Background Electronic prescribing (EP) systems have been recognised as successful in reducing chemotherapy prescribing errors. However, electronic prescriptions are unlikely to prevent all errors, and new types of errors may emerge.

Purpose To assess prescribing error rates and identify new error types and their causes with the implementation of a electronic prescribing system for ambulatory cancer patients at a London Cancer Centre.

Materials and Methods A service evaluation was conducted in two parts, covering two different strategies for interception of prescribing errors – prospectively by pharmacists during a 2-week period, and retrospectively using data from the pharmacy EP telephone helpline service, over 41 weeks.

Results The overall rate of error-containing prescriptions was estimated to be 6%.

In the prospective part, 32 errors were identified from 571 electronic chemotherapy prescriptions. Most commonly committed errors were chemotherapy drug dose adjustments (13; 41%) and weight omissions (11; 34%).

In the retrospective analysis, 95 of 141 errors (67%) were ‘e-selection errors’, classified mainly as ‘work-arounds’ (26; 18%), ‘wrong commands’ (35; 25%), or ‘wrong fields’ (27; 19%). 63 errors (45%) were related to scheduling a chemotherapy or supportive drug or regimen.

Electronic system-related causes of prescribing errors were recognised in 4 of 32 cases (13%) in the prospective part, and in 89 of 141 cases (63%) in the retrospective part. It was estimated that with implementation of technical solutions and additional prescriber training, 85% of these errors could be prevented in the future.

Conclusions The estimated rate of chemotherapy prescribing errors was 6%. A number of different errors, specific for electronic prescribing, were identified, with a thorough explanation of how various errors may have occurred. Future larger scale studies are needed to confirm prescribing error rates, and to possibly identify other, previously unrecognised, types of chemotherapy prescribing errors.

No conflict of interest.