To conduct an updated assessment of the validity and reliability of administrative coded data () in identifying hospital‐acquired infections (s).
We systematically searched three libraries for studies on detecting s compared to manual chart review. Meta‐analyses were conducted for prosthetic and nonprosthetic surgical site infections (s), infections (s), ventilator‐associated pneumonias/events (s/s) and non‐s/s, catheter‐associated urinary tract infections (s), and central venous catheter‐related bloodstream infections (s). A random‐effects meta‐regression model was constructed.
Of 1,906 references found, we retrieved 38 documents, of which 33 provided ( = 567,826 patients). identified incidence with high specificity (≥93 percent), prosthetic s with high sensitivity (95 percent), and both s and nonprosthetic s with moderate sensitivity (65 percent). exhibited substantial agreement with traditional surveillance methods for ( = 0.70) and provided strong diagnostic odds ratios (s) for the identification of s ( = 772.07) and s ( = 78.20). performance in identifying nosocomial pneumonia depended on the coding system ( = 0.05; = .036). Algorithmic coding improved 's sensitivity for s up to 22 percent. Overall, high heterogeneity was observed, without significant publication bias.
Administrative coded data may not be sufficiently accurate or reliable for the majority of s. Still, subgrouping and algorithmic coding as tools for improving validity deserve further investigation, specifically for prosthetic s. Analyzing a potential lower discriminative ability of ‐10 coding system is also a pending issue.
