Prediction and rate of infections in diabetes mellitus patients with diabetes ketoacidosis in Penang, Malaysia

Abstract

Study aimed to determine the rate and prediction of infection in diabetes mellitus patients ≥ 18 year, with diabetic ketoacidosis (DKA). Retrospective cohort study design was adopted to achieve the objectives. Universal sampling technique was employed for data collection among Diabetes ketoacidosis patients, over a period of 3 years (Jan 2008-Dec 2010). Statistical package for social sciences used to analyze data. Over a 3-year period, total of 967 admissions were identified. Of it, 112 (11.6%) with no infection, 679 (70.2%) with bacterial infection and 176 (18.2%) with presumed viral infection. The mean WBC for all the patients was 18,177 (±9431). 721 (74.6%) had leukocytosis, as defined by a WBC ≥ 15,000/mm 3 . WBC, differential, leukocytosis, as well as sex, temperature were not significant predictors (p > 0.05) of bacterial infection. There was significant (p < 0.05) difference of age between the 3 groups, age above 57 years have high rate of infection as compared to age below and equal 57 years. The infection rate in elderly patients with DKA was high and majority of them had lack of clinical evidence. Major bacterial infections with potential serious sequel were particularly common (33.3%), among every third patient being presumed to have serious consequences.

Antimicrobial stewardship in diabetic ketoacidosis: a single‐centre experience

“…However, there was an association between both pH and serum bicarbonate, and an elevated leucocyte count. 15 Our data corroborate the lack of correlation between neutrophilia and infection. The severity of DKA at presentation resulting in a proportionately higher total leucocyte count may contribute to the likelihood of antimicrobial prescription as clinicians may attribute this value to occult infection.…”

“…15,16 A study performed by Slovis et al 16 in 1987 on 153 patients reported a correlation between neutrophil band count >10% and infection, with a sensitivity of 100% and specificity of 80%. 15,16 A study performed by Slovis et al 16 in 1987 on 153 patients reported a correlation between neutrophil band count >10% and infection, with a sensitivity of 100% and specificity of 80%.…”

“…Haematological and biochemical analysis to predict likelihood of infection in DKA have been studied. 15,16 A study performed by Slovis et al 16 in 1987 on 153 patients reported a correlation between neutrophil band count >10% and infection, with a sensitivity of 100% and specificity of 80%. It is important to note that many laboratories do not report a separate neutrophil band count, and, therefore, this finding may have limited value in these settings.…”

“…There was a low correlation with regard to suspected infection on presentation and probable or proven infection on admission. 15 This may contribute to an overestimation of incidence. These findings are consistent with international emergency department (ED) studies demonstrating inappropriate prescription and broad spectrum use of antimicrobial therapy for RTI, UTI, and skin and soft tissue infection.…”

“…The current literature on DKA precipitation does not appear to use consistent standards regarding classification of infection, with one study specifically using antimicrobial prescription as an indicator of bacterial infection. 15 This may contribute to an overestimation of incidence.…”