Skip to main content
Key Specialties
Cardiology Diabetology Endocrinology Gastroenterology Geriatrics and Gerontology Gynecology and Obstetrics Hematology Infectious Disease Internal Medicine Nephrology Neurology Oncology Pediatrics Respiratory Medicine Rheumatology Surgery
Congress Coverage
2024 ACC Congress 2023 ESMO Congress 2023 EASD Congress 2023 ERS Congress 2023 ESC Congress
Clinical Collections
Advances in Alzheimer’s

Keynote webinar | Spotlight on medication adherence

LIVE: Thursday 27th June 2024, 18:00-19:30 (CEST)
Join our expert panel to discover why you need to understand the drivers of non-adherence in your patients, and how you can optimize medication adherence in your clinics to drastically improve patient outcomes.
ADVANCED SEARCH
Log in
Top
BMC Medical Research Methodology
Published in: BMC Medical Research Methodology 1/2016

Open Access 01-12-2016 | Research article

The case-crossover design via penalized regression

Authors: Sam Doerken, Maja Mockenhaupt, Luigi Naldi, Martin Schumacher, Peggy Sekula

Published in: BMC Medical Research Methodology | Issue 1/2016

Login to get access

Abstract

Background

The case-crossover design is an attractive alternative to the classical case–control design which can be used to study the onset of acute events if the risk factors of interest vary in time. By comparing exposures within cases at different time periods, the case-crossover design does not rely on control subjects which can be difficult to acquire. However, using the standard method of maximum likelihood, resulting risk estimates can be heavily biased when the prevalence to risk factors is very low (or very high).

Methods

To overcome the problem of low risk factor prevalences, penalized conditional logistic regression via the lasso (least absolute shrinkage and selection operator) has been proposed in the literature as well as related methods such as the Firth correction. We apply and compare several penalized regression approaches in the context of a case-crossover analysis of the European Study of Severe Cutaneous Adverse Reactions (EuroSCAR; 1997–2001).

Results

Out of 30 drugs, standard methods only correctly classified 17 drugs (including some highly implausible risk estimates), while penalized methods correctly classified 22 drugs.

