The use of bioelectrical impedance vector analysis in the assessment of refeeding syndrome on severely acute malnourished children

Abstract

Severe acute malnutrition remains a major cause of morbidity and mortality among children under five in low- and middle-income countries, including South Africa. Despite the existence of World Health Organisation inpatient severe acute malnutrition management guidelines, mortality in many public hospitals remains unacceptably high, and preventable complications continue to occur during the early stabilisation phase. Refeeding syndrome is a potentially fatal metabolic complication that arises when nutritional support is initiated or increased after a period of undernutrition. In routine paediatric practice, however, refeeding syndrome is often poorly recognised, inconsistently defined and rarely documented as a distinct clinical entity. Conventional anthropometric and clinical indicators may also fail to detect the rapid physiological shifts that precede clinical deterioration. This thesis addresses these gaps by examining the identification and management of refeeding syndrome in children with severe acute malnutrition. This was realised by evaluating the potential role of bioelectrical impedance vector analysis as an additional tool to assess hydration status and risk of refeeding syndrome during inpatient stabilisation. The overall aim of the thesis was to improve understanding of how refeeding syndrome is conceptualised, identified and managed in children who are undergoing severe acute malnutrition care and to explore whether bioelectrical impedance vector analysis can support early risk stratification. A three-phase, mixed-methods design was used. Phase 1 comprised a systematic review and meta-analysis of evidence on the identification and management of refeeding syndrome among hospitalised children aged 6 to 59 months with severe acute malnutrition in Sub-Saharan Africa. Phase 2 was a quantitative study conducted in public hospitals in the City of Tshwane, which admitted children with severe acute malnutrition and followed them over the first five days of stabilisation using bioelectrical impedance vector analysis, anthropometric, biochemical and clinical assessments. Phase 3 was a qualitative descriptive study that explored healthcare practitioners’ experiences of assessing and identifying refeeding syndrome in hospitalised children with severe acute malnutrition in the same setting. Phase 1 systematically reviewed published and grey literature to describe how refeeding syndrome is defined, monitored and treated in inpatient severe acute malnutrition programmes in Sub-Saharan African countries. Several electronic databases and relevant organisational websites were searched. The review demonstrated marked heterogeneity in refeeding syndrome definitions, with most studies either not defining refeeding syndrome explicitly or using non-standard criteria that did not consistently include hypophosphatemia. Biochemical monitoring during refeeding was variably reported, and routine phosphate measurement was uncommon. The review found descriptive variation in management strategies for refeeding syndrome in children with severe acute malnutrition across studies, although most adhered to the World Health Organisation guidelines for severe acute malnutrition treatment. The reported refeeding syndrome incidence varied widely, reflecting both true variation and inconsistent recognition and documentation. Overall, Phase 1 highlighted substantial evidence gaps, particularly the lack of context-appropriate operational definitions and structured monitoring strategies for refeeding syndrome in children undergoing severe acute malnutrition treatment. Phase 2 was a prospective quantitative observational longitudinal repeated-measures study involving 64 hospitalised children with severe acute malnutrition during the stabilisation phase of treatment. to selected public hospitals in the City of Tshwane. Anthropometry, clinical variables, laboratory markers and the bioelectrical impedance vector analysis-derived parameters were recorded at admission and followed over the five days of inpatient stabilisation. Refeeding syndrome risk was classified using phosphate-based criteria adapted from international guidelines and applied to the South African context of severe acute malnutrition. Clinically, vomiting and diarrhoea improved rapidly in both groups; however, children classified as at risk of refeeding syndrome had delayed oedema resolution (higher oedema grades on Days 2 to 4) and minimal net weight gain by Day 5, whereas children not at risk showed progressive weight gain with a significant between‑group difference by Day 5. Mid upper arm circumference remained comparatively stable across Days 1 to 5 in both groups, supporting the interpretation that early weight changes during stabilisation largely reflect fluid dynamics rather than true tissue accretion. At admission, the refeeding syndrome ‑risk group showed lower resistance values and shorter