Flexible derivative time-varying model in matched case-crossover studies for a small number of geographical locations among the participants

2020 ◽  
pp. 096228022096817
Author(s):  
Ana M Ortega-Villa ◽  
Inyoung Kim
Keyword(s):  
Logistic Regression ◽  
Conditional Logistic Regression ◽  
Effect Modification ◽  
Fixed Number ◽  
Time Varying ◽  
Model Framework ◽  
Conditional Logistic Regression Model ◽  
Case Crossover ◽  
Crossover Studies ◽  
Matched Case

In matched case-crossover studies, any stratum effect is removed by conditioning on the fixed number of case–control sets in the stratum, and hence, the conditional logistic regression model is not able to detect any effects associated with matching covariates. However, some matching covariates such as time and location often modify the effect of covariates, making the estimations obtained by conditional logistic regression incorrect. Therefore, in this paper, we propose a flexible derivative time-varying coefficient model to evaluate effect modification by time and location, in order to make correct statistical inference, when the number of locations is small. Our proposed model is developed under the Bayesian hierarchical model framework and allows us to simultaneously detect relationships between the predictor and binary outcome and between the predictor and time. Inference is proposed based on the derivative function of the estimated function to determine whether there is an effect modification due to time and/or location, for a small number of locations among the participants. We demonstrate the accuracy of the estimation using a simulation study and an epidemiological example of a 1–4 bidirectional case-crossover study of childhood aseptic meningitis with drinking water turbidity.

scholarly journals A novel weighting method to remove bias from within-subject exposure dependency in case-crossover studies

2021 ◽  
Vol 21 (1) ◽  
Author(s):  
Kiyoshi Kubota ◽  
Thu-Lan Kelly ◽  
Tsugumichi Sato ◽  
Nicole Pratt ◽  
Elizabeth Roughead ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Crossover Study ◽  
Odds Ratio ◽  
Time Trends ◽  
Rate Ratio ◽  
Conditional Logistic Regression ◽  
Time Varying ◽  
Weighting Method ◽  
Case Crossover ◽  
Crossover Studies

Abstract Background Case-crossover studies have been widely used in various fields including pharmacoepidemiology. Vines and Farrington indicated in 2001 that when within-subject exposure dependency exists, conditional logistic regression can be biased. However, this bias has not been well studied. Methods We have extended findings by Vines and Farrington to develop a weighting method for the case-crossover study which removes bias from within-subject exposure dependency. Our method calculates the exposure probability at the case period in the case-crossover study which is used to weight the likelihood formulae presented by Greenland in 1999. We simulated data for the population with a disease where most patients receive a cyclic treatment pattern with within-subject exposure dependency but no time trends while some patients stop and start treatment. Finally, the method was applied to real-world data from Japan to study the association between celecoxib and peripheral edema and to study the association between selective serotonin reuptake inhibitor (SSRI) and hip fracture in Australia. Results When the simulated rate ratio of the outcome was 4.0 in a case-crossover study with no time-varying confounder, the proposed weighting method and the Mantel-Haenszel odds ratio reproduced the true rate ratio. When a time-varying confounder existed, the Mantel-Haenszel method was biased but the weighting method was not. When more than one control period was used, standard conditional logistic regression was biased either with or without time-varying confounding and the bias increased (up to 8.7) when the study period was extended. In real-world analysis with a binary exposure variable in Japan and Australia, the point estimate of the odds ratio (around 2.5 for the association between celecoxib and peripheral edema and around 1.6 between SSRI and hip fracture) by our weighting method was equal to the Mantel-Haenszel odds ratio and stable compared with standard conditional logistic regression. Conclusion Case-crossover studies may be biased from within-subject exposure dependency, even without exposure time trends. This bias can be identified by comparing the odds ratio by the Mantel-Haenszel method and that by standard conditional logistic regression. We recommend using our proposed method which removes bias from within-subject exposure dependency and can account for time-varying confounders.

scholarly journals A Novel Weighting Method to Remove Bias from Within-subject Exposure Dependency in Case-crossover Studies

