Abstract Background As more indicators are monitored by the infection control service, automating the interpretation of results becomes increasingly important; here, we propose a method that uses both short-term (monthly) and long-term (annualized) data to provide a comprehensive understanding of infection trends, including the detection of outbreaks.Figure 1Criticality matrix for the indicator and automatic critical analysis of the rate.Criticality matrix for the indicator and automatic critical analysis of the rate. Methods The automated critical analysis of a rate is conducted by simultaneously evaluating both the monthly rate and its moving average. While the month-by-month rate shows the indicator's current monthly results, the moving average indicates the overall trend, smoothing out noise from occasional aberrant values. A criticality matrix for the indicator is constructed: Trend Analysis: Examine both the monthly rate and the moving average for trends. A p-value of up to 0.05 indicates a significant increase or decrease, while a p-value above 0.05 and below 0.10 suggests a probable increase or decrease. A p-value of 0.10 or higher indicates stability. Last Month's Status: Compare the indicator's monthly value and moving average to their respective targets (90th percentile). Generate risk scores for rate loss of control by combining the results of four analyses, yielding a total score S=S1+S2+S3+S4, and risk of loss of control is (S−4)/36×100%. This approach allows for automated analysis and interpretation of the epidemiological scenario (as illustrated in Fig. 1).Figure 2Automated critical analysis of the rate of hospital-acquired infections (HAI) per 1,000 patient-days caused by Acinetobacter baumannii resistant to carbapenems. Automated critical analysis of the rate of hospital-acquired infections (HAI) per 1,000 patient-days caused by Acinetobacter baumannii resistant to carbapenems. Results Figure 2 shows the rate of hospital-acquired infections (HAI) per 1,000 patient-days caused by Acinetobacter baumannii resistant to carbapenems. The automated analysis indicates a 56% risk of loss of control. Figure 3 illustrates the rate of central line-associated bloodstream infections (CLABSI) per 1,000 central-line days in an intensive care unit. The analysis indicates a 3% risk of loss of control. Figure 4 demonstrates the automated analysis of ventilator-associated pneumonia (VAP), with the number of VAP incidents per 1,000 ventilator-days, indicating an 11% risk of loss of control.Figure 3Automated critical analysis of the central line bloodstream infection (CLABSI) rate (#CLABSI/1,000 central-line days) in an intensive care unit. Automated critical analysis of the central line bloodstream infection (CLABSI) rate (#CLABSI/1,000 central-line days) in an intensive care unit. Conclusion The method allows automated critical analysis to effectively evaluate infection rates in healthcare settings. By combining short-term and long-term data, including monthly rates and moving averages, it offers a comprehensive view of infection trends.Figure 4Automated critical analysis of ventilator-associated pneumonia (number of VAP incidents per 1,000 ventilator-days) in an intensive care unit. Automated critical analysis of ventilator-associated pneumonia (number of VAP incidents per 1,000 ventilator-days) in an intensive care unit. Disclosures All Authors: No reported disclosures
Read full abstract