Diagnostic Accuracy of Tryptase Levels for Pediatric Anaphylaxis: A Case-Control Study.

Introduction: Anaphylaxis is a severe allergic reaction which can be difficult to diagnose. Two strategies evaluating changes in tryptase levels were proposed for diagnosing anaphylaxis. Strategy 1 established a threshold of tryptase levels during reaction exceeding 2 ng/mL + 1.2* (baseline tryptase levels) as a rule for detecting anaphylaxis, while strategy 2 established the ratio of tryptase levels during reaction versus baseline tryptase exceeding a threshold of 1.685. We aimed to compare the diagnostic test accuracy of the two strategies in pediatric anaphylaxis.

Methods: We conducted a case-control study. Cases consisted of 89 patients with anaphylaxis who had reaction tryptase and subsequent baseline tryptase measured. Controls consisted of 25 patients with chronic urticaria who had two tryptase measurements. Sensitivity and specificity for each of the strategies were computed and compared using McNemar test. The area under the curve (AUC) between the two strategies was compared using the DeLong test.

Results: The sensitivity and specificity for strategy 1 was 53.3% and 95.0%, respectively. For strategy 2, the sensitivity and specificity was 54.4% and 85.0%, respectively. There was no significant difference between both strategies' sensitivity and specificity. The Delong test determined that the AUC was significantly (p < 0.05) higher for strategy 1 (0.69) than strategy 2 (0.64).

Conclusion: The Delong test determined that strategy 1 was slightly better in validating anaphylaxis diagnosis than strategy 2. However, both strategies demonstrated a low sensitivity <55%.