROUND DKA is the most serious hyperglycemic emergency in patients with both Type 1 & 2 DM. The mortality rate is still high in developing countries. Standard DKA management guidelines ensure optimal management. OBJECTIVES · To assess the level of adherence to DKA management guidelines in KNH. · To assess outcomes of DKA patients in KNH. · To assess the diabetes-related knowledge of doctors and nurses in KNH. STUDY DESIGN A prospective descriptive study. STUDY SETTING The Accident and Emergency Department, Medical ICU and General medical wards in KNH. METHODOLOGY Patients with DKA were screened for eligibility and recruited upon signing informed consent. Patients’ socio-demographic and clinical characteristics were captured in a study proforma. The Principal Investigator used a checklist, based on standard DKA management guidelines to assess patient management as documented in the patients’ inpatient files, treatment and monitoring charts, and assessed the extent of conformity to DKA management guidelines. Key informants were interviewed on systems in place to manage DKA patients in KNH. Doctors’ diabetesrelated knowledge was assessed using a pre-tested questionnaire. Nurses’ perceived diabetesrelated knowledge was assessed using the Diabetes Self-Report Tool (DSRT); their actual diabetes-related knowledge was assessed using the Diabetes Basic Knowledge Test (DBKT). RESULTS Forty two patients with clinico-laboratory diagnosis of DKA over a four month period were recruited. Eight key informants, 49 doctors, and 70 nurses were also interviewed. xiv Prompt management according to guidelines was not observed. There were delays in patient review by SHOs or ICU doctors upon arrival in hospital; the median time was 10 hours. There were also delays in initiation of management. Guidelines were adhered to in initial patient clinical and laboratory assessment. However, subsequent clinical and laboratory monitoring was suboptimal; 64.3% of patients had less than 6 vital signs assessments done in the first 24 hours. All prescriptions for intravenous fluids (IVF), insulin, and potassium were according to guidelines. All patients had the correct type of IVF administered. However, only 11.9% had the correct amounts of IVF administered; the median amount of IVF administered in 24 hours was 3.0 litres. All received subcutaneous insulin as prescribed, while 74.3% received insulin infusion as prescribed. 23.7% did not receive potassium as prescribed. DKA resolution was not confirmed according to guidelines in 46.2% of patients. The median time to resolution was 59 hours. The all-cause inpatient mortality rate at 2 weeks was 11.9%. The doctors’ mean score in the questionnaire was 59.1%. Most respondents had sufficient knowledge on practical inpatient diabetes management and diabetes pharmacology. There was insufficient knowledge on diabetes diagnosis and targets. The nurses had a mean (SD) of 72.46 (12.01) points out of 88 on the DSRT. Generally, the respondents had a high perception of diabetes knowledge. On the DBKT, they had mean (SD) score of 26.96 (4.64) points out of 45. CONCLUSIONS This study demonstrated lack of adherence to guidelines, in terms of delays in initiation of management, insufficient administration of fluids, insulin and potassium, and inadequate monitoring. This may have contributed to the prolonged DKA resolution times and the high allcause mortality rate. Areas to improve on, therefore, are prompt patient review and initiation of management, consistent patient monitoring during care, and guideline adherence in terms of fluid, insulin and potassium treatment. Healthcare workers, both doctors and nurses, had sufficient knowledge on various aspects of DKA management. However they had insufficient knowledge on certain aspects of diabetes pharmacology and glucose dynamics. The areas of insufficient diabetes related knowledge are important areas to focus on in rolling educational programs for the healthcare workers.