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

Abstract

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.

Identification of Cyclase-Associated Protein-2 as a Novel Biomarker for Early- Stage Hepatocellular Carcinoma

“…For all statistical studies, P<0.05 was considered to be statistically significant. Sensitivity, specificity, and predictive values were calculated to study the overall predictability according to the following equations: [29] • …”

Reproducibility of Acute Mountain Sickness in Children and Adults: A Prospective Study

“…Predictive values and confidence intervals (CIs) were calculated as previously described. 7 Data are expressed as mean (SD). A P value Ͻ 0.05 was considered to indicate statistical significance.…”

Interval Estimates for Predictive Values in Diagnostic Testing with Three Outcomes

“…Similarly, the NPV of a test is the probability that, given a negative test result, the patient is actually disease-free. Mercaldo et al (2007) derived Wald-type interval estimates of both PPV and NPV for binary diagnostic tests in case-control studies where the true disease prevalence is assumed known. Stamey and Holt (2010) derived analogous Bayesian interval estimates that allow for uncertainty relative to disease prevalence.…”

“…Likewise n i represents the total number of people in disease state i. Mercaldo et al (2007) and Stamey and Holt (2010) described studies of this type that were not designed to reflect the true disease prevalence in the target population. The same situation is considered here.…”