Xiao H Zhou | PubMed Health

PubMed2007

Confidence intervals for predictive values with an emphasis to case-control studies.

The accuracy of a binary-scale diagnostic test can be represented by sensitivity (Se), specificity (Sp) and positive and negative predictive values (PPV and NPV). Although Se and Sp measure the intrinsic accuracy of a diagnostic test that does not depend on the prevalence rate, they do not provide information on the diagnostic accuracy of a particular patient. To obtain this information we need to use PPV and NPV. Since PPV and NPV are functions of both the accuracy of the test and the prevalence of the disease, constructing their confidence intervals for a particular patient is not straightforward. In this paper, a novel method for the estimation of PPV and NPV, as well as their confidence intervals, is developed. For both predictive values, standard, adjusted and their logit transformed-based confidence intervals are compared using coverage probabilities and interval lengths in a simulation study. These methods are then applied to two case-control studies: a diagnostic test assessing the ability of the e4 allele of the apolipoprotein E gene (ApoE.e4) on distinguishing patients with late-onset Alzheimer's disease (AD) and a prognostic test assessing the predictive ability of a 70-gene signature on breast cancer metastasis.

Alzheimer Disease↗

PubMed2004

Alcohol-related advice for Veterans Affairs primary care patients: Who gets it? Who gives it?

OBJECTIVE: Most patients who misuse alcohol do not receive alcohol counseling from their providers. This study evaluated primary care patient and provider characteristics associated with receipt of alcohol-related advice and whether patients were advised to drink less or to abstain. METHOD: Outpatients from seven Veterans Affairs (VA) general medicine clinics were eligible if they screened positive for alcohol misuse, completed the Alcohol Use Disorders Identification Test (AUDIT) and answered questions about alcohol-related treatment and advice. Hierarchical logistic regression was used to evaluate patient and provider characteristics associated with patient reports of alcohol-related advice from a primary care provider in the past year. RESULTS: Among 5191 patients with alcohol misuse in the past year, 1554 (30%) reported receiving alcohol-related advice from their primary care provider during that time. Of patients advised, 73% reported advice to abstain. The likelihood of reporting advice increased as AUDIT scores increased: from 13% of patients with AUDIT scores <8 to 71% of those with scores > or =20. After adjustment for important confounders, measures reflecting the severity of alcohol misuse were most strongly associated with receipt of alcohol-related advice. Adjusted analyses also revealed increased odds of receiving advice among patients who reported liver disease, hypertension, current smoking or continuity of care. No measured provider characteristic was associated with giving advice in the fully adjusted model. CONCLUSIONS: This multisite VA study found that most patients with alcohol misuse did not receive alcohol counseling from a primary care provider. Moreover, providers predominantly offered advice to abstain, and they appeared to focus on patients with the most severe problems due to drinking or medical contraindications to drinking.

Aged↗