N-way sensitivity analysis: a verification and stability-assessment technique for completely subjective decision analysis trees

The purpose of this paper is to describe the development of N-way sensitivity analysis, a modified version of traditional sensitivity analysis that was created for the purpose of verifying the stability of decisions made by a completely subjective decision-analytic tree. The technique was developed during research that addressed whether nurses make clinical decisions that coincide with those recommended by a decision-analytic model. Since all parameters of the model were derived from subjective assessment, traditional one-way or two-way sensitivity analysis was deemed inappropriate. Consequently, N-way sensitivity analysis was developed and used for the verification of the decision model's results.     

The use of efficiency linear programs for sensitivity analysis in medical decision making

Sensitivity analysis in most medical problems is a complex process involving repeated calculations that can be computationally cumbersome, and its results are only approximate. The authors present a linear program-based approach that reveals the optimum strategies in a decision problem when event probabilities are not known exactly but their value ranges are available. Its application in a clinical decision-making situation is demonstrated. The approach promises to provide a flexible, precise, and computationally efficient technique for sensitivity analysis in medical decision making.     

Stochastic trees: a new technique for temporal medical decision modeling.

This paper introduces stochastic trees, a new modeling approach for the class of medical decision problems in which risks of mortality and morbidity may extend over time. A stochastic tree may be regarded as a continuous-time version of a Markov-cycle tree, or alternately, as a multi-state DEALE model. Optimal decisions in stochastic trees can be determined by rollback, much in the same fashion as decision trees. The author discusses how age-dependent mortality rates and declining incidence rates may be modeled using stochastic trees. Concepts are illustrated using examples from the medical literature. It is argued that stochastic trees possess important advantages over Markov-cycle trees for medical decision modeling.     

Probabilistic sensitivity analysis incorporating the bootstrap: an example comparing treatments for the eradication of Helicobacter pylori.

Decision-analytic models are frequently used to evaluate the relative costs and benefits of alternative therapeutic strategies for health care. Various types of sensitivity analysis are used to evaluate the uncertainty inherent in the models. Although probabilistic sensitivity analysis is more difficult theoretically and computationally, the results can be much more powerful and useful than deterministic sensitivity analysis. The authors show how a Monte Carlo simulation can be implemented using standard software to perform a probabilistic sensitivity analysis incorporating the bootstrap. The method is applied to a decision-analytic model evaluating the cost-effectiveness of Helicobacter pylori eradication. The necessary steps are straightforward and are described in detail. The use of the bootstrap avoids certain difficulties encountered with theoretical distributions. The probabilistic sensitivity analysis provided insights into the decision-analytic model beyond the traditional base-case and deterministic sensitivity analyses and should become the standard method for assessing sensitivity.     

Stratified cost-effectiveness analysis: a framework for establishing efficient limited use criteria

The cost-effectiveness of new health care technologies is conditional upon who receives what therapy and under what circumstances. Understanding this heterogeneity in cost-effectiveness, health care payers often limit reimbursement of therapies to a more restrictive sub-group of patients than that indicated in a product's licensing. Such limits may be based upon clinical or demographic criteria that are prognostic of costs, outcomes or both. However, there is little guidance on how to estimate and interpret stratified cost-effectiveness analysis. In this paper we present a framework for estimating the benefits from stratification that permits consideration of both the opportunity cost resulting from a lack of adherence with criteria and the efficiency loss associated with incorporating equity concerns.     

Developing a pesticide policy for individuals with multiple chemical sensitivity: considerations for institutions.

Institutions are increasingly being asked to accommodate individuals with multiple chemical sensitivity (MCS). Most establishments have chosen to provide such accommodations on a case-by-case basis only. This paper investigates feasible actions that may be taken by institutions to reduce exposure of MCS individuals as well as the general institutional population to pesticides and other substances. Emphasis is placed on procedures that can be instituted on a regular basis and may be combined with case-by-case management for better resolution of problems.     

