An ordinal logistic regression model with misclassification of the outcome variable and categorical covariate

Ordinal outcomes are frequently employed in diagnosis and clinical trials. Clinical trials of Alzheimer's disease (AD) treatments are a case in point using the status of mild, moderate or severe disease as outcome measures. As in many other outcome oriented studies, the disease status may be misclassified. This study estimates the extent of misclassification in an ordinal outcome such as disease status. Also, this study estimates the extent of misclassification of a predictor variable such as genotype status. An ordinal logistic regression model is commonly used to model the relationship between disease status, the effect of treatment, and other predictive factors. A simulation study was done. First, data based on a set of hypothetical parameters and hypothetical rates of misclassification was created. Next, the maximum likelihood method was employed to generate likelihood equations accounting for misclassification. The Nelder-Mead Simplex method was used to solve for the misclassification and model parameters. Finally, this method was applied to an AD dataset to detect the amount of misclassification present. The estimates of the ordinal regression model parameters were close to the hypothetical parameters. β1 was hypothesized at 0.50 and the mean estimate was 0.488, β2 was hypothesized at 0.04 and the mean of the estimates was 0.04. Although the estimates for the rates of misclassification of X1 were not as close as β1 and β2, they validate this method. X 1 0-1 misclassification was hypothesized as 2.98% and the mean of the simulated estimates was 1.54% and, in the best case, the misclassification of k from high to medium was hypothesized at 4.87% and had a sample mean of 3.62%. In the AD dataset, the estimate for the odds ratio of X 1 of having both copies of the APOE 4 allele changed from an estimate of 1.377 to an estimate 1.418, demonstrating that the estimates of the odds ratio changed when the analysis includes adjustment for misclassification. ^

Assessment of the effect on statistical power of regression model misspecification by using techniques of mathematical statistics and simulation study

Objectives. This paper seeks to assess the effect on statistical power of regression model misspecification in a variety of situations. ^ Methods and results. The effect of misspecification in regression can be approximated by evaluating the correlation between the correct specification and the misspecification of the outcome variable (Harris 2010).In this paper, three misspecified models (linear, categorical and fractional polynomial) were considered. In the first section, the mathematical method of calculating the correlation between correct and misspecified models with simple mathematical forms was derived and demonstrated. In the second section, data from the National Health and Nutrition Examination Survey (NHANES 2007-2008) were used to examine such correlations. Our study shows that comparing to linear or categorical models, the fractional polynomial models, with the higher correlations, provided a better approximation of the true relationship, which was illustrated by LOESS regression. In the third section, we present the results of simulation studies that demonstrate overall misspecification in regression can produce marked decreases in power with small sample sizes. However, the categorical model had greatest power, ranging from 0.877 to 0.936 depending on sample size and outcome variable used. The power of fractional polynomial model was close to that of linear model, which ranged from 0.69 to 0.83, and appeared to be affected by the increased degrees of freedom of this model.^ Conclusion. Correlations between alternative model specifications can be used to provide a good approximation of the effect on statistical power of misspecification when the sample size is large. When model specifications have known simple mathematical forms, such correlations can be calculated mathematically. Actual public health data from NHANES 2007-2008 were used as examples to demonstrate the situations with unknown or complex correct model specification. Simulation of power for misspecified models confirmed the results based on correlation methods but also illustrated the effect of model degrees of freedom on power.^

Effects of informative censoring on the hazard function: Application to the proportional hazards regression model

