%diag_test: a generic SAS macro for evaluating diagnostic accuracy measures for multiple diagnostic tests.

Area of Science:

• Biostatistics
• Medical Informatics
• Health Services Research

Background:

• Diagnostic test accuracy measures (DAMs) like sensitivity, specificity, AUROC, and AUPRC are crucial for evaluating medical tests.
• Existing analysis tools often focus on single diagnostic tests using summarized data.
• There is a need for efficient methods to analyze multiple diagnostic tests using individual-level data.

Purpose of the Study:

• To develop and present a SAS macro for the comprehensive evaluation of multiple diagnostic tests.
• To automate the analysis of individual-level diagnostic test data, reducing time and errors.
• To provide researchers with publication-quality outputs, including various accuracy measures and graphical representations.

Main Methods:

• A SAS macro was developed to process individual-level data for diagnostic test evaluation.
• The macro automates the creation of 2x2 summary tables, AUROC, and AUPRC.
• It requires users to specify the input dataset, standard and test variables, and threshold values.

Main Results:

• The macro was validated by reproducing published results for dried blood spot (DBS) testing for HIV viral load monitoring.
• It was also used to replicate findings on machine learning algorithms for coronary artery disease prediction.
• The output includes over 15 accuracy measures and overlaid AUROC/AUPRC graphics.

Conclusions:

• The SAS macro is a powerful tool for analyzing multiple diagnostic tests, enhancing efficiency and accuracy.
• Automation of analysis saves time, minimizes transcription errors, and yields publication-ready results.
• The macro's source code can be modified to incorporate additional diagnostic measures and variance estimation methods.