impedance vectors. Hydration-related indices, such as percentage total body water, were generally higher in the refeeding syndrome risk group, consistent with apparent anthropometric adequacy in oedematous or fluid-overloaded children. Multivariable logistic regression identified lower resistance as a predictor of refeeding syndrome risk. These findings suggest that bioelectrical impedance vector analysis-derived markers may help to identify children at higher risk of refeeding syndrome at admission and could complement existing anthropometric and clinical assessments. Phase 3 employed a qualitative descriptive design to explore the experiences of healthcare practitioners involved in the care of children with severe acute malnutrition. Semi-structured interviews were conducted with a purposive sample of paediatricians, medical doctors, dieticians and nurses working in paediatric wards that manage severe acute malnutrition cases in public hospitals in the City of Tshwane. Generally, participants recognised refeeding syndrome as a theoretical or textbook concept but reported that it was seldom explicitly documented in patients' records and was rarely labelled as a distinct diagnosis. Instead, refeeding syndrome was included under more general severe acute malnutrition complications such as oedema, sepsis, electrolyte imbalance or “the child collapsing/crashing”. Themes that emerged included uncertainty and variation on how refeeding syndrome and risk are defined, reliance on informal clinical cues rather than structured monitoring tools, and the perception that refeeding syndrome is “not normally recognised” despite its potential severity. Participants described significant constraints on systematic biochemical monitoring, including delayed laboratory results and limited after-hours access. Multidisciplinary team communication was highlighted as both a strength and a challenge. This is to say, where regular joint ward rounds occurred, care tended to be more consistent, but after-hours and shift-to-shift handovers were seen as vulnerable points. Practitioners emphasised the need for simple, practical tools and focused training to support early recognition, documentation and standardised management of refeeding syndrome within existing resource limitations. The integration of all three phases provides a coherent account of why refeeding syndrome remains under-recognised and difficult to manage in paediatric severe acute malnutrition settings. The systematic review showed that refeeding syndrome is poorly defined and inconsistently reported in the regional literature, reflecting the absence of clear operational guidance in many programmes. The quantitative study demonstrated that clinically important biochemical shifts compatible with refeeding syndrome are common in hospitalised children with severe acute malnutrition and identified a bioelectrical impedance vector analysis-derived indicator that signals increased risk. The qualitative findings, in turn, revealed how conceptual uncertainty, documentation practices, laboratory constraints and communication challenges limit the translation of this theoretical knowledge into consistent bedside practice. The findings across these phases highlight the importance of strengthening health‑system protocols to support clinical decision‑making, while also expanding the use of technological methods for risk assessment to improve outcomes related to refeeding syndrome. This thesis makes several contributions to paediatric nutrition and critical care. It consolidates and appraises existing evidence on refeeding syndrome in severe acute malnutrition in Sub-Saharan Africa, thereby clarifying current knowledge gaps. It provides pilot data on the feasibility and potential use of bioelectrical impedance vector analysis as a bedside method for assessing hydration status and refeeding syndrome risk in children with severe acute malnutrition during the early stabilisation phase. It also offers an in-depth account of healthcare practitioners’ perspectives on the barriers and enablers to recognising and documenting refeeding syndrome in public hospital paediatric medical wards. Based on these findings, the thesis recommends: (i) adopting clear, context-appropriate definitions and monitoring for refeeding syndrome within national and facility-level severe acute malnutrition guidelines; (ii) prioritising routine electrolyte monitoring, particularly phosphate and potassium, during the first week of inpatient stabilisation; (iii) further evaluating bioelectrical impedance vector analysis-informed risk stratification in larger, multi-centre cohorts; and (iv) implementing targeted in-service training, standardised documentation prompts and strengthened multidisciplinary communication to support earlier recognition and management of refeeding syndrome. Overall, the work supports a more proactive, physiologically informed and system-sensitive approach to preventing refeeding syndrome and improving outcomes for hospitalised children with severe acute malnutrition.