Probabilistic sensitivity analysis using Monte Carlo simulation. A practical approach

The data for medical decision analyses are often unreliable. Traditional sensitivity analysis--varying one or more probability or utility estimates from baseline values to see if the optimal strategy changes--is cumbersome if more than two values are allowed to vary concurrently. This paper describes a practical method for probabilistic sensitivity analysis, in which uncertainties in all values are considered simultaneously. The uncertainty in each probability and utility is assumed to possess a probability distribution. For ease of application we have used a parametric model that permits each distribution to be specified by two values: the baseline estimate and a bound (upper or lower) of the 95 percent confidence interval. Following multiple simulations of the decision tree in which each probability and utility is randomly assigned a value within its distribution, the following results are recorded: (a) the mean and standard deviation of the expected utility of each strategy; (b) the frequency with which each strategy is optimal; (c) the frequency with which each strategy "buys" or "costs" a specified amount of utility relative to the remaining strategies. As illustrated by an application to a previously published decision analysis, this technique is easy to use and can be a valuable addition to the armamentarium of the decision analyst.     

Using full probability models to compute probabilities of actual interest to decision makers

The objective of this paper is to illustrate the advantages of the Bayesian approach in quantifying, presenting, and reporting scientific evidence and in assisting decision making. Three basic components in the Bayesian framework are the prior distribution, likelihood function, and posterior distribution. The prior distribution describes analysts' belief a priori, the likelihood function captures how data modify the prior knowledge; and the posterior distribution synthesizes both prior and likelihood information. The Bayesian approach treats the parameters of interest as random variables, uses the entire posterior distribution to quantify the evidence, and reports evidence in a "probabilistic" manner. Two clinical examples are used to demonstrate the value of the Bayesian approach to decision makers. Using either an uninformative or a skeptical prior distribution, these examples show that the Bayesian methods allow calculations of probabilities that are usually of more interest to decision makers, e.g., the probability that treatment A is similar to treatment B, the probability that treatment A is at least 5% better than treatment B, and the probability that treatment A is not within the "similarity region" of treatment B, etc. In addition, the Bayesian approach can deal with multiple endpoints more easily than the classic approach. For example, if decision makers wish to examine mortality and cost jointly, the Bayesian method can report the probability that a treatment achieves at least 2% mortality reduction and less than $20,000 increase in costs. In conclusion, probabilities computed from the Bayesian approach provide more relevant information to decision makers and are easier to interpret.     

Decentralization of health systems in Latin America.

Decentralization is often a major part of health reform policies. However, there have been few attempts to comparatively study the degree of decentralization and the effects of decentralization on equity of allocations to health, so we do not know how best to implement this reform. This article uses an innovative comparative analysis of the "decision space" that was allowed to local municipalities in the health reforms of Bolivia and Chile, two countries that have had several years of experience in implementing decentralization. The studies found that relatively little decision space was allowed to local authorities over key functions of health care systems. The studies also found that central authorities often reduce the decision space in order to direct more resources to health or to restrict local choice over human resources issues. The studies found that more equitable allocations of health funding were achieved through a common equalization fund for the municipalities in Chile and by forcing the assignment to health of a specific percentage of the central government transfers to municipalities in Bolivia.     

The Expert and the Lay Public: Reflections on Influenza A (H1N1) and the Risk Society

