adverse events, mechanical restraints, nurse staffing



  1. Whitman, Gayle R.
  2. Davidson, Lynda J.
  3. Sereika, Susan M.
  4. Rudy, Ellen B.


Background: In an effort to enhance patient safety in acute care settings, governmental and regulatory agencies have established initiatives aimed at limiting the use of mechanical restraints. Concurrently, hospital staffing levels are undergoing changes raising concerns about the impact these changes may have on restraint use. No studies to date have described the impact these two initiatives have had on restraint use in acute care hospitals.


Objectives: To determine across a multiple hospital system: (a) the rates, frequencies, duration, and timing of restraint use, and (b) the relationship between restraint use and staffing.


Methods: This was a secondary analysis of prospective, observational data from a large outcomes database for 10 acute care hospitals. Monthly data were obtained from 94 patient care units for periods ranging from 1-12 months for a total of 566 cumulative months during 1999.


Results: The system restraint application duration rate (total restraint hours/total possible hours) was 2.8% (hospital ranges: 0.3-4.4%). More restraints were applied on night shifts (48.8%;n = 5,296) than on day (33.5%;n = 3,634) or evening shifts (17.7%;n = 1,926) (p < .0001) and most applied at midnight (31.7%;n = 3,441) followed by 0600-0900 (33.3%;n = 3,614). There was a weak positive relationship between staffing and restraint use (r = 0.276, p = .0001) at the system level and units with higher staffing levels also had higher baseline restraint use (p < .0001).


Conclusions: Restraint frequency, duration, and timing may have been altered by recent initiatives, and there is beginning evidence that differences exist between community, rural, and tertiary hospitals. While there is a weak positive relationship between higher staffing and restraint use at the system and unit level, further exploration of the influence of other factors, specifically patient acuity, are in order. The finding of unit variability and consistent restraint application times provides a starting point for further quality initiatives or research interventions aimed at restraint reduction.