Sampling methods and sensitivity analysis for large parameter sets

Models with large parameter (i.e., hundreds or thousands of parameters) often behave as if they depend upon only a few parameters, with the rest having comparatively little influence. One challenge of sensitivity analysis with such models is screening parameters to identify the influential ones, and then characterizing their influences.

Large models often require significant computing resources to evaluate their output, and so a good screening mechanism should be efficient: it should minimize the number of times a model must be exercised. This paper describes an efficient procedure to perform sensitivity analysis on deterministic models with specified ranges or probability distributions for each parameter.

It is based on repeated exercising of the model, which can be treated as a black box. Statistical checks can ensure that the screening identified parameters that account for the bulk of the model variation. Subsequent sensitivity analysis can use the screening information to reduce the investment required to characterize the influence of influential and other parameters.

The procedure exploits simplifications in the dependence of a model output on model inputs. It works best where a small number of parameters are much more influential than all the rest. The method is much more sensitive to the number of influential parameters than to the total number of parameters. It is most effective when linear or quadratic effects dominate higher order effects and complex interactions.

The paper presents a set of M athematica functions that can be used to create a variety of types of experimental designs useful for sensitivity analysis, including simple random, latin hypercube and fractional factorial sampling. Each sampling method can use discretization, folding, grouping and replication to create composite designs. These techniques have beencombined in a composite approach called Iterated Fractional Factorial Design (IFFD).

The procedure is applied to model of nuclear fuel waste disposal, and to simplified example models to demonstrate the concepts involved.     

Bayesian Interval Estimation for the Difference in TPRs and FPRs of Two Diagnostic Tests with Unverified Negatives

We derive Bayesian interval estimators for the differences in the true positive rates and false positive rates of two dichotomous diagnostic tests applied to the members of two distinct populations. The populations have varying disease prevalences with unverified negatives. We compare the performance of the Bayesian credible interval to the Wald interval using Monte Carlo simulation for a spectrum of different TPRs, FPRs, and sample sizes. For the case of a low TPR and low FPR, we found that a Bayesian credible interval with relatively noninformative priors performed well. We obtain similar interval comparison results for the cases of a high TPR and high FPR, a high TPR and low FPR, and of a high TPR and mixed FPR after incorporating mildly informative priors.     

From Street‐level Bureaucrats to Street‐level Policy Entrepreneurs? Central Policy and Local Action in Lottery‐funded Community Cancer Care

Recent contributions to the policy implementation literature have applied Kingdon's model of ‘policy windows’ to the implementation of policy on health inequalities in the UK, and have identified the key role played by ‘policy entrepreneurs’ at local as well as national level. Despite this, the picture that emerges is of frustration of central policy intentions at the local level, alongside frustration of local aspirations by the centre. This article explores the relationship between central policy and local implementation in the context of a Lottery‐funded initiative to develop community cancer care in the UK. We examine the relationships: between the BIG Lottery Fund and central government; between BIG and the cancer care projects it funded; and between the projects and their local economy of cancer care. We found evidence of success both in vertical cascading of policy and in local policy innovation; 83 per cent of projects succeeded in obtaining continuation funding at the end of their Lottery funding. We suggest that this was due, in part, to two features of Lottery funding and accountability that combined to differentiate it from the other policy initiatives studied. They meant that projects were ‘buffered’ not just from national policy churn, but also from competing local priorities. In the ‘protected space’ that was thus formed, ‘street‐level policy entrepreneurs’ played a key role in developing cancer care innovations for adoption by mainstream funding agencies.     

Gambling among young Thai people in Melbourne,Australia: an exploratory study

The purpose of the study reported in this paper was to explore the gambling behaviours of young Thai people resident in Melbourne, Australia; the meaning that gambling had for this group; whether the attitudes of these young Thai people towards gambling changed as a result of their location in a society with more relaxed attitudes towards gambling than in Thailand; whether any such changes of attitude were reflected in gambling participation; and whether such participation had demonstrably negative impacts. It was also hoped to gain some insight into whether pre-existing beliefs and attitudes provided a protective factor in relation to the respondent's gambling. Fifty young people aged 18 to 25, forming five natural friendship networks, were observed on over 180 occasions in relation to their gambling and leisure behaviours. A number of major themes were identified from the observational data including money, fortune and luck, consumerism, entertainment, spirituality and superstition, gambling and family values. It was found that despite high levels of gambling participation there was little, if any, evidence of problematic play or its consequences. Strong adherence to values espoused in families of origin and to complex beliefs about luck in the context of beliefs about spirituality and the location of gambling within the context of a range of leisure pursuits served as protective factors for these young people.     

Repression and other Forms of Cognitive Evasiveness: Implications for Stress Symptoms in College Students

Abstract

Lazarus has proposed that palliative defenses against threat that interfere with direct efforts to alleviate the threat may prove maladaptive. Three evasive cognitive attributes that circumvent awareness of threat were studied to determine whether they related to the prevalence of stress symptoms in college students. Repression (selective forgetting of threatening material), awareness of the repressive defense, and internal scanning (breadth of association to cues) were considered. College students of both sexes who engaged in evasive cognitive activity reported more stress symptoms than those who did not, with a progressive increase in stress as increasingly homogeneous groups were considered. Students who defended themselves by repression experienced more stress than those who did not. Repressors who were the least aware of their use of this defense reported even more stress. Unconscious repressors who were the least reflective (narrow scanners) showed the highest frequency of stress symptoms.     

