Biomedical subjects

Neil Klar

Publications and source records attributed to Neil Klar.

At least 19 recordsLinked to original sources

PubMed2007

Accounting for expected attrition in the planning of community intervention trials.

Trials in which intact communities are the units of randomization are increasingly being used to evaluate interventions which are more naturally administered at the community level, or when there is a substantial risk of treatment contamination. In this article we focus on the planning of community intervention trials in which k communities (for example, medical practices, worksites, or villages) are to be randomly allocated to each of an intervention and a control group, and fixed cohorts of m individuals enrolled in each community prior to randomization. Formulas to determine k or m may be obtained by adjusting standard sample size formulas to account for the intracluster correlation coefficient rho. In the presence of individual-level attrition however, observed cohort sizes are likely to vary. We show that conventional approaches of accounting for potential attrition, such as dividing standard sample size formulas by the anticipated follow-up rate pi or using the average anticipated cohort size m pi, may, respectively, overestimate or underestimate the required sample size when cluster follow-up rates are highly variable, and m or rho are large. We present new sample size estimation formulas for the comparison of two means or two proportions, which appropriately account for variation among cluster follow-up rates. These formulas are derived by specifying a model for the binary missingness indicators under the population-averaged approach, assuming an exchangeable intracluster correlation coefficient, denoted by tau. To aid in the planning of future trials, we recommend that estimates for tau be reported in published community intervention trials.

Canada↗

PubMed2007

The merits of breaking the matches: a cautionary tale.

Matched-pair cluster randomization trials are frequently adopted as the design of choice for evaluating an intervention offered at the community level. However, previous research has demonstrated that a strategy of breaking the matches and performing an unmatched analysis may be more efficient than performing a matched analysis on the resulting data, particularly when the total number of communities is small and the matching is judged as relatively ineffective. The research concerning this question has naturally focused on testing the effect of intervention. However, a secondary objective of many community intervention trials is to investigate the effect of individual-level risk factors on one or more outcome variables. Focusing on the case of a continuous outcome variable, we show that the practice of performing an unmatched analysis on data arising from a matched-pair design can lead to bias in the estimated regression coefficient, and a corresponding test of significance which is overly liberal. However, for large-scale community intervention trials, which typically recruit a relatively small number of large clusters, such an analysis will generally be both valid and efficient. We also consider other approaches to testing the effect of an individual-level risk factor in a matched-pair cluster randomization design, including a generalized linear model approach that preserves the matching, a two-stage cluster-level analysis, and an approach based on generalized estimating equations.

Asthma↗

PubMed2006

A blended knowledge translation initiative to improve colorectal cancer staging [ISRCTN56824239].

BACKGROUND: A significant gap has been documented between best practice and the actual practice of surgery. Our group identified that colorectal cancer staging in Ontario was suboptimal and subsequently developed a knowledge translation strategy using the principles of social marketing and the influence of expert and local opinion leaders for colorectal cancer. METHODS/DESIGN: Opinion leaders were identified using the Hiss methodology. Hospitals in Ontario were cluster-randomized to one of two intervention arms. Both groups were exposed to a formal continuing medical education session given by the expert opinion leader for colorectal cancer. In the treatment group the local Opinion Leader for colorectal cancer was detailed by the expert opinion leader for colorectal cancer and received a toolkit. Forty-two centres agreed to have the expert opinion leader for colorectal cancer come and give a formal continuing medical education session that lasted between 50 minutes and 4 hours. No centres refused the intervention. These sessions were generally well attended by most surgeons, pathologists and other health care professionals at each centre. In addition all but one of the local opinion leaders for colorectal cancer met with the expert opinion leader for colorectal cancer for the academic detailing session that lasted between 15 and 30 minutes. DISCUSSION: We have enacted a unique study that has attempted to induce practice change among surgeons and pathologists using an adapted social marketing model that utilized the influence of both expert and local opinion leaders for colorectal cancer in a large geographic area with diverse practice settings.

Attitude of Health Personnel↗

PubMed2005

Predictors of self-reported confidence ratings for adult recall of early life sun exposure.

