Abstract
Despite advances in identifying malnutrition at hospital admission, decline in nutritional status of well-nourished patients can be overlooked. The aim of this systematic review was to investigate the incidence of hospital-acquired malnutrition (HAM), diagnostic criteria and health-related outcomes. PubMed, CINAHL, Embase and Cochrane Library were searched up to July 2021. Studies were included if changes in nutritional status was assessed with a validated nutrition assessment tool in acute and subacute adult (≥18 yrs) hospitalised patients. A random-effects method was used to pool the incidence proportion of HAM in prospective studies. The certainty of evidence was appraised using the Grading of Recommendation Assessment, Development and Evaluation system. We identified 12 observational cohort studies (10 prospective and 2 retrospective), involving 35,324 participants from acute (9 studies) and subacute settings (3 studies). Retrospective studies reported a lower incidence of HAM (<1.4%) than prospective studies (acute: 9-38%; subacute: 0-7%). The pooled incidence of HAM in acute care was 25.9% (95% confidence interval (CI): 17.3-34.6). Diagnostic criteria varied, with use of different nutrition assessment tools and timeframes for assessment (retrospective studies: >14 days; prospective studies: ≥7 days). Nutritional decline is probably associated with longer length of stay and higher 6-month readmission (moderate certainty of evidence) and may be association with higher complications and infections (low certainty of evidence). The higher incidence of HAM in the acute setting, where nutritional assessments are conducted prospectively, highlights the need for consensus regarding diagnostic criteria and further studies to understand the impact of HAM.