The standard analyses of survival data involve the assumption that survival and censoring are independent. When censoring and survival are related, the phenomenon is known as informative censoring. This paper examines the effects of an informative censoring assumption on the hazard function and the estimated hazard ratio provided by the Cox model.^ The limiting factor in all analyses of informative censoring is the problem of non-identifiability. Non-identifiability implies that it is impossible to distinguish a situation in which censoring and death are independent from one in which there is dependence. However, it is possible that informative censoring occurs. Examination of the literature indicates how others have approached the problem and covers the relevant theoretical background.^ Three models are examined in detail. The first model uses conditionally independent marginal hazards to obtain the unconditional survival function and hazards. The second model is based on the Gumbel Type A method for combining independent marginal distributions into bivariate distributions using a dependency parameter. Finally, a formulation based on a compartmental model is presented and its results described. For the latter two approaches, the resulting hazard is used in the Cox model in a simulation study.^ The unconditional survival distribution formed from the first model involves dependency, but the crude hazard resulting from this unconditional distribution is identical to the marginal hazard, and inferences based on the hazard are valid. The hazard ratios formed from two distributions following the Gumbel Type A model are biased by a factor dependent on the amount of censoring in the two populations and the strength of the dependency of death and censoring in the two populations. The Cox model estimates this biased hazard ratio. In general, the hazard resulting from the compartmental model is not constant, even if the individual marginal hazards are constant, unless censoring is non-informative. The hazard ratio tends to a specific limit.^ Methods of evaluating situations in which informative censoring is present are described, and the relative utility of the three models examined is discussed. ^

The number and timing of time -dependent covariate measurements for the Cox regression model

The problem of analyzing data with updated measurements in the time-dependent proportional hazards model arises frequently in practice. One available option is to reduce the number of intervals (or updated measurements) to be included in the Cox regression model. We empirically investigated the bias of the estimator of the time-dependent covariate while varying the effect of failure rate, sample size, true values of the parameters and the number of intervals. We also evaluated how often a time-dependent covariate needs to be collected and assessed the effect of sample size and failure rate on the power of testing a time-dependent effect.^ A time-dependent proportional hazards model with two binary covariates was considered. The time axis was partitioned into k intervals. The baseline hazard was assumed to be 1 so that the failure times were exponentially distributed in the ith interval. A type II censoring model was adopted to characterize the failure rate. The factors of interest were sample size (500, 1000), type II censoring with failure rates of 0.05, 0.10, and 0.20, and three values for each of the non-time-dependent and time-dependent covariates (1/4,1/2,3/4).^ The mean of the bias of the estimator of the coefficient of the time-dependent covariate decreased as sample size and number of intervals increased whereas the mean of the bias increased as failure rate and true values of the covariates increased. The mean of the bias of the estimator of the coefficient was smallest when all of the updated measurements were used in the model compared with two models that used selected measurements of the time-dependent covariate. For the model that included all the measurements, the coverage rates of the estimator of the coefficient of the time-dependent covariate was in most cases 90% or more except when the failure rate was high (0.20). The power associated with testing a time-dependent effect was highest when all of the measurements of the time-dependent covariate were used. An example from the Systolic Hypertension in the Elderly Program Cooperative Research Group is presented. ^

Burden of invasive pneumococcal disease in Harris County, Texas (2003--2010) and the implication for enhanced public health surveillance of isolates

