Conditional logistic analysis of case-control studies with complex sampling

Methods for the analysis of unmatched case-control data based on a finite population sampling model are developed. Under this model, and the prospective logistic model for disease probabilities, a likelihood for case-control data that accommodates very general sampling of controls is derived. This likelihood has the form of a weighted conditional logistic likelihood. The flexibility of the methods is illustrated by providing a number of control sampling designs and a general scheme for their analyses. These include frequency matching, counter-matching, case-base, randomized recruitment, and quota sampling. A study of risk factors for childhood asthma illustrates an application of the counter-matching design. Some asymptotic efficiency results are presented and computational methods discussed. Further, it is shown that a 'marginal' likelihood provides a link to unconditional logistic methods. The methods are examined in a simulation study that compares frequency and counter-matching using conditional and unconditional logistic analyses and indicate that the conditional logistic likelihood has superior efficiency. Extensions that accommodate sampling of cases and multistage designs are presented. Finally, we compare the analysis methods presented here to other approaches, compare counter-matching and two-stage designs, and suggest areas for further research.To whom correspondence should be addressed.     

The design and analysis of case-control studies with biased sampling

A design is proposed for case-control studies in which selection of subjects for full variable ascertainment is based jointly on disease status and on easily obtained "screening" variables that may be related to the disease. Recruitment of subjects follows an independent Bernoulli sampling scheme, with recruitment probabilities set by the investigator in advance. In particular, the sampling can be set up to achieve, on average, frequency matching, provided prior estimates of the disease rates or odds ratios associated with screening variables such as age and sex are available. Alternatively--for example, when studying a rare exposure--one can enrich the sample with certain categories of subject. Following such a design, there are two valid approaches to logistic regression analysis, both of which allow for efficient estimation of effects associated with the screening variables that were allowed to bias the recruitment. The statistical properties of the estimators are compared, both for large samples, based on asymptotics, and for small samples, based on simulations.     

Evaluating the use of friend or family controls in epidemiologic case-control studies

BackgroundTraditional methodologies for identifying and recruiting controls in epidemiologic case-control studies, such as random digit dialing or neighborhood walk, suffer from declining response rates. Here, we revisit the feasibility and comparability of using alternative sources of controls, specifically friend and family controls.MethodsWe recruited from a recently completed case-control study of non-Hodgkin lymphoma (NHL) among women in Los Angeles County where controls from the parent study were ascertained by neighborhood walk. We calculated participation rates and compared questionnaire responses between the friend/family controls and the original matched controls from the parent study.ResultsOf the 182 NHL case patients contacted, 111 (61%) agreed to participate in our feasibility study. 70 (63%) provided contact information for potential friend and/or family member controls. We were able to successfully contact and recruit a friend/family member for 92% of the case patients. This represented 46 friend controls and 54 family controls. Family controls significantly differed from original matched controls by sex and household income. Other characteristics were similar between friend controls and the original study’s neighborhood controls.ConclusionThe apparent comparability of neighborhood controls to friend and family controls among respondents in this study suggests that these alternative methods of control identification can serve as a complementary source of eligible controls in epidemiologic case-control studies.     

Basic statistical analysis in genetic case-control studies

This protocol describes how to perform basic statistical analysis in a population-based genetic association case-control study. The steps described involve the (i) appropriate selection of measures of association and relevance of disease models; (ii) appropriate selection of tests of association; (iii) visualization and interpretation of results; (iv) consideration of appropriate methods to control for multiple testing; and (v) replication strategies. Assuming no previous experience with software such as PLINK, R or Haploview, we describe how to use these popular tools for handling single-nucleotide polymorphism data in order to carry out tests of association and visualize and interpret results. This protocol assumes that data quality assessment and control has been performed, as described in a previous protocol, so that samples and markers deemed to have the potential to introduce bias to the study have been identified and removed. Study design, marker selection and quality control of case-control studies have also been discussed in earlier protocols. The protocol should take ~1 h to complete.     

Sensitivity analysis for matched case-control studies

A sensitivity analysis in an observational study indicates the degree to which conclusions would be altered by hidden biases of various magnitudes. A method of sensitivity analysis previously proposed for cohort studies is extended for use in matched case-control studies with multiple controls, where slightly different derivations and calculations are required. Also discussed is a sensitivity analysis for case-control studies that have two distinct types of controls, say hospital and neighborhood controls, where the two types may be affected by different biases. For illustration, the method is applied to five case-control studies, including a study of herniated lumbar disc in which there are three types of cases, and a study of breast cancer with two types of controls.     

Logistic regression methods for retrospective case-control studies using complex sampling procedures

There are a number of possible designs for case-control studies. The simplest uses two separate simple random samples, but an actual study may use more complex sampling procedures. Typically, stratification is used to control for the effects of one or more risk factors in which we are interested. It has been shown (Anderson, 1972, Biometrika 59, 19-35; Prentice and Pyke, 1979, Biometrika 66, 403-411) that the unconditional logistic regression estimators apply under stratified sampling, so long as the logistic model includes a term for each stratum. We consider the case-control problem with stratified samples and assume a logistic model that does not include terms for strata, i.e., for fixed covariates the (prospective) probability of disease does not depend on stratum. We assume knowledge of the proportion sampled in each stratum as well as the total number in the stratum. We use this knowledge to obtain the maximum likelihood estimators for all parameters in the logistic model including those for variables completely associated with strata. The approach may also be applied to obtain estimators under probability sampling.     

