Bernard C. K. Choi

Can scientists and policy makers work together

This paper addresses a fundamental question in evidence based policy making—can scientists and policy makers work together? It first provides a scenario outlining the different mentalities and imperatives of scientists and policy makers, and then discusses various issues and solutions relating to whether and how scientists and policy makers can work together. Scientists and policy makers have different goals, attitudes toward information, languages, perception of time, and career paths. Important issues affecting their working together include lack of mutual trust and respect, different views on the production and use of evidence, different accountabilities, and whether there should be a link between science and policy. The suggested solutions include providing new incentives to encourage scientists and policy makers to work together, using knowledge brokers (translational scientists), making organisational changes, defining research in a broader sense, re-defining the starting point for knowledge transfer, expanding the accountability horizon, and finally, acknowledging the complexity of policy making. It is hoped that further discussion and debate on the partnership idea, the need for incentives, recognising the incompatibility problems, the role of civil society, and other related themes will lead to new opportunities for further advancing evidence based policy and practice.

Slopes of a Receiver Operating Characteristic Curve and Likelihood Ratios for a Diagnostic Test

This paper clarifies two important concepts in clinical epidemiology: the slope of a receiver operating characteristic (ROC) curve and the likelihood ratio. It points out that there are three types of slopes in an ROC curve--the tangent at a point on the curve, the slope between the origin and a point on the curve, and the slope between two points on the curve. It also points out that there are three types of likelihood ratios that can be defined for a diagnostic test that produces results on a continuous scale--the likelihood ratio for a particular single test value, the likelihood ratio for a positive test result, and the likelihood ratio for a test result in a particular level or category. It further illustrates mathematically and empirically the following three relations between these various definitions of slopes and likelihood ratios: 1) the tangent at a point on the ROC curve corresponds to the likelihood ratio for a single test value represented by that point; 2) the slope between the origin and a point on the curve corresponds to the positive likelihood ratio using the point as a criterion for positivity; and 3) the slope between two points on the curve corresponds to the likelihood ratio for a test result in a defined level bounded by the two points. The likelihood ratio for a single test value is considered an important parameter for evaluating diagnostic tests, but it is not easily estimable directly from laboratory data because of limited sample size. However, by using ROC analysis, the likelihood ratio for a single test value can be easily measured from the tangent. It is suggested that existing ROC analysis software be revised to provide estimates for tangents at various points on the ROC curve.

Risk factors for diabetes mellitus by age and sex: results of the National Population Health Survey.

Abstract.Aims/hypothesis: We aimed to assess the risk factors for diabetes mellitus, by age and sex in Canada and to recommend prevention and control strategies. Methods: This study was based on the Canadian 1996–1997 National Population Health Survey which comprised 69 494 participants aged 12 years and over. The prevalence of diabetes mellitus was analysed in relation to age, sex, body mass index, overweight status, energy expenditure, physical activity, smoking, drinking, income, marital status, education and rural or urban residence. Results: The prevalence of diabetes increased with age and body mass index and increased inversely with energy expenditure in both males and females. Current and former smokers were associated with a higher prevalence of diabetes. No effect was observed in regular or former drinkers. Prevalence of diabetes increased inversely with income, especially among women. Women who were single and 35 to 64 years old had a higher prevalence of diabetes than women of the same age who were married. The prevalence of diabetes was not found to be related to the level of education. Urban or rural residence was not found to have an effect on the prevalence of diabetes. Conclusion/interpretation: Women and men of all ages should avoid becoming overweight, by maintaining their body mass index below 25 kg/m2 and 27 kg/m2, respectively. They should maintain a moderate level of physical activity. Patients with diabetes should give up smoking completely. Diabetes prevention and control strategies should be targeted for women in low income groups. [Diabetologia (2001) 44: 1221–1231]