Use of self-reported confidence ratings may be an efficient method for assessing recall bias. In this exploratory application of the method, the authors examined the relation between case-control status and self-reported confidence ratings. In 2002 and 2003, melanoma cases (n = 141) and controls (n = 143) aged 20-44 years residing in Ontario, Canada, estimated the amounts of time they had spent outdoors in summer activities when they were 6-18 years of age and indicated their confidence in the accuracy of each estimate. The generalized estimating equations extension of logistic regression was used to examine dichotomized confidence ratings (more confident vs. less confident) for activities reported for ages 6-11 years and 12-18 years. Types of activity were associated with more confident reporting for both age strata; as the number of stable outdoor activity periods (total number of similar outdoor periods within each activity) reported by respondents increased, confidence decreased. Cumulative time spent outdoors was also associated with more confidence but reached statistical significance only for the age stratum 12-18 years. There was no statistically significant association between case-control status and self-reported confidence for either age stratum (6-11 years: odds ratio = 0.91; 12-18 years: odds ratio = 1.32), which suggests an absence of recall bias for reported time spent outdoors.

Adolescent↗

PubMed2005

A non-iterative confidence interval estimating procedure for the intraclass kappa statistic with multinomial outcomes.

We obtain the asymptotic sample variance of the intraclass kappa statistic for multinomial outcome data. A modified Wald type procedure based on this theory is then used for confidence interval construction. The results of a simulation study show that the proposed non-iterative approach performs very well in terms of confidence interval coverage and width for samples as small as 50. The procedure is illustrated with two examples from previously published medical studies.

Carotid Stenosis↗

PubMed2005

Accuracy of breast screening among women with and without a family history of breast and/or ovarian cancer.

OBJECTIVES: To compare interval cancer rates, sensitivity and specificity of breast cancer screening between women with moderate or strong family history and women without a family history. METHODS: From 1996 to 1997, 115,460 women aged 50 to 69 screened by the Ontario Breast Screening Program, offering eligible women screening with mammography and clinical breast examination, were examined. Women were followed for up to 12 months after their screening examination. Family history definitions were based on the number of affected first degree relatives and their ages at diagnosis. Multivariate analysis was conducted to adjust for potential confounding variables. RESULTS: Interval cancer rates increased across family history groups and were greatest in women with a strong family history. The rate ratio (RR) for interval cancer rate in women with a strong family history compared to women without a family history approached significance (RR=2.28, 95% confidence interval (CI) 0.97-5.34), while for women with a moderate family history it did not (RR=1.37, 95% CI 0.62-3.04). A slightly but not significantly lower sensitivity was observed in women with a strong family history compared to women without a family history. There was little variation in specificity across family history groups. CONCLUSIONS: Screening was able to detect a large proportion of invasive breast cancers in women with a family history, indicating their potential to benefit from regular breast cancer screening. However, due to increased interval cancer rates, screening with one-year intervals may be important even in an older population of women with a family history.

Age Factors↗

PubMed2005

Colorectal screening is associated with reduced colorectal cancer risk: a case-control study within the population-based Ontario Familial Colorectal Cancer Registry.

OBJECTIVE: This is the first study to evaluate the association between colonic screening and colorectal cancer risk among Canadians. METHODS: A case-control study was conducted. Cases were diagnosed with cancer of the colorectum, between 1997 and 2000, aged 20 to 74 years, identified through the population-based Ontario Cancer Registry and recruited by the Ontario Familial Colorectal Cancer Registry. Controls were a sex- and age-matched random sample of the population of Ontario. 971 cases and 1944 controls completed questionnaires (including colorectal screening history and many risk factors). Multivariate logistic regression analysis was used to obtain adjusted odds ratios (OR) estimates. RESULTS: Having had a fecal occult blood screen was associated with reduced colorectal cancer risk (OR=0.76; 95% confidence interval (CI): 0.59, 0.97). Having had a screening sigmoidoscopy was associated with a halving of colorectal cancer risk (OR = 0.52; 95% CI: 0.34, 0.80). Having had a screening colonoscopy did not significantly reduce colorectal cancer risk (OR = 0.69; 95% CI: 0.44, 1.07); however, having had either screening endoscopy was associated with a significant reduction in colorectal cancer risk (OR = 0.62; 95% CI: 0.44, 0.87). Findings differed slightly by anatomic sub-site (proximal and distal colorectum). CONCLUSIONS: We report a reduction in colorectal cancer risk among persons who underwent colorectal cancer screening; in particular, sigmoidoscopy. Findings are of great importance for the prevention of colorectal cancer.

