The major goal of the proposed research is the development of accurate statistical methodologies for analyzing industrial hygiene data in the presence of non-detects. Statistical procedures will be developed in a comprehensive manner for a variety of practical scenarios: univariate and bivariate normal and lognormal distributions, random e(R)ects models, and regression models. The emphasis of the proposed research is to come up with accurate pro- cedures for exposure data analysis, applicable regardless of the sample size. The problems to be addressed include the interval estimation of the arithmetic and geometric means of lognormal distributions, and of the di(R)erence and ratios of two lognormal means, the com- putation of tolerance intervals (intervals that are expected to contain a certain proportion of the exposure measurements), and the computation of upper tolerance limits. Sample size de- termination will also be addressed when detection limits are present?a problem that has not been addressed so far in the statistics literature, or in the industrial hygiene literature. All of these will be accomplished using novel methodologies that have recently become available in the statistics literature: methodologies based on the concepts of generalized con?dence intervals and generalized p-values, and modi?ed likelihood ratio procedures. A second goal of the proposed research is the development of reference limits that are suitable for mak- ing decisions concerning acceptable exposure levels in occupational health. A ?nal goal is the investigation of accuracy criteria for the quanti?cation of the measurement accuracy of industrial hygiene data. In order to facilitate the easy application of the methodologies that will be developed, computational algorithms and software codes will be provided. The proposed work will be carried out in collaboration with an industrial hygienist. Further- more, the proposed research will be disseminated among the industrial hygiene community through professional development courses, webinars, professional publications and conference presentations.

Public Health Relevance

The proposed research aims to develop a comprehensive set of statistical methodologies for analyzing industrial hygiene data when samples include values below a detection limit. Reference limits will be developed, suitable for making decisions concerning acceptable exposure levels in occupational health. Finally, accuracy criteria will be investigated for the quanti?cation of the measurement accuracy of industrial hygiene data.

Agency
National Institute of Health (NIH)
Institute
National Institute for Occupational Safety and Health (NIOSH)
Type
Research Project (R01)
Project #
5R01OH003628-09
Application #
8300689
Study Section
Safety and Occupational Health Study Section (SOH)
Program Officer
Frederick, Linda J
Project Start
2010-09-01
Project End
2014-08-31
Budget Start
2012-09-01
Budget End
2013-08-31
Support Year
9
Fiscal Year
2012
Total Cost
$365,488
Indirect Cost
$80,362
Name
University of Maryland Balt CO Campus
Department
Biostatistics & Other Math Sci
Type
Schools of Arts and Sciences
DUNS #
061364808
City
Baltimore
State
MD
Country
United States
Zip Code
21250
Krishnamoorthy, K; Mathew, Thomas; Xu, Zhao (2014) Comparison of means of two lognormal distributions based on samples with multiple detection limits. J Occup Environ Hyg 11:538-46
Krishnamoorthy, Kalimuthu; Mathew, Thomas; Xu, Zhao (2013) Tests for an upper percentile of a lognormal distribution based on samples with multiple detection limits and sample-size calculation. Ann Occup Hyg 57:1200-12
Krishnamoorthy, Kalimuthu; Mathew, Thomas (2013) The symmetric-range accuracy under a one-way random model with balanced or unbalanced data. Ann Occup Hyg 57:953-61
Krishnamoorthy, Kalimuthu; Xu, Zhao (2011) Confidence limits for lognormal percentiles and for lognormal mean based on samples with multiple detection limits. Ann Occup Hyg 55:495-509
Bebu, Ionut; Mathew, Thomas (2008) Comparing the means and variances of a bivariate log-normal distribution. Stat Med 27:2684-96