Explore the vast range of titles available.

BackgroundMalnutrition in children is widely prevalent around the world. It has been observed that malnourished children with multiple anthropometric deficits have higher mortality. However, adequate studies are not available on the outcome and recovery of these children.Nandurbar, a tribal district from Maharashtra, India, shows high prevalence of all three forms of malnutrition, often occurring simultaneously. A project previously undertaken in Nandurbar from July 2014 to June 2016 studied the effect of various therapeutic feeds in treatment of children with uncomplicated severe acute malnutrition (SAM). In this study, we analyse secondary data from it to correlate effects of stunting, wasting and underweight on treatment recovery.MethodsAnalysis was done on 5979 children with SAM using linear and logistic regression on R software for recovery rates and weight gain in children with SAM with single versus multiple anthropometric deficits, their relation to age, sex, and recovery from severe stunting by gain in height.ResultsThe mean age of children was 35 months and 53.1% of the children were males. 2346 (39.2%) children recovered at the end of the 8-week treatment. 454 (7.6%) had single anthropometric deficit (SAM only), 3164 (52.9%) had two anthropometric deficits (SAM and severe underweight (SUW)) and 2355 (39.4%) children had three anthropometric deficits (SAM, SUW and severe stunting). Out of the 5979 children with SAM, only 52 (0.9%) of children were not underweight (severe or moderate).44.94% of children with SAM who were severely stunted recovered, compared with 35.52% of children who were not (p<0.001). After controlling for confounders, severe stunting was found to increase the odds of recovery by 1.49. Severely stunted children with SAM also showed faster recovery and weight gain by 1.93 days (p<0.012) and 0.29 g/kg/day (p<0.001), respectively. Recovery was higher in females and younger age group. Recovery was also found to depend on the therapeutic feed, with children receiving medical nutrition therapy showing better recovery for severely stunted children.ConclusionOur findings corroborate previous literature that stunting is a way for the body to deal with chronic stress of nutritional deprivation and provides a survival advantage to a child.

Objectives To assess the gaps and trends in child immunization coverage among urban and rural areas in India, and compare the success of immunisation program in each. Methods PubMed, Scopus, and Crossref, and Google Scholar electronic databases were searched on October 9, 2019, and March 21, 2020, for studies that measured and reported immunization coverage indicators in India. Random-effects meta-analyses and meta-regressions were conducted. Results The authors' search identified 545 studies, and 2 were obtained by expert suggestion. Among these 68 studies and 6 surveys were included. They found that full immunization coverage has grown yearly at 2.65% and 0.82% in rural and urban areas, respectively whereas partial immunization coverage declined by −2.44% and −0.69%, respectively. Percentage of nonimmunized children did not show a statistically significant trend in either. Conclusion While rural immunization coverage has seen a large increase over the past two decades, the progress in urban areas is weak and negligible. This was largely attributable to a focus on minimizing dropouts in rural areas. However, a lack of significant reduction in unimmunized children may indicate left-out children or pockets in both rural and urban areas. The poor performance of immunization programs in urban areas, coupled with a larger impact of COVID-19, warrants that India urgently adopts urban-sensitive and urban-focused policies and programs.

Global cloud service providers handle inference workloads for Large Language Models (LLMs) that span latency-sensitive (e.g., chatbots) and insensitive (e.g., report writing) tasks, resulting in diverse and often conflicting Service Level Agreement (SLA) requirements. Managing such mixed workloads is challenging due to the complexity of the inference serving stack, which encompasses multiple models, GPU hardware, and global data centers. Existing solutions often silo such fast and slow tasks onto separate GPU resource pools with different SLAs, but this leads to significant under-utilization of expensive accelerators due to load mismatch. In this article, we characterize the LLM serving workloads at Microsoft Office 365, one of the largest users of LLMs within Microsoft Azure cloud with over 10 million requests per day, and highlight key observations across workloads in different data center regions and across time. This is one of the first such public studies of Internet-scale LLM workloads. We use these insights to propose SageServe, a comprehensive LLM serving framework that dynamically adapts to workload demands using multi-timescale control knobs. It combines short-term request routing to data centers with long-term scaling of GPU VMs and model placement with higher lead times, and co-optimizes the routing and resource allocation problem using a traffic forecast model and an Integer Linear Programming (ILP) solution. We evaluate SageServe through real runs and realistic simulations on 10 million production requests across three regions and four open-source models. We achieve up to 25% savings in GPU-hours compared to the current baseline deployment and reduce GPU-hour wastage due to inefficient auto-scaling by 80%, resulting in a potential monthly cost savings of up to $2.5 million, while maintaining tail latency and meeting SLAs.