Probabilistic sensitivity analysis in cost-effectiveness. An application from a study of vaccination against pneumococcal bacteremia in the elderly

OBJECTIVES: We explore the policy implications of probabilistic sensitivity analysis in cost-effectiveness analysis by applying simulation methods to a decision model. METHODS: We present the multiway sensitivity analysis results of a study of the cost-effectiveness of vaccination against pneumococcal bacteremia in the elderly. We then execute a probabilistic sensitivity analysis of the cost-effectiveness ratio by specifying posterior distributions for the uncertain parameters in our decision analysis model. In order to estimate probability intervals, we rank the numerical values of the simulated incremental cost-effectiveness ratios (ICERs) to take into account preferences along the cost-effectiveness plane. RESULTS: The 95% probability intervals for the ICER were generally much narrower than the difference between the best case and worst case results from a multiway sensitivity analysis. Although the multiway sensitivity analysis had indicated that, in the worst case, vaccination in the 85 and older age group was not acceptable from a policy standpoint, probabilistic methods indicated that the cost-effectiveness of vaccination was below $50,000 per quality-adjusted life-year in greater than 92% of the simulations and below $100,000 in greater than 95% of the simulations. CONCLUSIONS: Probabilistic methods can supplement multiway sensitivity analyses to provide a more comprehensive picture of the uncertainty associated with cost-effectiveness ratios and thereby inform policy decisions.     

Foot-and-mouth disease non-structural protein serology in cattle: use of a Bayesian framework to estimate diagnostic sensitivity and specificity of six ELISA tests and true prevalence in the field

The diagnostic performance of six foot-and-mouth disease (FMD) assays for detection of antibodies to the non-structural proteins (NSP) of the FMD virus (FMDV) was estimated using a Bayesian analysis on field sera from cattle of unknown infection status originating from post-FMDV outbreak situations in Israel and Zimbabwe. Estimations of the disease prevalence in both populations were also obtained. The diagnostic sensitivity estimates did not differ between both field studies, although overall Bayesian estimates were markedly higher than those previously reported based on sera from comparable experimentally infected (vaccinated) cattle populations. All NSP-based assays demonstrated a lower diagnostic specificity when applied to the Zimbabwean sera compared to both published specificities and similar Bayesian specificity estimates derived for the Israeli dataset. In Israel, the disease prevalence was estimated at 23.9% (95% credibility interval: 19.5-28.8%), whereas 65.4% (59.0-72.5%) was found in Zimbabwe. The need for reliable diagnostic test performance estimates and the benefits of Bayesian analysis in obtaining them are also addressed.     

Confirmatory adaptive designs with Bayesian decision tools for a targeted therapy in oncology

The ability to select a sensitive patient population may be crucial for the development of a targeted therapy. Identifying such a population with an acceptable level of confidence may lead to an inflation in development time and cost. We present an approach that allows to decrease these costs and to increase the reliability of the population selection. It is based on an actual adaptive phase II/III design and uses Bayesian decision tools to select the population of interest at an interim analysis. The primary endpoint is assumed to be the time to some event like e.g. progression. It is shown that the use of appropriately stratified logrank tests in the adaptive test procedure guarantees overall type I error control also when using information on patients that are censored at the adaptive interim analysis. The use of Bayesian decision tools for the population selection decision making is discussed. Simulations are presented to illustrate the operating characteristics of the study design relative to a more traditional development approach. Estimation of treatment effects is considered as well. Copyright © 2009 John Wiley & Sons, Ltd.     

Pew Memorial Trust policy synthesis: 5. State coverage for organ transplantation: a framework for decision making.

Transplantation of hearts and livers for both adults and children is increasingly viewed as therapeutic and lifesaving, but access to these procedures is impeded by their high cost as well as by a limited supply of organs. In the absence of comprehensive federal coverage, pressure is being brought to bear on states to provide broader access to these procedures. This synthesis provides a framework for the consideration of coverage decisions at the state level. While there are no "right" answers about whether a state should support such coverage, the analytic tools of cost analysis, demand estimation, and assessment of capacity described in this synthesis can better inform the decision-making process.     

