Vítor Sá Martins1, Teresa Adragão2, Leila Aguiar3, Iola Pinto4, Catarina Dias3, Rita Figueiredo3, Pedro Lourenço3, Tânia Pascoal3, Juliana Pereira3, Tatiana Pinheiro3, Inês Ramião3, Brígida Velez3, Ana Luisa Papoila5, Nuno Borges6, Conceição Calhau7, Fernando Macário3. 1. Medical Department DIAVERUM Portugal, Sintra, Portugal; Faculdade de Ciências da Nutrição e Alimentação, Universidade do Porto, Porto, Portugal; CINTESIS, Center for Health Technology Services Research, Rua Doutor Plácido da Costa, Porto, Portugal. Electronic address: vitor.martins@diaverum.com. 2. Medical Department DIAVERUM Portugal, Sintra, Portugal; Nephrology Department, Santa Cruz Hospital, Carnaxide, Portugal. 3. Medical Department DIAVERUM Portugal, Sintra, Portugal. 4. CMA, Faculdade de Ciências e Tecnologia da Universidade Nova de Lisboa, Lisboa, Portugal; ISEL, Instituto Superior de Engenharia de Lisboa, Lisboa, Portugal. 5. CEAUL, Centro de Estatística e Aplicações da Universidade de Lisboa, Lisboa, Portugal; NOVA Medical School|Faculdade de Ciências Médicas da Universidade Nova de Lisboa, Lisboa, Portugal. 6. Faculdade de Ciências da Nutrição e Alimentação, Universidade do Porto, Porto, Portugal; CINTESIS, Center for Health Technology Services Research, Rua Doutor Plácido da Costa, Porto, Portugal. 7. CINTESIS, Center for Health Technology Services Research, Rua Doutor Plácido da Costa, Porto, Portugal; NOVA Medical School|Faculdade de Ciências Médicas da Universidade Nova de Lisboa, Lisboa, Portugal; Unidade Universitária Lifestyle Medicine José de Mello Saúde by NOVA Medical School, Lisboa, Portugal.
Abstract
OBJECTIVE: Since its development, cumulative evidence has accumulated regarding the prognostic value of the Malnutrition-Inflammation Score (MIS/Kalantar score) prognostic value; however, there is a shortage of recent and large studies with comprehensive statistical methodologies that contribute to support a higher level of evidence and a consensual cutoff. The aim of this study was to assess the strength of MIS association with hospitalization and mortality in a nationwide cohort. METHODS: This was a historical cohort study of hemodialysis patients from 25 outpatient centers followed up for 48 months. Univariable and multivariable Cox additive regression models were used to analyze the data. The C-index was estimated to assess the performance of the final model. RESULTS: Two thousand four hundred forty-four patients were analyzed, 59.0% males, 32.0% diabetic, and median age of 71 years (P25 = 60, P75 = 79). During a median period of 45-month follow-up, with a maximum of 48 months (P25 = 31; P75 = 48), 875 patients presented an MIS <5 (35.8%) and 860 patients (35.2%) died. The proportion of deaths was 23.1% for patients with the MIS <5 and 41.9% if the MIS ≥5 (P < .001). A total of 1,528 patients (62.5%) were hospitalized with a median time to the first hospitalization of 26 months (P25 = 9; P75 = 45). A new cutoff point regarding the risk of death, MIS ≥6, was identified for this study data set. In multivariable analysis for hospitalization risk, a higher MIS, higher comorbidity index, and arteriovenous graft or catheter increased the risk, whereas higher Kt/V and higher albumin had a protective effect. In multivariable analysis for mortality risk, adjusting for age, albumin, normalized protein catabolic rate, Charlson comorbidity index, interdialytic weight gain, Kt/V, diabetes, hematocrit, and vascular access, patients with the MIS ≥6 showed a hazard ratio of 1.469 (95% confidence interval: 1.262-1.711; P < .001). Higher age, higher interdialytic weight gain, higher comorbidity index, and catheter increased significantly the risk, whereas higher Kt/V, higher albumin, and higher normalized protein catabolic rate (≥1.05 g/kg/d) reduced the risk. CONCLUSION: The MIS maintains its relevant and significant association with hospitalization and mortality.
OBJECTIVE: Since its development, cumulative evidence has accumulated regarding the prognostic value of the Malnutrition-Inflammation Score (MIS/Kalantar score) prognostic value; however, there is a shortage of recent and large studies with comprehensive statistical methodologies that contribute to support a higher level of evidence and a consensual cutoff. The aim of this study was to assess the strength of MIS association with hospitalization and mortality in a nationwide cohort. METHODS: This was a historical cohort study of hemodialysis patients from 25 outpatient centers followed up for 48 months. Univariable and multivariable Cox additive regression models were used to analyze the data. The C-index was estimated to assess the performance of the final model. RESULTS: Two thousand four hundred forty-four patients were analyzed, 59.0% males, 32.0% diabetic, and median age of 71 years (P25 = 60, P75 = 79). During a median period of 45-month follow-up, with a maximum of 48 months (P25 = 31; P75 = 48), 875 patients presented an MIS <5 (35.8%) and 860 patients (35.2%) died. The proportion of deaths was 23.1% for patients with the MIS <5 and 41.9% if the MIS ≥5 (P < .001). A total of 1,528 patients (62.5%) were hospitalized with a median time to the first hospitalization of 26 months (P25 = 9; P75 = 45). A new cutoff point regarding the risk of death, MIS ≥6, was identified for this study data set. In multivariable analysis for hospitalization risk, a higher MIS, higher comorbidity index, and arteriovenous graft or catheter increased the risk, whereas higher Kt/V and higher albumin had a protective effect. In multivariable analysis for mortality risk, adjusting for age, albumin, normalized protein catabolic rate, Charlson comorbidity index, interdialytic weight gain, Kt/V, diabetes, hematocrit, and vascular access, patients with the MIS ≥6 showed a hazard ratio of 1.469 (95% confidence interval: 1.262-1.711; P < .001). Higher age, higher interdialytic weight gain, higher comorbidity index, and catheter increased significantly the risk, whereas higher Kt/V, higher albumin, and higher normalized protein catabolic rate (≥1.05 g/kg/d) reduced the risk. CONCLUSION: The MIS maintains its relevant and significant association with hospitalization and mortality.