Objective The Pharmacy Paediatric Intensive Care (PICU) team record interventions that affect a patient's episode on PICU. These interventions are not always reported via the hospital's incident reporting route. The aim of this report is to highlight information that may benefit or influence a number of healthcare professionals.

Methods A database of PICU pharmacist intervention records was designed. Categories were allocated to each intervention as indicated on the intervention proforma. A single intervention can be assigned into more than one category. Categories not included on the proforma were added to the database, these included premature neonate, pharmacokinetic issues, compatibility, and education/advice. Each intervention was coded according to the Medication Error Reporting and Prevention (MERP) algorithm.1 Drugs listed in the guidance from the National Patient Safety Agency (NPSA) or Institute of Safe Medication Practice (ISMP) were coded on the database.

Results 287 drug interventions were added to the database, involving 102 different drugs. Over 30% of the interventions involved drugs listed as high risk of causing potential harm by the ISMP or NPSA and 62% were drugs given via the intravenous route. Numbers of intervention corresponding to MERP medication categories are shown in the table below

Pharmacists prevented more errors reaching patients on Fridays than other days of the week, whereas patient harm had occurred more often before pharmacists intervened on Mondays. Following increased pharmacist time on the unit, there was an increase in numbers of interventions per unit of time.

One in 10 interventions that did not reach the patient were made during the attended morning ward round, one in five since introduction of ad hoc visits to the unit in the afternoon.

The most common reasons for intervening was found to be altered renal handling, monitoring issues, wrong frequency, drug not prescribed. Sixteen per cent required the pharmacists knowledge of altered drug handling, for example, prematurity or renal impairment. Antimicrobials were the class of drug requiring most intervention, most commonly during to altered handling. The most frequent drugs involved were ranitidine, vancomycin, parenteral nutrition, aciclovir, caffeine and gentamicin. Non-PICU doctor initiated prescription accounted for 17% or the interventions. In 8.3% of interventions the opportunity was also taken to provide education or advice to the doctor. Missed therapy or need for a drug accounted for 14.7% of interventions, whereas only 2.7% involved stopping therapy. One in 20 interventions included problems with drug compatibilities.

Conclusion Analysis from the intervention records are now used for targeted education for medical staff; individuals and general education programme. Problem drugs can be prioritised for guideline production and alerts are added into the drug reference books to alert staff. Interventions by the PICU team should be recorded via the incident reporting system.