Hospital-acquired complications: A comparison of surveillance and coding data for hospital-acquired pneumonia and urinary tract infections.

Authors

Jaimie M Yamada, Pauline Bass, Denise Del Rosario-Kelly, Shirley Leong, Philip J Rawson-Harris, Kirsty Sim, Stephanie J Curtis, Andrew J Stewardson

Published in

Infection, disease & health. Volume 31. Issue 4. Pages 100449. Jul 02, 2026. Epub Jul 02, 2026.

Abstract

Hospital-acquired complications (HACs) have recently been introduced in Australia as a quality-of-care metric. These use International Classification of Diseases (ICD) codes to identify potentially avoidable complications, including healthcare-associated infections. We sought to determine the positive predictive value (PPV) of the HAC algorithms in detecting hospital-acquired pneumonia (HAP) and healthcare-associated urinary tract infections (HA-UTI) using traditional surveillance as the standardised comparator.
We conducted a retrospective analysis at Alfred Health. For both HAP and UTI, we selected 50 admitted episodes with ICD codes that satisfied the HAC definition. We applied standardised surveillance definitions to these episodes, based on European Centre for Disease Prevention and Control (ECDC) guidance. This involved a manual review of documentation and results by an Infection Prevention and Control Nurse Consultant. PPVs were calculated by comparing ICD-code diagnoses with surveillance-confirmed infections.
The PPV for the HAP algorithm was 20% (95% CI: 11.0-33.2%). Among non-confirmed cases, 23% (9/40) had no radiological evidence of pneumonia, and 23% (9/40) had symptoms or consolidation on admission. Of the ten true HAP cases, 50% (5/10) were ventilator-associated. The PPV for the HA-UTI algorithm was 58% (95% CI: 44.2-70.6%), with 71% (15/21) of false positives due to asymptomatic bacteriuria.
We found that current HAC algorithms based on ICD coding data did not reliably predict HAP and HA-UTI as defined by standardised surveillance. Given their use in healthcare funding, further validation and improvement of current algorithms should be explored.

PMID:
42391928
Bibliographic data and abstract were imported from PubMed on 03 Jul 2026.

Read full publication at:
Please sign in to see all details.

Stats

1-terrible, 9-excellent. How would you rate this publication? Sign in in to submit your rating.

Post a comment

You need to be signed in to post comments. You can sign in here.

Comments

There are no comments yet.