Adult↗

PubMed2005

Group sequential methods for cluster randomization trials with binary outcomes.

BACKGROUND: Cluster randomization trials in which intact social units are randomly assigned to different intervention groups have become very popular in recent years, particularly for the evaluation of innovations in the delivery of health care. An extensive literature dealing with the associated methodological challenges has also appeared. Although the monitoring of such trials using formal stopping rules is clearly indicated when the outcomes are irreversible and individual-level data are available sequentially, simple and reliable statistical methods that may be used for this purpose are currently not available. PURPOSE: To investigate the validity of standard group sequential methods when applied to cluster randomization trials having binary outcomes. METHODS: The large sample distributions for each of five test statistics computed from sequentially accumulated data are derived. A simulation study is performed to evaluate the finite sample properties of these statistics when applied to the interim analysis of cluster randomization trials. Data from the World Health Organization antenatal care trial are used to illustrate the methods. RESULTS: Each of the joint distributions is shown to be characterized by a covariance structure that asymptotically satisfies an independent increments structure, a foundation that simplifies group sequential methods. The simulation study reveals that four of the five test statistics evaluated provide satisfactory performance with as few as 10 clusters allocated to each of two interventions. LIMITATIONS: The applicability of our results to effect estimation following a group sequential cluster randomization trial is not investigated, although a theoretical foundation which may be used for this purpose is presented. CONCLUSIONS: Standard group sequential methods can be applied to cluster randomization trials when interim analyses are warranted.

Clinical Trials as Topic↗

PubMed2004

Antidepressant medication use and non-Hodgkin's lymphoma risk: no association.

Animal and human studies have suggested that antidepressant medications may be associated with several cancers. The authors evaluated the association between antidepressant medication use and the risk of non-Hodgkin's lymphoma using a Canadian population-based case-control study, the National Enhanced Cancer Surveillance Study. Non-Hodgkin's lymphoma cases (n=638) diagnosed in 1995-1996 were identified using the Ontario Cancer Registry, and controls (n=1,930) were identified from the Ontario Ministry of Finance Property Assessment Database. Antidepressant medication use was ascertained using a self-administered questionnaire. Multivariate logistic regression was used to estimate odds ratios. "Ever" use of antidepressant medications was not associated with non-Hodgkin's lymphoma risk. The odds ratio for non-Hodgkin's lymphoma with 25 or more months of tricyclic antidepressant medication use was 1.6; however, this was nonsignificant. Duration or history of use or individual types of antidepressant medications were not associated with non-Hodgkin's lymphoma risk. These findings do not support an increased risk of non-Hodgkin's lymphoma with antidepressant medication use.

Age Distribution↗

PubMed2004

Methods for modelling change in cluster randomization trials.

Randomized trials are often designed to assess an intervention's ability to change patient knowledge, behaviour or health. The study outcome will then need to be measured at least twice for each subject--prior to random assignment and following implementation of the intervention. In this paper we consider methods for modelling change when data are obtained from cluster randomization trials where the unit of allocation is a family, school or community. Attention focuses on mixed effects linear regression extensions of (i) two-sample t-tests and (ii) analysis of covariance, in both cases accounting for dependencies among cluster members. Algebraic expressions for tests of the intervention effect are derived for the special case where there are a fixed number of subjects per cluster while simulation studies are used to compare the power of these procedures in the more realistic case where there is variability in cluster size. A key conclusion is that there can be considerable gains in power when allowing for different individual-level and cluster-level associations between the baseline and follow-up assessments. The discussion is illustrated using data from a school-based smoking prevention trial.

