سیستم پشتیبانی از تصمیم گیری برای زمانبندی بیمار در اداره واکسن مسافرتی

The administration of travel vaccines presents a number of operations management challenges. The interplay between shared consumption of multi-dose vaccine packages, rapid spoilage upon opening, the high cost of wastage, and the unique vaccination needs of the patients makes for a very interesting and complex scheduling problem that could benefit from computerized decision support. We compare the performance of a novel binary integer programming model and a genetic algorithm solution technique with conventional scheduling approaches. Computational results show that significant cost savings can be achieved with the DSS while simultaneously considering scheduling preferences of patients and mitigating scheduling inconvenience.

Health care costs are one of the most pressing economic problems of our time. As such, computerized decision support that leverages sound operations research and management science principles has a unique opportunity to assist medical professionals in ensuring the delivery of services in the most efficient means possible. This paper considers one such opportunity. With the rising cost of prescription medicines, reducing waste in the delivery and administration of pharmaceutical products has become of increasing concern [31]. Management of pharmaceutical stock is often regarded simply as an issue of perishable inventory management, a field well-studied by management scientists over many decades—for a review, see Ref. [40]. However, while conventional perishable inventory management may apply for many prescription drugs, there is a field of medical practice where the inventory management problem is atypical: the field of travel vaccine administration. The United States Centers for Disease Control (CDC) recommends a variety of routine, recommended, and required vaccines for travelers to foreign destinations [7]. For example, when traveling to South Africa, the CDC recommends travelers be brought up-to-date on all routine vaccinations (including measles, mumps and rubella; diphtheria, pertussis, and tetanus; polio, and others), and also receive additional immunizations for hepatitis A, hepatitis B, typhoid, and rabies. Travel vaccine administration presents peculiar challenges to management scientists for numerous reasons. Firstly, many vaccines are packaged in multi-dose vials such that the contents are administered to several patients (shared consumption). Secondly, each patient typically requires a particular combination of vaccines, depending on their planned destination of travel and prior immunization history; therefore, consumer assortment demands are particularly stringent. Thirdly, and most critically, once opened or reconstituted, many vaccines must be used within a very short time-span, or discarded, due to rapid spoilage when exposed to air and room temperatures [2] and [60]. Finally, vaccines are costly to produce, making wasted doses quite expensive. For instance, YF-Vax® (yellow fever vaccine), is available in multi-dose vials with a shelf-life-once-opened of 60 min [44]. YF-Vax® wholesales for over $340 per multi-dose vial [51], with each wasted dose costing the medical practice almost $70. A single dose formulation is also available, as an alternative to the multi-dose vial, but the per-dose cost is 25% higher for the single dose formulation [51], so clinics should prefer the multi-dose formulation if no doses will be wasted. Simulation results from [34] indicate the severity of poor vaccine format stocking decisions: incorrect selection of single-dose vs. multi-dose format for the clinic's formulary can cost clinics between $8000 and $24,000 per vaccine, per year. For a clinic stocking 5 common vaccines, total costs of poor choice of single vs. multi-dose formats can exceed $65,000 per clinic, per year. The interplay between shared consumption of single packages, rapid spoilage once opened, stringent consumer assortment demands, and high cost of wastage creates a difficult scheduling problem for the travel clinic administrator. With multiple vials of different vaccines open and deteriorating rapidly, and numerous patients each potentially requiring a different assortment of vaccines, determining the optimal scheduling of patients is non-trivial. In this paper, we show that traditional patient scheduling methods in travel vaccine clinics, such as first-in, first-out (FIFO), are sub-optimal and can lead to significant vaccine wastage. We propose and compare an integer programming model and a genetic algorithm solution procedure in Excel, that can significantly reduce a clinic's vaccine costs. It is possible to develop a patient scheduling DSS using a variety of software engines instead of Excel (e.g. MATLAB, LINDO, CPLEX), and some DSS designers will prefer one engine to another. Our objective was to design a decision support tool for clerical staff in a health care clinic using a familiar, affordable, and accessible software platform. We begin with a discussion of related work in perishable inventory management, job shop scheduling, patient scheduling, and vaccine management and describe how our approach differs from past work. We then provide descriptions and experimental evaluations for the solution procedures we tested. Finally, we conclude with a discussion of recommendations, limitations, and areas for future research.

Optimal choice of single versus multi-dose vaccine formulations within a travel vaccine clinic is highly complex given the short shelf-life-once-opened of multi-dose formulations and the unique vaccine assortment requirements of each patient. In this paper, we introduced a spreadsheet-based decision support tool for the travel vaccination problem. Our tool incorporated a novel binary integer programming model and a GA. In experiments comparing these approaches to alternative patient scheduling mechanisms (such as vaccinating patients on a FIFO basis, or sorting patients so that those requiring the highest-cost vaccinations are grouped together), we have shown that the integer programming and GA methods produce meaningful cost savings, at mild implementation complexity. Importantly, we were able to implement a GA solution to the travel vaccination problem that produces a rapid solution, of similar quality to the integer programming method, within only a few minutes of runtime. We have demonstrated that our practical decision support tool presents a useful new operations management mechanism that supplements traditional stock issuance and job scheduling policies and could allow health care managers to significantly lower vaccine administration costs. Gupta and Denton [24] observe that while operations research models have been used successfully to improve efficiency in many service industries, the same degree of success has not yet occurred in the health care; and note this as an important challenge for the research community. We believe that the DSS proposed in this paper is one small step in meeting this challenge.