Comparison of racial differences in childhood cancer risk in case-control studies and population-based cancer registries

Introduction: Although selection bias in case-control studies has been studied extensively, little is known about selection of cases and controls among various ethnic groups. This study compares racial differences in childhood cancer rates as estimated by case-control studies with various design features. It also compares estimates of racial distribution among cases as reported by case-control studies to those observed for an ideal case series with complete ascertainment of cases for these studies or in population-based cancer registries in corresponding geographic regions and calendar periods. Methods: Peer-reviewed publications on childhood leukemia and brain tumors from North America, published between 1980 and 2007, were reviewed. Incidence data by race/ethnicity were compiled from research publications, federal cancer statistics, and cancer registries. Meta-analysis was conducted to assess racial/ethnic differences by study characteristics. Racial distributions of cases from published case-control studies were compared to those of a presumably noncensored case distribution (i.e. include both participating and non-participating cases in a case-control study) or cases recorded by cancer registries. Results: In interview-based case-control studies of childhood cancer, the proportion of Whites compared to non-Whites tended to be higher among controls than among cases; however, the opposite was true for record-based case-control studies. Additionally, the proportion of Whites tended to be higher among the participating cases in the published case-control studies compared to the proportion of Whites among the non-participating cases or in cancer registries. Conclusions: Investigators need to consider differential participation by racial group as a potential source of bias in the interpretation of case-control study results.     

Reducing mean squared error in the analysis of stratified epidemiologic studies

Kalish (1990, Biometrics 46, 493-499) proposed an approximately optimal estimator of a common odds ratio for pair-matched case-control studies. His approach is easily extended to general stratified studies. For unmatched stratified studies, the form of the approximately optimal estimator is very simple, and may often correspond to no more than a small correction to the ordinary stratified estimator.     

Accounting for cluster sampling in constructing enumerative sequential sampling plans

Green's sequential sampling plan is widely used in applied entomology. Green's equation can be used to construct sampling stop charts, and a crop can then be surveyed using a simple random sampling (SRS) approach. In practice, however, crops are rarely surveyed according to SRS. Rather, some type of hierarchical design is usually used, such as cluster sampling, where sampling units form distinct groups. This article explains how to make adjustments to sampling plans that intend to use cluster sampling, a commonly used hierarchical design, rather than SRS. The methodologies are illustrated using diamondback moth, Plutella xylostella (L.), a pest of Brassica crops, as an example.     

Genetic model selection in two-phase analysis for case-control association studies

The Cochran-Armitage trend test (CATT) is well suited for testing association between a marker and a disease in case-control studies. When the underlying genetic model for the disease is known, the CATT optimal for the genetic model is used. For complex diseases, however, the genetic models of the true disease loci are unknown. In this situation, robust tests are preferable. We propose a two-phase analysis with model selection for the case-control design. In the first phase, we use the difference of Hardy-Weinberg disequilibrium coefficients between the cases and the controls for model selection. Then, an optimal CATT corresponding to the selected model is used for testing association. The correlation of the statistics used for selection and the test for association is derived to adjust the two-phase analysis with control of the Type-I error rate. The simulation studies show that this new approach has greater efficiency robustness than the existing methods.     

Incorporating single-locus tests into haplotype cladistic analysis in case-control studies

In case-control studies, genetic associations for complex diseases may be probed either with single-locus tests or with haplotype-based tests. Although there are different views on the relative merits and preferences of the two test strategies, haplotype-based analyses are generally believed to be more powerful to detect genes with modest effects. However, a main drawback of haplotype-based association tests is the large number of distinct haplotypes, which increases the degrees of freedom for corresponding test statistics and thus reduces the statistical power. To decrease the degrees of freedom and enhance the efficiency and power of haplotype analysis, we propose an improved haplotype clustering method that is based on the haplotype cladistic analysis developed by Durrant et al. In our method, we attempt to combine the strengths of single-locus analysis and haplotype-based analysis into one single test framework. Novel in our method is that we develop a more informative haplotype similarity measurement by using p-values obtained from single-locus association tests to construct a measure of weight, which to some extent incorporates the information of disease outcomes. The weights are then used in computation of similarity measures to construct distance metrics between haplotype pairs in haplotype cladistic analysis. To assess our proposed new method, we performed simulation analyses to compare the relative performances of (1) conventional haplotype-based analysis using original haplotype, (2) single-locus allele-based analysis, (3) original haplotype cladistic analysis (CLADHC) by Durrant et al., and (4) our weighted haplotype cladistic analysis method, under different scenarios. Our weighted cladistic analysis method shows an increased statistical power and robustness, compared with the methods of haplotype cladistic analysis, single-locus test, and the traditional haplotype-based analyses. The real data analyses also show that our proposed method has practical significance in the human genetics field.     

Reducing mean squared error in the analysis of pair-matched case-control studies

The standard estimator of the common odds ratio for pair-matched case-control studies, the stratified estimate, is consistent but it ignores all information from the concordant pairs. At the other extreme, the pooled estimator is more efficient as it uses all the data, but is not consistent. In order to trade between bias and precision, Liang and Zeger (1988, Biometrics 44, 1145-1156) proposed an estimator that is a compromise between the stratified and pooled estimates. In the current paper, the possibility of optimizing the trade-off is explored. Specifically, the family of weighted averages of the stratified and pooled estimates is considered, and the weight that minimizes an asymptotic approximation of mean squared error is derived. In practice, the optimal weight must be estimated from the data so that the estimator is only approximately optimal. Small-sample properties are evaluated via simulations.     

A note on case-control sampling to estimate kappa coefficients

The feasibility and cost-effectiveness of estimation of kappa using a case-control method of sampling, proposed by Jannarone, Macera, and Garrison (1987, Biometrics 43, 433-437), is provided support. However, in this article unrealistic assumptions in their presentation are identified and more general results for more realistic settings are provided.