Method: All admitted patients to Tan Tock Seng Hospital in 2012 were included. A risk prediction model was developed and validated to select high risk patients for screening, using logistic regression and Bayesian Information Criteria. Markov decision analysis was applied to identify the most cost-effective screening strategy. The five strategies were compared in terms of the cost per infection prevented: PCR screening for all; PCR screening for selected high risk patients; no screening. The modeling cycle (time length of transition) is 1 hour. The total modeled exposure time in hospital is about 120 hours (5 days). Costs to hospital will be used as the primary cost measure. We will also measure the cost from the perspectives of patients.
Result: In the risk stratification model, the important predictors identified were MRSA colonization history; elder age; infection or hospitalization in last 3 months; admitted from nursing homes; with kidney diseases, or stroke. The c-statistics of the ROC of the prediction model was 0.82 (95%CI: 0.81-0.83). The MRSA prevalence at admission was about 7.3% in 2012. Considering the cost of infection treatment, the incidence rate of hospital infection, the sensitivity and specificity of predicting the high risk patients, the most cost effective screening strategy was selective screening, which cost about $15.8K (95%CI: $7.8K - $ 21.9K) per infection prevented compared with no screening.
Conclusion: The study provides an evidence-based decision tool for policy makers to standardize care and set guidelines on cost effective infectious disease control in hospitals.