Instrumental variables in the evaluation of diagnostic test procedures when the true disease state is unknown

Abstract

We explore the estimation of sensitivity and specificity of diagnostic tests when the true disease state is unknown. Instrumental variables which subdivide the patient population are used. A logistic model, relating these instrumental variables to the (unknown) true disease state is proposed. It is shown that this procedure allows the goodness-of-fit to the resulting model to be tested.

Measurement of circulating transcript levels (NETest) to detect disease recurrence and improve follow‐up after curative surgical resection of well‐differentiated pancreatic neuroendocrine tumors

“…were derived to assess the strength of the association or "relatedness" of each factor or combination of factors to recurrence. 37 Predictive feature importance analysis (FIA) was undertaken to define the "importance value" for each factor (biomarker or clinical criterion) alone or in combination. Importance values were derived using a random forest approach that evaluates the output from decision tree algorithms used to define the relationship of a variable for example, a biomarker, to an output for example, recurrence.…”

“…Multiple regression and logistic regression analyses were undertaken to identify which parameters were associated with recurrence. The odds ratio (OR), χ 2 value, and Nagelkerke R 2 coefficient (coefficient of determination) were derived to assess the strength of the association or “relatedness” of each factor or combination of factors to recurrence . Predictive feature importance analysis (FIA) was undertaken to define the “importance value” for each factor (biomarker or clinical criterion) alone or in combination.…”

13C-Urea breath test threshold calculation and evaluation for the detection of Helicobacter pylori infection in children

“…The resulting threshold Δδ values were evaluated in the complementary 50% of the data on the basis of the observed H. pylori detection rate. Coefficients of determination (R 2 ) were obtained according to [21]. …”

Simultaneous inference of a misclassified outcome and competing risks failure time data

“…In model (1) for the IBTR status, the covariate vector x P (i.e., age, distant recurrence, and tumor stage in the MDACC dataset) plays the role of instrumental variables and it makes the sensitivity and specificity parameters identifiable when it has sufficient numbers of different realizations 12 . In addition, the survival information included in model (6) is an important determinant of IBTR status, manifested by the clear dichotomy in the cumulative incidence curves in Figures 1 and 2.…”

“…To this end, Nagelkerke et al 12 suggested modeling the unobserved true disease status as a function of an instrumental variable, which is an additional parameter to increase the outcome degrees of freedom. This instrumental variable framework can be extended to utilize the available correlated survival time to gain additional information regarding the disease status classification.…”