چه کسی سر نخی برای جلوگیری از آنفلوآنزا دارد ؟ موشکافی اثرات عرضه و تقاضا بر اساس واکسیناسیون آنفولانزا

Influenza is a serious disease, especially for older people, and incomplete vaccination take-up poses a major public health challenge. On both the side of physicians and patients, there could be promising channels for increasing immunization rates, but no attempt has yet been made to empirically unravel their respective influences. Using exclusion restrictions implied by an economic model of physician–patient interactions, our study quantifies the particular effects of supply and demand on influenza immunization. On the supply side, our estimates highlight the importance of physician agency and physician quality, while a patient’s education and health behaviors are key demand side factors.

Influenza is an infectious disease that can have severe consequences for those affected. Older people and individuals with specific health conditions, such as heart or respiratory diseases, run a particularly high risk of suffering complications from an infection with one of the influenza viruses. Every year, influenza leads to a large number of excess hospitalizations and deaths worldwide (WHO, 2003). Even though vaccination can considerably reduce the incidence and severity of influenza, its take-up is often far from complete. Even specifically targeted high-risk groups, such as the older population, often feature substantial gaps in vaccine take-up, with take-up rates below 50% at times Mattke et al., 2006 and Pohl, 2006. As a consequence, increasing influenza vaccination is one of the top public health priorities in many countries (WHO, 2005). Asymmetric information is one of the key features characterizing the market for health care, and patients’ perceptions of their own care need are often inaccurate Arrow, 1963 and Kenkel, 1990. Influenza is no exception in this regard, and even individuals at high risk of severe complications tend to have considerable misconceptions with respect to the seriousness of influenza and their own resistance (Kroneman et al., 2006). For this reason, physicians often need to act as agents for their less-informed patients, which leads to an important role for supply-side factors in determining actual patterns of health care use. In this way, physician agency may also offer an important supply-side channel for increasing the take-up rate of influenza vaccinations. The main objective of this paper is to unravel supply and demand factors in the determination of vaccination take-up and assess their relative importance quantitatively. Disentangling the separate influences of supply and demand is particularly informative for the design of health policies targeting either market side. One important issue in this regard is the role of physician agency for vaccination take-up among high-risk individuals. Do high-risk patients exhibit sufficient health literacy to independently demand influenza vaccination or do they critically rely on their family physicians to obtain indicated immunizations? We propose a simple economic model for vaccination take-up that highlights the role of physicians, patients and their interactions in the administration of influenza vaccines to illustrate key subject matters and inform our subsequent empirical analysis. Particularly, our model points out important simultaneity issues implied by physician agency in the physician–patient relationship. At the same time, it offers some guidance on potential exclusion restrictions that we can use to separate the respective influences of supply and demand on vaccination take-up. To this end, we estimate a semiparametric double index model for influenza vaccine take-up using novel survey data on older individuals in Germany. Specifically, our econometric model features two distinct indices, one for supply and one for demand. Using exclusion restrictions implied by an illustrative theoretical model, we are able to separate structural supply and demand effects as well as quantify the impact of various micro-level factors on vaccination take-up. Importantly, our model also allows us to identify the exact pathway through which key health-related risk characteristics of the patients, such as age or background health, affect the conditional probability of getting vaccinated. We can therefore gauge the extent of physician agency in vaccination decisions based on our estimation results. The remainder of the paper is organized as follows: Section 2 briefly reviews some non-technical background material on influenza, influenza vaccination and potential barriers to comprehensive immunization coverage. Section 3 presents our illustrative model of patient–physician interactions, which guides our empirical investigation. The corresponding econometric framework is detailed in Section 4. This section describes the most important aspects of our semiparametric estimation procedure and gives a detailed discussion of how we define structural effects of supply and demand. Section 5 describes the data underlying our analysis as well as the exact specification of our empirical model. Section 6 presents our estimation results, with Section 7 concluding the paper.