2021 ◽  
Author(s):  
Kiyoshi Kubota ◽  
Lan Kelly ◽  
Tsugumichi Sato ◽  
Nicole Pratt ◽  
Elizabeth Roughead ◽  
...  
Keyword(s):  
Logistic Regression ◽  
Crossover Study ◽  
Odds Ratio ◽  
Time Trends ◽  
Rate Ratio ◽  
Conditional Logistic Regression ◽  
Time Varying ◽  
Weighting Method ◽  
Case Crossover ◽  
Crossover Studies

Abstract Background:Case-crossover studies have been widely used in various fields including pharmacoepidemiology. Vines and Farrington indicated in 2001 that when within-subject exposure dependency exists, conditional logistic regression can be biased. However, this bias has not been well studied.Methods:We have extended findings by Vines and Farrington to develop a weighting method for the case-crossover study which removes bias from within-subject exposure dependency. Our method calculates the exposure probability at the case period in the case-crossover study which is used to weight the likelihood formulae presented by Greenland in 1999. We simulated data for the population with a disease where most patients receive a cyclic treatment pattern with within-subject exposure dependency but no time trends while some patients stop and start treatment. Finally, the method was applied to real-world data from Japan to study the association between celecoxib and peripheral edema and to study the association between selective serotonin reuptake inhibitor (SSRI) and hip fracture in Australia.Results:When the simulated rate ratio of the outcome was 4.0 in a case-crossover study with no time-varying confounder, the proposed weighting method and the Mantel-Haenszel odds ratio reproduced the true rate ratio. When a time-varying confounder existed, the Mantel-Haenszel method was biased but the weighting method was not. When more than one control period was used, standard conditional logistic regression was biased either with or without time-varying confounding and the bias increased (up to 9.4) when the study period was extended. In real-world analysis with a binary exposure variable in Japan and Australia, the point estimate of the odds ratio (around 2.5 for the association between celecoxib and peripheral edema and around 1.6 between SSRI and hip fracture) by our weighting method was equal to the Mantel-Haenszel odds ratio and stable compared with standard conditional logistic regression. Conclusion:Case-crossover studies may be biased from within-subject exposure, even without exposure time trends. This bias can be identified by comparing the odds ratio calculated by the Mantel-Haenszel method and that by standard conditional logistic regression. Our proposed method will remove bias from within-subject exposure dependency and can account for time-varying confounders.

Semiparametric time varying coefficient model for matched case-crossover studies

2016 ◽  
Vol 36 (6) ◽  
pp. 998-1013 ◽  
Author(s):  
Ana Maria Ortega-Villa ◽  
Inyoung Kim ◽  
H. Kim
Keyword(s):  
Time Varying ◽  
Varying Coefficient Model ◽  
Varying Coefficient ◽  
Case Crossover ◽  
Crossover Studies ◽  
Matched Case

scholarly journals Association of Infrastructure and Route Environment Factors with Cycling Injury Risk at Intersection and Non-Intersection Locations: A Case-Crossover Study of Britain

2021 ◽  
Vol 18 (6) ◽  
pp. 3060
Author(s):  
Rachel Aldred ◽  
Georgios Kapousizis ◽  
Anna Goodman
Keyword(s):  
Injury Risk ◽  
Conditional Logistic Regression ◽  
Traffic Signals ◽  
Control Groups ◽  
Class A ◽  
Case Crossover ◽  
Main Class ◽  
Road Intersections ◽  
Matched Case ◽  
Safety In Numbers

Objective: This paper examines infrastructural and route environment correlates of cycling injury risk in Britain for commuters riding in the morning peak. Methods: The study uses a case-crossover design which controls for exposure. Control sites from modelled cyclist routes (matched on intersection status) were compared with sites where cyclists were injured. Conditional logistic regression for matched case–control groups was used to compare characteristics of control and injury sites. Results: High streets (defined by clustering of retail premises) raised injury odds by 32%. Main (Class A or primary) roads were riskier than other road types, with injury odds twice that for residential roads. Wider roads, and those with lower gradients increased injury odds. Guard railing raised injury odds by 18%, and petrol stations or car parks by 43%. Bus lanes raised injury odds by 84%. As in other studies, there was a ‘safety in numbers’ effect from more cyclists. Contrary to other analysis, including two recent studies in London, we did not find a protective effect from cycle infrastructure and the presence of painted cycle lanes raised injury odds by 54%. At intersections, both standard and mini roundabouts were associated with injury odds several times higher than other intersections. Presence of traffic signals, with or without an Advanced Stop Line (‘bike box’), had no impact on injury odds. For a cyclist on a main road, intersections with minor roads were riskier than intersections with other main roads. Conclusions: Typical cycling environments in Britain put cyclists at risk, and infrastructure must be improved, particularly on busy main roads, high streets, and bus routes.

