Is heart period variability associated with the administration of lifesaving interventions in individual prehospital trauma patients with normal standard vital signs?

Caroline A. Rickards, Kathy L. Ryan, David Ludwig, Victor A. Convertino

Research output: Contribution to journalArticle

23 Citations (Scopus)

Abstract

Objective: To determine whether heart period variability provides added value in identifying the need for lifesaving interventions (LSI) in individual trauma patients with normal standard vital signs upon early medical assessment. Design: Retrospective database review. Setting: Helicopter transport to Level 1 trauma center and first 24 hrs of in-hospital care. Patients: Prehospital trauma patients requiring helicopter transport to Level 1 trauma center. Measurements and Main Results: Heart period variability was analyzed from electrocardiographic recordings collected from 159 prehospital trauma patients with normal standard vital signs (32 LSI patients, 127 No-LSI patients). Although 13 of the electrocardiogram derived metrics demonstrated simple (i.e., univariate) discrimination between groups, at the multivariate level, only fractal dimension by curve length (FD-L) was uniquely associated with group membership (LSI vs. No-LSI, p =.0004). Whereas the area under the receiver operating characteristics curve for FD-L was 0.70, the overall correct classification rate (true positives and true negatives) of 82% was only 2% higher than the baseline prediction rate of 80% (i.e., no information except for the known proportion of overall No-LSI cases, 127 of 159 patients). Furthermore, 84% of the individual FD-L values for the LSI group were within the range of the No-LSI group. Conclusions: Only FD-L was uniquely able to distinguish patient groups based on mean values when standard vital signs were normal. However, the accuracy of FD-L in distinguishing between patients was only slightly better than the baseline prediction rate. There was also very high overlap of individual heart period variability values between groups, so many LSI patients could be incorrectly classified as not requiring an LSI if a single heart period variability value was used as a triage tool. Based on this analysis, heart period variability seems to have limited value for prediction of LSIs in prehospital trauma patients with normal standard vital signs.

Original languageEnglish
Pages (from-to)1666-1673
Number of pages8
JournalCritical Care Medicine
Volume38
Issue number8
DOIs
StatePublished - Aug 1 2010

Fingerprint

Vital Signs
Wounds and Injuries
Aircraft
Trauma Centers
Fractals
Triage
ROC Curve
Electrocardiography
Demography
Databases

Keywords

  • decision support
  • heart period variability
  • heart rate
  • R-R interval
  • trauma
  • triage

ASJC Scopus subject areas

  • Critical Care and Intensive Care Medicine

Cite this

Is heart period variability associated with the administration of lifesaving interventions in individual prehospital trauma patients with normal standard vital signs? / Rickards, Caroline A.; Ryan, Kathy L.; Ludwig, David; Convertino, Victor A.

In: Critical Care Medicine, Vol. 38, No. 8, 01.08.2010, p. 1666-1673.

Research output: Contribution to journalArticle

Rickards, Caroline A. ; Ryan, Kathy L. ; Ludwig, David ; Convertino, Victor A. / Is heart period variability associated with the administration of lifesaving interventions in individual prehospital trauma patients with normal standard vital signs?. In: Critical Care Medicine. 2010 ; Vol. 38, No. 8. pp. 1666-1673.
@article{6fb0d9dac5a6442fa545bffa72912ebe,
title = "Is heart period variability associated with the administration of lifesaving interventions in individual prehospital trauma patients with normal standard vital signs?",
abstract = "Objective: To determine whether heart period variability provides added value in identifying the need for lifesaving interventions (LSI) in individual trauma patients with normal standard vital signs upon early medical assessment. Design: Retrospective database review. Setting: Helicopter transport to Level 1 trauma center and first 24 hrs of in-hospital care. Patients: Prehospital trauma patients requiring helicopter transport to Level 1 trauma center. Measurements and Main Results: Heart period variability was analyzed from electrocardiographic recordings collected from 159 prehospital trauma patients with normal standard vital signs (32 LSI patients, 127 No-LSI patients). Although 13 of the electrocardiogram derived metrics demonstrated simple (i.e., univariate) discrimination between groups, at the multivariate level, only fractal dimension by curve length (FD-L) was uniquely associated with group membership (LSI vs. No-LSI, p =.0004). Whereas the area under the receiver operating characteristics curve for FD-L was 0.70, the overall correct classification rate (true positives and true negatives) of 82{\%} was only 2{\%} higher than the baseline prediction rate of 80{\%} (i.e., no information except for the known proportion of overall No-LSI cases, 127 of 159 patients). Furthermore, 84{\%} of the individual FD-L values for the LSI group were within the range of the No-LSI group. Conclusions: Only FD-L was uniquely able to distinguish patient groups based on mean values when standard vital signs were normal. However, the accuracy of FD-L in distinguishing between patients was only slightly better than the baseline prediction rate. There was also very high overlap of individual heart period variability values between groups, so many LSI patients could be incorrectly classified as not requiring an LSI if a single heart period variability value was used as a triage tool. Based on this analysis, heart period variability seems to have limited value for prediction of LSIs in prehospital trauma patients with normal standard vital signs.",
keywords = "decision support, heart period variability, heart rate, R-R interval, trauma, triage",
author = "Rickards, {Caroline A.} and Ryan, {Kathy L.} and David Ludwig and Convertino, {Victor A.}",
year = "2010",
month = "8",
day = "1",
doi = "10.1097/CCM.0b013e3181e74cab",
language = "English",
volume = "38",
pages = "1666--1673",
journal = "Critical Care Medicine",
issn = "0090-3493",
publisher = "Lippincott Williams and Wilkins",
number = "8",

