Article Text

Download PDFPDF
CP-046 Introducing computerised physician order entry systems: does following a protocol make prescribing safer?
  1. E Davidau,
  2. H Junot,
  3. I Peyron,
  4. D Oumeziane,
  5. P Tilleul
  1. Groupe Hospitalier Pitie Salpetriere AP HP, Pharmacy, Paris, France


Background A computerised physician order entry system (CPOE) has been implemented to improve the safety of medicines prescribed in three surgical wards (orthopaedics, urology, abdominal surgery). To facilitate its adoption, protocols have been created to make ordering prescriptions easier.

Purpose To estimate the impact of these standardised treatments on patient safety and the quality of prescribing, one month after implementation.

Materials and methods For two weeks, hospital pharmacists analysed all patient prescriptions and evaluated the rates of protocol adherence and of modifications of these protocols. Pharmacists listed errors made when the prescribing protocol was followed, with or without modifications. They also listed the number of prescriptions that should have been prescribed according to a protocol, and estimated the rate of prescription errors in them.

Results 50 protocols were created (for premedication prescriptions, post-operative analgesia prescriptions, etc.). Of 415 orders created, 71.6% contained at least one drug prescribed with reference to a protocol (297 orders; total of 577 protocols). 27% of all protocols prescribed (n = 156) were not followed as originally envisaged, due to which 6.4% (n = 10) of prescriptions contained an error. To sum up, prescription errors occurred on 5.3% of orders with a prescription protocol (n = 22 orders, 24 protocols). 2.4% of order errors were due to a protocol modification (e.g. forget to stop after change of dose), and 3.4% of order errors were due to a protocol prescription without a change on it (lack of switch to oral route or switch to oral route without stopping infusion, redundancies on the prescription, etc.). On the other hand, some physicians refer to the protocols (28.4% of prescriptions): errors were identified in 43.2% of those orders (forget to stop, infusions wrongly prescribed etc.).

Conclusions These results show that prescribing without following a protocol increases the risk of order error and that following a protocol makes for safer prescribing. Nevertheless, a residual risk of drug error still remains with protocols, showing that pharmaceutical checking is necessary. To minimise this risk, protocols should be regularly updated to reflect current medical practice.

No conflict of interest.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.