[Pharmacists' interventions before and after prescription computerization in an internal medicine department].

Objective: To compare and analyze the number and types of pharmacist interventions when prescriptions were handwritten (period 1) or entered on 2 different computerized physician order entry systems (CPOE): Phedra (period 2) and Actipidos (period 3).

Methods: This study took place over 54 weeks (18 weeks for each period) in a 46-bed internal medicine department. Pharmacist interventions were categorized as either simple substitution to an available drug or "complex" interventions such as changing dosage (D), changing drug (M), stopping or substituting in cases of contraindication or overdosing (CO), new medication (N) and identifying errors due to CPOE (C).

Results: The study analyzed 12420 prescriptions that led to 1420 interventions. There were 720 substitutions and 98 interventions in period 1, 40 and 238 in period 2, and 97 and 227 in period 3. The percentage of interventions by types for each of the three periods, respectively were D: 52, 37 and 34%; M: 21, 22 and 35%; CO: 16, 12 and 16%; N: 11, 5 and 2%; and C: 0, 24 and 13%. The errors due to CPOE were mainly wrong dosage units and duplicate orders.

Discussion: Introduction of CPOE drastically reduced the number of simple substitutions and significantly increased the complex interventions. CPOE introduced new risks of serious errors.

Conclusion: In our study, CPOE did not prevent medication errors and led to new types of errors. The presence and intervention of clinical pharmacists remained necessary.