DIAGNOSTIC ACCURACY OF POINT-OF-CARE ULTRASOUND IN ACUTE ABDOMINAL PAIN AT MARDAN MEDICAL COMPLEX: A CROSS-SECTIONAL VALIDATION STUDY

Abstract

Background: Acute abdominal pain is a frequent emergency-department presentation for which rapid diagnostic stratification is essential. Point-of-care ultrasound (POCUS) offers bedside, non-invasive, radiation-free imaging and may be particularly useful in resource-constrained settings.

Objective: To evaluate the diagnostic performance of POCUS in adults presenting with acute non-traumatic abdominal pain to the Emergency Department of Mardan Medical Complex, Pakistan.

Methods: This cross-sectional validation study enrolled 237 consecutive adults from August 2025 to February 2026. POCUS was performed by emergency physicians with at least two years of ultrasound experience using a portable Mindray M7® system. Findings were compared with a composite reference standard comprising radiology-performed ultrasonography, contrast-enhanced computed tomography when clinically indicated, surgical and histopathological findings, and seven-day clinical follow-up. Diagnostic parameters were recalculated from the supplied 2×2 contingency counts with Wilson 95% confidence intervals (CIs). Pearson chi-square tests were used for categorical comparisons.

Results: The mean age was 38.2 ± 12.7 years and 128/237 (54.0%) participants were male. Appendicitis (58/237, 24.5%), renal colic (45/237, 19.0%), and cholecystitis (40/237, 16.9%) were the most frequent final diagnoses. In the supplied overall binary cross-tabulation, POCUS yielded 153 true positives, 41 false positives, 24 false negatives, and 19 true negatives. Accordingly, sensitivity was 86.4% (95% CI 80.6–90.7), specificity 31.7% (21.3–44.2), positive predictive value 78.9% (72.6–84.0), negative predictive value 44.2% (30.4–58.9), and overall accuracy 72.6% (66.6–77.9). POCUS and the reference classification were significantly associated (Pearson χ²=9.893, p=0.002). Condition-specific accuracy was 91.6% for appendicitis, 94.1% for cholecystitis, 92.4% for renal colic, and 94.9% for intestinal obstruction. POCUS identified 57/67 (85.1%; 95% CI 74.7–91.7) patients requiring urgent surgery. Diagnostic correctness did not differ by sex (χ²=0.044, p=0.834).

Conclusion: POCUS showed high sensitivity for the overall target-pathology classification and strong condition-specific performance for common abdominal emergencies; however, the low overall specificity and negative predictive value in the supplied binary dataset indicate that POCUS should not be used as a stand-alone rule-out test. Integration with clinical assessment and confirmatory imaging remains important when diagnostic uncertainty persists.