Header

UZH-Logo

Maintenance Infos

How to detect a polytrauma patient at risk of complications: a validation and database analysis of four published scales


Halvachizadeh, Sascha; Baradaran, Larissa; Cinelli, Paolo; Pfeifer, Roman; Sprengel, Kai; Pape, Hans-Christoph (2020). How to detect a polytrauma patient at risk of complications: a validation and database analysis of four published scales. PLoS ONE, 15(1):e0228082.

Abstract

INTRODUCTION: Early accurate assessment of the clinical status of severely injured patients is crucial for guiding the surgical treatment strategy. Several scales are available to differentiate between risk categories. They vary between expert recommendations and scores developed on the basis of patient data (level II). We compared four established scoring systems in regard to their predictive abilities for early (e.g., hemorrhage-induced mortality) versus late (Multiple Organ Failure (MOF), sepsis, late death) in-hospital complications.
METHODS: A database from a level I trauma center was used. The inclusion criteria implied an injury severity score (ISS) of ≥16 points, primary admission, and a complete data set from admission to hospital-day 21. The following four scales were tested: the clinical grading scale (CGS; covers acidosis, shock, coagulation, and soft tissue injuries), the modified clinical grading scale (mCGS; covers CGS with modifications), the polytrauma grading score (PTGS; covers shock, coagulation, and ISS), and the early appropriate care protocol (EAC; covers acid-base changes). Admission values were selected from each scale and the following endpoints were compared: mortality, pneumonia, sepsis, death from hemorrhagic shock, and multiple organ failure.
STATISTICS: Shapiro-Wilk test for normal distribution, Pearson Chi square, odds ratios (OR) for all endpoints, 95% confidence intervals. Fitted, generalized linear models were used for prediction analysis. Krippendorff was used for comparison of CGS and mCGS. Alpha set at 0.05.
RESULTS: In total, 3668 severely injured patients were included (mean age, 45.8±20 years; mean ISS, 28.2±15.1 points; incidence of pneumonia, 19.0%; incidence of sepsis, 14.9%; death from hem. shock, 4.1%; death from multiple organ failure (MOF), 1.9%; mortality rate, 26.8%). Our data show distinct differences in the prediction of complications, including mortality, for these scores (OR ranging from 0.5 to 9.1). The PTGS demonstrated the highest predictive value for any late complication (OR = 2.0), sepsis (OR = 2.6, p = 0.05), or pneumonia (OR = 2.0, p = 0.2). The EAC demonstrated good prediction for hemorrhage-induced early mortality (OR = 7.1, p<0.0001), but did not predict late complications (sepsis, OR = 0.8 and p = 0.52; pneumonia, OR = 1.1 and p = 0.7) CGS and mCGS are not comparable and should not be used interchangeably (Krippendorff α = 0.045).
CONCLUSION: Our data show that prediction of complications is more precise after using values that covers different physiological systems (coagulation, hemorrhage, acid-base changes, and soft tissue damage) when compared with using values of only one physiological system (e.g., acidosis). When acid-base changes alone were tested in terms of complications, they were predictive of complications within 72 hours but failed to predict late complications. These findings should be considered when performing early assessment of trauma patients or for the development of new scores.

Abstract

INTRODUCTION: Early accurate assessment of the clinical status of severely injured patients is crucial for guiding the surgical treatment strategy. Several scales are available to differentiate between risk categories. They vary between expert recommendations and scores developed on the basis of patient data (level II). We compared four established scoring systems in regard to their predictive abilities for early (e.g., hemorrhage-induced mortality) versus late (Multiple Organ Failure (MOF), sepsis, late death) in-hospital complications.
METHODS: A database from a level I trauma center was used. The inclusion criteria implied an injury severity score (ISS) of ≥16 points, primary admission, and a complete data set from admission to hospital-day 21. The following four scales were tested: the clinical grading scale (CGS; covers acidosis, shock, coagulation, and soft tissue injuries), the modified clinical grading scale (mCGS; covers CGS with modifications), the polytrauma grading score (PTGS; covers shock, coagulation, and ISS), and the early appropriate care protocol (EAC; covers acid-base changes). Admission values were selected from each scale and the following endpoints were compared: mortality, pneumonia, sepsis, death from hemorrhagic shock, and multiple organ failure.
STATISTICS: Shapiro-Wilk test for normal distribution, Pearson Chi square, odds ratios (OR) for all endpoints, 95% confidence intervals. Fitted, generalized linear models were used for prediction analysis. Krippendorff was used for comparison of CGS and mCGS. Alpha set at 0.05.
RESULTS: In total, 3668 severely injured patients were included (mean age, 45.8±20 years; mean ISS, 28.2±15.1 points; incidence of pneumonia, 19.0%; incidence of sepsis, 14.9%; death from hem. shock, 4.1%; death from multiple organ failure (MOF), 1.9%; mortality rate, 26.8%). Our data show distinct differences in the prediction of complications, including mortality, for these scores (OR ranging from 0.5 to 9.1). The PTGS demonstrated the highest predictive value for any late complication (OR = 2.0), sepsis (OR = 2.6, p = 0.05), or pneumonia (OR = 2.0, p = 0.2). The EAC demonstrated good prediction for hemorrhage-induced early mortality (OR = 7.1, p<0.0001), but did not predict late complications (sepsis, OR = 0.8 and p = 0.52; pneumonia, OR = 1.1 and p = 0.7) CGS and mCGS are not comparable and should not be used interchangeably (Krippendorff α = 0.045).
CONCLUSION: Our data show that prediction of complications is more precise after using values that covers different physiological systems (coagulation, hemorrhage, acid-base changes, and soft tissue damage) when compared with using values of only one physiological system (e.g., acidosis). When acid-base changes alone were tested in terms of complications, they were predictive of complications within 72 hours but failed to predict late complications. These findings should be considered when performing early assessment of trauma patients or for the development of new scores.

Statistics

Citations

Dimensions.ai Metrics
2 citations in Web of Science®
2 citations in Scopus®
Google Scholar™

Altmetrics

Downloads

6 downloads since deposited on 20 Feb 2020
6 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Journal Article, refereed, original work
Communities & Collections:04 Faculty of Medicine > University Hospital Zurich > Department of Trauma Surgery
Dewey Decimal Classification:610 Medicine & health
Scopus Subject Areas:Life Sciences > General Biochemistry, Genetics and Molecular Biology
Life Sciences > General Agricultural and Biological Sciences
Health Sciences > Multidisciplinary
Language:English
Date:2020
Deposited On:20 Feb 2020 15:06
Last Modified:11 May 2020 19:37
Publisher:Public Library of Science (PLoS)
ISSN:1932-6203
OA Status:Gold
Free access at:PubMed ID. An embargo period may apply.
Publisher DOI:https://doi.org/10.1371/journal.pone.0228082
PubMed ID:31978109

Download

Gold Open Access

Download PDF  'How to detect a polytrauma patient at risk of complications: a validation and database analysis of four published scales'.
Preview
Content: Published Version
Language: English
Filetype: PDF
Size: 810kB
View at publisher
Licence: Creative Commons: Attribution 4.0 International (CC BY 4.0)