Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury
Recent advances in emergency medicine and the co-ordinated delivery of trauma care mean more critically-injured patients now reach the hospital alive and survive life-saving operations. Indeed, between 2008 and 2017, the odds of surviving a major traumatic injury in the UK increased by nineteen perc...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-12-01
|
| Series: | Metabolites |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2218-1989/12/1/29 |
| _version_ | 1797492096990445568 |
|---|---|
| author | Animesh Acharjee Jon Hazeldine Alina Bazarova Lavanya Deenadayalu Jinkang Zhang Conor Bentley Dominic Russ Janet M. Lord Georgios V. Gkoutos Stephen P. Young Mark A. Foster |
| author_facet | Animesh Acharjee Jon Hazeldine Alina Bazarova Lavanya Deenadayalu Jinkang Zhang Conor Bentley Dominic Russ Janet M. Lord Georgios V. Gkoutos Stephen P. Young Mark A. Foster |
| author_sort | Animesh Acharjee |
| collection | DOAJ |
| description | Recent advances in emergency medicine and the co-ordinated delivery of trauma care mean more critically-injured patients now reach the hospital alive and survive life-saving operations. Indeed, between 2008 and 2017, the odds of surviving a major traumatic injury in the UK increased by nineteen percent. However, the improved survival rates of severely-injured patients have placed an increased burden on the healthcare system, with major trauma a common cause of intensive care unit (ICU) admissions that last ≥10 days. Improved understanding of the factors influencing patient outcomes is now urgently needed. We investigated the serum metabolomic profile of fifty-five major trauma patients across three post-injury phases: acute (days 0–4), intermediate (days 5–14) and late (days 15–112). Using ICU length of stay (LOS) as a clinical outcome, we aimed to determine whether the serum metabolome measured at days 0–4 post-injury for patients with an extended (≥10 days) ICU LOS differed from that of patients with a short (<10 days) ICU LOS. In addition, we investigated whether combining metabolomic profiles with clinical scoring systems would generate a variable that would identify patients with an extended ICU LOS with a greater degree of accuracy than models built on either variable alone. The number of metabolites unique to and shared across each time segment varied across acute, intermediate and late segments. A one-way ANOVA revealed the most variation in metabolite levels across the different time-points was for the metabolites lactate, glucose, anserine and 3-hydroxybutyrate. A total of eleven features were selected to differentiate between <10 days ICU LOS vs. >10 days ICU LOS. New Injury Severity Score (NISS), testosterone, and the metabolites cadaverine, urea, isoleucine, acetoacetate, dimethyl sulfone, syringate, creatinine, xylitol, and acetone form the integrated biomarker set. Using metabolic enrichment analysis, we found valine, leucine and isoleucine biosynthesis, glutathione metabolism, and glycine, serine and threonine metabolism were the top three pathways differentiating ICU LOS with a <i>p</i> < 0.05. A combined model of NISS and testosterone and all nine selected metabolites achieved an AUROC of 0.824. Differences exist in the serum metabolome of major trauma patients who subsequently experience a short or prolonged ICU LOS in the acute post-injury setting. Combining metabolomic data with anatomical scoring systems allowed us to discriminate between these two groups with a greater degree of accuracy than that of either variable alone. |
| first_indexed | 2024-03-10T00:57:43Z |
| format | Article |
| id | doaj.art-6ee5acbf497e4e849c74cc42f73b08fb |
| institution | Directory Open Access Journal |
| issn | 2218-1989 |
| language | English |
| last_indexed | 2024-03-10T00:57:43Z |
| publishDate | 2021-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Metabolites |
| spelling | doaj.art-6ee5acbf497e4e849c74cc42f73b08fb2023-11-23T14:39:48ZengMDPI AGMetabolites2218-19892021-12-011212910.3390/metabo12010029Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic InjuryAnimesh Acharjee0Jon Hazeldine1Alina Bazarova2Lavanya Deenadayalu3Jinkang Zhang4Conor Bentley5Dominic Russ6Janet M. Lord7Georgios V. Gkoutos8Stephen P. Young9Mark A. Foster10Microbiology Research Centre, National Institute for Health Research Surgical Reconstruction, Queen Elizabeth Hospital Birmingham, Birmingham B15 2GW, UKMicrobiology Research Centre, National Institute for Health Research Surgical Reconstruction, Queen Elizabeth Hospital Birmingham, Birmingham B15 2GW, UKInstitute of Cancer and Genomic Sciences, University of Birmingham, Birmingham B15 2TT, UKInstitute of Cancer and Genomic Sciences, University of Birmingham, Birmingham B15 2TT, UKInstitute of Inflammation and Ageing, University of Birmingham, Birmingham B15 2TT, UKMicrobiology Research Centre, National Institute for Health Research Surgical Reconstruction, Queen Elizabeth Hospital Birmingham, Birmingham B15 2GW, UKInstitute of Cancer and Genomic Sciences, University of Birmingham, Birmingham B15 2TT, UKMicrobiology Research Centre, National