Skin-temperature gradients are a validated measure of fingertip perfusion

House and Tipton’s evaluation of forearm-minus-fingertip, skin-temperature gradients (House and Tipton 2002) is predicated on the assumption that ‘‘the Tsk-diff method has not been validated’’. In fact, the method has been extensively validated (Rubinstein and Sessler 1990). There is an excellent relationship between gradients and fingertip volume plethysmography – which is widely considered the gold standard for arterio-venous shunt perfusion (Fig. 1). This article has been cited more than 130 times and is well known among thermoregulatory researchers. A subsequent article demonstrated that a virtually identical relationship between plethysmography and forearm-minus-fingertip gradients remains even during anesthesia (Sessler et al. 1992). A major purpose of measuring fingertip blood flow is to determine the threshold for vasoconstriction (triggering core temperature). House and Tipton make the point that skin-temperature gradients depend critically on fingertip cooling, which takes longer than real-time assessments of finger flow. There is thus some danger that the slow response of gradients may result in an artifactually low threshold. However, the original validation article (Rubinstein and Sessler 1990) contained a detailed analysis of the method’s rate sensitivity and showed that the time constant for finger cooling is only 6.6 (1.2) min. Thresholds determined with skin-temperature gradients are thus reliable unless the core-cooling rate is extraordinarily rapid. House and Tipton take us to task for using a skintemperature gradient of 4 C in our initial studies. Those studies were the first to demonstrate thermoregulatory responses in anesthetized patients. We thus intentionally required severe vasoconstriction to avoid improperly concluding that mild vasoconstriction (due to stress or hypovolemia, for example) was a specific thermoregulatory defense. However, we recognize that thermoregulatory thresholds are generally characterized by response initiation. A gradient of 0 C indicates the initiation of vasoconstriction; furthermore, we have shown that a gradient of 0 C corresponds to the core-temperature plateau that results from thermoregulatory vasoconstriction (Kurz et al. 1995a). In dozens of subsequent publications, others and we have thus taken a gradient of 0 C to indicate the initiation of thermoregulatory vasoconstriction. Figures similar to those of House and Tipton have been published previously. Plethysmographically determined flow, for example, starts to decrease almost exactly at a skin-temperature gradient of 0 C (Sessler et al. 1992). There is an excellent relationship between skin-temperature gradients and the perfusion index, a measurement based on pulse oximetry (Kurz et al. 1995b). Similarly, there is an excellent relationship between skin-temperature gradients and laser Doppler flowmetry (Kurz et al. 1995b). Even the relationship between gradients and distal cutaneous heat flux has previously been published (Sessler et al. 1992). The average age of the references that House and Tipton cite is 24 years, and only one reference is more recent than 1988. This leads to a curious lack of detail in the Introduction to their article. For example, they cite a paper from 1961 and a review from 1983 to support the statement that anesthesia ‘‘suppresses thermoregulatory responses such as shivering and vasoconstriction.’’ In fact, since 1995, the dose-dependent effects of all major intravenous (Matsukawa et al. 1995) and volatile (Annadata et al. 1995) anesthetics on sweating, vasoconstriction, and shivering thresholds have been quantified. Furthermore, the effects of volatile anesthetics on the Eur J Appl Physiol (2003) 89: 401–402 DOI 10.1007/s00421-003-0812-8