}

TY - JOUR

T1 - Is heart period variability associated with the administration of lifesaving interventions in individual prehospital trauma patients with normal standard vital signs?

AU - Rickards, Caroline A.

AU - Ryan, Kathy L.

AU - Ludwig, David

AU - Convertino, Victor A.

PY - 2010/8/1

Y1 - 2010/8/1

N2 - Objective: To determine whether heart period variability provides added value in identifying the need for lifesaving interventions (LSI) in individual trauma patients with normal standard vital signs upon early medical assessment. Design: Retrospective database review. Setting: Helicopter transport to Level 1 trauma center and first 24 hrs of in-hospital care. Patients: Prehospital trauma patients requiring helicopter transport to Level 1 trauma center. Measurements and Main Results: Heart period variability was analyzed from electrocardiographic recordings collected from 159 prehospital trauma patients with normal standard vital signs (32 LSI patients, 127 No-LSI patients). Although 13 of the electrocardiogram derived metrics demonstrated simple (i.e., univariate) discrimination between groups, at the multivariate level, only fractal dimension by curve length (FD-L) was uniquely associated with group membership (LSI vs. No-LSI, p =.0004). Whereas the area under the receiver operating characteristics curve for FD-L was 0.70, the overall correct classification rate (true positives and true negatives) of 82% was only 2% higher than the baseline prediction rate of 80% (i.e., no information except for the known proportion of overall No-LSI cases, 127 of 159 patients). Furthermore, 84% of the individual FD-L values for the LSI group were within the range of the No-LSI group. Conclusions: Only FD-L was uniquely able to distinguish patient groups based on mean values when standard vital signs were normal. However, the accuracy of FD-L in distinguishing between patients was only slightly better than the baseline prediction rate. There was also very high overlap of individual heart period variability values between groups, so many LSI patients could be incorrectly classified as not requiring an LSI if a single heart period variability value was used as a triage tool. Based on this analysis, heart period variability seems to have limited value for prediction of LSIs in prehospital trauma patients with normal standard vital signs.

AB - Objective: To determine whether heart period variability provides added value in identifying the need for lifesaving interventions (LSI) in individual trauma patients with normal standard vital signs upon early medical assessment. Design: Retrospective database review. Setting: Helicopter transport to Level 1 trauma center and first 24 hrs of in-hospital care. Patients: Prehospital trauma patients requiring helicopter transport to Level 1 trauma center. Measurements and Main Results: Heart period variability was analyzed from electrocardiographic recordings collected from 159 prehospital trauma patients with normal standard vital signs (32 LSI patients, 127 No-LSI patients). Although 13 of the electrocardiogram derived metrics demonstrated simple (i.e., univariate) discrimination between groups, at the multivariate level, only fractal dimension by curve length (FD-L) was uniquely associated with group membership (LSI vs. No-LSI, p =.0004). Whereas the area under the receiver operating characteristics curve for FD-L was 0.70, the overall correct classification rate (true positives and true negatives) of 82% was only 2% higher than the baseline prediction rate of 80% (i.e., no information except for the known proportion of overall No-LSI cases, 127 of 159 patients). Furthermore, 84% of the individual FD-L values for the LSI group were within the range of the No-LSI group. Conclusions: Only FD-L was uniquely able to distinguish patient groups based on mean values when standard vital signs were normal. However, the accuracy of FD-L in distinguishing between patients was only slightly better than the baseline prediction rate. There was also very high overlap of individual heart period variability values between groups, so many LSI patients could be incorrectly classified as not requiring an LSI if a single heart period variability value was used as a triage tool. Based on this analysis, heart period variability seems to have limited value for prediction of LSIs in prehospital trauma patients with normal standard vital signs.

KW - decision support

KW - heart period variability

KW - heart rate

KW - R-R interval

KW - trauma

KW - triage

UR - http://www.scopus.com/inward/record.url?scp=77954954886&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77954954886&partnerID=8YFLogxK

U2 - 10.1097/CCM.0b013e3181e74cab

DO - 10.1097/CCM.0b013e3181e74cab

M3 - Article

C2 - 20562705

AN - SCOPUS:77954954886

VL - 38

SP - 1666

EP - 1673

JO - Critical Care Medicine

JF - Critical Care Medicine

SN - 0090-3493

IS - 8

ER -