Use of an Electronic Adverse Event Reporting System in the Ambulatory Clinic Setting

Presenters: Kathryn Nelson, MHA; Karen Cox, PhD

Institution: University of Missouri Health Care, Office of Clinical Effectiveness

Introduction: The Institute of Medicine's report, To Err is Human, encouraged the development of confidential, voluntary error reporting systems at the organizational level. In response, the University of Missouri Health Care built the Patient Safety Net (PSN). The PSN is a software program and database, accessible via the Internet, for "Safety Report" entry by health care workers and clinicians (adverse events, medical errors, near-misses). The PSN also allows reporting of "Comments" (compliments, complaints, and suggestions) by patients, family members or visitors. Reports are completed using simple pull-down lists and text boxes for narrative. Staff members complete Safety Reports in the categories of Medication, Drug Reaction, Equipment/Device, Falls, Therapeutic/Diagnostic Interventions and Miscellaneous. Each report also includes error types that are specific to each category of Safety Report, such as "Wrong Dose" or "Wrong Patient" for the Medication category. Safety Reports also include contributing factors leading up to the error such as "Staffing Level", "Communication Breakdown", etc. A narrative section is also available for additional event descriptions not captured by drop-down categories. Once reports are completed, they are immediately viewable to clinic managers who are then responsible for timely investigation, resolution and documentation of findings/actions taken. Report entry and resolution for Comments works similarly.

Methods: In January 2002, after implementing the PSN in the hospital setting, an interdisciplinary team convened to explore features of an ideal adverse event reporting system for the ambulatory setting. The team developed a list of priority criteria for an ideal system and appropriate taxonomy. The system was implemented in 59 primary and specialty care clinics in late July 2002.

Results: In the first year following implementation (8/1/02-7/31/03), 664 reports from 59 clinics were submitted into the PSN, 211 of these were Safety Reports and 453 were Comments. The monthly reporting frequency varied from a high of 31 safety reports in the first month to a low of 12 in the 11th month. Compared with the baseline year of "paper-based" safety reports, there was a 197% increase in safety reports (from 71 to 211). Forty-two percent of reports were entered by RN/LPNs, 27% by Pharmacy staff, 7% by physicians, and 4% by "managers". In the baseline year, only six "medication" incidents were reported, compared with 78 in the post PSN implementation year. Over 50% of reports documented impact to the patient ranging from "inconvenience" to "hospital admission". In terms of the 453 comments, baseline data was not available as there had not been a central location for complaints, compliments, or suggestions. Of the 453 comments following PSN implementation, 12% were entered directly by patients using the web; all other entry was done on behalf of patients by staff. The comments consisted of 65% complaints, 8% suggestions, and 27% compliments. For both compliments and complaints, the highest-ranking categories were identical: staff behavior/attitude and quality of care.

Discussion: In the spirit of patient-centered care, our objective was to encourage reporting by anyone who was aware of any quality of care concerns. The 59 clinics had approximately 270,000 visits. If mandated to produce an "error-rate", the rate would have been 0.08% (211/270,000). We do not expect "error rates" to be a reliable indicator of safe versus unsafe clinics. Rather, we have focused on the development of a process to rapidly receive and respond to reports, to protect reporters from retaliation, and to promote identification and reporting of unsafe conditions. The PSN provides the unique ability to monitor error and complaint trends, for particular types of errors, or for any given patient population. For clinic managers and physicians to improve safety, safety report "resolution" must move from blaming individuals to studying flaws and re-designing systems. Unfortunately, systematically studying and correcting system deficiencies occurred in less than 6% of resolutions. Developing "system" skills will require focused study, practice, and reflection, experiences not generally included in today's curriculum.