Trust between the lay public and scientific experts is a key element to ensuring the efficient implementation of emergency public health measures.In modern risk societies, the management and elimination of risk have become preeminent drivers of public policy. In this context, the protection of public trust is a complex task. Those actors involved in public health decision-making and implementation (e.g., mass vaccination for influenza A virus) are confronted with growing pressures and responsibility to act. However, they also need to accept the limits of their own expertise and recognize the ability of lay publics to understand and be responsible for public health.Such a shared responsibility for risk management, if grounded in participative public debates, can arguably strengthen public trust in public health authorities and interventions.The influenza A (H1N1) virus pandemic was not as devastating as expected, so the preventive health measures that were deployed to cope with the outbreak are now being challenged.1,2 Questions remain about the appropriateness of large-scale population vaccination programs, such as those promoted as the best response to the expected influenza epidemic in spring 2009. Large-scale vaccination involves considerable financial (and other resource) costs for governments, and the political decision to make such an investment in public health is not without repercussions. In particular, if such decision-making processes are not fully transparent and well justified—for example, if accusations of conflict of interest arise, as was the case with World Health Organization (WHO) recommendations2,3—public trust in the resulting public health program or intervention can be threatened. Clearly, an erosion of public trust in the judgments of public health authorities (whether they be local, national, or international) can have serious negative consequences on the future implementation of other emergency response programs.4To respond to this problem of a loss of (or weakened) public trust, we must understand its sociocultural and historical origins. The examination of past implementations of emergency programs can hopefully help us understand our strengths and faults and eventually serve as tools for continuously improving our management of public health in such emergency situations. Neustadt and Fineberg’s book on the 1976 swine flu “affair” is a good example of how critical of ourselves we should be to react better to such crises in the future.5 For this same purpose, and by placing a sociological macroscopic lens over a particular recent crisis, we are presenting a case analysis of the 2009 H1N1 flu pandemic.Drawing on the literature in the social sciences, we have affirmed that the management of health crises is necessarily also the management of human crises. We integrate reflections from contemporary bioethics and political philosophy, in line with views about the responsibility of decision makers in democratic states. First, we argue that the concept of trust (e.g., by the public in health experts) should be situated in the context of modernity—namely, in a risk society in which the public and policymakers are increasingly concerned with safety and the maximal reduction of certain risks. Second, we highlight a close relation between risk perception (known to be subjective) and risk assessment (expected to be objective), which deserves special attention, given the important role played by experts in the management of public health. Third, we suggest that public health actors (professionals, science advisers, policymakers) need to accept the limits of their own expertise (and of its objectivity) and responsibility and recognize the ability of lay publics to understand and to take responsibility for their public health. As a consequence, we argue that public health actors should engage more actively in ongoing participative and deliberative public debates both to preserve and to strengthen public trust toward public health authorities and interventions.We do not aim to judge or to hold accountable the individuals for the decisions that were made during the crisis, which is beyond the scope of this article and, to our minds, less interesting than examining the structural elements that make such behavior “the norm” for experts and decision makers and thus lead repeatedly to situations like H1N1. If we simply focus on pointing fingers at a few individuals to be held personally responsible, we miss the larger social dynamics that arguably generated the problematic inconsistencies between the messages put forward by public health experts and what was understood or accepted by the general public.     

Standards for chemical quality of drinking water: A critical assessment

Summary The author critically reviews present standards for the chemical quality of drinking water, particularly the limits proposed by the Commission of the European Communities (CEC) in 1979. Particularly, the general principles of standard setting are discussed. It appears that there exists a surprisingly high similarity in drinking water limits, issued by various national and international authorities, although for other environmental compartments important discrepancies exist. Usually, drinking water limits lack adequate documentation, and appear often to be copied from other existing lists. There is an apparent lack of logical consistency in limits set for food, ambient or workroom air, and drinking water, probably due to lack of communication between health experts and decision-making authorities. Moreover, there is a lack of toxicologic studies, explicitly aimed at setting limits. Extrapolation from the acceptable daily intakes (ADI) for food or the Threshold Limit Value (TLV)-Maximum Acceptable Concentration (MAC) for workroom air could be undertaken to derive tentative drinking water limits, as long as explicitly designed studies for drinking water are not yet available.     

Bayesian methods for health-related decision making

This article reviews the basics of Bayesian decision theory, and comments on its use in medical decision making. It emphasizes the subjectivity of the probability and utility inputs, and the desirability, in certain contexts, of representing several decision makers, each with his or her own probabilities and utilities. Applications and ethical considerations are also discussed. A brief bibliography gives pointers to the literature.     

Assessment of guidelines for limiting exposures to emf using methods of probabilistic risk analysis

Allowable limits of human exposure to radiofrequency fields commonly include a "factor of safety," typically between 10 to 50, which is somewhat arbitrary. The broad objective in our work is to assess radiofrequency exposure limits, hazard thresholds, and safety factors using methods of probabilistic risk analysis. We focus our analysis on the variables affecting peak radiofrequency specific energy absorption rate (SAR) values in the brain from digital mobile telephones operating at approximately 900 MHz. As SAR is defined as a product of positive random variables, it is not unreasonable to assume that SAR has a lognormal distribution. Our analysis of component SAR variables such as conductivity and permittivity of grey brain matter and radiated field strengths using experimental and numerical modeling data strongly supports our hypothesis that SAR values are distributed lognormally. It then follows that the probability that the SAR exceeds a certain threshold can be derived directly and is shown to be very low for handset SARs relative to presently allowable standard limits.     

