Abstract
BACKGROUND: Malnutrition can lead to adverse clinical outcomes in hospitalized patients, timely and accurate diagnosis of malnutrition is crucial for initiating early nutritional support programs. To assess the correlation between malnutrition diagnosed by different malnutrition diagnostic tools and patients' clinical outcomes. METHODS: Meta-analyses of the associations between malnutrition and patients' clinical outcomes were screened and included by searching databases. For each association, this study used fixed and random effects models, calculated 95% CI (confidence intervals) and 95% PI (prediction intervals), and assessed heterogeneity, evidence of small-study effects, and excess significance bias. RESULTS: A total of 138 meta-analyses were included in this study, and 407 associations were evaluated. For oncology patients, malnutrition diagnosed by eight tools was associated with oncological survival, with three evidence scores of PNI (prognostic nutritional index), GNRI (geriatric nutritional risk index), and CONUT (controlling nutritional status) being highly recommended (Class II). For nontumor patients, malnutrition diagnosed by nine tools was associated with poor clinical outcomes, with four tools with high evidence scores (Class II) of PNI, BMI (body mass index) < 18.5 kg/m(2), GNRI, and CONUT being highly recommended. CONCLUSION: This study demonstrated a significant correlation (Class II) between malnutrition diagnosed by four tools, the PNI, BMI < 18.5 kg/m(2), GNRI, and CONUT, and clinical outcomes, and the other tools need to be validated in future high-quality studies despite their correlation. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42024586175.