Impact of an electronic health record transition on chemotherapy error reporting

Background: Incident reporting systems allow for frontline employees to report errors and are a critical component of healthcare patient safety programs. Although incident reporting systems cannot quantify total errors, organizations can utilize incident reporting systems to help identify risks and trends to act upon. The objective of this article is to utilize incident reporting systems to evaluate trends in medication error reporting before and after implementation of a new electronic health record system. Methods: A five-month pre- and post-analysis was completed in a cancer hospital following electronic health record conversion by reviewing medication errors reported via the institution's voluntary incident reporting systems. Error reports included medication error category, date error was reported/occurred, patient location at time of error, harm severity score, medication(s) involved, medication use system node error originated/discovered in, medication source, narrative summary, and contributing factors. Data were analyzed using descriptive statistics within Office Excel. Results: Oncology medication error reports submitted pre- and post-electronic health record were 68 vs. 57, respectively. During the pre- and post-electronic health record conversion, a majority of errors had a harm severity index of 0 or 1; 12 (18%) in pre-electronic health record and 3 (5%) in post-electronic health record were level 2, and one (1%) in pre-electronic health record vs. 0 in post-electronic health record were level 3. Reported medication errors originated most commonly during the prescribing, administration, and preparation/dispensing phase and were primarily identified in the administration phase of the medication use process. The most frequently reported error category was 'wrong dose' followed by 'other' and 'overdose' in the pre-electronic health record phase and 'missing dose/delayed delivery' and 'order incorrect' in the post-electronic health record phase. The most frequently reported medications included methotrexate, chemotherapy (unspecified), and cisplatin. Conclusion: Analyzing data from incident reporting system reports allowed our institution to understand different trends of reporting in the cancer hospital following electronic health record adoption. Utilization of incident reporting systems must be combined with proactive risk identification approaches to enable systems-focused improvements to improve patient safety.