Conclusion

Penalized methods generally yield better risk classifications and much more plausible risk estimates for the EuroSCAR study than standard methods. As these novel techniques can be easily implemented using available R packages, we encourage routine use of penalized conditional logistic regression for case-crossover data.
Appendix
Available only for authorised users
Literature
1.
Breslow NE, Day NE. Statistical Methods in Cancer Research. Vol. 1. The Analysis of Case–control Studies. (IARC Scientific Publication no. 32). Lyon: International Agency for Research on Cancer; 1980.
2.
Maclure M. The case-crossover design: A method for studying transient effects on the risk of acute event. Am J Epidemiol. 1991;133(2):144–53.CrossRefPubMed
3.
Greenland S. Invited commentary: variable selection versus shrinkage in the control of multiple confounders. Am J Epidemiol. 2008;167:523–9.CrossRefPubMed
4.
Cole SR, Chu H, Greenland S. Maximum likelihood, profile likelihood, and penalized likelihood: a primer. Am J Epidemiol. 2014;179(2):252–60.CrossRefPubMed
5.
Avalos M, Grandvalet Y, Adroher ND, et al. Analysis of multiple exposures in the case-crossover design via sparse conditional likelihood. Stat Med. 2012;31(21):2290–302.CrossRefPubMed
6.
Avalos M, Orriols L, Pouyes H, et al. Variable selection on large case-crossover data: application to a registry-based study of prescription drugs and road traffic crashes. Pharmacoepidemiol Drug Saf. 2014;23:140–51.CrossRefPubMed
7.
Mockenhaupt M, Viboud C, Dunant A, et al. Stevens-Johnson syndrome and toxic epidermal necrolysis: assessment of medication risks with emphasis on recently marketed drugs. The EuroSCAR study. J Invest Dermatol. 2008;128(1):35–44.CrossRefPubMed
8.
Viboud C, Boëlle PY, Kelly J, et al. Comparison of the statistical efficiency of case-crossover and case–control designs: Application to severe cutaneous adverse reactions. J Clin Epidemiol. 2001;54:1218–27.CrossRefPubMed
9.
Hosmer DW, Lemeshow S. Applied Logistic Regression. 3rd ed. Hoboken: Wiley-Interscience; 2013.CrossRef
10.
Hastie T, Tibshirani R, Friedman J. The Elements of Statistical Learning: Data Mining, Inference and Prediction. New York: Springer Publishing Company; 2009.CrossRef
11.
Bach F. Bolasso: model consistent lasso estimation through the bootstrap. McCallum A, Roweis S, Cohen W, (eds). In Proceedings of the 25th International Conference on Machine Learning (ICML 2008). Helsinki, Finland; 2008.
12.
De Bin R, Janitza S, Sauerbrei W, et al. Subsampling versus bootstrapping in resampling-based model selection for multivariable regression. Biom. 2016;72(1):272–280.
13.
14.
Firth D. Bias reduction of maximum likelihood estimates. Biometrika. 1993;80(1):27–38.CrossRef
15.
Zeng C, Newcomer SR, Glanz JM, et al. Bias correction of risk estimates in vaccine safety studies with rare adverse events using a self-controlled case series design. Am J Epidemiol. 2013;178(12):1750–9.CrossRefPubMed
16.
Heinze G, Puhr R. Bias-reduced and separation-proof conditional logistic regression with small or sparse data sets. Stat Med. 2010;29(7–8):770–7.CrossRefPubMed
17.
Sun JX, Sinha S, Wang S, et al. Bias reduction in conditional logistic regression. Stat Med. 2011;30(4):348–55.CrossRefPubMed
18.
Reid S, Tibshirani R. clogitL1. R package version 1.4. 2014.
19.
Reid S, Tibshirani R. Regularization paths for conditional logistic regression: the clogitL1 package. J Stat Softw. 2014;58(12):1–23.CrossRef
20.
21.
Avalos M, Grandvalet Y, Pouyes H, et al. High-dimensional sparse matched case–control and case-crossover data: A review of recent works, description of an R tool and an illustration of the use in epidemiological studies. In: Formenti E, Tagliaferri R, Wit E, editors. Computational Intelligence Methods for Bioinformatics and Biostatistics, vol. 8452. Lecture Notes in Computer Science. Cham, Switzerland: Springer International Publishing; 2014. p. 109–24.
22.
Sun H, Wang S. Network-based regularization for matched case–control analysis of high-dimensional DNA methylation data. Stat Med. 2013;32(21):2127–39.CrossRefPubMed
23.
Ploner M, Heinze G. coxphf: Cox regression with Firth's penalized likelihood. R package version 1.11. 2015.
24.
Papay J, Yuen N, Mockenhaupt M, et al. Spontaneous adverse event reports of Stevens-Johnson syndrome/toxic epidermal necrolysis: detecting associations with medications. Pharmacoepidemiol Drug Saf. 2012;21(3):289–96.CrossRefPubMed
25.
Herring AH. Nonparametric bayes shrinkage for assessing exposures to mixtures subject to limits of detection. Epidemiology. 2010;21 Suppl 4:71–6.CrossRef
26.
Chatterjee A, Lahiri SN. Bootstrap lasso estimators. JASA. 2011;106(494):608–25.CrossRef
27.
Greenland S, Schwartzbaum JA, Finkle WD. Problems due to small samples and sparse data in conditional logistic regression analysis. Am J Epidemiol. 2000;151(5):531–9.CrossRefPubMed
28.
Rose S. Mortality risk score prediction in an elderly population using machine learning. Am J Epidemiol. 2013;177(5):443–52.CrossRefPubMed
29.
Burgette LF, Reiter JP, Miranda ML. Exploratory quantile regression with many covariates: an application to adverse birth outcomes. Epidemiology. 2011;22(6):859–66.CrossRefPubMed
30.
31.
Mostofsky E, Schwartz J, Coull BA, et al. Modeling the association between particle constituents of air pollution and health outcomes. Am J Epidemiol. 2012;176(4):317–26.CrossRefPubMedPubMedCentral
32.
Sullivan SG, Greenland S. Bayesian regression in SAS software. Int J Epidemiol. 2013;42(1):308–17.CrossRefPubMed
33.
34.
Farrington CP. Relative incidence estimation from case series for vaccine safety. Biometrics. 1995;51(1):228–35.CrossRefPubMed
35.
Avalos M, Pouyes H, Grandvalet Y, et al. Sparse conditional logistic regression for analyzing large-scale matched data from epidemiological studies: a simple algorithm. BMC Bioinformatics. 2015;16 Suppl 6:51.CrossRef
36.
Avalos M, Adroher ND, Lagarde E, et al. Prescription-drug-related risk in driving: comparing conventional and lasso shrinkage logistic regression. Epidemiology. 2012;23(5):706–12.CrossRefPubMed
Metadata
Title
The case-crossover design via penalized regression
Authors
Sam Doerken
Maja Mockenhaupt
Luigi Naldi
Martin Schumacher
Peggy Sekula
Publication date
01-12-2016
Publisher
BioMed Central
Published in
BMC Medical Research Methodology / Issue 1/2016
Electronic ISSN: 1471-2288
DOI
https://doi.org/10.1186/s12874-016-0197-0

Other articles of this Issue 1/2016

BMC Medical Research Methodology 1/2016 Go to the issue

Research article

How large are the consequences of covariate imbalance in cluster randomized trials: a simulation study with a continuous outcome and a binary covariate at the cluster level

Research article

Identification of confounder in epidemiologic data contaminated by measurement error in covariates

Research article

Utilizing distributional analytics and electronic records to assess timeliness of inpatient blood glucose monitoring in non-critical care wards

Research article

How to deal with missing longitudinal data in cost of illness analysis in Alzheimer’s disease—suggestions from the GERAS observational study

Research article

How do researchers determine the difference to be detected in superiority trials? Results of a survey from a panel of researchers

Research article

Eliciting parental support for the use of newborn blood spots for pediatric research