Rethinking the Paradigm of Using Ps for Diagnosing Compartment Syndrome.

Abstract

Background: To evaluate the predictive power of 7 clinical signs and symptoms associated with acute compartment syndrome (ACS) of the leg, namely pain, paresthesia, paralysis, pallor, poikilothermia, pulselessness, and pressure on palpation (7P's).

Methods: Retrospective data of 357 patients were obtained from the databases of 5-level one trauma centers in Canada, the United States, and France. Inclusion criteria were patients with tibia injuries that received fasciotomies in adults with documented serial clinical assessments. All possible combinations of signs/symptoms used were generated. The combinations were tested for predictive power using 2 machine learning algorithms.

Results: Pressure on palpation was the strongest clinical predictor of ACS while pain was the weakest. Using any single P to assess for ACS yields a poor prediction. Increasing the number of Ps improves the performance up to 4Ps, regardless of the composition of the combination. None of the combinations had a perfect predictive power which means that the use of single or multiple Ps does not guarantee diagnosis. Predictive performance indicated that poikilothermia, pallor, and paralysis are not significantly informative.

Conclusion: The presence of specific patterns of clinical signs/symptoms associated with ACS seems to influence a surgeon's decision to perform fasciotomy. These data question the gold standard of clinical signs for diagnosis of ACS. The reliance on the Ps classically taught in medical school does not seem to be sufficient for accurate diagnosis. Objective measures such as continuous pressure or a physiologic marker of ischemia may be better indications for compartment syndrome.

Level of evidence: Level III. See Instructions for Authors for a complete description of levels of evidence.