A retrospective evaluation of the risk of bias in perioperative temperature metrics
- 33 Downloads
The prevention and treatment of hypothermia is an important part of routine anesthesia care. Avoidance of perioperative hypothermia was introduced as a quality metric in 2010. We sought to assess the integrity of the perioperative hypothermia metric in routine care at a single large center. Perioperative temperatures from all anesthetics of at least 60 min duration between January 2012 and 2017 were eligible for inclusion in analysis. Temperatures were displayed graphically, assessed for normality, and analyzed using paired comparisons. Automatically-recorded temperatures were obtained from several monitoring sites. Provider-entered temperatures were non-normally distributed, exhibiting peaks at temperatures at multiples of 0.5 °C. Automatically-acquired temperatures, on the other hand, were more normally distributed, demonstrating smoother curves without peaks at multiples of 0.5 °C. Automatically-acquired median temperature was highest, 36.8 °C (SD = 0.8 °C), followed by the three manually acquired temperatures (nurse-documented postoperative temperature, 36.5 °C [SD = 0.6 °C]; intraoperative manual temperature, 36.5 °C [SD = 0.6 °C]; provider-documented postoperative temperature, 36.1 °C [SD = 0.6 °C]). Provider-entered temperatures exhibit values that are unlikely to represent a normal probability distribution around a central physiologic value. Manually-entered perioperative temperatures appear to cluster around salient anchoring values, either deliberately, or as an unintended result driven by cognitive bias. Automatically-acquired temperatures may be superior for quality metric purposes.
KeywordsPerioperative informatics Hypothermia Quality improvement Electronic health records Temperature
Patrick Jablonski, Ph.D., provided statistical recommendations to assist in the preparation and refinement of the statistical analysis.
Dr. Freundlich receives Grant support from an NIH-NCATS KL2 (KL2 TR002245).
Compliance with ethical standards
Conflict of interest
Dr. Freundlich has received grant support and consulting fees from Medtronic for work unrelated to the content of this manuscript.
- 4.Poveda VB, Nascimento AS. The effect of intraoperative hypothermia upon blood transfusion needs and length of stay among gastrointestinal system cancer surgery. Eur J Cancer Care (Engl), 2017. 26(6).Google Scholar
- 7.Alderson P, et al. Thermal insulation for preventing inadvertent perioperative hypothermia. Cochrane Database Syst Rev, 2014(6): Cd009908.Google Scholar
- 9.PQRS Measure #193: Perioperative Temperature Management. 2015. https://www.pqrspro.com/cmsmeasures/2015/perioperative_temperature_management/. Accessed 25 Mar 2017.
- 10.PQRS Measure #424: Perioperative Temperature Management. 2016. https://www.pqrspro.com/cmsmeasures/2016/perioperative_temperature_management-2/. Accessed 25 Mar 2017.
- 11.Medicare CF, Services M. Physician quality reporting system. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/PQRS/index.html. Accessed 21 June 2012.
- 16.Phelps M, et al. Comparison of minute distribution frequency for anesthesia start and end times from an anesthesia information management system and paper records. J Clin Monit Comput. 2016: 31: 1–6.Google Scholar