METHODS: This study used available under-five nutritional secondary data from the Demographic and Health Surveys performed in sub-Saharan African countries. The research used bagging, boosting, and voting algorithms, such as random forest, decision tree, eXtreme Gradient Boosting, and k-nearest neighbors machine learning methods, to generate the MVBHE model.
RESULTS: We evaluated the model performances in contrast to each other using different measures, including accuracy, precision, recall, and the F1 score. The results of the experiment showed that the MVBHE model (96%) was better at predicting malnutrition than the random forest (81%), decision tree (60%), eXtreme Gradient Boosting (79%), and k-nearest neighbors (74%).
CONCLUSIONS: The random forest algorithm demonstrated the highest prediction accuracy (81%) compared with the decision tree, eXtreme Gradient Boosting, and k-nearest neighbors algorithms. The accuracy was then enhanced to 96% using the MVBHE model. The MVBHE model is recommended by the present study as the best way to predict malnutrition in under-five children.
METHODS: Children enrolled in the TREAT Asia Pediatric HIV Observational Database who had SM (weight-for-height or body mass index-for-age Z score less than -3) at ART initiation were analyzed. Generalized estimating equations were used to investigate poor weight recovery (weight-for-age Z score less than -3) and poor CD4% recovery (CD4% <25), and competing risk regression was used to analyze mortality and toxicity-associated treatment modification.
RESULTS: Three hundred fifty-five (11.9%) of 2993 children starting ART had SM. Their median weight-for-age Z score increased from -5.6 at ART initiation to -2.3 after 36 months. Not using trimethoprim-sulfamethoxazole prophylaxis at baseline was associated with poor weight recovery [odds ratio: 2.49 vs. using; 95% confidence interval (CI): 1.66-3.74; P < 0.001]. Median CD4% increased from 3.0 at ART initiation to 27.2 after 36 months, and 56 (15.3%) children died during follow-up. More profound SM was associated with poor CD4% recovery (odds ratio: 1.78 for Z score less than -4.5 vs. -3.5 to less than -3.0; 95% CI: 1.08-2.92; P = 0.023) and mortality (hazard ratio: 2.57 for Z score less than -4.5 vs. -3.5 to less than -3.0; 95% CI: 1.24-5.33; P = 0.011). Twenty-two toxicity-associated ART modifications occurred at a rate of 2.4 per 100 patient-years, and rates did not differ by malnutrition severity.
CONCLUSION: Trimethoprim-sulfamethoxazole prophylaxis is important for the recovery of weight-for-age in severely malnourished children starting ART. The extent of SM does not impede weight-for-age recovery or antiretroviral tolerability, but CD4% response is compromised in children with a very low weight-for-height/body mass index-for-age Z score, which may contribute to their high rate of mortality.
METHODS: A literature search was performed on six databases using the terms "malnutrition", "hospitalised elderly", "nutritional assessment", "Mini Nutritional Assessment (MNA)", "Geriatric Nutrition Risk Index (GNRI)", and "Subjective Global Assessment (SGA)".
RESULTS: According to the previous studies, the prevalence of malnutrition among hospitalized elderly shows an increasing trend not only locally but also across the world. Under-recognition of malnutrition causes the number of malnourished hospitalized elderly to remain high throughout the years. Thus, the development of nutritional screening and assessment tools has been widely studied, and these tools are readily available nowadays. SGA, MNA, and GNRI are the nutritional assessment tools developed specifically for the elderly and are well validated in most countries. However, to date, there is no single tool that can be considered as the universal gold standard for the diagnosis of nutritional status in hospitalized patients.
CONCLUSION: It is important to identify which nutritional assessment tool is suitable to be used in this group to ensure that a structured assessment and documentation of nutritional status can be established. An early and accurate identification of the appropriate treatment of malnutrition can be done as soon as possible, and thus, the malnutrition rate among this group can be minimized in the future.
DESIGN: The 2011 Bangladesh Demographic and Health Survey, conducted from 8 July to 27 December 2011.
SETTING: Selected urban and rural areas of Bangladesh.
SUBJECTS: A total of 2197 ever-married women living with at least one child younger than 5 years. Exposure was determined from maternal reports of high-risk fertility behaviours. We considered three parameters, maternal age at the time of delivery, birth order and birth interval, to define the high-risk fertility behaviours. Chronic undernutrition, anaemia and the coexistence of anaemia and undernutrition among women were the outcome variables.
RESULTS: A substantial percentage of women were exposed to have a high-risk fertility pattern (41·8 %); 33·0 % were at single high-risk and 8·8 % were at multiple high-risk. After adjusting for relevant covariates, high-risk fertility behaviours were associated with increased likelihood of chronic undernutrition (adjusted relative risk; 95 % CI: 1·22; 1·03, 1·44), anaemia (1·12; 1·00, 1·25) and the coexistence of anaemia and undernutrition (1·52; 1·17, 1·98). Furthermore, multiple high-risk fertility behaviours appeared to have more profound consequences on the outcome measured.
CONCLUSIONS: Maternal high-risk fertility behaviours are shockingly frequent practices among women in Bangladesh. High-risk fertility behaviours are important predictors of the increased likelihood of women's chronic undernutrition, anaemia and the coexistence of anaemia and undernutrition.