Relationship of restraint use, patient injury, and assaults on EMS personnel

We hypothesized that the assaults on EMS personnel by patients requiring restraints can be correlated with demographic information, patient condition, and other scene information such as presence the of law enforcement.

The study was a one-year cross-sectional study of paramedic restraint use and assault on EMS personnel in an urban area. A data collection form was completed by EMS for each patient placed in restraints. Study outcome variable was "Assault on EMS personnel." Predictor variables included demographic and EMS call information, patient condition, law-enforcement related variables, and the paramedic's perception of the need for chemical restraints. To compare predictor and outcome variables, a multivariable model with odds ratios and 95% confidence intervals was used.

The study included 271 restrained patients over a 12-month period from April 2002 to April 2003. Seventy-seven (28%) cases were positive for assaults on EMS personnel. Multivariable analysis including 8 variables, indicated the following 6 variables were associated with assault on EMS personnel: time of day between midnight and 6 am (OR = 4.4, 95% CI = 1.6-12.7); female patient (OR for males 0.6, 95% CI = 0.3-1.0); violent patient (OR = 10.1, 95%CI = 2.3-48.2); patient injured under supervision (OR = 3.9, 95% CI = 1.1-13.8); arrested patient (OR = 4.4, 95% CI = 1.1-18.5); and perceived need for chemical restraint (OR = 2.1, 95% CI = 1.2-3.9).

Multiple factors are correlated with assaults on EMS personnel by patients requiring restraints. By specifically targeting patients exhibiting these factors, EMS providers can help prevent injury to themselves. Patients not exhibiting these factors may be less dangerous.