A novel integrated Markov-based model for vaccine cost-effectiveness in Lyme borreliosis and tick-borne encephalitis: implications for further research

Introduction In 2012, a pioneering Markov model was developed to calculate the cost-effectiveness of TBE vaccination in Slovenia (1,2). This groundbreaking model has served as a foundation for subsequent research and has provided a platform to demonstrate some of the policy implications to date (1,2). Furthermore, based on the original article (1), Finland (3), Sweden (4,5,6), Estonia (7), and Germany (8) have performed cost-effectiveness analyses on TBE. For the first time, the full burden of TBE, measured in disability-adjusted life years (DALYs), was estimated (9). A correction factor of 4.5 was calculated to account for the underestimation of reported TBE cases. This underestimation includes two calculations: a factor of 2.5 to correct for under-ascertainment and a factor of 1.8 to address under-reporting (9). The 1999 study evaluated the cost-effectiveness of LB vaccination in the U.S, developing a decision tree (10). Subsequent analyses in 2001 and 2002 used a Markov model to further explore the vaccine's cost-effectiveness (11,12). In 2019, Mihajlović J. et al. examined the cost-effectiveness of a potential single combined anti-tick vaccine targeting LB and TBE (13), using a Markov model based on previous research on LB and TBE. They estimated the combined incremental cost-effectiveness ratio (ICER) and propose that their model could be useful for assessing future vaccination strategies. This article aims to contribute to the scientific community by critically evaluating both the strengths and limitations of Mihajlović J. et al.'s model and showing how these may influence its conclusions. It also proposes ways to improve and identifies gaps that could guide future research efforts. Strengths of the study by Mihajlović J. et al. (2019) (13) Mihajlović J. et al used a Markov model to calculate the combined incremental cost-effectiveness ratio (ICER) for a potential single, combined anti-tick vaccine. This calculation is based on two established disease pathways: LB and TBE, which are becoming increasingly significant in regions where both diseases are endemic and share overlapping tick-borne disease burdens. This study opens the way for meaningful discussions about combined vaccine development and vaccination strategy planning by presenting a model that can be enhanced and expanded upon. While the study provides a useful starting point for assessing combined vaccination strategies, its methodological limitations significantly impact the reliability and policy relevance of its results. These limitations are categorized into three groups: structural and modeling assumptions; epidemiological and clinical validity; and economic inputs and cost data. Limitations of the study by Mihajlović J. et al. (2019) (13) 1. Structural and modeling assumptions Independent modeling of LB and TBE without co-infection The model calculates the ICER for a combined vaccine by simulating LB and TBE independently, using separate probabilities of infection. It does not account for co-infection (simultaneous infections with LB and TBE), despite evidence that co-infection occurs in 9.1% to 16.7% of cases (14,15). Impact on results: This approach likely underestimates the value of the vaccine. By excluding the possibility of co-infection, the model may fail to capture the combined vaccine's full protective benefit. Lack of disease-specific ICERs The study presents only a combined ICER for LB and TBE, without disaggregating the results by disease. Impact on results: Without disease-specific ICERs, it is difficult to evaluate the relative contribution of each disease to the overall cost-effectiveness. This limits the model's utility for targeted policy decisions and vaccine prioritization. Tornado diagrams based on unreliable inputs The study uses tornado diagrams to illustrate key ICER drivers, but these are based on uncertain or speculative input parameters. Impact on results: Flawed base-case assumptions may distort the prioritization of cost drivers, leading to misleading conclusions about which variables most influence vaccine value. Absence of probabilistic sensitivity analysis The authors state that a standard probabilistic sensitivity analysis (PSA) was not performed because of "potentially imprecise results". Impact on results: PSA is essential for quantifying uncertainty, and its absence prevents assessment of ICER variability and undermines confidence in the robustness of the findings. The lack of 95% confidence intervals further limits reliability. Absence of base-case and worst-case scenarios The study does not present base-case or worst-case scenarios to explore the range of possible outcomes. Impact on results: This approach may result in an incomplete risk assessment, lacking a full representation of the potential vaccine impact, leading to less reliable conclusions. 2. Epidemiological and clinical validity Incidence rate uncertainty The study uses incidence rates ranging from 9 to 30.3 /100,000 for LB and 1.51 to 3.53 /100,000 for TBE, without specifying the time period. Impact on results: These rates are significantly lower than historical data (e.g., 260/100,000 for LB and 10/100,000 for TBE), which may lead to underestimation of disease burden and bias the ICER (16-18). Misclassification of EM as a chronic sequela EM is modeled as a potentially chronic condition lasting up to one year and leading to complications. Impact on results: This contradicts clinical guidance (18-20), which defines EM as self-limiting. Misclassification inflates disease burden and distorts ICER estimates. Uniform utility weights for LB forms and their sequelae The model applies identical utility weights to active LB and its sequelae. Impact on results: This fails to capture differences in quality of life impact, potentially misrepresenting disease burden and biasing ICER calculations. Unrealistic categorization of TBE sequelae The model assumes that sequelae severity corresponds strictly to initial disease severity (13), unlike other studies (1,3,4,5,7,8,21) that recognize variability. Impact on results: This rigid classification overlooks (13) clinical realities (21)—such as moderate TBE leading to severe sequelae—and may misrepresent vaccine benefits. Misalignment of the distribution of TBE sequelae across the severity level The proportion of mild sequelae resulting from mild TBE is 52.62%, the proportion of moderate sequelae stemming from moderate TBE is 9.25%, while 50.26% of patients with severe TBE experience severe sequelae, imply the highest recovery rate in moderate TBE (90.75%), followed by severe (49.74%) and mild (46.38%) cases. Impact on results: This contradicts clinical expectations, where mild TBE typically has the highest recovery rate (21). The lack of supporting evidence raises concerns about model validity. 