Influenza is a serious illness that can be prevented by annual vaccination. Infection with one of the influenza viruses may have severe consequences for those affected, to the point of hospitalization or even premature death, both of which tend to be concentrated among older people. Vaccination decreases the risk of infection substantially and largely alleviates its adverse consequences in case of influenza contraction. As a result, increasing the take-up of influenza vaccinations is one of the chief public health concerns in many countries. Asymmetric information is particularly widespread in health care markets. For that reason, expert physicians often need to act on behalf of their less informed patients. We analyze a semiparametric double index model for influenza vaccination use that allows us to disentangle the distinct effects of supply and demand on take-up, resolving important simultaneity issues as implied by widespread physician agency. Particularly, our empirical model delivers important insights on the relative importance of various micro-determinants of vaccination usage as well as regarding the exact pathways through which they affect take-up. Our analysis establishes that both supply and demand have important effects on influenza vaccination take-up among individuals aged 50+50+ in Germany. Specifically, our estimates for the structural effects of supply and demand indicate that vaccination rates more than double over the respective supports of either market side. Yet, the exact nature of these structural effects differs substantially between supply and demand. While we find evidence for relatively strong supply effects for most respondents, any large demand effects seem concentrated in the tails of its distribution. Hence, our results indicate larger average supply than demand responses to marginal changes in their respective micro-determinants. Analyzing the effects of each micro-determinant individually, we find that key risk factors for complications, such as age or specific health conditions, lead to considerably higher vaccination rates, and mainly do so through their effect on the supply side. We interpret this finding as evidence for the importance of physician agency in vaccination take-up, highlighting the important role of family physicians for the delivery of responsive health care. Beyond the physician’s role as agent for her patients, our estimation results also identify physician quality as a key supply-side factor underlying vaccination take-up. Having a family physician who generally complies with indicated geriatric assessments has a strongly significant positive supply effect. Specifically, our estimates indicate on average 12% points higher vaccination rates among respondents whose family physician performs all geriatric assessments relative to those whose doctor does not undertake any evaluation. On the demand side, a patient’s educational attainment and general preventive health behaviors emerge as the most important determinants for vaccination take-up. While the former suggests a significant role for health literacy in explaining influenza immunization take-up, the additional effect of health behaviors – conditional on education – may point to the importance of preference heterogeneity for health in general. Beyond it, we also estimate significant effects of gender and partnership status on vaccination demand. Interestingly, we do not find any significant or quantitatively important employment effects on vaccination take-up, at least for the older population studies here. In terms of health policy implications, our analysis suggests that both supply and demand interventions may provide suitable leverage for increasing influenza vaccination rates. On the one hand, our estimates of strong education effects indicates that campaigns to improve the health literacy of older patients may indeed represent an important health policy tool for increasing their take-up of influenza vaccines. At the same time, our evidence for (imperfect) physician agency and the important role of quality of care indicators for take-up further highlight the potential effectiveness of supply-side interventions aimed at increasing awareness among physicians regarding the benefits of influenza vaccinations as well as better compliance with official treatment recommendations. In particular, enhancing the quality of physician agency seems especially important for protecting individuals at high risk from complications, such as the older population or people with chronic diseases. While our research has established the importance of both demand and supply channels in determining vaccination take-up, it does not yet offer a very precise picture of the exact mechanisms at work. Although the established effects of education, patient health behavior and physician quality are highly suggestive, our data do not allow us to measure more disease-specific aspects of health literacy such as knowledge about influenza risk, vaccine effectiveness or insurance coverage. Nor does the data permit a more detailed description of the exact nature of patient–physician interactions with regard to influenza vaccination take-up. A better understanding of these mechanisms, especially the role of physician agency for alleviating asymmetic information among high-risk patients, is clearly important for the design of actionable health policy interventions for increasing vaccination take-up, and thus an important topic for future research. Despite the lack of detailed data on the nature of physician–patient interactions with regard to influenza vaccination, we nonetheless consider our analysis of the simultaneous effects of supply and demand an important first step in integrating supply characteristics and physician agency into microeconometric models of health care use in general and influenza vaccination in particular.