Objective: To identify the prevalence of clinically significant prescription errors in a Brazilian university hospital compared with their occurrence in 2003 and 2007.
Methods: Variables and group of variables, such as readability, compliance with legal and institutional procedures of prescription, and prescription errors analysis were analyzed.
Results: When the prevalence rates of clinically significant prescription errors were calculated, a statistically significant decrease was shown [year of 2003 (29.25%), year of 2007 (24.20%); (z = 2.99; p = 0.03)], reflecting on the safety rate [year of 2003 (70.75%), year of 2007 (75.80%); (z = 3.30; p =0.0001)].
Conclusion: Despite significant, the increased safety rate reflected the quantitative reduction of errors, with no observed difference in severity between the studied periods. Our results suggest the institutional steps taken could reduce the number of errors, but they were ineffective in reducing the severity of the errors.