Diseases of comfort: primary cause of death in the 22nd century

Context: The world has started to feel the impact of a global chronic disease epidemic, which is putting pressure on our health care systems. If uncurbed, a new generation of “diseases of comfort” (such as those chronic diseases caused by obesity and physical inactivity) will become a major public health problem in this and the next century. Objective: To describe the concept, causes, and prevention and control strategies of diseases of comfort. Methods: Brokered by a senior research scientist specialised in knowledge translation, a chair, a president, and a past president of national public health associations contributed their views on the subject. Results: Diseases of comfort have emerged as a price of living in a modern society. It is inevitable that these diseases will become more common and more disabling if human “progress” and civilisation continue toward better (more comfortable) living, without necessarily considering their effects on health. Modern technology must be combined with education, legislation, intersectoral action, and community involvement to create built and social environments that encourage, and make easy, walking, physical activity, and nutritious food choices, to reduce the health damaging effects of modern society for all citizens and not only the few. Conclusions: Public health needs to be more passionate about the health issues caused by human progress and adopt a health promotion stance, challenging the assumptions behind the notion of social “progress” that is giving rise to the burden of chronic disease and developing the skills to create more health promoting societies in which individual health thrives.

Likelihood Approach to the First Dark Matter Results from XENON100

Many experiments that aim at the direct detection of dark matter are able to distinguish a dominant background from the expected feeble signals, based on some measured discrimination parameter. We develop a statistical model for such experiments using the profile likelihood ratio as a test statistic in a frequentist approach. We take data from calibrations as control measurements for signal and background, and the method allows the inclusion of data from Monte Carlo simulations. Systematic detector uncertainties, such as uncertainties in the energy scale, as well as astrophysical uncertainties, are included in the model. The statistical model can be used to either set an exclusion limit or to quantify a discovery claim, and the results are derived with the proper treatment of statistical and systematic uncertainties. We apply the model to the first data release of the XENON100 experiment, which allows one to extract additional information from the data, and place stronger limits on the spin-independent elastic weakly interacting massive particles nucleon scattering cross section. In particular, we derive a single limit, including all relevant systematic uncertainties, with a minimum of 2.4×10-44  cm2 for weakly interacting massive particles with a mass of 50  GeV/c2. © 2011 American Physical Society

Finding the real case-fatality rate of H5N1 avian influenza

Background: Accurate estimation of the case-fatality (CF) rate, or the proportion of cases that die, is central to pandemic planning. While estimates of CF rates for past influenza pandemics have ranged from about 0.1% (1957 and 1968 pandemics) to 2.5% (1918 pandemic), the official World Health Organization estimate for the current outbreak of H5N1 avian influenza to date is around 60%. Methods and results: The official estimate of the H5N1 CF rate has been described by some as an over-estimate, with little relevance to the rate that would be encountered under pandemic conditions. The reasons for such opinions are typically: (i) numerous undetected asymptomatic/mild cases, (ii) under-reporting of cases by some countries for economic or other reasons, and (iii) an expected decrease in virulence if and when the virus becomes widely transmitted in humans. Neither current data nor current literature, however, adequately supports these scenarios. While the real H5N1 CF rate could be lower than the current estimate of 60%, it is unlikely that it will be at the 0.1–0.4% level currently embraced by many pandemic plans. We suggest that, based on surveillance and seroprevalence studies conducted in several countries, the real H5N1 CF rate should be closer to 14–33%. Conclusions: Clearly, if such a CF rate were to be sustained in a pandemic, H5N1 would present a truly dreadful scenario. A concerted and dedicated effort by the international community to avert a pandemic through combating avian influenza in animals and humans in affected countries needs to be a global priority.

