Why Match? Investigating Matched Case-Control Study Designs with Causal Effect Estimation

Rose, Sherri; Laan, Mark J. van der

doi:10.2202/1557-4679.1127

Cited by 253 publications

(193 citation statements)

References 27 publications

Supporting

Mentioning

193

Contrasting

Order By: Relevance

“…Typically, cases are matched to controls on a limited number of variables to control for effects of variables that prior work has demonstrated affect the outcome variable. 13 Our approach also allowed for estimation of the association of the non-matched variables on the outcome using multivariable analysis. 13 For each case, we first identified the primary care clinician seen most often in the 6 months prior to death.…”

Section: Subjectsmentioning

confidence: 99%

“…13 Our approach also allowed for estimation of the association of the non-matched variables on the outcome using multivariable analysis. 13 For each case, we first identified the primary care clinician seen most often in the 6 months prior to death. Then, we identified potential controls as all patients who had received care from the same clinician in the same 3-month time period (quarter).…”

Section: Subjectsmentioning

confidence: 99%

See 1 more Smart Citation

Correlates of Suicide Among Veterans Treated in Primary Care: Case–Control Study of a Nationally Representative Sample

et al. 2014

View full text Add to dashboard Cite

BACKGROUND: Veterans receiving Veterans Affairs (VA) healthcare have increased suicide risk compared to the general population. Many patients see primary care clinicians prior to suicide. Yet little is known about the correlates of suicide among patients who receive primary care treatment prior to death. OBJECTIVE: Our aim was to describe characteristics of veterans who received VA primary care in the 6 months prior to suicide; and to compare these to characteristics of control patients who also received VA primary care. DESIGN: This was a retrospective case-control study. SUBJECTS: The investigators partnered with VA operations leaders to obtain death certificate data from 11 states for veterans who died by suicide in 2009. Cases were matched 1:2 to controls based on age, sex, and clinician. MAIN MEASURES: Demographic, diagnosis, and utilization data were obtained from VA's Corporate Data Warehouse. Additional clinical and psychosocial context data were collected using manual medical record review. Multivariate conditional logistic regression was used to examine associations between potential predictor variables and suicide. KEY RESULTS: Two hundred and sixty-nine veteran cases were matched to 538 controls. Average subject age was 63 years; 97 % were male. Rates of mental health conditions, functional decline, sleep disturbance, suicidal ideation, and psychosocial stressors were all significantly greater in cases compared to controls. In the final model describing men in the sample, non-white race (OR=0.51; 95 % CI=0.27-0.98) and VA service-connected disability (OR=0.54; 95 % CI=0.36-0.80) were associated with decreased odds of suicide, while anxiety disorder (OR=3.52; 95 % CI=1.79-6.92), functional decline (OR=2.52; 95 % CI=1.55-4.10), depression (OR =1.82; 95 % CI=1.07-3.10), and endorsement of suicidal ideation (OR=2.27; 95 % CI=1.07-4.83) were associated with greater odds of suicide. CONCLUSIONS: Assessment for anxiety disorders and functional decline in addition to suicidal ideation and depression may be especially important for determining suicide risk in this population. Continued development of interventions that support identifying and addressing these conditions in primary care is indicated.KEY WORDS: mental health; veterans; primary care; health services research.

show abstract

Section: Subjectsmentioning

confidence: 99%

Section: Subjectsmentioning

confidence: 99%

Correlates of Suicide Among Veterans Treated in Primary Care: Case–Control Study of a Nationally Representative Sample

et al. 2014

View full text Add to dashboard Cite

show abstract

“…In addition, patients undergoing EC:TORS for primary surgical resection of OPSCC or SGSCC from March 2007 through August 2010 were used as a subject pool from which matched controls were selected. Criteria were chosen to decrease potential bias across confounding variables 11,12 . Patients were matched using the following criteria: 1) tumor location; 2) clinical T stage; 3) gender; and 4) age ( 12 yrs).…”

Section: Discussionmentioning

confidence: 99%

Transoral Robotic Surgery using the Thulium‐YAG Laser: A Prospective Study

et al. 2011

View full text Add to dashboard Cite

“…As reviewed in detail elsewhere, the primary purpose of matching in nested case-control studies is not to avoid confounding, but to enhance study efficiency. [2][3][4] In fact, matching can introduce a selection bias, that must be accounted for in the analysis by control of the matching factors. 4 Our goal was to provide an unbiased estimate of the relative risk of HCC that would be expected in a general population for the metabolic and inflammatory biomarkers.…”

Section: Issue In Statistical Strategy In Case-control Studymentioning

confidence: 99%

Reply

Aleksandrova¹,

Pischon

2015

Hepatology

View full text Add to dashboard Cite

Issue in Statistical Strategy In Case-Control StudyTo the Editor:We read with great interest the article by Aleksandrova and colleagues in HEPATOLOGY, 1 which gave us a new view of biomarkers of inflammation and hyperinsulinemia as predictors for hepatocellular carcinoma (HCC) among the general population. Indeed, inflammation biomarkers presented promising predictive capability in malignant disease, according to mounting evidence including our studies. 2 However, the population epidemic characteristics were not parallel among the groups. This study utilized risk-set sampling to identify the control group with paired factors including matched sex, age, date of blood collection, fasting status, time of the day at blood collection, menopausal status, and exogenous hormone use at blood donation. However, chronic hepatitis B virus (HBV) infection and chronic hepatitis C virus (HCV) infection were included as matching characteristics. It is widely accepted that chronic HBV or HCV infection is a major causes of HCC. 3 Thus, it might be more appropriate to include HBV and HCV infection in paired parameters. As a result, patients in the HCC group and the control group presented different status of HBV/HCV infection, which might lead to the different incidence rates of HCC among the population. The authors used a multivariable study to exclude the influence of the epidemic bias on their targets and reached acceptable results. However, the unparalleled patient characteristics might limit the practical value of this study.In summary, as suggested by Aleksandrova and colleagues, 1 higher circulating concentrations of interleukin-6, C-reactive protein, C-peptide, non-high-molecular weight adiponectin, and glutamate dehydrogenase were potentially associated with higher risk of HCC. Further study is needed to identify their value in practical usage.

show abstract

Why Match? Investigating Matched Case-Control Study Designs with Causal Effect Estimation

Cited by 253 publications

References 27 publications

Correlates of Suicide Among Veterans Treated in Primary Care: Case–Control Study of a Nationally Representative Sample

Correlates of Suicide Among Veterans Treated in Primary Care: Case–Control Study of a Nationally Representative Sample

Transoral Robotic Surgery using the Thulium‐YAG Laser: A Prospective Study

Reply

Contact Info

Product

Resources

About