Background: Compartmental syndrome (CS) is difficult to diagnose in intensive care unit patients. Compartment perfusion pressure (CPP) is an invasive, indirect measure of ischemia. Near-infrared spectroscopy is noninvasive, and directly measures ischemia by transmitting light through tissues at wavelengths that react with hemoglobin to provide percent tissue oxygen saturation (Sto(2)). Animal studies demonstrate that Sto(2) is superior to CPP for detecting CS. However, there are no studies in humans comparing Sto(2) with CPP. We hypothesized that Sto(2) can reliably detect CS, and is superior to CPP.
Methods: CS was induced in 15 human volunteers using a standard calf compression model. At 30-minute intervals, compression was increased to reduce Sto(2) from baseline (86% +/- 4%) to 60%, 40%, 20%, and < 10%, with simultaneous recording of CPP. Outcome variables included deep peroneal nerve conduction assessed by electromyography, cutaneous peroneal nerve sensitivity using Semmes-Weinstein monofilaments, and pain (visual analog scale).
Results: Both Sto(2) and CPP significantly correlated with all ischemia outcome variables (p < 0.001). Receiver operating characteristic curves of deep peroneal nerve conduction demonstrated that Sto(2) had higher sensitivity than CPP for detecting > 50% block. For example, when specificity was 83% for Sto(2) and 84% for CPP, sensitivity was 85% versus 56%, respectively (p = 0.02). When specificity for both was 72%, sensitivity was 94% for Sto(2) versus 76% for CPP (p = 0.04).
Conclusion: In intensive care unit patients who cannot alert physicians to symptoms, near-infrared spectroscopy may help clinicians to avoid delayed or unnecessary prophylactic fasciotomy, and provides the benefits of a continuous, noninvasive monitoring technique.