Loading...
Publication
GLIM criteria: use of C-reactive protein as a biomarker of inflammation in hospitalized patients
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 39594 | 2.42 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Background: O Global Leadership Initiative on Malnutrition (GLIM) é uma ferramenta baseada na avaliação de dois tipos de critérios: fenotípico e etiológico, com o objetivo de padronizar o diagnóstico de desnutrição em adultos. A inflamação é um dos critérios etiológicos utilizados nesta ferramenta.
Objetivo: O presente estudo tem como objetivo avaliar a validade concorrente e preditiva do GLIM utilizando a Proteína C-Reativa como critério para inflamação (GLIM-CRP), além dos critérios de inflamação baseados na doença, propostos na ferramenta original (GLIM).
Metodologia: Foi realizado um estudo observacional prospectivo em pacientes hospitalizados com idade ≥18 anos. O risco nutricional foi avaliado com o Nutritional Risk Screening (NRS-2002). A validade concorrente do GLIM e GLIM-CRP foi avaliada usando o Patient Generated Subjective Global Assessment (PG-SGA) como método de referência. A sensibilidade e especificidade foram calculadas. A concordância foi quantificada usando o coeficiente Kappa ponderado (K). A validade preditiva foi avaliada por meio de riscos relativos ajustados (HR) e intervalos de confiança a 95% (IC95%) para alta hospitalar, através da análise de regressão de Cox.
Resultados: A concordância entre PG-SGA e o GLIM foi 50,2%, K=0,391, enquanto que no GLIM-CRP foi de 41,3%, K=0,284. A sensibilidade para o GLIM foi de 72,8% e para o GLIM-CRP foi de 79,3%. A especificidade foi de 68,9% para o GLIM e de 60,5% para o GLIM-CRP. A desnutrição grave avaliada pelas duas ferramentas estava associada de forma independente a uma menor probabilidade de terem alta para domicílio, HR=0,672 (IC 95%: 0,494;0,914) e HR=0,249 (IC 95%: 0,244;0,995), respetivamente. No entanto, no que diz respeito à desnutrição moderada, essa associação foi significativa para o GLIM Criteria, mas não significativa para o GLIM-CRP HR=0,598 (IC 95%: 0,426;0,839) e HR=0,814 (IC 95%: 0,547;1,211), respetivamente.
Conclusões: Embora o GLIM utilizando a PCR como critério alternativo para inflamação mantenha a validade concorrente e preditiva, substituir o critério de inflamação baseado em doenças do GLIM pela PCR não melhorou a ferramenta. No entanto, na ausência de Diagnóstico clínico que permita a identificação de inflamação, a PCR pode ser usada como critério alternativo.
Background: The Global Leadership Initiative on Malnutrition is a tool based on the evaluation of two criteria: phenotypic and etiological, aiming to standardize the diagnosis of malnutrition in adults. Inflammation is one of the etiologic criteria used in this tool. Aims: The present study aims to evaluate the concurrent and predictive validity of GLIM using C-Reactive Protein as a criterion for inflammation (GLIM-CRP) other than diseasebased inflammation criteria proposed in the original tool (GLIM). Methodology: A prospective observational study was conducted among hospitalized patients aged ≥18 years. Nutritional risk was assessed with NRS-2002. Concurrent validy of both GLIM and GLIM-CRP was evaluated using PG-SGA as the reference method. Sensitivity and specificity were calculated. Agreement between these tools was quantified using weighed Kappa (K). Predictive validity was assessed by adjusted hazard ratios (HR) and 95% confidence intervals (95%CI) for being discharged home, trough Cox regression analysis. Results: The agreement between PG-SGA and GLIM criteria was 50.2%, K=0.391 and between PG-SGA GLIM-CRP was 41.3% K=0.284. Sensitivity for GLIM criteria was 72.8% and 79.0% for GLIM CRP. Specificity was 68.9% for GLIM and 60.5% for GLIMCRP. Severe undernutrition assessed by GLIM and GLIM-CRP was independently associated with lower probability of being discharged home, HR=0.672 (95% CI: 0.494;0.914) and HR=0.244 (95% CI: 0.244; 0.995), respectively. However, regarding moderate undernutrition, this association was significant for GLIM but not significant for GLIM-CRP HR=0.598 (95% CI: 0.426;0.839) and HR=0.814 (95% CI: 0.547; 1.211), respectively. Conclusions: Although GLIM using CRP as an alternative criterion for inflammation maintains concurrent validity and predictive validity, replacing the disease-based GLIM inflammation criterion with CRP did not improve the tool. Notwithstanding this, in the absence of clinical diagnosis that allow for the identification of inflammation, CRP can be used as an alternative criterion.
Background: The Global Leadership Initiative on Malnutrition is a tool based on the evaluation of two criteria: phenotypic and etiological, aiming to standardize the diagnosis of malnutrition in adults. Inflammation is one of the etiologic criteria used in this tool. Aims: The present study aims to evaluate the concurrent and predictive validity of GLIM using C-Reactive Protein as a criterion for inflammation (GLIM-CRP) other than diseasebased inflammation criteria proposed in the original tool (GLIM). Methodology: A prospective observational study was conducted among hospitalized patients aged ≥18 years. Nutritional risk was assessed with NRS-2002. Concurrent validy of both GLIM and GLIM-CRP was evaluated using PG-SGA as the reference method. Sensitivity and specificity were calculated. Agreement between these tools was quantified using weighed Kappa (K). Predictive validity was assessed by adjusted hazard ratios (HR) and 95% confidence intervals (95%CI) for being discharged home, trough Cox regression analysis. Results: The agreement between PG-SGA and GLIM criteria was 50.2%, K=0.391 and between PG-SGA GLIM-CRP was 41.3% K=0.284. Sensitivity for GLIM criteria was 72.8% and 79.0% for GLIM CRP. Specificity was 68.9% for GLIM and 60.5% for GLIMCRP. Severe undernutrition assessed by GLIM and GLIM-CRP was independently associated with lower probability of being discharged home, HR=0.672 (95% CI: 0.494;0.914) and HR=0.244 (95% CI: 0.244; 0.995), respectively. However, regarding moderate undernutrition, this association was significant for GLIM but not significant for GLIM-CRP HR=0.598 (95% CI: 0.426;0.839) and HR=0.814 (95% CI: 0.547; 1.211), respectively. Conclusions: Although GLIM using CRP as an alternative criterion for inflammation maintains concurrent validity and predictive validity, replacing the disease-based GLIM inflammation criterion with CRP did not improve the tool. Notwithstanding this, in the absence of clinical diagnosis that allow for the identification of inflammation, CRP can be used as an alternative criterion.
Description
Trabalho Complementar apresentado à Universidade Fernando Pessoa como parte dos requisitos para obtenção do grau de Licenciada em Ciências da Nutrição
Keywords
GLIM Inflamação Desnutrição Validade concorrente Validade preditiva Inflammation Malnutrition Concurrent validity Predictive validity