Injuries from domestic violence in the Autonomous Community of Valencia (Spain)

The lack of homogeneous and reliable epidemiological data on domestic violence greatly limits public decision making on the help that should be provided to victims of this form of abuse. Health professionals are obliged to report cases of domestic violence in adults to the judicial authorities and a unified, easily completed model for reporting injuries from domestic violence has been established in the Autonomous Community of Valencia. From June to October 2005, 500 cases of domestic violence were received and processed, most of which were reported by primary care physicians (68%). Cases of domestic violence occurred mainly in young women (aged, 20-39 years), either married or cohabiting, and with primary or secondary level education. The most frequent findings were physical injury or psychological damage (62%). Eighty-three percent of victims had at least one antecedent of abuse.     

A pitfall in utility assessment--patients' undisclosed investment decisions.

Among those decisions that may be made by a patient in response to an illness, the authors single out a certain class: contingent investment decisions. They are characterized by the patient's committing him- or herself, on the basis of prognostic counseling, to a certain action or non-action that he or she may regret in retrospect. Examples show that, when assessing utilities, the decision analyst runs a risk of handling such investment decisions incorrectly, unless they are made explicit and incorporated into the medical decision process. The anomaly is explained as a violation of the structural rules for decision trees and is also interpreted in terms of "the price of prognostic ignorance," a quantity closely related to the expected utility value of perfect information.     

Can primary care be both patient-centred and community-led?

Examines the relationships between the macro-, meso-, and micro-levels in the NHS at the end of the fundholding period and considers their contemporary implications for primary care groups (PCGs) and local health care co-operatives (LHCCs). Fundholding achieved some success in challenging the way in which services were provided at the micro-level (the practice), but had a less marked effect in terms of changing service provision at the health authority (meso-) level or in developing collaborative working with trusts and health authorities in strategic decision making. The health authorities prioritized alternative models of devolved commissioning. Trusts regarded fundholders as a distraction who exerted influence and commanded trust management time disproportionate to their "market share". PCGs and LHCCs represent a shift back to the meso-level in service planning and purchasing. As such there is a risk that the micro-level benefits of fundholding and other forms of devolved commissioning will be lost, while uncertainties remain regarding the capacity of PCGs and LHCCs to incorporate GPs into a collaborative approach to strategic decision making.     

Decision makers' and scientists' opinion about contingent valuation and choice experiments for measuring willingness to pay in health care: results from a survey in Germany

OBJECTIVES: Assessment of willingness to pay (WTP) by contingent valuation (CV) and choice experiments (CE) is increasingly performed in economic evaluation of health care. However, the question of whether the methods for measuring WTP are acceptable to decision makers and scientists has remained largely unacknowledged. The aim of this study was to learn more about decision makers' and scientists' opinion concerning these methods. METHODS: An expert group developed a questionnaire consisting of key items that may influence the opinion about CV and CE according to the constructs "attitude toward behavior," "subjective norm," and "behavioral intention" as defined by the Theory of Reasoned Action by Ajzen and Fishbein. In a survey, seventy-seven decision makers representing key institutions in the German healthcare system and forty-two scientists in health economics completed the questionnaire. RESULTS: Scientists and decision makers in particular did not show a high intention to use methods for measuring WTP. Skepticism regarding precision of the methods and subjects' capability to imagine paying an amount of money for a certain health commodity were stated along with the assertion that the hypothetical decision-making scenario was rather a distant reality. Nevertheless, the majority of scientists and decision makers did not state rejection of the methods. CONCLUSIONS: Increasing the probability of using methods for measuring WTP in health care, the hypothetical scenarios should be made more realistic and payment vehicles should be used to help patients relate payment to a real health benefit. Moreover, an intensive discussion on the potential usefulness of CV/CE without excluding ethical concerns in comparison to existing alternatives has to be resumed.     

Stochastic thresholds

Thresholds have traditionally been represented by a single number; the optimal management of the patient depends on whether his probability of disease is above or below this number. The concept of a threshold as a single number, however, inadequately represents the treatment approach of a group of physicians who do not have all the same threshold or a single physician who is uncertain about the exact value of the threshold. An alternative to a single valued threshold is to consider the threshold as having a probability distribution: for every probability that the patient has the disease there is a probability that the threshold is exceeded. This "stochastic" threshold model contains information about the uncertainty of the threshold estimation. Stochastic thresholds can be useful for testing the sensitivity of a management decision to the patient's probability of disease. They can also be used for comparing the standards of practice of individual physicians or comparing the practice of an individual physician with that of a group.     