Invasive pneumococcal disease (IPD) causes significant health burden in the US, is responsible for the majority of bacterial meningitis, and causes more deaths than any other vaccine preventable bacterial disease in the US. The estimated National IPD rate is 14.3 cases per 100,000 population with a case-fatality rate of 1.5 cases per 100,000 population. Although cases of IPD are routinely reported to the local health department in Harris County Texas, the incidence (IR) and case-fatality (CFR) rates have not been reported. Additionally, it is important to know which serotypes of S. pneumoniae are circulating in Harris County Texas and to determine if ‘replacement disease’ is occurring. ^ This study reported incidence and case-fatality rates from 2003 to 2009, and described the trends in IPD, including the IPD serotypes circulating in Harris County Texas during the study period, particularly in 2008 and 2010. Annual incidence rates were calculated and reported for 2003 to 2009, using complete surveillance-year data. ^ Geographic information system (GIS) software was used to create a series of maps of the data reported during the study period. Cluster and outlier analysis and hot spot analysis were conducted using both case counts by census tract and disease rate by census tract. ^ IPD age- and race-adjusted IR for Harris County Texas and their 95% confidence intervals (CIs) were 1.40 (95% CI 1.0, 1.8), 1.71 (95% CI 1.24, 2.17), 3.13 (95% CI 2.48, 3.78), 3.08 (95% CI 2.43, 3.74), 5.61 (95% CI 4.79, 6.43), 8.11 (95% CI 7.11, 9.1), and 7.65 (95% CI 6.69, 8.61) for the years 2003 to 2009, respectively (rates were age- and race-adjusted to each year's midyear US population estimates). A Poisson regression model demonstrated a statistically significant increasing trend of about 32 percent per year in the IPD rates over the course of the study period. IPD age- and race-adjusted case-fatality rates (CFR) for Harris County Texas were also calculated and reported. A Poisson regression model demonstrated a statistically significant increasing trend of about 26 percent per year in the IPD case-fatality rates from 2003 through 2009. A logistic regression model associated the risk of dying from IPD to alcohol abuse (OR 4.69, 95% CI 2.57, 8.56) and to meningitis (OR 2.42, 95% CI 1.46, 4.03). ^ The prevalence of non-vaccine serotypes (NVT) among IPD cases with serotyped isolates was 98.2 percent. In 2008, the year with the sample more geographically representative of all areas of Harris County Texas, the prevalence was 96 percent. Given these findings, it is reasonable to conclude that ‘replacement disease’ is occurring in Harris County Texas, meaning that, the majority of IPD is caused by serotypes not included in the PCV7 vaccine. Also in conclusion, IPD rates increased during the study period in Harris County Texas.^

Hazardous air pollutants and childhood lymphohematopoietic cancer in Southeast Texas, 1995--2004

Southeast Texas, including Houston, has a large presence of industrial facilities and has been documented to have poorer air quality and significantly higher cancer rates than the remainder of Texas. Given citizens’ concerns in this 4th largest city in the U.S., Mayor Bill White recently partnered with the UT School of Public Health to determine methods to evaluate the health risks of hazardous air pollutants (HAPs). Sexton et al. (2007) published a report that strongly encouraged analytic studies linking these pollutants with health outcomes. In response, we set out to complete the following aims: 1. determine the optimal exposure assessment strategy to assess the association between childhood cancer rates and increased ambient levels of benzene and 1,3-butadiene (in an ecologic setting) and 2. evaluate whether census tracts with the highest levels of benzene or 1,3-butadiene have higher incidence of childhood lymphohematopoietic cancer compared with census tracts with the lowest levels of benzene or 1,3-butadiene, using Poisson regression. The first aim was achieved by evaluating the usefulness of four data sources: geographic information systems (GIS) to identify proximity to point sources of industrial air pollution, industrial emission data from the U.S. EPA’s Toxic Release Inventory (TRI), routine monitoring data from the U.S. EPA Air Quality System (AQS) from 1999-2000 and modeled ambient air levels from the U.S. EPA’s 1999 National Air Toxic Assessment Project (NATA) ASPEN model. Further, once these four data sources were evaluated, we narrowed them down to two: the routine monitoring data from the AQS for the years 1998-2000 and the 1999 U.S. EPA NATA ASPEN modeled data. We applied kriging (spatial interpolation) methodology to the monitoring data and compared the kriged values to the ASPEN modeled data. Our results indicated poor agreement between the two methods. Relative to the U.S. EPA ASPEN modeled estimates, relying on kriging to classify census tracts into exposure groups would have caused a great deal of misclassification. To address the second aim, we additionally obtained childhood lymphohematopoietic cancer data for 1995-2004 from the Texas Cancer Registry. The U.S. EPA ASPEN modeled data were used to estimate ambient levels of benzene and 1,3-butadiene in separate Poisson regression analyses. All data were analyzed at the census tract level. We found that census tracts with the highest benzene levels had elevated rates of all leukemia (rate ratio (RR) = 1.37; 95% confidence interval (CI), 1.05-1.78). Among census tracts with the highest 1,3-butadiene levels, we observed RRs of 1.40 (95% CI, 1.07-1.81) for all leukemia. We detected no associations between benzene or 1,3-butadiene levels and childhood lymphoma incidence. This study is the first to examine this association in Harris and surrounding counties in Texas and is among the first to correlate monitored levels of HAPs with childhood lymphohematopoietic cancer incidence, evaluating several analytic methods in an effort to determine the most appropriate approach to test this association. Despite recognized weakness of ecologic analyses, our analysis suggests an association between childhood leukemia and hazardous air pollution.^