3. Economic inputs and cost data Use of outdated cost estimates for TBE sequelae Cost inputs for mild (€70), moderate (€122), and severe (€28,592) sequelae are sourced from the 2012 study (1). Impact on results: Outdated cost inputs may lead to inaccurate ICER estimates and misinformed conclusions about vaccine cost-effectiveness. Speculative vaccine parameters The study relies on assumptions and external projections regarding the anti-tick vaccine's efficacy, dosing regimen, and cost (1,10,11,12) without empirical validation. Impact on results: These speculative inputs introduce uncertainty and may mislead stakeholders about the vaccine's potential value, affecting funding and development decisions. Discussion The study by Mihajlović J. et al. (2019), represents an early attempt to assess the cost-effectiveness of a single, combined anti-tick vaccine targeting LB and TBE. While it is a good starting point for evaluating combined vaccination strategies, the study has significant methodological limitations. These limitations include structural and epidemiological inaccuracies, outdated cost inputs, and speculative vaccine parameters. These limitations diminish the reliability of the study's findings and limit its relevance for policy and funding decisions. The model's utility is further weakened by the absence of probabilistic sensitivity analysis, disease-specific ICERs, and validation against real-world data, particularly in regions where co-infections are clinically relevant. By failing to account for co-infection of LB and TBE, the model overlooks a key dimension of disease burden and vaccine benefit. As cost-effectiveness models increasingly influence public health priorities, future research must adopt more rigorous and transparent approaches (22). This includes: - Restructuring the model to simulate both individual infections and co-infections. - Applying validated utility weights and clinically accurate disease classifications. - Updating cost inputs to reflect current healthcare expenditures. - Incorporating robust sensitivity analyses, including probabilistic and scenario-based techniques. - Disaggregating ICERs to enable disease-specific policy evaluation. A revised Markov model that captures single and co-infection pathways while accounting for underascertainment and underreporting would allow for more accurate comparisons of disease burden and vaccine impact. This framework would support evidence-based decision-making as an anti-tick vaccine is developed and evaluated (22). Mihajlović J. et al. identified a cost-effectiveness threshold of €50,000/QALY gained in Slovenia (13). From a societal perspective, a combined anti-tick vaccine strategy with two doses is favorable, with an ICER of €46,061/QALY gained for protection against both diseases (13). However, assuming adults receive 18 doses results in a sharply increased ICER of €246,456/QALY gained for combined protection, which far exceeds the acceptable threshold (13). In contrast, other studies have shown that TBE-only vaccinations remain cost-effective with multiple doses (1,3,6,7). From a Slovenian healthcare payer's perspective, the ICER for TBE vaccination ranges from €15,128 to €20,099/QALY gained for adults receiving 18 doses, which is well within acceptable limits (1). In Sweden, ICERs for TBE vaccination (including multiple doses) are estimated at 27,761 SEK/QALY, 99,527 SEK/QALY, and 160,827 SEK/QALY when starting at ages 3, 40, and 50, respectively, from the healthcare payer's perspective (6). In Estonia, the ICER is €60,572/QALY for the general population and €24,576/QALY for the elderly (50+), both including boosters and assessed from the healthcare payer's view (7). A Finnish study found that TBE vaccination is cost-saving at an incidence of 15/100,000, cost-effective at 10/100,000, but not cost-effective at 5/100,000, where the cost/QALY exceeded €100,000—all scenarios included boosters for the general population (3). A recent German study showed that, from a healthcare perspective, the ICER for TBE vaccination using only the primary series of three doses is €253,529/QALY for the 1-85 age group and €82,358/QALY for the 60-85 age group when the incidence is lower than 5/100,000 (8). The cost-effectiveness of LB in the US was estimated at $62,300&/QALY gained for individuals in endemic areas who received three doses. When annual boosters were included, the ICER increased to $72,700/QALY gained. However, when indirect costs (e.g., productivity losses) were considered alongside annual boosters, the ICER increased to $55,000/QALY gained. The analysis was conducted over a 10-year time horizon (11). A single combined anti-tick vaccine that protects against multiple tick-borne diseases, including both LB and TBE (23), could provide several advantages, such as having one shot for various diseases. This is especially beneficial for children and individuals who receive other vaccines, as it improves vaccination adherence and protection. Investing in the development of a potential human anti-tick vaccine is therefore advantageous from individual, health, and societal perspectives. Conclusion The study by Mihajlović (13) is an important step toward evaluating the cost-effectiveness of a single combined anti-tick vaccine. However, to realize its full potential, future research should use robust, evidence-based data and enhanced modeling approaches. These improvements are essential for accurately assessing the benefits of a combined vaccine and will provide valuable insights for future health economic planning for LB, TBE, and combined prevention strategies.