Infections may not play as important a role in precipitating diabetic ketoacidosis (DKA) as previously suggested, according to a Queensland study.

In findings that may have implications for antibiotic stewardship, clinicians at Townsville Hospital say that antibiotics may not be appropriate for three out of four DKA cases where infection is suspected.

A review of 249 episodes of DKA in adults with type 1 diabetes presenting to the hospital found 40% were documented as suspected infections or treated with antimicrobial therapy on admission.

However on review, only 9.2% involved microbiologically-proven infections and an additional 5.2% were probable infections involving clinical features of an infective process and corroboration from two senior treating clinicians.

Of the proven and probable infections diagnosed, the most common were soft tissue infections such as cellulitis (25%), urinary tract infections (22%), respiratory tract infections (19%), diabetic foot infections (11%), gastrointestinal infections (11%), and dental infections (8%).

“In total 80 antimicrobial prescriptions were dispensed for empirical therapy, of which only 30/80 (38%) were in the setting of proven or probable bacterial infection requiring systemic antimicrobials,” the study said.

“Only 20 (25%) antimicrobial prescriptions were appropriate for the suspected infection, as classified by national guidelines.”

The study found the laboratory tests that predicted proven or probable infection were lymphopenia (OR 3.1, 95% CI, 1.14 to 8.2) and an elevated CRP (OR: 10.1 (95% CI, 1.05 to 97.0). Neutrophilia was not associated with infection.

Coauthor Dr Ian Gassiep, a microbiologist and infectious diseases physician with Pathology Queensland, told the limbic that the most common cause of DKA in the study was medication non-compliance rather than infection.

“Patients do present with features potentially suggestive of infections such as fever or elevated white cell counts but that can also be seen specifically in the setting of DKA without an infection trigger.”

“People get overtreated [with antibiotics] by presuming that everyone coming in with a fever and DKA has an infection whereas it seems that is not the case.”

Other precipitants of DKA included alcohol and drug intoxication.

He said prospective multicentre studies on presentations of DKA may be able to identify specific risk factors or develop a predictive scoring system to diagnose infection acutely.

“Certain markers are much more associated with infection – for example, CRP could be an easy marker if that result was available in a timely fashion.”

The study said the findings were consistent with international emergency department (ED) studies showing high levels of inappropriate prescription and flagged an ongoing need for antimicrobial stewardship (AMS) programs in EDs.

The findings are published in Internal Medicine Journal.