Comparison of hospital surgical site infection rates and rankings using claims versus National Healthcare Safety Network surveillance data

Abstract

National policies target healthcare-associated infections using medical claims and National Healthcare Safety Network surveillance data. We found low concordance between the 2 data sources in rates and rankings for surgical site infection following colon surgery in 155 hospitals, underscoring the limitations in evaluating hospital quality by claims data.

Figure 1.. Bubble plot of hospital rankings by quartiles for rates of surgical site infection after colon surgery by National Healthcare Safety Network data versus claims billing data (2012–2014 combined)

Bubble sizes are proportional to the number of hospitals in each matched quartile. The actual number of hospitals in each category is denoted within the bubbles. The cohort included 155 hospitals; the total number of hospitals reflected in the graph equals the number of unique hospital-years (N=394) during the study period. Lower quartiles indicate better performance (i.e., quartile 1 = lowest SSI rates, quartile 4 = highest SSI rates). The solid line indicates where all hospitals / bubbles would lie if concordance was perfect between claims and NHSN data. Bubbles above the solid line indicate quartiles that are worse by claims vs NHSN data. Bubbles below the dotted line indicate quartiles that are better by claims data versus NHSN data. τ = Kendall’s tau, a measure of the ordinal association between two measures (in this case, hospital rankings using NHSN vs claims data). We considered a tau of <0.40 as poor agreement, 0.41–0.75 as fair-good, and >0.75 as excellent.