Canada↗

PubMed2004

Accuracy of cancer family histories: comparison of two breast cancer syndromes.

Cancer risk programs rely on accurately reported family history information. This study compares the accuracy with which cancer sites and ages at diagnosis are reported by Li-Fraumeni syndrome (LFS) and hereditary breast-ovarian cancer syndrome (HBOCS) families undergoing genetic testing. We analyzed the accuracy of 191 cancer diagnoses among first-degree (FDRs) and second-degree (SDRs) relatives reported by 32 LFS and 52 HBOCS participants in genetic testing programs. Cancer diagnoses of relatives were more accurately reported in the HBOCS cohort (78%) than in the LFS cohort (52%). Almost all breast cancer diagnoses were accurately reported, whereas 74% of ovarian cancer diagnoses and only 55% of other LFS-related cancers were accurately reported. Age at diagnosis was accurate within 5 years for 60% of LFS relatives and 53% of HBOCS relatives. Factors correlating with accurate reporting of cancer history included: being member of BRCA1 family, higher education level, female historian, degree of closeness to affected relative, and having fewer than 5 affected FDRs and SDRs. Relying on verbal histories would not have altered eligibility for genetic testing among HBOCS historians, but fewer than half of LFS historians provided information that would have led to TP53 testing. Our data suggest that it may not be necessary to confirm breast cancer diagnoses routinely; however, documentation of other cancer types remains important for appropriate risk assessment and follow-up.

Age Factors↗

PubMed2004

Methods for the statistical analysis of binary data in split-cluster designs.

Split-cluster designs are frequently used in the health sciences when naturally occurring clusters such as multiple sites or organs in the same subject are assigned to different treatments. However, statistical methods for the analysis of binary data arising from such designs are not well developed. The purpose of this article is to propose and evaluate a new procedure for testing the equality of event rates in a design dividing each of k clusters into two segments having multiple sites (e.g., teeth, lesions). The test statistic proposed is a generalization of a previously published procedure based on adjusting the standard Pearson chi-square statistic, but can also be derived as a score test using the approach of generalized estimating equations.

Anti-Infective Agents, Local↗

PubMed2004

Pitfalls of and controversies in cluster randomization trials.

It is now well known that standard statistical procedures become invalidated when applied to cluster randomized trials in which the unit of inference is the individual. A resulting consequence is that researchers conducting such trials are faced with a multitude of design choices, including selection of the primary unit of inference, the degree to which clusters should be matched or stratified by prognostic factors at baseline, and decisions related to cluster subsampling. Moreover, application of ethical principles developed for individually randomized trials may also require modification. We discuss several topics related to these issues, with emphasis on the choices that must be made in the planning stages of a trial and on some potential pitfalls to be avoided.

Bias↗

PubMed2003

A Canadian national survey of attitudes and knowledge regarding preventive vaccines.

BACKGROUND: Vaccines have virtually eliminated many diseases, but public concerns about their safety could undermine future public health initiatives. OBJECTIVE: To determine Canadians' attitudes and knowledge about vaccines, particularly in view of increasing public concern about bioterrorism and the possible need for emergency immunizations after weaponized anthrax incidents and the events of September 11, 2001. METHOD: A 20-question survey based on well-researched dimensions of vaccine responsiveness was telephone-administered to a random sample of N = 1330 adult Canadians in January, 2002. RESULTS: 1057 (79.5%) completed the survey. Respondents perceived vaccines to be highly effective and demonstrated considerable support for further vaccine research. However, results also indicate a lack of knowledge about vaccines and uncertainty regarding the safety. CONCLUSIONS: Support for vaccines is broad but shallow. While Canadians hold generally positive attitudes about vaccines, support could be undermined by widely publicized adverse events. Better public education is required to maintain support for future public health initiatives.

Journal Article↗

PubMed2002

Issues in the meta-analysis of cluster randomized trials.