A community-based assessment of the spatiotemporal distribution of influenza in a U.S.-Mexico border county during the spring 2009 novel H1N1 swine-origin influenza A pandemic

This study retrospectively evaluated the spatial and temporal disease patterns associated with influenza-like illness (ILI), positive rapid influenza antigen detection tests (RIDT), and confirmed H1N1 S-OIV cases reported to the Cameron County Department of Health and Human Services between April 26 and May 13, 2009 using the space-time permutation scan statistic software SaTScan in conjunction with geographical information system (GIS) software ArcGIS 9.3. The rate and age-adjusted relative risk of each influenza measure was calculated and a cluster analysis was conducted to determine the geographic regions with statistically higher incidence of disease. A Poisson distribution model was developed to identify the effect that socioeconomic status, population density, and certain population attributes of a census block-group had on that area's frequency of S-OIV confirmed cases over the entire outbreak. Predominant among the spatiotemporal analyses of ILI, RIDT and S-OIV cases in Cameron County is the consistent pattern of a high concentration of cases along the southern border with Mexico. These findings in conjunction with the slight northward space-time shifts of ILI and RIDT cluster centers highlight the southern border as the primary site for public health interventions. Finally, the community-based multiple regression model revealed that three factors—percentage of the population under age 15, average household size, and the number of high school graduates over age 25—were significantly associated with laboratory-confirmed S-OIV in the Lower Rio Grande Valley. Together, these findings underscore the need for community-based surveillance, improve our understanding of the distribution of the burden of influenza within the community, and have implications for vaccination and community outreach initiatives.^

Identfying predictors of St. Louis encephalitis activity in Houston Texas

Generalized linear Poisson and logistic regression models were utilized to examine the relationship between temperature and precipitation and cases of Saint Louis encephalitis virus spread in the Houston metropolitan area. The models were investigated with and without repeated measures, with a first order autoregressive (AR1) correlation structure used for the repeated measures model. The two types of Poisson regression models, with and without correlation structure, showed that a unit increase in temperature measured in degrees Fahrenheit increases the occurrence of the virus 1.7 times and a unit increase in precipitation measured in inches increases the occurrence of the virus 1.5 times. Logistic regression did not show these covariates to be significant as predictors for encephalitis activity in Houston for either correlation structure. This discrepancy for the logistic model could be attributed to the small data set.^ Keywords: Saint Louis Encephalitis; Generalized Linear Model; Poisson; Logistic; First Order Autoregressive; Temperature; Precipitation. ^

Selection bias and the cross-validation of regression models for prediction

Strategies are compared for the development of a linear regression model with stochastic (multivariate normal) regressor variables and the subsequent assessment of its predictive ability. Bias and mean squared error of four estimators of predictive performance are evaluated in simulated samples of 32 population correlation matrices. Models including all of the available predictors are compared with those obtained using selected subsets. The subset selection procedures investigated include two stopping rules, C$\sb{\rm p}$ and S$\sb{\rm p}$, each combined with an 'all possible subsets' or 'forward selection' of variables. The estimators of performance utilized include parametric (MSEP$\sb{\rm m}$) and non-parametric (PRESS) assessments in the entire sample, and two data splitting estimates restricted to a random or balanced (Snee's DUPLEX) 'validation' half sample. The simulations were performed as a designed experiment, with population correlation matrices representing a broad range of data structures.^ The techniques examined for subset selection do not generally result in improved predictions relative to the full model. Approaches using 'forward selection' result in slightly smaller prediction errors and less biased estimators of predictive accuracy than 'all possible subsets' approaches but no differences are detected between the performances of C$\sb{\rm p}$ and S$\sb{\rm p}$. In every case, prediction errors of models obtained by subset selection in either of the half splits exceed those obtained using all predictors and the entire sample.^ Only the random split estimator is conditionally (on $\\beta$) unbiased, however MSEP$\sb{\rm m}$ is unbiased on average and PRESS is nearly so in unselected (fixed form) models. When subset selection techniques are used, MSEP$\sb{\rm m}$ and PRESS always underestimate prediction errors, by as much as 27 percent (on average) in small samples. Despite their bias, the mean squared errors (MSE) of these estimators are at least 30 percent less than that of the unbiased random split estimator. The DUPLEX split estimator suffers from large MSE as well as bias, and seems of little value within the context of stochastic regressor variables.^ To maximize predictive accuracy while retaining a reliable estimate of that accuracy, it is recommended that the entire sample be used for model development, and a leave-one-out statistic (e.g. PRESS) be used for assessment. ^

