Procedures
Data were collected from August 2012 to January 
2013, and the sampling method was by convenience. 
An attempt was made to include all inpatients that met 
inclusion criteria in those 6 months. Some of the authors 
visited the PED daily, gathered data about hospitalizations 
in the last 24 hours, interviewed patients and reviewed 
medical records to fill out the study form.
One of the authors was exclusively assigned to apply 
the OAS scale by interviewing the patient and a family 
member, as well as talking to the nursing staff and the 
attending physicians. As several aggressive incidents 
may be part of the same episode, which may last for 
hours, we considered each 24-hour period as an episode.
Six resident physicians were responsible for filling 
out the scales after interviewing the patients, which 
should be done within 24 hours of admission, taking into 
account patient condition at admission.
Diagnosis were assessed by the PED and PU medical 
teams according to their service routines. During patient 
stay in the hospital, diagnosis were discussed with 
the resident´s supervisor and recorded in the chart at 
discharge using the 10th revision of the International 
Classification of Diseases (ICD-10). No specific diagnostic 
instruments were used because of the practical difficulties 
of their application, such as the time necessary to fill 
in the questionnaire and the patient conditions in the 
emergency department, sometimes agitated and 
sometimes sedated. In addition, our purpose was not to 
associate diagnosis and aggression, because diagnoses 
were less important than clinical and psychopathological 
characteristics in our study.
Each admission was counted separately, regardless 
of how many times the patient had been admitted in 
the period. The same patients may have a different 
psychopathological condition each time they are 
admitted, and may be either aggressive or not.
When data collection was concluded, all the forms were 
reviewed. After that, data were entered into Excel by two 
operators, to avoid typing errors. This study was approved 
by the Research Ethics Committee of Universidade 
Federal de Santa Maria and was registered in the Brazilian 
Research Ethics Database (Plataforma Brasil; CAAE