Sequential test selection in the analysis of abdominal pain

Area of Science:

• Medical Decision Making
• Health Informatics
• Diagnostic Imaging

Background:

• Physician decision-making tools often lack methods to incorporate ambiguous clinical information.
• Existing analytical models do not fully integrate physician judgment with probability analysis.
• Accurate diagnosis management requires tools that handle uncertain data effectively.

Observation:

• A case study involved a patient with upper abdominal pain requiring sequential diagnostic testing.
• Physicians evaluated four tests: upper gastrointestinal series (GI), abdominal ultrasonography (US), abdominal computed tomography (CT), and upper gastrointestinal endoscopy (END).
• Test selection criteria included minimizing risk, discomfort, and cost, while maximizing diagnostic capability.

Findings:

• The Analytic Hierarchy Process (AHP) model identified the upper GI series as the optimal first diagnostic test.
• Bayes' theorem was used to update diagnostic probabilities after initial test results.
• Subsequent AHP analysis, incorporating updated probabilities, recommended abdominal ultrasonography as the optimal second test.

Implications:

• This integrated model enhances diagnostic accuracy by combining subjective physician judgment with objective probability analysis.
• The sequential testing approach optimizes resource allocation and patient management in complex diagnostic scenarios.
• Sensitivity analysis highlights the dynamic nature of optimal test sequencing based on evolving clinical data and criteria preferences.