scholarly journals Sparse estimation for case–control studies with multiple disease subtypes

Biostatistics ◽  
2020 ◽  
Author(s):  
Nadim Ballout ◽  
Cedric Garcia ◽  
Vivian Viallon
Keyword(s):  
Logistic Regression ◽  
Regression Model ◽  
Logistic Regression Model ◽  
Multinomial Logistic Regression ◽  
Conditional Logistic Regression ◽  
Case Control ◽  
Case Control Studies ◽  
Conditional Logistic Regression Model ◽  
Symmetric Formulation ◽  
Disease Subtypes

Summary The analysis of case–control studies with several disease subtypes is increasingly common, e.g. in cancer epidemiology. For matched designs, a natural strategy is based on a stratified conditional logistic regression model. Then, to account for the potential homogeneity among disease subtypes, we adapt the ideas of data shared lasso, which has been recently proposed for the estimation of stratified regression models. For unmatched designs, we compare two standard methods based on $L_1$-norm penalized multinomial logistic regression. We describe formal connections between these two approaches, from which practical guidance can be derived. We show that one of these approaches, which is based on a symmetric formulation of the multinomial logistic regression model, actually reduces to a data shared lasso version of the other. Consequently, the relative performance of the two approaches critically depends on the level of homogeneity that exists among disease subtypes: more precisely, when homogeneity is moderate to high, the non-symmetric formulation with controls as the reference is not recommended. Empirical results obtained from synthetic data are presented, which confirm the benefit of properly accounting for potential homogeneity under both matched and unmatched designs, in terms of estimation and prediction accuracy, variable selection and identification of heterogeneities. We also present preliminary results from the analysis of a case–control study nested within the EPIC (European Prospective Investigation into Cancer and nutrition) cohort, where the objective is to identify metabolites associated with the occurrence of subtypes of breast cancer.

scholarly journals Matched versus Unmatched Analysis of Matched Case-Control Studies

2021 ◽  
Author(s):  
Fei Wan ◽  
Graham A Colditz ◽  
Siobhan Sutcliffe
Keyword(s):  
Logistic Regression ◽  
Functional Form ◽  
Conditional Logistic Regression ◽  
Control Sample ◽  
Target Population ◽  
Case Control ◽  
Model Specification ◽  
Exact Matching ◽  
Case Control Studies ◽  
Matched Case

Abstract Although the need for addressing matching in the analysis of matched case-control studies is well established, debate remains as to the most appropriate analytic method when matching on at least one continuous factor. We compare the bias and efficiency of unadjusted and adjusted conditional logistic regression (CLR) and unconditional logistic regression (ULR) in the setting of both exact and non-exact matching. To demonstrate that case-control matching distorts the association between the matching variables and the outcome in the matched sample relative to the target population, we derive the logit model for the matched case-control sample under exact matching. We conduct simulations to validate our theoretical conclusions and to explore different ways of adjusting for the matching variables in CLR and ULR to reduce biases. When matching is exact, CLR is unbiased in all settings. When matching is not exact, unadjusted CLR tends to be biased and this bias increases with increasing matching caliper size. Spline smoothing of the matching variables in CLR can alleviate biases. Regardless of exact or non-exact matching, adjusted ULR is generally biased unless the functional form of the matched factors is modelled correctly. The validity of adjusted ULR is vulnerable to model specification error. CLR should remain the primary analytic approach.

Sign in / Sign up

Export Citation Format

Share Document