Diagnostic Testing Macro

Diagnostic Testing Macro evaluates diagnostic accuracy measures for multiple diagnostic tests using individual-level data, producing publication-quality outputs with overlaid AUROC and AUPRC graphics.

Key Features:

• Multiple Test Evaluation: Analyzes multiple diagnostic tests simultaneously using individual-level data rather than summarized data.
• Comprehensive Accuracy Measures: Calculates more than 15 diagnostic accuracy measures including sensitivity, specificity, predictive values, likelihood ratios, AUROC, AUPRC, diagnostic effectiveness, disease prevalence, and diagnostic odds ratio.
• Automated Analysis: Automates the analysis process to reduce transcription errors and analysis time.
• Publication-Quality Outputs: Generates output in Microsoft Word and Excel with overlaid AUROC and AUPRC graphics.
• Variance Estimation Methods: Provides additional variance estimation methods beyond normal distribution approximation.
• Customizable Source Code: Allows SAS programmers to modify source code to include other diagnostic measures and variance estimation methods.

Scientific Applications:

• HIV Viral Load Monitoring: Evaluates multiple types of dried blood spots (DBS) from venous, microcapillary and direct spotting as alternatives to plasma for HIV viral load monitoring in resource-limited settings.
• Machine Learning Algorithm Evaluation: Assesses performance of multiple classification machine learning algorithms for predicting coronary artery disease.

Methodology:

The macro requires specification of the input dataset, "standard" and "test" variables, and threshold values. It creates a 2 x 2 summary table and calculates AUROC and AUPRC as part of output.