Mapping the knowledge base for maritime health: 2. a framework for analysis

The knowledge base for maritime health has a number of constant features that have become apparent over the last 150 years. These can be used to structure an analysis of the current state of knowledge and to identify where there is sound evidence about the nature and scale of risks and about the effectiveness of intervention to reduce harm. It can also show where there are deficiencies in knowledge and point to the ways in which these could be remedied. Past events, as discussed in the first article, also indicate the dynamics of the political, economic and human interactions that are central to improving knowledge and to its application to improve the health of seafarers. The sources of useful knowledge about seafarer's health range from single case reports of an unusual disease to long-term studies of common chronic disease incidence. The most accessible events to record are clinically apparent illness, injury, or cause of death, but active investigative studies may look at risks in the environment, personal risk factors, or pre-clinical phases of disease. Comparisons between subsets of a population are needed to look rigorously at health risks or at the effectiveness of intervention. This is best done if information on the at risk population can be used as the basis for deriving the incidence or prevalence of illness and if the populations compared are as similar as possible in every way, except that being studied. Sometimes large studies in onshore populations can provide information that it is not feasible to collect on seafarers. Information on seafarers' health can be collected in several settings: at sea, on arrival in port, during leave periods, or after retirement. For acute illness and for injury a single setting can provide the basis for estimating risks, but for chronic conditions cases arising in several settings have to be included and the at risk population calculated to enable the incidence to be studied. Knowledge about the health of seafarers can be used to improve prevention both by attention to the conditions of living and working at sea and by selection of seafarers who are considered 'fit' for work. It is also important for defining the needs for emergency care at sea and in port. The overall patterns of illness and injury in seafarers and how these compare with other workers are important inputs to regulatory decisions on the measures to be taken to reduce harm from illness and injury. Markers of improved seafarer health can confirm the effectiveness of measures taken with this goal in mind. Reducing the contribution of health-related impairment to accidents and other risks at sea requires knowledge of the effects of such impairments on performance and safety in the routine and emergency tasks of a seafarer. This information can then be used to determine whether someone with an impairment can safely work at sea.     

Risk-benefit analysis of the AstraZeneca COVID-19 vaccine in Australia using a Bayesian network modelling framework

Thrombosis and Thrombocytopenia Syndrome (TTS) has been associated with the AstraZencea (AZ) COVID-19 vaccine (Vaxzevria). Australia has reported low TTS incidence of < 3/100,000 after the first dose, with case fatality rate (CFR) of 5–6%. Risk-benefit analysis of vaccination has been challenging because of rapidly evolving data, changing levels of transmission, and variation in rates of TTS, COVID-19, and CFR between age groups. We aim to optimise risk–benefit analysis by developing a model that enables inputs to be updated rapidly as evidence evolves. A Bayesian network was used to integrate local and international data, government reports, published literature and expert opinion. The model estimates probabilities of outcomes under different scenarios of age, sex, low/medium/high transmission (0.05%/0.45%/5.76% of population infected over 6 months), SARS-CoV-2 variant, vaccine doses, and vaccine effectiveness. We used the model to compare estimated deaths from AZ vaccine-associated TTS with i) COVID-19 deaths prevented under different scenarios, and ii) deaths from COVID-19 related atypical severe blood clots (cerebral venous sinus thrombosis & portal vein thrombosis). For a million people aged ≥ 70 years where 70% received first dose and 35% received two doses, our model estimated < 1 death from TTS, 25 deaths prevented under low transmission, and > 3000 deaths prevented under high transmission. Risks versus benefits varied significantly between age groups and transmission levels. Under high transmission, deaths prevented by AZ vaccine far exceed deaths from TTS (by 8 to > 4500 times depending on age). Probability of dying from COVID-related atypical severe blood clots was 58–126 times higher (depending on age and sex) than dying from TTS. To our knowledge, this is the first example of the use of Bayesian networks for risk–benefit analysis for a COVID-19 vaccine. The model can be rapidly updated to incorporate new data, adapted for other countries, extended to other outcomes (e.g., severe disease), or used for other vaccines.     

Screening for Chlamydia trachomatis in adolescent males: a cost-based decision analysis.

To evaluate the cost and benefits of screening tests for Chlamydia trachomatis in adolescent males, we developed a decision analysis model and compared the leukocyte esterase urine dipstick test with culture, with direct-smear fluorescent antibody (DFA), and with the option of no screening (no treatment). The leukocyte esterase test has the lowest average cost-per-cure ($51) compared with direct-smear fluorescent antibody ($192) and culture ($414). Compared with the DFA, we estimate that the leukocyte esterase test saves over $9,727 per cohort of 1,000 sexually active adolescent males screened. Sensitivity analyses show the leukocyte esterase test results in a lower cost-per-cure and lower overall costs (per cohort) than culture and direct-smear fluorescent antibody at any prevalence of C. trachomatis infection, and lower overall costs (per cohort) than no screening at prevalences above 21 percent.     

