Frequentie van en determinanten voor een juiste afhandeling van voorschrijffouten in Nederlandse ziekenhuisapotheken

Objective
To examine the frequency of correct handling of prescription errors using test patients and to identify features associated with correct handling.
Design
Prospective observational study with test patients in Dutch hospital pharmacies, using a case-control design to identify features associated with correct handling.
Methods
Test patients were supposed to be treated as real patients by the participating pharmacies while conducting medication surveillance. Their medication records contained intentionally added prescription errors (total 40 errors in 9 medication records divided among three test patients). An expert group had selected the errors in view of relevance and unequivocality. The pharmacies had to detect these errors and consequently report these errors to the investigator.
Results and conclusions
57% of the prescription errors were detected and handled correctly by the participating hospital pharmacies. The use of a computer system was independently associated with correct handling [OR 15.4 (CI95 3.6-65.5) for computerized physician order entry; OR 15.4 (CI95 3.6-65.7) for order entry by pharmacy technicians]. Furthermore, manual screening of dosage errors in children with or without computerized surveillance (OR 2.0; CI95 1.1-3.8), the qualification of pharmacy technicians (OR 1.3; CI95 1.0-1.7) and the use of protocols by pharmacy technicians (OR 1.3; CI95 1.0-1.6) were independently associated with correct handling. Handling of prescription errors can be subject to further improvement and the identified determinants can give direction to such improvements