Information bias and lifetime mortality risks of radiation-induced cancer

Additive and multiplicative models of relative risk were used to measure the effect of cancer misclassification and DS86 random errors on lifetime risk projections in the Life Span Study (LSS) of Hiroshima and Nagasaki atomic bomb survivors. The true number of cancer deaths in each stratum of the cancer mortality cross-classification was estimated using sufficient statistics from the EM algorithm. Average survivor doses in the strata were corrected for DS86 random error ($\sigma$ = 0.45) by use of reduction factors. Poisson regression was used to model the corrected and uncorrected mortality rates with covariates for age at-time-of-bombing, age at-time-of-death and gender. Excess risks were in good agreement with risks in RERF Report 11 (Part 2) and the BEIR-V report. Bias due to DS86 random error typically ranged from $-$15% to $-$30% for both sexes, and all sites and models. The total bias, including diagnostic misclassification, of excess risk of nonleukemia for exposure to 1 Sv from age 18 to 65 under the non-constant relative projection model was $-$37.1% for males and $-$23.3% for females. Total excess risks of leukemia under the relative projection model were biased $-$27.1% for males and $-$43.4% for females. Thus, nonleukemia risks for 1 Sv from ages 18 to 85 (DRREF = 2) increased from 1.91%/Sv to 2.68%/Sv among males and from 3.23%/Sv to 4.02%/Sv among females. Leukemia excess risks increased from 0.87%/Sv to 1.10%/Sv among males and from 0.73%/Sv to 1.04%/Sv among females. Bias was dependent on the gender, site, correction method, exposure profile and projection model considered. Future studies that use LSS data for U.S. nuclear workers may be downwardly biased if lifetime risk projections are not adjusted for random and systematic errors. (Supported by U.S. NRC Grant NRC-04-091-02.) ^

REGRESSION ANALYSIS FOR STANDARDIZED MORTALITY RATIOS WITH APPLICATION TO OCCUPATIONAL FOLLOW-UP STUDIES

Traditional comparison of standardized mortality ratios (SMRs) can be misleading if the age-specific mortality ratios are not homogeneous. For this reason, a regression model has been developed which incorporates the mortality ratio as a function of age. This model is then applied to mortality data from an occupational cohort study. The nature of the occupational data necessitates the investigation of mortality ratios which increase with age. These occupational data are used primarily to illustrate and develop the statistical methodology.^ The age-specific mortality ratio (MR) for the covariates of interest can be written as MR(,ij...m) = ((mu)(,ij...m)/(theta)(,ij...m)) = r(.)exp (Z('')(,ij...m)(beta)) where (mu)(,ij...m) and (theta)(,ij...m) denote the force of mortality in the study and chosen standard populations in the ij...m('th) stratum, respectively, r is the intercept, Z(,ij...m) is the vector of covariables associated with the i('th) age interval, and (beta) is a vector of regression coefficients associated with these covariables. A Newton-Raphson iterative procedure has been used for determining the maximum likelihood estimates of the regression coefficients.^ This model provides a statistical method for a logical and easily interpretable explanation of an occupational cohort mortality experience. Since it gives a reasonable fit to the mortality data, it can also be concluded that the model is fairly realistic. The traditional statistical method for the analysis of occupational cohort mortality data is to present a summary index such as the SMR under the assumption of constant (homogeneous) age-specific mortality ratios. Since the mortality ratios for occupational groups usually increase with age, the homogeneity assumption of the age-specific mortality ratios is often untenable. The traditional method of comparing SMRs under the homogeneity assumption is a special case of this model, without age as a covariate.^ This model also provides a statistical technique to evaluate the relative risk between two SMRs or a dose-response relationship among several SMRs. The model presented has application in the medical, demographic and epidemiologic areas. The methods developed in this thesis are suitable for future analyses of mortality or morbidity data when the age-specific mortality/morbidity experience is a function of age or when there is an interaction effect between confounding variables needs to be evaluated. ^

