Short Graphic Values History Tool for decision making during serious illness.

OBJECTIVES: To develop and validate a values clarification tool, the Short Graphic Values History Tool (GVHT), designed to support person-centred decision making during serious illness. METHODS: The development phase included input from experts and laypersons and assessed acceptability with patients/family members. In the validation phase, we recruited additional participants into a before-after study. Our primary validation hypothesis was that the tool would reduce scores on the Decisional Conflict Scale (DCS) at 1-2 weeks of follow-up. Our secondary validation hypotheses were that the tool would improve values clarity (reduce scores) more than other DCS subscales and increase engagement in advance care planning (ACP) processes related to identification and discussion of one's values. RESULTS: In the development phase, the tool received positive overall ratings from 22 patients/family members in hospital (mean score 4.3; 1=very poor; 5=very good) and family practice (mean score 4.5) settings. In the validation phase, we enrolled 157 patients (mean age 71.8 years) from family practice, cancer clinic and hospital settings. After tool completion, decisional conflict decreased (-6.7 points, 95% CI -11.1 to -2.3, p=0.003; 0-100 scale; N=100), with the most improvement seen in the values clarity subscale (-10.0 points, 95% CI -17.3 to -2.7, p=0.008; N=100), and the ACP-Values process score increased (+0.4 points, 95% CI 0.2 to 0.6, p=0.001; 1-5 scale; N=61). CONCLUSIONS: The Short GVHT is acceptable to end users and has some measure of validity. Further study to evaluate its impact on decision making during serious illness is warranted.

Complementariedade do escore NUTRIC modificado com ou sem proteína C-reativa e avaliação subjetiva global na predição de mortalidade em pacientes críticos / Complementarity of modified NUTRIC score with or without C-reactive protein and subjective global assessment in predicting mortality in critically ill patients

RESUMO Objetivo: Avaliar a concordância entre o escore NUTRIC modificado e o escore NUTRIC com proteína C-reativa na identificação de pacientes em risco nutricional e na predição da mortalidade entre pacientes críticos. Avaliou-se também o risco de óbito com agrupamento dos pacientes segundo o risco nutricional e a desnutrição detectada pela avaliação subjetiva global. Métodos: Estudo de coorte em pacientes admitidos em uma unidade de terapia intensiva. O risco nutricional foi avaliado por meio do escore NUTRIC modificado e uma versão do escore NUTRIC com proteína C-reativa. Aplicou-se avaliação subjetiva global para diagnóstico de desnutrição. Calculou-se a estatística de Kappa e construiu-se uma curva ROC considerando o NUTRIC modificado como referência. A validade preditiva foi avaliada considerando a mortalidade em 28 dias (na unidade de terapia intensiva e após a alta) como desfecho. Resultados: Estudaram-se 130 pacientes (63,05 ± 16,46 anos, 53,8% do sexo masculino). Segundo o NUTRIC com proteína C-reativa, 34,4% foram classificados como escore alto, enquanto 28,5% dos pacientes tiveram esta classificação com utilização do NUTRIC modificado. Segundo a avaliação subjetiva global, 48,1% dos pacientes estavam desnutridos. Observou-se concordância excelente entre o NUTRIC modificado e o NUTRIC com proteína C-reativa (Kappa = 0,88; p < 0,001). A área sob a curva ROC foi igual a 0,942 (0,881 - 1,000) para o NUTRIC com proteína C-reativa. O risco de óbito em 28 dias estava aumentado nos pacientes com escores elevados pelo NUTRIC modificado (HR = 1,827; IC95% 1,029 - 3,244; p = 0,040) e pelo NUTRIC com proteína C-reativa (HR = 2,685; IC95% 1,423 - 5,064; p = 0,002). Observou-se elevado risco de óbito nos pacientes com alto risco nutricional e desnutrição, independentemente da versão do NUTRIC aplicada. Conclusão: A concordância entre o escore NUTRIC modificado e o NUTRIC com proteína C-reativa foi excelente. Além disto, a combinação da avaliação com um escore NUTRIC mais avaliação subjetiva global pode aumentar a precisão para predição de mortalidade em pacientes críticos.

ABSTRACT Objective: To evaluate the concordance between the modified NUTRIC and NUTRIC with C-reactive protein instruments in identifying nutritional risk patients and predicting mortality in critically ill patients. The risk of death in patient groups was also investigated according to nutritional risk and malnutrition detected by subjective global assessment. Methods: A cohort study of patients admitted to an intensive care unit. Nutritional risk was assessed by modified NUTRIC and a version of NUTRIC with C-reactive protein. Subjective global assessment was applied to diagnose malnutrition. Kappa statistics were calculated, and an ROC curve was constructed considering modified NUTRIC as a reference. The predictive validity was assessed considering mortality in 28 days (whether in the intensive care unit or after discharge) as the outcome. Results: A total of 130 patients were studied (63.05 ± 16.46 years, 53.8% males). According to NUTRIC with C-reactive protein, 34.4% were classified as having a high score, while 28.5% of patients had this classification with modified NUTRIC. According to SGA 48.1% of patients were malnourished. There was excellent agreement between modified NUTRIC and NUTRIC with C-reactive protein (Kappa = 0.88, p < 0.001). The area under the ROC curve was equal to 0.942 (0.881 - 1.000) for NUTRIC with C-reactive protein. The risk of death within 28 days was increased in patients with high modified NUTRIC (HR = 1.827; 95%CI 1.029 - 3.244; p = 0.040) and NUTRIC with C-reactive protein (HR = 2.685; 95%CI 1.423 - 5.064; p = 0.002) scores. A high risk of death was observed in patients with high nutritional risk and malnutrition, independent of the version of the NUTRIC score applied. Conclusion: An excellent agreement between modified NUTRIC and NUTRIC with C-reactive protein was observed. In addition, combining NUTRIC and subjective global assessment may increase the accuracy of predicting mortality in critically ill patients.