PS-072 Does the computerised physician order entry system reduce prescribing errors for inpatients? A before and after study. (14th February 2016)

• Record Type: Journal Article
• Authors: Rouayroux, N; Divol, E; Sallerin, B; Delzor, F; Bachelet, B; Cestac, P; Calmels, V
• Is Part Of: European journal of hospital pharmacy. Volume 23(2016)Supplement 1
• Page Start: A245
• Page End: A245
• Publication Date: 2016-02-14
• Subjects: Pharmacy -- Periodicals; Hospital pharmacies -- Periodicals; 615.1
• DOI: 10.1136/ejhpharm-2016-000875.556
• Languages: English
• ISSNs: 2047-9956

Abstract

Abstract : Background: The Health Institution recommends the use of health information technology to reduce the risk of iatrogenesis errors. While many publications highlight the benefits of computerised physician order entry (CPOE) system, others worry about the unintended consequences of such a system on healthcare quality. Purpose: The aim of this study was to measure the impact of computerisation on the quality of drug prescriptions. Material and methods: An observational before and after study was carried out in two medical units (diabetology and cardiology). It included all patients admitted during a 30 day pre- and a 30 day post-CPOE (ORBIS) implementation. The pharmacists analysed the drug prescriptions according to the methodology of the French Clinical Pharmacy Society. Medication errors due to the CPOE system were analysed quantitatively and qualitatively. Results: In the pre-CPOE period, 121 pharmacist Interventions (PI) recorded in the handwritten prescriptions of 321 patients were analysed. In the post-CPOE period, 144 PI recorded in the CPOE system of 282 patients were analysed. The ratio of PI per patient was 0.38 without the CPOE system and 0.51 with it (χ 2, p = 0.001). This ratio was increased significantly by 34% with computerisation. The CPOE system itself generated 27% of the errors. Among them, 30% were errors of dose units, 23% errors of prescription redundancies and 15% dosage errors. These prescribing errors were not reported with handwritten prescriptions, except for the dosage errors (2% of 121 PI). Without the errors linked to computerisation, the ratio remained unchanged. Conclusion: With the use of the CPOE system, the iatrogenic risk seemed to increase. A new type of error was observed: errors linked to the CPOE system. These errors can be due to a lack of ergonomics (poor readability of the prescriptions, complex functionality) or a misuse of the software by the physicians. However, they are avoidable. In order to reduce them, it is important to raise the level of awareness of the prescribers, to improve their training and to promote pharmacists' and nurses' vigilance. A partnership with the software publisher is essential to secure the CPOE system and make it evolve. References and/or Acknowledgements: We thank the pharmacists, physicians and nurses of the diabetology and cardiology departments. No conflict of interest.