Efficient and accurate measurement of physical restraint use in acute care.

Objectives: Describe and evaluate a method for assessing whether physical restraint prevalence differs by timing and frequency of data collection and to determine the minimum period of observation necessary to provide accurate prevalence estimates on both Intensive Care Unit (ICU) and medical-surgical units.

Design: Two-period, cross-sectional design with repeated observations in year 1 for 18 consecutive days and in year 2 for 21 consecutive days with method modifications.

Setting: 400-bed urban teaching hospital.

Participants: All beds on general medical, surgical, and intensive care units.

Measurement: Direct observation of patients, nurse interview, and medical record review conducted by trained observers.

Results: There were no significant differences in mean restraint use prevalence rates comparing: (a) morning and evening periods; (b) weekdays and weekend days; and (c) observation periods of 7, 14, or 21 consecutive days or for 7 days using every 3rd day on either medical-surgical units or ICUs. Analyses using data from an increasing number of days of observation indicates that the mean prevalence rate stabilizes after 16 days. There were larger mean differences for comparisons on ICU-ventilator units and lack of significant differences may be due to low statistical power.

Conclusion: Direct observation by trained observers, supplemented by nurse report and medical record documentation over brief monitoring periods, results in accurate, nonintrusive, cost-efficient estimates of physical restraint prevalence. As few as seven consecutive or nonconsecutive days in measuring restraint prevalence is sufficient to obtain accurate estimates, although the number of days may vary depending on patient mix and unit type.