Accelerometry Shows Inpatients With Acute Medical or Surgical Conditions Spend Little Time Upright and Are Highly Sedentary: Systematic Review

Abstract

Background: Physical inactivity and sedentary behaviors have significant and independent effects on health. The use of wearable monitors to measure these constructs in people who are hospitalized with an acute illness is rapidly expanding, but has not been systematically described.

Purpose: The purpose of this study was to review the use of accelerometer monitoring with inpatients who are acutely ill, including what activity and sedentary behaviors have been measured and how active or sedentary inpatients are.

Data sources: Databases used were MEDLINE, EMBASE, CINAHL, and Scopus.

Study selection: Quantitative studies of adults with an acute medical or surgical hospital admission, on whom an accelerometer was used to measure a physical activity or sedentary behavior, were selected.

Data extraction and data synthesis: Procedures were completed independently by 2 reviewers, with differences resolved and cross-checked by a third reviewer. Forty-two studies were identified that recruited people who had medical diagnoses (n = 10), stroke (n = 5), critical illness (n = 3), acute exacerbations of lung disease (n = 7), cardiac conditions (n = 7), or who were postsurgery (n = 10). Physical activities or sedentary behaviors were reported in terms of time spent in a particular posture (lying/sitting, standing/stepping), active/inactive, or at a particular activity intensity. Physical activity was also reported as step count, number of episodes or postural transitions, and bouts. Inpatients spent 93% to 98.8% (range) of their hospital stay sedentary, and in most studies completed <1,000 steps/day despite up to 50 postural transitions/day. No study reported sedentary bouts. Many studies controlled for preadmission function as part of the recruitment strategy or analysis or both.

Limitations: Heterogeneity in monitoring devices (17 models), protocols, and variable definitions limited comparability between studies and clinical groups to descriptive synthesis without meta-analysis.

Conclusions: Hospitalized patients were highly inactive, especially those with medical admissions, based on time and step parameters. Accelerometer monitoring of sedentary behavior patterns was less reported and warrants further research.