|ognizant Communication Corporation|
A Journal of Science Serving Legislative, Regulatory, and Judicial Systems
Human Advancement · Environmental Protection · Industrial Development
Volume 9, Numbers 1-2
Technology, Vol. 9, pp. 3-14
1072-9240/02 $20.00 + 00.
Copyright © 2002 Cognizant Communication Corp.
Printed in the USA. All rights reserved.
Clues to the Possible Viral Etiology of Childhood Leukemia
Gabor Mezei and Leeka I. Kheifets
Electric Power Research Institute, Palo Alto, CA
This paper discusses possible etiologic factors, with a focus on clues to a possible viral etiology. It gives a description of the epidemiology of childhood leukemia; summarizes genetic and environmental risk factors; and weighs the evidence regarding a viral etiology. Special consideration is given to the two main hypotheses in the development of childhood leukemia proposed by Greaves and Kinlen. The viral etiology hypothesis is a popular one supported by a growing body of evidence. Conclusive evidence, however, remains elusive
Key words: Childhood leukemia; Viral etiology
An Outbreak of Leptospirosis in North Queensland, Australia: January to May, 1999
Lee Smythe, Leonie Barnett, Meegan Symonds, and Michael Dohnt
WHO/FAO Collaborating Center for Reference & Research on Leptospirosis, Queensland Health Scientific Services, Brisbane, Queensland, Australia
Peter Baade and Christine McClintock
Health Information Center, Queensland Health, Brisbane, Queensland, Australia
Michael R. Moore
National Research Center for Environmental Toxicology, University of Queensland, Brisbane, Australia
The World Health Organization/Food and Agriculture Organization Collaborating Center for Reference and Research on Leptospirosis, Western Pacific Region, is part of Queensland Health Scientific Services, Queensland Health. In 1999 there was a marked increase in human leptospirosis notifications in North Queensland, Australia, over the period of January to May. There were 216 notifications for the state of Queensland in 1999, compared with 108 in 1998 and 64 in 1997. By April 1999, the Reference Laboratory and Queensland Health's Tropical Public Health Unit (TPHU), Cairns, had reported an outbreak of leptospirosis in North Queensland. Notifications reported for the North Queensland Health districts of Cairns, Tablelands and Innisfail were the highest, with the majority involving workers in the banana industry. The outbreak resulted from infections with a number of serovars, mostly attributed to serovars australis and zanoni. Sixty leptosprial isolates were obtained from human cases in North Queensland during the outbreak. The hospitalization rate was in excess of 50%. The Queensland Health's TPHU has implemented a program to educate target industries and the community on the risks and prevention strategies associated with leptospirosis.
Key words: Leptospirosis; Outbreak; Banana industry; Hospitalization;
Occupational risk; Queensland health districts; Rainfall; Notifications;
Empirical Causation and Biases in Epidemiology: Issues and Solutions
Paolo F. Ricci
University of San Francisco, San Francisco, CA
L.A. Cox, Jr.
Cox Associates, Denver, CO
This paper reviews the epidemiological literature and suggests practical ways to reach valid conclusions about empirical causation in epidemiology. It deals with a form of causation that is based on statistical and probabilistic models that relate risk factors to the probability of adverse response. Specifically, it discusses issues that relate to statistical models used in epidemiology and study the principal issues affecting them, including: 1. Model-form selection; 2. Variable selection; 3. Heterogeneity; 4. Variable coding; 5. Errors in the independent variables; 6. Selection and information biases (diagnostic suspicion bias and recall bias); and 7. Measurement and classification errors. Each of these is discussed and often illustrated with suggestions for remedies and alternatives. Establishing defensible empirical causation requires using a portfolio of statistical methods because biases and other issues crossover and can often arise in a single study. Some of these methods can be computationally demanding. Nonetheless, if epidemiological results are used to justify policy, the analytical rigor required to support causal assertions cannot be dismissed on practical or other ad hoc grounds. Epidemiological investigations are of critical importance to determine both how and why exposure to microbial, chemical or viral agents causes adverse response to those at risk. Epidemiological results used in public health and health policy can lead to either over or under protection; however, the actual level of protection can remain unknown. The appearance of accuracy can be misleading and result in the misallocation of scarce resources that can be unjust and place undue burdens on society.
