 |
| Source: mhcc.maryland.gov |
Overall, 114 patient records were reviewed by 18 infection prevention specialists, the majority of whom specified they followed National Healthcare Safety Network (NHSN) criteria. The overall agreement amongst infection preventionists by kappa was 0.42 (SE 0.06). Better agreement was observed with a simple laboratory-based definition with an average kappa of 0.55 (SE 0.05). The proportion of patient records that 18 reviewers reported with CLABSI ranged from 14% to 39% (overall mean 28% with a CV of 25%). Again when simple laboratory-based methods were applied to patient records, classification was more consistent with CLABSI assigned in a proportion ranging from 36% to 42% (overall mean 39%).
The findings are significant for several reasons. First, despite training, certification and the use of a standardized CLABSI definition, application can be nuanced and is subject to interpretation based on individual cases with complex clinical conditions. At play here is the concept of inter-observer reliability. This is not a new concept. Here is an interesting study highlighting how the assessment of chest x-rays for pneumonia can differ between radiologists.
More importantly, the above CLABSI finding calls into question the methodological limitation of publicly reporting hospital acquired bloodstream infections. Despite a standardized CLABSI definition, the reliability of surveillance to appears not be ideal for the public goal of inter-hospital comparisons.