Value-based partnering in healthcare: a framework for analysis

Value-based partnering is designed to move the healthcare system beyond cost-based competition. It recognizes that the healthcare "product" is not a commodity and that much of the value in the system comes from relationships between and among four stakeholders: consumers, providers, health plans, and employers. Given the difficulty of measuring such benefits as quality of care, improved health status, and increased employee productivity, stakeholders within the system traditionally have focused on easily measurable financial considerations such as premium rates. This focus has led to a system that defines relationships in purely financial terms. In contrast, the value-based partnering model presented in this article recognizes the range of factors that stakeholders consider in their relationships with each other. This approach has the potential to change the nature of competition and presents opportunities for those organizations that can effectively partner with other stakeholders and demonstrate value, rather than just lower cost. Moreover, by recognizing the interdependencies among stakeholder groups, the approach creates a strategic reason for employers, health plans, providers, and consumers to exchange information and create long-term alliances.     

Goodness-of-fit measures for individual-level models of infectious disease in a Bayesian framework

In simple models there are a variety of tried and tested ways to assess goodness-of-fit. However, in complex non-linear models, such as spatio-temporal individual-level models, less research has been done on how best to ascertain goodness-of-fit. Often such models are fitted within a Bayesian statistical framework, since such a framework is ideally placed to account for the many areas of data uncertainty. Within a Bayesian context, a major tool for assessing goodness-of-fit is the posterior predictive distribution. That is, a distribution for a test statistic is found through simulation from the posterior distribution and then compared with the observed test statistic for the data. Here, we examine different test statistics and ascertain how well they can detect model misspecification via a simulation study.     

Spatial multicriteria decision analysis of flood risks in aging-dam management in China: a framework and case study

Approximately 30,000 dams in China are aging and are considered to be high-level risks. Developing a framework for analyzing spatial multicriteria flood risk is crucial to ranking management scenarios for these dams, especially in densely populated areas. Based on the theories of spatial multicriteria decision analysis, this report generalizes a framework consisting of scenario definition, problem structuring, criteria construction, spatial quantification of criteria, criteria weighting, decision rules, sensitivity analyses, and scenario appraisal. The framework is presented in detail by using a case study to rank dam rehabilitation, decommissioning and existing-condition scenarios. The results show that there was a serious inundation, and that a dam rehabilitation scenario could reduce the multicriteria flood risk by 0.25 in the most affected areas; this indicates a mean risk decrease of less than 23%. Although increased risk (<0.20) was found for some residential and commercial buildings, if the dam were to be decommissioned, the mean risk would not be greater than the current existing risk, indicating that the dam rehabilitation scenario had a higher rank for decreasing the flood risk than the decommissioning scenario, but that dam rehabilitation alone might be of little help in abating flood risk. With adjustments and improvement to the specific methods (according to the circumstances and available data) this framework may be applied to other sites.     

HIV-related stigma: adapting a theoretical framework for use in India

Stigma complicates the treatment of HIV worldwide. We examined whether a multi-component framework, initially consisting of enacted, felt normative, and internalized forms of individual stigma experiences, could be used to understand HIV-related stigma in Southern India. In Study 1, qualitative interviews with a convenience sample of 16 people living with HIV revealed instances of all three types of stigma. Experiences of discrimination (enacted stigma) were reported relatively infrequently. Rather, perceptions of high levels of stigma (felt normative stigma) motivated people to avoid disclosing their HIV status. These perceptions often were shaped by stories of discrimination against other HIV-infected individuals, which we adapted as an additional component of our framework (vicarious stigma). Participants also varied in their acceptance of HIV stigma as legitimate (internalized stigma). In Study 2, newly developed measures of the stigma components were administered in a survey to 229 people living with HIV. Findings suggested that enacted and vicarious stigma influenced felt normative stigma; that enacted, felt normative, and internalized stigma were associated with higher levels of depression; and that the associations of depression with felt normative and internalized forms of stigma were mediated by the use of coping strategies designed to avoid disclosure of one's HIV serostatus.