Many hospital emergency departments (EDs) in the United States have implemented the use of the five-level Emergency Severity Index (ESI) as their clinical decision support method to enhance clinical decision making in the triage process. The ESI designates the most acutely ill patients as level 1 or 2 and those who do not meet these criteria are assigned to levels 3–5 based on estimated resource utilization. Although the number of resources is the primary decision rule to determine levels 3–5, physiological and descriptive variables can also be used to predict the ESI level. This study uses several physiological and descriptive variables as predictors to determine the ESI value. The physiological variables include heart rate, blood pressure, temperature, respiration rate, and oxygen level, whereas the descriptive variables include age, gender, pain level, and patient complaint. An ordered probit model was developed for ESI prediction. In addition, a linear regression model was also developed to demonstrate the necessity of having a decision making tool that allows for non-integer values. The results of this research can be used to enhance the precision of the ESI and the nurse's ability to prioritize treatment based on triage acuity. The decision making tool can also be used to stratify patients who are classified in the same priority group and may eliminate the necessity of grouping patients into different categories.

Prevention of infectious diseases is an important concern for managing public health. Although vaccines are the most effective means for preventing infectious diseases, the existence of a negative network externality often makes it difficult for vaccine coverage to reach a level that is socially optimal. In this research, we consider how a subsidy program can induce a socially optimal vaccine coverage. We consider an oligopoly market with identical vaccine producers and derive a subsidy that leads to a socially efficient level of coverage. We also derive a tax-subsidy combination that is revenue neutral, but achieves the same effect. Overall, our results provide useful insights for governments and policy makers with respect to an important issue related to public health.

Blood products are an essential component of any health system, and their safety, in terms of being free of “transfusion-transmitted infections” (TTIs), i.e., diseases that include Human Immunodeficiency Virus, Hepatitis Viruses, Human T-cell Lymphotropic Virus, Syphilis, West Nile Virus, and Chagas’ Disease, is crucial. However, blood screening tests are imperfectly reliable, with the possibility of false-negatives and false-positives. The budget-constrained decision-maker needs to (i) select a set of screening tests to administer to each unit of donated blood, and (ii) construct a “decision rule” with which to classify each blood unit as infection-free versus infected. The objective is to minimize the TTI risk for blood classified as infection-free, which depends on the efficacies of the entire set of tests selected and the decision rule adopted. This risk structure leads to a nonlinear optimization problem. Our analysis provides efficient optimal algorithms for a special case where only mono-infections are possible, and an effective heuristic and lower bounds for the general case with co-infection possibility. Our case study for the sub-Saharan Africa, Ghana, Thailand, and the United States illustrates the value of our optimization-based approach, which generates region-specific test composition by explicitly considering the regional TTI prevalence rates, and provides insights.

The data used in health care and public health monitoring and surveillance are likely to be nonstationary and highly auto-correlated. These data may come from clinical and nonclinical sources, the latter being increasingly used in an attempt to supplement traditional surveillance. We propose an alternative approach applicable to multiple sets of health data that combines a method of preconditioning (ARMA filtering), a forecasting method (the method of analogues) and a monitoring method (the change-point). This procedure is applied to Influenza-Like Illness (ILI) surveillance in France using both clinical data (time series of weekly ILI incidence at the national level and in 22 regions) and nonclinical data (over-the-counter and prescribed weekly medication sales data at the national level). The outcomes show that multivariate monitoring provides an improved ability to recognize warning signals and drug sales data provide an indirect but earlier indication of influenza outbreaks. The proposed method has the advantage of being both practical and relatively simple to implement and could be easily adapted to develop adequate surveillance tools for specific diseases.