Kathryn Daveson 1,2, Caroline Wilson 1, Andrea Menzies 1
1 Canberra Hospital & Health Services, Garran, ACT, Australia
2 Australian National University, Canberra, ACT, Australia
Introduction: Australian guidelines for the diagnosis of Clostridium difficile infection (CDI) recommends laboratory initiated testing (LIT) on all unformed stools in patients hospitalised > 72hrs irrespective of physician’s request. This is currently applied inconsistently across Australia. The aim of this study is to assess the effect of LIT on the surveillance rates of CDI.
Methods: Cases identified between 2012-2014 from a prospectively, collected CDI surveillance database, stratified by hospital-identified (HI) and extended surveillance definitions were used as the cohort for analysis. A laboratory information system extraction was used to identify those within the dataset diagnosed by LIT (performed at greater than 48 hours of admission to hospital in our institution). The proportion diagnosed by LIT was determined. Sensitivity analyses were performed to assess the impact of this process on low or high healthcare-associated infection (HCA) incidence institutions.
Results: The average proportion of HCA-CDI cases during this period was 66%. Cases of HCA-CDI and HI-CDI identified due to LIT ranged from 36-48% and 24-29% respectively over the years. Excess cases occurred solely in the HCA-healthcare facility onset class due to the definition of LIT. The impact was more pronounced in high HCA-CDI institutions.
Conclusion: The impact of LIT on CDI surveillance figures in our institution is large and varies based on institutional incidence of HCA-CDI. Given the magnitude of impact, mandatory laboratory initiated testing as part of mandatory surveillance benchmarking would decrease variation between hospitals and increase the rigour of the process.