Letters to the Editor: Letter to the Editor
Thank for your interest in our article.1 In response, regarding the cases excluded because of missing values, as noted in the Limitations section of our article, both transfusion and mortality rates were higher in the excluded group. Among the 480 excluded patients, 60 (12.5%) received packed red blood cells (pRBC) within 3 hours, 77 (16.0%) received pRBC within 6 hours, and 84 (17.5%) received pRBC within 12 hours. One patient was among the 34 who had incomplete laboratory data, and 1 of the 34 patients was also among the 48 (10.0%) patients who died. Because the priority is to save patients’ lives rather than gather additional data (all these patients already had pulse oximetry monitored by a separate sensor), the pulse oximeter-derived hemoglobin (SpHb) sensors and finger shield were not placed if there were concerns about interruption of emergent patient care. In some instances, SpHb sensors were able to be placed in such unstable trauma patients, but these data were only intermittent (maybe because of inadequate time to ensure correct sensor and shield placement) and these data were not included. Our study cohort therefore represents a population of trauma patients whose signals were sufficient to potentially benefit from the SpHb monitoring. Increasing sensor stability and ease of placement in such a resuscitation environment would improve data collection rate and could potentially have strengthened the study conclusions.
Rice et al2 appear to have misread our article; we did not exclude patients with shock index (SI) >0.62, as they comment. As stated in the Abstract and Methods of the article, “we enrolled direct trauma patient admissions ≥18 years with prehospital shock index (SI) ≥0.62.” The enrollment cutoff for SI using prehospital SI ≥0.623 to enroll trauma patients in mild shock and SI ≥0.62 in patients donating blood4 has been used by others. Such a cutoff had an overall intent to ensure that enrollment criteria focused on the group who need transfusion prediction most and included hemorrhaging patients but excluded patients with no shock.
Rice et al2 report data collected by Macknet et al5 in 20 volunteers undergoing hemodilution and conclude that SpHb for such patients in the range of hemoglobin 6 to 10 g/dL does not help guide transfusion decisions. We agree that few of our patients had hemoglobin levels in this 6 to 10 g/dL range. Transfusion decisions are quite different in unstable trauma patients and more difficult to make than for elective hemodiluted or elective surgical patients because of uncertainty about the site and the extent of bleeding. With our report of 667 unstable trauma patients, we show that SpHb oximetry was no better than a transfusion prediction algorithm based on the photoplethysmograph of a conventional pulse oximeter requiring no special sensor. We have also previously shown that the conventional pulse oximeter transfusion prediction algorithm was equally as good or better than expert clinical judgment on the need for life-saving interventions during trauma patient resuscitation.6 To address Rice et al letter title statement specifically, our study does not provide evidence that SpHb adds value.
Colin F. Mackenzie, MBChBShiming Yang, PhDPeter F. Hu, PhDDepartment of AnesthesiologyUniversity of Maryland School of MedicineBaltimore, Marylandcmack003@umaryland.edu
1. Yang S, Hu P, Anazoda A, Gao C, et al. Comparison of continuous noninvasive hemoglobin and laboratory blood tests in the prediction of blood transfusion. Anesth Analg. 2016;122:115125.
2. Rice MJ, Gravenstein N, Morey TE. Noninvasive hemoglobin monitoring: how accurate is enough? Anesth Analg. 2013;117:902907.
3. Mutschler M, Nienaber U, Munzberg M, et al.; The TraumaRegisterDGU. The Shock Index revisited—a fast guide to transfusion requirement? A retrospective analysis on 21,853 patients derived from the Trauma Register DGU. Crit Care. 2013;17:R 172.
4. Zarzaur BL, Croce MA, Fischer PE, Magnotti LJ, Fabian TC. New vitals after injury: shock index for the young and age x shock index for the old. J Surg Res. 2008;147:229236.
5. Macknet MR, Allard M, Applegate RL 2nd, Rook JThe accuracy of noninvasive and continuous total hemoglobin measurement by pulse CO-Oximetry in human subjects undergoing hemodilution. Anesth Analg. 2010;111:14241426.
6. Mackenzie CF, Gao C, Hu PF, et al.; ONPOINT Study Group. Comparison of decision-assist and clinical judgment of experts for prediction of lifesaving interventions. Shock. 2015;43:238243.