Testing Nurse's Medication Errors Model based on Reason Human Error Model

Medication errors are one of the most common medical errors. Since the performance of a theory in nursing is the extent of its application for description or prediction of evidence and also its potential in handling of the results as an occasional theory; this research was done to test the theoretical model of Reason human errors in nursing medication error. It was a descriptive, correlational study that was conducted at a large teaching hospital in Tehran, I.R. The nursing staffs (n = 150) were chosen in different wards affiliated to Shahid Beheshti University of Medical Sciences through convenience sampling. A researcher made questionnaire was utilized to collect demographic data, Velasquez nursing care complexity scale, Salyer work dynamic scale, Gittel et al., nurse-physician communication scale, Minick et al., work commitment scale, Wakefield et al., reporting medication error scale and Rybowiak et al., learning climate questionnaire were used to test the hypotheses related to individual and organizational factors based on Reasons Human Error Model. Using SPSS-PC (v.16), and LISREL 8.8 the data were analyzed by descriptive and path analysis tests (P_ value<0.01). The age of the majority of the respondents ranged between 31 to 40 years old (44.7%) and they had 10.60 ± 7.24 years of work experience. Model fit indices (CFI =1, GFI=0.99, RSMEA=0.00, = 4.83 RMR=0.65) showed that the paths considered for relationship between variables was appropriately predicted. Learning climate correlated with work commitment with standardized coefficient of (r =0.40) and learning climate correlated with physician - nurse communication with standardized coefficient of (r =0.20). There was no significant relationship between other variables. Based on the nursing medication error model, encouraging nurses to report error and using the model as a way to self-learning and other colleagues’ learning would be very helpful in the future. Also by creation of positive learning climate and modifying punishment for unintentional errors and proposing approaches to prevent recurrent errors in future occasions, mangers would be very influential.