Intensive care units in the triage mode: an organizational perspective

Decisions to admit and discharge patients to and from the intensive care unit (ICU) when beds are scarce should be made in accordance with the triage principle--that is, allocate resources on the basis of the ability to benefit from intensive care. However, uncertainty over resource capacity and patient prognosis limits the ability of decision makers to use this prioritization principle and results in ICUs containing inappropriately placed patients who are denying or delaying care to patients who could benefit more. Using Jay Galbraith's "information processing" model, ICU admission and discharge decision making is described. Organizational strategies to reduce uncertainty and improve decision making are discussed, including strengthening the management role of the ICU physician director and employing prognostic instruments (e.g., mortality prediction models) to share and process information.     

Risk management by labelling 26 fragrances?: Evaluation of Article 10 (1) of the seventh Amendment (Guideline 2003/15/EC) of the Cosmetic Directive

Background

Some fragrance compounds are severe contact allergens. According to the so-called “26 allergens rule” (Article 1 (10) of Directive 2003/15/EC) (EC, 2003), 26 supposedly allergenic fragrances must be listed on the containers of cosmetics products if they are present above certain mass percentages in the product. This declaration is meant to inform the consumer of potential risks of skin sensitizers in the products.

Objective

The objective of this paper is to validate whether “the 26 allergens rule” meets the expectations to improve consumer protection.

Methods

The method used for this validation was on one hand a reflection on the elements of the approach used in “the 26 allergens rule” and on the other hand a product analysis of 742 products by 4 large producers of cosmetic products on the German market.

Results

It was found that more than 50% of these cosmetic and washing and cleansing products contain at least one of the 26 substances above the thresholds for labelling and that there are 14% of all products which contain strong allergens. Many consumers apparently still buy these products. The indirect effect that producers reduce the amounts of these fragrances to avoid declaration seems to be small.

Discussion and recommendations

Several arguments were assembled which show that other instruments are needed to ensure consumer protection or protection of the environment. This paper recommends different approaches. The use of a list of single substances in such a directive is not in line with scientific standards. It is recommended to base decision making on comprehensive risk assessments or at least on valid and strong criteria. More parameters need to be involved, not only contact allergy. As illustrated in this article, the roles taken over by authorities and manufacturers in risk management of the “26 allergens” are relatively small compared with the responsibility carried by consumers. However, consumers are only able to take over their part properly if they are sufficiently trained and have the necessary infrastructure, capability and time to inform themselves. Regulations are not effective if they load the major responsibility for risk management on consumers, instead of on authorities and manufacturers. A successful risk management would include bans and restrictions of especially hazardous substances issued by governments, as well as efficient surveys to control the implementation of regulations by the responsible authorities. It would also include that producers meet the legal standards and take over voluntary action to make products safer. The evaluation of “the 26 allergens rule” is an example which can be transferred to other regulations and which could help to improve future regulatory approaches, with a focus on the roles authorities, manufacturers and consumers play in a promising risk management.     

Recent Swedish experiences in 222Rn control

Swedish local authorities are responsible for decreasing 222Rn progeny concentrations in homes in their municipalities. To obtain an overall view of their experiences, concerned national authorities sent a questionnaire in 1986 to local authorities. The results were intended to form one basis for decisions by the government regarding revised statements on financial contributions, limits, etc. The results were also intended to be of use to national authorities in determining limits and recommendations and to local authorities in their field work. One result of the survey was an enhanced interest in the Rn problem among Swedish politicians and the mass media. This increased attention resulted in new plans for continued work to decrease Rn levels indoors during 1987-1989, on both a national and a local level. The experiences of the local authorities show that Rn progeny concentrations decreased to below the design level in 95% of newly built houses investigated. It was also found that Rn progeny concentrations were below the limit for reconstruction in 53% of existing homes that previously had levels exceeding the limit.     

Dynamic probability control limits for risk‐adjusted CUSUM charts based on multiresponses

For a patient who has survived a surgery, there could be several levels of recovery. Thus, it is reasonable to consider more than two outcomes when monitoring surgical outcome quality. The risk‐adjusted cumulative sum (CUSUM) chart based on multiresponses has been developed for monitoring a surgical process with three or more outcomes. However, there is a significant effect of varying risk distributions on the in‐control performance of the chart when constant control limits are applied. To overcome this disadvantage, we apply the dynamic probability control limits to the risk‐adjusted CUSUM charts for multiresponses. The simulation results demonstrate that the in‐control performance of the charts with dynamic probability control limits can be controlled for different patient populations because these limits are determined for each specific sequence of patients. Thus, the use of dynamic probability control limits for risk‐adjusted CUSUM charts based on multiresponses allows each chart to be designed for the corresponding patient sequence of a surgeon or a hospital and therefore does not require estimating or monitoring the patients' risk distribution. Copyright © 2017 John Wiley & Sons, Ltd.     