Epidemiologic Evidence for Chloroprene Carcinogenicity: Review of Study Quality and its Application to Risk Assessment

This article evaluates the quality and weight of evidence associated with epidemiologic studies of cancer among occupational cohorts exposed to chloroprene. The focus is on liver, lung, and lymphohematopoietic cancers, which had been increased in early studies. Literature searches identified eight morbidity/mortality studies covering seven chloroprene-exposed cohorts from six countries. These studies were summarized and their quality was assessed using the 10 criteria suggested by the U.S. Environmental Protection Agency. The limitations within this literature (primarily the early studies) included crude exposure assessment, incomplete follow-up, uncertain baseline rates, and uncontrolled confounding by factors such as smoking, drinking, and co-exposure to benzene and vinyl chloride. Four cohorts were studied by the same group of investigators, who reported no overall increased associations for any cancers. This four-cohort study was by far the most rigorous, having the most comprehensive exposure assessment and follow-up and the most detailed documentation. This study also contained the two largest cohorts, including an American cohort from Louisville, Kentucky, that ranked at or near the top for each of the 10 quality criteria. There was evidence of a strong healthy worker effect in the four-cohort study, which could have hidden small excess risks. Small increased risks were suggested by internal or company-specific analyses, but these were most likely caused by uncontrolled confounding and low baseline rates. Overall, the weight of evidence does not support any substantial link between chloroprene exposure and cancer, but inconsistencies and a lack of control for major confounders preclude drawing firmer conclusions.     

Validation of an Arabic ADAM questionnaire for androgen deficiency screening in the Arab community

Background.?It is well documented that testosterone levels decline with age, this decline is associated with symptoms which could be assessed denoting androgen deficiency. We investigated the validity of an Arabic version of the Saint Louis University androgen deficiency in ageing men (ADAM) questionnaire to screen for androgen deficiency in Saudi and non Saudi Arabic speaking men.

Methods.?It was a cross sectional study of ambulatory community-based Arabic Saudi men recruited from Volunteers in Riyadh city, Capital of Saudi Arabia, aged 18–80 years. Seven hundred thirty men agreed to fill the Arabic ADAM questionnaire, they were invited to a morning blood sample for total testosterone and sex hormone binding globulin and those who agreed to complete the whole study were only 407 men. Low serum bioavailable testosterone (BT) levels (androgen deficiency) were defined as <10th percentile of serum BT levels in young healthy Saudi men (18–30 years).

Results.?Cronbach's Alpha of 0.71 (n?=?730) showed a good internal consistency of the Arabic ADAM questionnaire. Among participants, 18.2% and 77.6% had low serum BT levels and a positive ADAM questionnaire, respectively. The prevalence of positive ADAM and low serum BT is increasing with age. The Arabic ADAM questionnaire had a high sensitivity of 86.5%, a low specificity of 24.3%, and positive predictive values (+PVs) and negative (?PVs) of 20.3% and 89%, respectively.

Conclusion.?The Arabic ADAM questionnaire has a very good sensitivity but very low specificity for screening of androgen deficiency in Saudi men, therefore biological confirmation is needed especially when clinical symptoms of androgen deficiency are present.     

糖尿病、降糖治疗和癌症

2型糖尿病与癌症之间的关系日益明显,虽然其中机制尚不十分明确,但对这种情况尽可能的及早调查是很重要的,也完全可能的。从一些观察报告显示二甲双胍对癌症有保护作用,而另一方面外源性胰岛素、胰岛素模拟物以及磺脲类药物都有潜在的增加癌症发生的危险。     

The Use of Reality Orientation in Adult Foster Care Homes:

The emphasis in recent years on providing long-term care in community-based alternatives to the institutional setting can help to facilitate the development of new and exciting opportunities for enrichment in the lives of individuals whose educational needs often go unmet. Reality orientation has been demonstrated to be an effective approach to the treatment and prevention of loss of memory, confusion, and disorientation among institutionalized adults. However, certain problems can arise that may limit the potential benefits of reality orientation programs in institutions. This paper explores the rationale for considering the use of reality orientation in one alternative living arrangement, the adult foster care home. The personalized, family-like environment of the adult foster care setting might help to minimize or even eliminate many of the problems that detract from reality orientation programs in larger institutions.     

Nonspecific Urethritis: Its Current Status

Abstract

The etiology, epidemiology, clinical presentation, diagnosis, and treatment of nonspecific urethritis (N.S.U.) is discussed. Chlamydia seems to be responsible for approximately 40–50% of cases while the etiology of the remainder remains obscrue. There seems to be reasonably good evidence for venereal transmission as well as for treatment of the female partner whether symptomatic or asymptomatic.

Tetracyline given for 21 days or oxytetracycline given for four days seem to be the most effective first line treatment regimens at this time.

Complications which include prostatitis, epididymitis, urethral stricture, and Reiter's Syndrome, are discussed. It appears as though there is a reasonable possibility that Reiter's Syndrome is a separate entity (or immunological sequela) rather than a complication.