Diagnostic accuracy of cerebrospinal fluid protein markers for sporadic Creutzfeldt-Jakob disease in Canada: a 6-year prospective study

BackgroundTo better characterize the value of cerebrospinal fluid (CSF) proteins as diagnostic markers in a clinical population of subacute encephalopathy patients with relatively low prevalence of sporadic Creutzfeldt-Jakob disease (sCJD), we studied the diagnostic accuracies of several such markers (14-3-3, tau and S100B) in 1000 prospectively and sequentially recruited Canadian patients with clinically suspected sCJD.MethodsThe study included 127 patients with autopsy-confirmed sCJD (prevalence = 12.7%) and 873 with probable non-CJD diagnoses. Standard statistical measures of diagnostic accuracy were employed, including sensitivity (Se), specificity (Sp), predictive values (PVs), likelihood ratios (LRs), and Receiver Operating Characteristic (ROC) analysis.ResultsAt optimal cutoff thresholds (empirically selected for 14-3-3, assayed by immunoblot; 976 pg/mL for tau and 2.5 ng/mL for S100B, both assayed by ELISA), Se and Sp respectively were 0.88 (95% CI, 0.81-0.93) and 0.72 (0.69-0.75) for 14-3-3; 0.91 (0.84-0.95) and 0.88 (0.85-0.90) for tau; and 0.87 (0.80-0.92) and 0.87 (0.84-0.89) for S100B. The observed differences in Sp between 14-3-3 and either of the other 2 markers were statistically significant. Positive LRs were 3.1 (2.8-3.6) for 14-3-3; 7.4 (6.9-7.8) for tau; and 6.6 (6.1-7.1) for S100B. Negative LRs were 0.16 (0.10-0.26) for 14-3-3; 0.10 (0.06-0.20) for tau; and 0.15 (0.09-0.20) for S100B. Estimates of areas under ROC curves were 0.947 (0.931-0.961) for tau and 0.908 (0.888-0.926) for S100B. Use of interval LRs (iLRs) significantly enhanced accuracy for patient subsets [e.g., 41/120 (34.2%) of tested sCJD patients displayed tau levels > 10,000 pg/mL, with an iLR of 56.4 (22.8-140.0)], as did combining tau and S100B [e.g., for tau > 976 pg/mL and S100B > 2.5 ng/mL, positive LR = 18.0 (12.9-25.0) and negative LR = 0.02 (0.01-0.09)].ConclusionsCSF 14-3-3, tau and S100B proteins are useful diagnostic markers of sCJD even in a low-prevalence clinical population. CSF tau showed better overall diagnostic accuracy than 14-3-3 or S100B. Reporting of quantitative assay results and combining tau with S100B could enhance case definitions used in diagnosis and surveillance of sCJD.

Using Health Utility Index (HUI) for Measuring the Impact on Health-Related Quality of Life (HRQL) Among Individuals with Chronic Diseases

Quality of life is an important indicator in assessing the burden of disease, especially for chronic conditions. The Health Utilities Index (HUI) is a recently developed system for measuring the overall health status and health-related quality of life (HRQL) of individuals, clinical groups, and general populations. Using the HUI (constructed based on eight attributes: vision, hearing, speech, mobility, dexterity, cognition, emotion, and pain/discomfort) to measure the HRQL for chronic disease patients and to detect possible associations between HUI system and various chronic conditions, this study provides information to improve the management of chronic diseases.This study is of interest to data analysts, policy makers, and public health practitioners involved in descriptive clinical studies, clinical trials, program evaluation, population health planning, and assessments. Based on the Canadian Community Health Survey (CCHS) for 2000–01, the HUI was used to measure the quality of life for individuals living with various chronic conditions (Alzheimer/other dementia, effects of stroke, urinary incontinence, arthritis/rheumatism, bowel disorder, cataracts, back problems, stomach/intestinal ulcers, emphysema/COPD, chronic bronchitis, epilepsy, heart disease, diabetes, migraine headaches, glaucoma, asthma, fibromyalgia, cancers, high blood pressure, multiple sclerosis, thyroid condition, and other remaining chronic diseases). Logistic Regression Model was employed to estimate the associations between the overall HUI scores and various chronic conditions. The HUI scores ranged from 0.00 (corresponding to a state close to death) to 1.00 (corresponding to perfect health); negative scores reflect health states considered worse than death. The mean HUI score by sex and age group indicated the typical quality of life for persons with various chronic conditions. Logistic Regression results showed a strong relationship between low HUI scores (≤ 0.5 and 0.06–1.0) and certain chronic conditions. Age- and sex-adjusted Odds Ratio (OR) and p values showed an effect among individuals diagnosed with each chronic disease on the overall HUI score. Results of this study showed that arthritis/rheumatism, heart disease, high blood pressure, cataracts, and diabetes had a severe impact on HRQL. Urinary incontinence, Alzheimer/other dementia, effects of stroke, cancers, thyroid condition, and back problems have a moderate impact. Food allergy, allergy other than food, asthma, migraine headaches, and other remaining chronic diseases have a relatively mild effect. It is concluded that major chronic diseases with significant health burden were associated with poor HRQL. The HUI scores facilitate the measurement and interpretation of results of health burden and the HRQL for individuals with chronic diseases and can be useful for development of strategies for the prevention and control of chronic diseases.