What is the meaning of threshold in laser injury experiments? Implications for human exposure limits

The derivations of human exposure limits for laser radiation rely heavily upon experimental ocular injury studies. The limits are derived by committees of ophthalmic experts through a review of all available threshold data and an understanding of mechanisms of laser/tissue interaction. A major point of discussion in this derivation process relates to the level of uncertainty of the threshold of injury. An indication of the level of uncertainty relates to the slope of the transformed dose-response curve, or the "probit plot" of the data. The most cited point on the probit plot is the exposure that represents a 50% probability of injury: the ED-50. This value is frequently referred to as the "threshold," even though some experimental damage points exist below this "threshold." An analysis of any number of example data sets reveals that the slope in most experiments cannot be explained by biological variation alone. The optical, thermophysical, and biological factors influencing the probit plot are critically analyzed to provide guidance for deriving exposure limits. By theoretically modeling an experiment, small errors in focus are shown to produce a substantial change in the ED-50 and the slope of the probit plot.     

Assessing elemental mercury vapor exposure from cultural and religious practices

Use of elemental mercury in certain cultural and religious practices can cause high exposures to mercury vapor. Uses include sprinkling mercury on the floor of a home or car, burning it in a candle, and mixing it with perfume. Some uses can produce indoor air mercury concentrations one or two orders of magnitude above occupational exposure limits. Exposures resulting from other uses, such as infrequent use of a small bead of mercury, could be well below currently recognized risk levels. Metallic mercury is available at almost all of the 15 botanicas visited in New York, New Jersey, and Pennsylvania, but botanica personnel often deny having mercury for sale when approached by outsiders to these religious and cultural traditions. Actions by public health authorities have driven the mercury trade underground in some locations. Interviews indicate that mercury users are aware that mercury is hazardous, but are not aware of the inhalation exposure risk. We argue against a crackdown by health authorities because it could drive the practices further underground, because high-risk practices may be rare, and because uninformed government intervention could have unfortunate political and civic side effects for some Caribbean and Latin American immigrant groups. We recommend an outreach and education program involving religious and community leaders, botanica personnel, and other mercury users.     

Data and decisions: Can mental health management be knowledge-based?

Three major factors suggest a healthy future for data-based decision making within mental health authorities:(1) the improved knowledge base related to the treatment and management of serious mental illness, (2) advances in data-processing technology and (3) conceptual advances in management information system design, most notably the National Institute of Mental Health (NIMH) Mental Health Statistics Improvement Package. This paper brifly outlines these three factors and goes on to examine information needed by state mental health authorities (SMHAs) to enhance decision making. The clientlevel data necessary for data-based policy decisions, while still scarce, are increasingly available and are increasingly finding homes within SMHA management information systems. As SMHAs improve their information systems to accommodate such data, they face substantial implementation challenges and substantial payoffs in terms of increased knowledge for decision making. This report is an expansion of a talk given by the authors at a conference sponsored by the National Institute of Mental Health on Oct. 30, 1989, for recipients of grants to enhance the information-processing capabilities of state mental health authorities.     

The application of risk assessment methods in making veterinary public health and animal health decisions.

The newly emerging discipline of quantitative risk assessment has wide application in the field of veterinary public health and animal health. Regulatory authorities are increasingly faced with public policy decisions that must assess the risks of new technology or practices relative to the potential benefits, thereby establishing a level of acceptable risk. The elements of risk are a choice of action, a probability of loss and a magnitude of loss. Perceived risk and actual risk are seldom equivalent; adoption of the methodologies used in technological and human health risk assessments will allow veterinary regulators to make better decisions. Determination of levels of acceptable risk are increasingly dependent on quantitative models, and examples are presented for evaluation of different post-mortem meat inspection systems, estimating disease risks associated with animal embryo transfer and formulating national border protection strategies. All models have some degree of subjectivity, and the decisions made by regulators and risk managers should incorporate a wide knowledge of the risk assessment process, as well as the conditions of use that will occur in the real world.