Key words: Epidemiology; Analytical epidemiology; Causation; Estimation; Biases; Remedies
A Simulation Model of Human Health Risks From Chicken-Borne Campylobacter Jejuni
Louis Anthony Cox, Jr. and Douglas Popken
Cox Associates, Denver, CO
Quantitative risk assessment (QRA) of microbial hazards poses challenges not encountered in chemical or radiation QRAs. The ability of microorganisms to reproduce under favorable circumstances, even after partial decontamination, invalidates the mass-balance calculations often used for chemical exposure and QRA calculations. Evolution of resistant strains under selection pressures from antibiotics and spread of infections from one victim to another have no parallels in the world of chemical risk assessment. Also dose-response relations for microorganisms are often unusual, with smaller doses of bacterial pathogens sometimes creating a higher probability of illness than larger doses (e.g., Teunis et al., 1999). This paper introduces a discrete-event simulation model for combining available information from many sources. The simulation can also incorporate assumptions driven by the current concerns and needs of risk managers, thus supporting assumption-based risk analysis for hypothesized risks. The resulting model can be used to identify effective risk-management strategies and to eliminate ineffective ones, even in complex situations with high uncertainty and variability in individual exposures and dose-response relations. This paper examines in detail how a discrete-event simulation model of human exposures to a common food-borne pathogen, Campylobacter jejuni, can be used to estimate the probable human health impacts of alternative risk management strategies. The simulation model and its uncertainty analysis reveal large, robust differences in expected health impacts between the most and least effective risk management strategies. Strategies aimed at reducing microbial load during processing and at the point of food consumption (e.g., restaurants) are most likely to create public health benefits.
Key words: Human health risk assessment; Antimicrobial; Fluoroquinolone; Enrofloxacin; Simulation model; Campylobacter; Chicken
Detecting and Explaining Rare Event-Clusters in Sparse Data Sets: Some Initial Results
Paolo F. Ricci
University of San Francisco, San Francisco, CA
Applied Biomathematics, Setauke, NY
Traditional statistical tests used to detect clustering, such as the chi-square (C2), can be either inappropriate or insufficient in working with sparse data and rare events because these tests are based on asymptotic assumptions. Using these methods, in those conditions, can either underestimate or overestimate the evidence for clustering. Moreover, those tests can be coarse and thus miss complex clustering patterns. Knowledge of such patterns can be quite useful in understanding the cause or causes of the observed adverse outcomes. This paper describes and applies seven alternative exact statistical tests that employ combinatorial formulations and are therefore exact under the null-hypothesis and thus are reliable, regardless of the size of the sample. These tests are especially useful when data sets are small or have different structures than the traditional rows by columns layout. The paper empirically describes the performance of those combinatorial tests, using random samples, relative to the chi-square and likelihood ratio tests, as well as account for the effect of multiple comparisons. The results indicate that exact tests are sensitive to different cluster's morphology, are useful in detecting clusters when traditional tests fail and are informative about the processes generating the clusters. This information can be critical in understanding and possibly preventing communicable and other diseases. In particular, these tests can help allay public concerns about the cause of the cluster by providing additional information.
Key words: Epidemiology; Cluster analysis; Structure of clusters; Exact statistical methods; Comparison of results
Low Cost Technology for Reliable Use of Mexico City's Wastewater for Agricultural Irrigation
Blanca Jiménez and Alma Chávez
Instituto de Ingenieria - UNAM, Grupo: Tratamiento y Reúso, Coyoacán, Mexico
Mexico City's metropolitan area generates 74.5 m3/s of wastewater containing both sewage and rainwater. Also 90% of this untreated wastewater is sent to a nearby valley where it is used for agricultural irrigation. Because of the increase in productivity associated with this practice, farmers have accepted the use of untreated water. It is therefore necessary to find a means of treatment which will manage to control associated health problems while maintaining the content of organic material - nitrogen and phosphorus. This new approach has led to the development of a treatment which is based on using a physicochemical process (advanced primary) together with filtration and a final disinfection stage. With the method, Helminth ova are reduced to less than 1 HO/L. The removal of total suspended solids is up to 92.3% (from 359 mg/L to 27.8 mg/L); 15.7% of toal Kjeldahl nitrogen is removed (from 20 mg/L to 16.8 mg/L); and also 71.6% of phosphorus (from 9.5 mg/L to 2.7 mg/L). In order to reduce the content of fecal coliforms to less than 1000 most probable number/100mL it was necessary to add 12 mg/L of chlorine for 3h. Stabilization of sludge was carried out by adding 20% CaO, such that class "B" biosolids were obtained in accordance with the Environmental Protection Agency standards.
Key words: Agricultural reuse; Advanced primary treatment; Helminth ova; High rate sedimentation