Cost-effectiveness analysis in infectious diseases

Abstract

Incorporating costs into healthcare decision-making is assuming increasing importance in infectious diseases and clinical microbiology, as in other medical domains. With the introduction of novel, costly molecular diagnostics and with an increasing need for tighter infection control, cost considerations will become even more important. There are special problems when attempting to assess costs or cost-effectiveness of interventions in infectious diseases, and a dearth of guidance on how to conduct such studies appropriately. The spectrum of disease ranges from acute infections occurring in the hospital to chronic, life-long conditions (e.g. human immunodeficiency virus (HIV) disease). The breadth of indirect costs surrounding infections in hospital or in the community is undefined. Resistance development following antibiotic treatment and its effects on the individual, the microenvironment and the macroenvironment is difficult to incorporate in cost-effectiveness analyses, just as in clinical studies. A review of studies whose primary aim was to assess costs, cost-effectiveness or other costs analyses in the top ten journals publishing original studies in infectious diseases in the last 5 years (Journal Citation Reports 2009, ranked by impact factor) revealed 95 studies, most of which were published in the last year. The large majority of the studies (47%) addressed HIV disease prevention (13 studies) and treatment (32 studies) in developing countries (32; 71% of studies on HIV). Other main fields were rotavirus vaccine or management (11 studies, eight in developing countries), influenza (mainly prevention, seven studies) and management of hepatitis (six studies), the last two being conducted solely in developed countries. Only seven studies assessed antibiotic treatment for various acute bacterial or fungal infections, and another seven studies addressed very different aspects related to hospital-acquired infections (methicillin-resistant Staphylococcus aureus (MRSA) screening (2), methicillin-sensitive S. aureus or MRSA bacteraemia (2), and one each on surgical site infections, Clostridium difficile and an antimicrobial stewardship team), all in developed counties. A single study addressed costs attributable to antimicrobial-resistant infections in hospitals in the USA [1]. In 56 studies whose design was based on a model, the reference case reported in most studies was primarily cost/ quality-adjusted life-years (QALYs; 55% of studies). Eight studies, all of which dealt with HIV management, reported primarily on costs per life-year. Clinical studies whose primary aim was cost analysis reported most commonly simply on the costs of an infection or a management strategy (27/ 39 studies; 69%). Fifteen studies overall (16%) reported on cost/outcome avoided (infection, treatment failure or disease-related death). These studies used intermediary measures of outcome that are difficult to compare with other studies and interventions. The perspective in the studies was most commonly the provider’s (the hospital, healthcare organization or government in 56/94 studies; 60%). The costs considered were those of medications, testing, personnel or clinic visits/hospital days. A societal perspective, considering some, although usually not all, indirect medical and non-medical costs was reported in 34% of studies, and more commonly in studies on HIV disease management (13/32, 40%). Five studies reported that the analysis was performed from the patient’s perspective. These studies added cost components, such as transportation, food, loss of income and assets, reduced productivity and other out-of-pocket expenditure of the family, that were not included in studies reporting on the societal perspective. Studies on antibiotic treatment and hospitalacquired infections reported more frequently only the provider’s perspective (13/16 studies; 81%) than did other studies dealing with HIV, vaccines, influenza and other viral infections (44/79; 56%). It is interesting to note that of 67 studies that aimed to assess the cost-effectiveness of a specific intervention (treatment, vaccine, prevention strategy), 53 (79%) reported the intervention to be cost-effective in the base case analysis (the basic analysis incorporating data and methods best representing the interventions and choices under consideration). Another seven studies concluded that the intervention was cost-effective under most assumptions. Only three studies reported a negative result for the base case, and another four studies gave negative conclusions for most of the scenarios assessed in sensitivity analyses. The information conveyed in the negative studies is rather interesting and useful