A technique to re-assess epidemiologic evidence in light of the healthy worker effect: the case of firefighting and heart disease.

The healthy worker effect (HWE) is a bias that is believed to have strongly affected the validity of previous cohort mortality studies on the relationship between firefighting and heart disease. There is a strong healthy hired effect (a component of the HWE) among firefighters, owing particularly to the recruitment of nondiabetic candidates. This is shown in previous studies in which the reported standardized mortality ratios for diabetes are much less than unity, generally around 0.3 to 0.5. Because diabetes is known to increase the risk of heart disease, a deficit of diabetes among firefighters is expected to lead to a deficit of heart injury and disease. This would make the cohort mortality studies incapable of detecting any increase in risk of heart injury and disease among firefighters. There is also a strong healthy worker survivor effect (another component of the HWE) among firefighters. In addition, heart disease is a classic example of the HWE because heart disease is chronic and its risk factors can be identified in the selection process. It is believed that (1) a major problem of previous studies on firefighting and heart disease is their failure to recognize the importance of the HWE when interpreting their results, and (2) a technique to re-assess results in light of the HWE is urgently needed. This article addresses the generally accepted principles relating to the HWE, including its definition and sources, and proposes a technique for re-assessing the literature in light of the HWE. The technique was applied to carefully re-assess 23 studies that provided direct evidence for the relationship between firefighting and heart disease. Before the re-assessment, 7 of the 23 studies showed positive evidence and 16 showed no evidence. After the re-assessment, 11 studies showed positive evidence and 12 showed no evidence. Based on the results of the re-assessment of the 23 studies, we concluded that (1) there is strong evidence of an increased risk of death overall from heart disease among firefighters; (2) there is insufficient evidence, even after considering the HWE, that there is an increased risk of death from aortic aneurysm among firefighters; and (3) there is insufficient evidence, even after considering the HWE, for a relationship between firefighting and any heart disease subtype, such as acute myocardial infarction.

Erythema multiforme (Stevens-Johnson Syndrome) – Chart Review of 123 Hospitalized Patients

The charts of 123 patients hospitalized with erythema multiforme were reviewed. Antecedent infections and the administration of drugs were associated with 15 and 32% of the cases, respectively. Pyrexia arthralgia, photosensitivity, abnormal liver function tests and abnormal urine sediment were common findings. The length of stay in hospital was 12.6 +/- 0.7 days (mean +/- SE), time to initial improvement after onset was 10.3 +/- 0.6 days and time to complete resolution was 23.0 +/- 2.4 days. The presence of bullous skin lesions and greater overall severity were associated with longer hospitalization. Antecedent respiratory tract symptoms were related to shorter hospitalization. Steroid therapy was not associated with a shorter time for initial improvement to occur, while the average length of stay in hospital was 4.2 days longer than that for the patients not treated with corticosteroid.