Melody Cheuk Yin Leung, Catherine Sherrington, Abby Haynes, Ana Helena Salles Dos Reis, Stephen Gilbert, Leanne Hassett, Lyn Olivetti, Peter Youssef, Charlotte McLennan, Minghua Cao, Daniel Treacy, Rhuta Gole, Braydon Finnerty, Marina B Pinheiro
108 included studies were mostly conducted in acute settings (n = 47), involving stroke survivors (n = 44) and patients over 65 years (n = 52). Behavioural mapping (n = 75) was the most common technique with eight others identified. Commonly recorded outcomes included activity type, location, and social context (n = 63), documented using paper or electronic tools. Observation periods commonly spanned five days (Median 5, Range 1 to 68), nine hours per session (Median 9, Range 2 to 34) over twelve weeks (Median 90 days, Range 1 day to 3 years), involving two observers (Median 2, Range 1 to 13) per ward.
BACKGROUND: Physical inactivity is prevalent among hospitalised patients despite effective interventions. Current literature lacks clarity on how inpatient movement behaviours are observed and recorded. Greater methodological clarity is needed to support evaluation and implementation of inpatient physical activity interventions.
OBJECTIVES: To synthesise direct observation methods for assessing inpatient movement behaviour at ward level, enhancing consistency and utility for evaluating and promoting mobilisation in hospital settings.
DATA SOURCES: MEDLINE, Embase, PsycINFO, CINHAL, Cochrane, Scopus, and Web of Science were searched (inception to 18/3/2026).
ELIGIBILITY CRITERIA: We included studies investigating inpatient movement behaviour (physical activity and/or sedentary behaviour) through ward-level direct observation. Exclusions were studies on infants, undertaken in emergency departments, or targeting fixed-position limb activities.
RESULTS: 108 included studies were mostly conducted in acute settings (n = 47), involving stroke survivors (n = 44) and patients over 65 years (n = 52). Behavioural mapping (n = 75) was the most common technique with eight others identified. Commonly recorded outcomes included activity type, location, and social context (n = 63), documented using paper or electronic tools. Observation periods commonly spanned five days (Median 5, Range 1 to 68), nine hours per session (Median 9, Range 2 to 34) over twelve weeks (Median 90 days, Range 1 day to 3 years), involving two observers (Median 2, Range 1 to 13) per ward.
LIMITATIONS: Evaluation of method performance and addressing methodological heterogeneity were beyond our scope, but future work could assess robustness and support standardisation across settings.
CONCLUSIONS AND IMPLICATIONS: This review provides methodological guidance on ward-based direct observation of inpatient movement behaviour, supporting its use in development and evaluation of physical rehabilitation and broader hospital-based services to promote patient outcomes. The scoping review is not funded and is registered at the Open Science Framework via https://osf.io/t9bh5. CONTRIBUTION OF THE PAPER.