J Llop-Talaveron1, M B Badia-Tahull2, E Leiva-Badosa1. 1. Pharmacy Department, Hospital Universitari Bellvitge, IDIBELL, Universitat Barcelona, L'Hospitalet de Llobregat, Barcelona, Spain. 2. Pharmacy Department, Hospital Universitari Bellvitge, IDIBELL, Universitat Barcelona, L'Hospitalet de Llobregat, Barcelona, Spain. Electronic address: mbadia@bellvitgehospital.cat.
Abstract
BACKGROUND: There is no "gold standard" score for predicting poor-nutrition-related outcomes. The objective of this study was to identify the optimal predictive score, based on inflammatory parameters, for the clinical outcomes of parenteral nutrition (PN). MATERIAL AND METHODS: This was a 4-year retrospective observational study of 460 patients treated with PN. C-reactive protein (CRP), prealbumin, albumin, CRP/prealbumin and CRP/albumin were studied as potential prognostic scores at the beginning of PN for clinical outcomes during PN. Three different statistical approaches were developed: 1) A univariate analysis of each of the 5 prognostic scores and 5 multivariate models for CRP/albumin and CRP/prealbumin to study their association with exitus, infection, sepsis, liver failure, renal impairment, cancer, intensive care unit stay, mechanical ventilation; 2) Univariate and multivariate survival analysis of PN length, intensive care unit (ICU) length of saty and days of mechanical ventilation vs CRP/albumin and CRP/prealbumin; 3) A ROC analysis of the prognostic accuracy of CRP/albumin and CRP/prealbumin over morbidity/mortality. RESULTS: 1) CRP, albumin and CRP/albumin gave more information about morbidity/mortality than prealbumin and CRP/prealbumin. CRP/albumin was statistically significant for exitus (OR 1.85; CI 95%: 1.00-3.45), infection (OR 2.15; CI 95%: 1.22-3.80), sepsis (OR 2.82; CI 95%: 1.69-4.70) and liver failure (OR 2.66; CI 95%: 1.55-4.58). CRP/prealbumin for sepsis was (OR 2.21; CI 95%: 1.34-3.64) and for liver failure (OR 2.04; CI 95%: 1.17-3.53); 2) CRP/albumin and CRP/prealbumin significantly predict PN duration, days in ICU and days on mechanical ventilation; 3) and are related to exitus, infection, sepsis and liver failure. CONCLUSION: The CRP/albumin score at the beginning of PN treatment has more prognostic capability than CRP/prealbumin, albumin or prealbumin. The systematic use of this score could help to identify those patients with higher risk.
BACKGROUND: There is no "gold standard" score for predicting poor-nutrition-related outcomes. The objective of this study was to identify the optimal predictive score, based on inflammatory parameters, for the clinical outcomes of parenteral nutrition (PN). MATERIAL AND METHODS: This was a 4-year retrospective observational study of 460 patients treated with PN. C-reactive protein (CRP), prealbumin, albumin, CRP/prealbumin and CRP/albumin were studied as potential prognostic scores at the beginning of PN for clinical outcomes during PN. Three different statistical approaches were developed: 1) A univariate analysis of each of the 5 prognostic scores and 5 multivariate models for CRP/albumin and CRP/prealbumin to study their association with exitus, infection, sepsis, liver failure, renal impairment, cancer, intensive care unit stay, mechanical ventilation; 2) Univariate and multivariate survival analysis of PN length, intensive care unit (ICU) length of saty and days of mechanical ventilation vs CRP/albumin and CRP/prealbumin; 3) A ROC analysis of the prognostic accuracy of CRP/albumin and CRP/prealbumin over morbidity/mortality. RESULTS: 1) CRP, albumin and CRP/albumin gave more information about morbidity/mortality than prealbumin and CRP/prealbumin. CRP/albumin was statistically significant for exitus (OR 1.85; CI 95%: 1.00-3.45), infection (OR 2.15; CI 95%: 1.22-3.80), sepsis (OR 2.82; CI 95%: 1.69-4.70) and liver failure (OR 2.66; CI 95%: 1.55-4.58). CRP/prealbumin for sepsis was (OR 2.21; CI 95%: 1.34-3.64) and for liver failure (OR 2.04; CI 95%: 1.17-3.53); 2) CRP/albumin and CRP/prealbumin significantly predict PN duration, days in ICU and days on mechanical ventilation; 3) and are related to exitus, infection, sepsis and liver failure. CONCLUSION: The CRP/albumin score at the beginning of PN treatment has more prognostic capability than CRP/prealbumin, albumin or prealbumin. The systematic use of this score could help to identify those patients with higher risk.
Authors: Markus Kieler; Paul Kössler; Matija Milovic; Elias Meyer; Kristína Križanová; Lea Kum; Alexander Friedrich; Eva Masel; Raimund Bauer; Matthias Unseld Journal: Palliat Med Date: 2022-02-20 Impact factor: 4.762
Authors: Lea Kum; Alexander Friedrich; Markus Kieler; Elias Meyer; Petar Popov; Paul Kössler; Anna Kitta; Feroniki Adamidis; Raimund Oberle; Eva Katharina Masel; Matthias Unseld Journal: Nutrients Date: 2022-02-11 Impact factor: 5.717