Early identification of sepsis in burn patients using compensatory reserve measurement: A prospective case series study

Victor A. Convertino*, Amanda R. Wagner, Kevin S. Akers, Christopher A. VanFosson, Leopoldo C. Cancio

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

2 Scopus citations


Background: Early identification of sepsis can be lifesaving, but current diagnostic approaches rely on standard vital signs and biochemistry analyses that delay recognition. We postulate that early identification of sepsis could be accomplished with a clinical assessment of real-time changes in physiological responses that reflect compensation to relative hypovolemia, or the compensatory reserve measurement (CRM). In this pilot study, we measured CRM in burn patients who transitioned from a non-infected state to sepsis to determine if CRM would be reduced prior to clinical identification of sepsis using traditional standard-of-care criteria. Methods: Each subject underwent placement of a radial arterial catheter upon arrival to the burn ICU. Morphology of analog arterial waveforms was recorded and analyzed based on a machine-learning algorithm, employing feature-extraction techniques that provided CRM values on a relative scale of 100% (normal) to 0% (onset of decompensated shock). CRM values one (D-1) or two (D-2) times prior to the day of sepsis diagnosis (D+0) were compared in time to standard diagnostic techniques. Results: CRM was lower in all patients on D+0 compared to ICU admission day. Complete analyzable arterial waveform data collected on D-1 and D-2 were captured from 5 subjects. These patients demonstrated an average reduction (p = 0.003) in CRM from 48 ± 14% on D-2 to 31 ± 14% on D-1 prior to sepsis diagnosis with standard criteria on D+0. Conclusion: The results of this clinical investigation conducted in burn patients provide the first data to substantiate a capability for early diagnosis of sepsis by measurement of the compensatory reserve. Our findings support the notion that this novel technology could offer caregivers of at-risk burn patients with a user-friendly decision-support indicator of infection for individualized triage and treatment. Level of evidence: Diagnostic, level IV.

Original languageEnglish
Pages (from-to)137-145
Number of pages9
JournalBurns Open
Issue number4
StatePublished - Oct 2022
Externally publishedYes


  • Burns
  • Infection
  • Machine-learning algorithm
  • Monitoring
  • Sepsis


Dive into the research topics of 'Early identification of sepsis in burn patients using compensatory reserve measurement: A prospective case series study'. Together they form a unique fingerprint.

Cite this