Institute for Health Research Surgical Reconstruction, Queen Elizabeth Hospital Birmingham, Birmingham B15 2GW, UKMicrobiology Research Centre, National Institute for Health Research Surgical Reconstruction, Queen Elizabeth Hospital Birmingham, Birmingham B15 2GW, UKInstitute of Inflammation and Ageing, University of Birmingham, Birmingham B15 2TT, UKMicrobiology Research Centre, National Institute for Health Research Surgical Reconstruction, Queen Elizabeth Hospital Birmingham, Birmingham B15 2GW, UKRecent advances in emergency medicine and the co-ordinated delivery of trauma care mean more critically-injured patients now reach the hospital alive and survive life-saving operations. Indeed, between 2008 and 2017, the odds of surviving a major traumatic injury in the UK increased by nineteen percent. However, the improved survival rates of severely-injured patients have placed an increased burden on the healthcare system, with major trauma a common cause of intensive care unit (ICU) admissions that last ≥10 days. Improved understanding of the factors influencing patient outcomes is now urgently needed. We investigated the serum metabolomic profile of fifty-five major trauma patients across three post-injury phases: acute (days 0–4), intermediate (days 5–14) and late (days 15–112). Using ICU length of stay (LOS) as a clinical outcome, we aimed to determine whether the serum metabolome measured at days 0–4 post-injury for patients with an extended (≥10 days) ICU LOS differed from that of patients with a short (<10 days) ICU LOS. In addition, we investigated whether combining metabolomic profiles with clinical scoring systems would generate a variable that would identify patients with an extended ICU LOS with a greater degree of accuracy than models built on either variable alone. The number of metabolites unique to and shared across each time segment varied across acute, intermediate and late segments. A one-way ANOVA revealed the most variation in metabolite levels across the different time-points was for the metabolites lactate, glucose, anserine and 3-hydroxybutyrate. A total of eleven features were selected to differentiate between <10 days ICU LOS vs. >10 days ICU LOS. New Injury Severity Score (NISS), testosterone, and the metabolites cadaverine, urea, isoleucine, acetoacetate, dimethyl sulfone, syringate, creatinine, xylitol, and acetone form the integrated biomarker set. Using metabolic enrichment analysis, we found valine, leucine and isoleucine biosynthesis, glutathione metabolism, and glycine, serine and threonine metabolism were the top three pathways differentiating ICU LOS with a <i>p</i> < 0.05. A combined model of NISS and testosterone and all nine selected metabolites achieved an AUROC of 0.824. Differences exist in the serum metabolome of major trauma patients who subsequently experience a short or prolonged ICU LOS in the acute post-injury setting. Combining metabolomic data with anatomical scoring systems allowed us to discriminate between these two groups with a greater degree of accuracy than that of either variable alone.https://www.mdpi.com/2218-1989/12/1/29metabolomicsomics integrationICU length of stayinflammation |
| spellingShingle | Animesh Acharjee Jon Hazeldine Alina Bazarova Lavanya Deenadayalu Jinkang Zhang Conor Bentley Dominic Russ Janet M. Lord Georgios V. Gkoutos Stephen P. Young Mark A. Foster Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury Metabolites metabolomics omics integration ICU length of stay inflammation |
| title | Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury |
| title_full | Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury |
| title_fullStr | Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury |
| title_full_unstemmed | Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury |
| title_short | Integration of Metabolomic and Clinical Data Improves the Prediction of Intensive Care Unit Length of Stay Following Major Traumatic Injury |
| title_sort | integration of metabolomic and clinical data improves the prediction of intensive care unit length of stay following major traumatic injury |
| topic | metabolomics omics integration ICU length of stay inflammation |
| url | https://www.mdpi.com/2218-1989/12/1/29 |
| work_keys_str_mv | AT animeshacharjee integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT jonhazeldine integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT alinabazarova integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT lavanyadeenadayalu integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT jinkangzhang integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT conorbentley integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT dominicruss integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT janetmlord integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT georgiosvgkoutos integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT stephenpyoung integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury AT markafoster integrationofmetabolomicandclinicaldataimprovesthepredictionofintensivecareunitlengthofstayfollowingmajortraumaticinjury |