Article Text

Download PDFPDF
CP-082 Cost benefit analisys of a computerised physician order entry project: a method of benefit evaluation based on error analysis
  1. N Jannitti,
  2. M Sansone,
  3. S Di Segni,
  4. M Pasquantonio,
  5. AM La Malfa
  1. Istituti Fisioterapici Ospitalieri, Pharmacy, Rome, Italy


Background The computerised physician order entry (CPOE) system is considered the single most effective technological solution for reducing medical errors, but high costs and limited data on its financial benefits are major barriers to adoption.

Purpose To conduct a cost benefit analysis (CBA) of implementing a CPOE system, using local error analysis and an estimate of avoided adverse drug events (ADEs) in order to monetize benefits.

Materials and methods In a cancer hospital all prescription errors related to injectable antineoplastic drugs were analysed by the Pharmacy (January–April 2012) in order to identify errors potentially leading to ADE. The value of ADE was obtained by a broad-based literature search. Only tangible costs were considered.

Results A CPOE system could save pharmacists 4 h/day of work, 1/4 of the work pharmacists dedicate to prescription analysis. This work avoids 50 ADEs/year. The benefits of CPOE would be 1/4 of the value of avoided ADEs. Costs of ADE in cancer are 1,300–3,500 €; good-quality CPOE software costs approximately 30,000 €. Our CBA shows return on investment (ROI) in the 4th, 3rd, 2nd year, depending on the chosen value for ADE. Compared to the literature, this earlier breakeven point seems due to much lower costs and not to an overestimation of avoided costs. Avoided costs are cautiously estimated on the basis of a 0.2% clinically relevant error rate; error rates of 1% and 2% were reported in two recent studies on prescribing error prevention in relation to injectable antineoplastic drugs.

Conclusions The local data analysis was useful in monetizing pharmacists’ activity and time saved by adopting CPOE. In our hospital, the ROI for CPOE project reached in the 4th year would be the worst case scenario. This new-found data was not supported in most of the literature covered, which showed a ROI occurring much later on.

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.