Slope estimation in a parametric measurement error model: A simulation study

In regression analysis, covariate measurement error occurs in many applications. The error-prone covariates are often referred to as latent variables. In this proposed study, we extended the study of Chan et al. (2008) on recovering latent slope in a simple regression model to that in a multiple regression model. We presented an approach that applied the Monte Carlo method in the Bayesian framework to the parametric regression model with the measurement error in an explanatory variable. The proposed estimator applied the conditional expectation of latent slope given the observed outcome and surrogate variables in the multiple regression models. A simulation study was presented showing that the method produces estimator that is efficient in the multiple regression model, especially when the measurement error variance of surrogate variable is large.^

Logistic regression with Markov chains as covariates

The tobacco-specific nitrosamine 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) is an obvious carcinogen for lung cancer. Since CBMN (Cytokinesis-blocked micronucleus) has been found to be extremely sensitive to NNK-induced genetic damage, it is a potential important factor to predict the lung cancer risk. However, the association between lung cancer and NNK-induced genetic damage measured by CBMN assay has not been rigorously examined. ^ This research develops a methodology to model the chromosomal changes under NNK-induced genetic damage in a logistic regression framework in order to predict the occurrence of lung cancer. Since these chromosomal changes were usually not observed very long due to laboratory cost and time, a resampling technique was applied to generate the Markov chain of the normal and the damaged cell for each individual. A joint likelihood between the resampled Markov chains and the logistic regression model including transition probabilities of this chain as covariates was established. The Maximum likelihood estimation was applied to carry on the statistical test for comparison. The ability of this approach to increase discriminating power to predict lung cancer was compared to a baseline "non-genetic" model. ^ Our method offered an option to understand the association between the dynamic cell information and lung cancer. Our study indicated the extent of DNA damage/non-damage using the CBMN assay provides critical information that impacts public health studies of lung cancer risk. This novel statistical method could simultaneously estimate the process of DNA damage/non-damage and its relationship with lung cancer for each individual.^

Indigent care and infant mortality in Texas

Infant Mortality has been made a point of emphasis by the Department of Health and Human Services in the Healthy People 2000, 2010, and 2020 priorities. This study used the Behavioral Model for Vulnerable Populations to consider a number of factors which impact infant mortality in the indigent populations in the State of Texas. The primary focus of this study was the enabling factor of community resources, specifically the program used by each county to provide care to their indigent population. The Legislature of the State of Texas requires that each state have a program set up within a Hospital District, Public Hospital or develop a County Indigent Health Care Program (CIHCP) in order to provide the basic health care needs of their most vulnerable residents. We sought to determine whether the development of a CICHP without an appointed hospital to provide the care would have an adverse effect on residents seeking care and increase infant mortality. A Poisson Regression Analysis was used to analyze incidence rate ratios adjusting for race/ethnicity and wealth/poverty variables. Our study showed that counties using a CIHCP had significantly lower infant mortality rates when compared to counties using a hospital district and were statistically equivalent to counties using a public hospital program or a combination of service programs. This relationship was maintained when adjusted incidence rate ratios were calculated. This may give evidence that counties struggling to fund a public hospital or hospital district may be able to find a more cost-effective alternative in the CIHCP without adversely affecting the health status of their residents. More cost-benefit analysis and controlling analysis must be done to further characterize this relationship. ^