Meta-analyses involving the synthesis of evidence from cluster randomization trials are being increasingly reported. These analyses raise challenging methodologic issues beyond those raised by meta-analyses which include only individually randomized trials. In this paper we review and comment on a selected number of these issues, including problems of study heterogeneity, difficulties in estimating design effects from individual trials and the choice of statistical methods.

Cluster Analysis↗

PubMed2002

Sample size re-estimation in cluster randomization trials.

Cluster randomization trials in which families are the unit of allocation are commonly adopted for the evaluation of disease prevention interventions. Sample size estimation for cluster randomization trials depends on parameters that quantify the variability within and between clusters and the variability in cluster size. Accurate advance estimates of these nuisance parameters may be difficult to obtain and misspecification may lead to an underpowered study. Since families are typically recruited over time, we propose using a portion of the data to estimate the nuisance parameters and to re-estimate sample size based on the estimates. This extends the standard internal pilot study methods to the setting of cluster randomization trials. The effect of this design on the power, significance level and sample size is analysed via simulation and is shown to provide a flexible and practical approach to cluster randomization trials.

Cluster Analysis↗

PubMed2002

Challenges to improving the impact of worksite cancer prevention programs: comparing reach, enrollment, and attrition using active versus passive recruitment strategies.

OBJECTIVE: The impact of worksite intervention studies is maximized when reach and enrollment are high and attrition is low. Differences in reach, enrollment, and retention were investigated by comparing 2 different employee recruitment methods for a home-based cancer-prevention intervention study. METHODS: Twenty-two worksites (N = 10,014 employees) chose either active or passive methods to recruit employees into a home-based intervention study. Reach (e.g., number of employees who gave permission to be called at home), Enrollment (e.g., number of employees who joined the home intervention study), and Attrition (e.g., number who did not complete the 12- and 24-month follow-ups) were determined. Analysis at the cluster level assessed differences between worksites that selected active (n = 12) versus passive (n = 10) recruitment methods on key outcomes of interest. Employees recruited by passive methods had significantly higher reach (74.5% vs. 24.4% for active) but significantly lower enrollment (41% vs. 78%) and retention (54% vs. 70%) rates (all ps < .0001). Passive methods also successfully enrolled a more diverse, high-risk employee sample. Passive (vs. active) recruitment methods hold advantages for increased reach and the ability to retain a more representative employee sample. Implications of these results for the design of future worksite studies that involve multilevel recruitment methods are discussed.

Adult↗

PubMedDate not supplied

A population-based study of Ashkenazi Jewish women's attitudes toward genetic discrimination and BRCA1/2 testing.

PURPOSE: The Human Genome Project continues to produce an increasing number of genetic susceptibility tests. Some of these genetic tests target social or ethnic groups who are at increased risk of developing a disease. The Ashkenazi Jewish community is one ethnic group that is an ongoing subject of genetic investigation. We assessed the attitudes of a population-based sample of Ashkenazi Jewish women toward breast-ovarian cancer susceptibility testing (BRCA 1/2). In particular, we assessed concerns about group discrimination, perceptions of the advantages and disadvantages of testing, and the relationship between concerns about discrimination and the potential benefits of genetic testing. METHODS: A telephone survey of a population-based sample of 200 Jewish women. RESULTS: A minority of women (17%) in this study expressed concern or discomfort with Jews being offered testing. Most women believed there were scientific reasons for testing Jews (71%), and only 5% of women felt that research that focused on Jews was bad for Jews as a group. Increased concern about genetic discrimination was associated with women who were highly educated (odds ratio 2.68). Forty percent of women surveyed were interested in testing, 40% were not interested, and 20% were uncertain about whether they would obtain testing. Increased interest in genetic testing was associated with a desire to obtain information about children's risk of disease and valuing information for its own sake. CONCLUSIONS: The majority of a population-based sample of Jewish women did not express concerns about group discrimination resulting from genetic testing. Women who are highly educated are more concerned about genetic discrimination. There is significant variation among Jewish women's interest in breast cancer susceptibility testing.

Adolescent↗