Use of the failure mode and effect analysis tool in the clinical medication process in an intensive care unit

Abstract

Objective: To describe failure modes and establish contingency measures related to the clinical medication process using medical prescriptions of patients admitted to an Intensive Respiratory Therapy Unit (UTIR), using the Failure Mode and Effects Analysis (FMEA) tool. Methods: This is a descriptive and cross-sectional study carried out in an Intensive Care Unit of a public hospital in Fortaleza, Brazil, from November/2015 to March/2016. Study population included adults aging ≥ 18 years in intensive care at the UTIR. The study included the medical prescriptions released on Mondays, Wednesdays, and Fridays. The study was divided in five phases: situational diagnosis, formation of a multiprofessional team, assessment of failure modes (FM), monitoring of FM and calculation of the priority coefficient (PC). In the FM assessment, scoring of the three indicators of the FMEA was used within a range of 1-10, whereas a score of 10 characterized the most concerning situation. Therefore, the indicators gravity (G), prevalence (P) and detection (D) were analyzed. The study was carried out with an active interaction between the subjects of the group and several in-person and virtual sessions were performed. Drugs used in the study were categorized for therapeutic class, according to the Anatomical Therapeutic Chemical Classification System. Data analysis was performed using Microsoft Office Excel® 2013 software. Results: 301 prescriptions were analyzed, with the identification of 452 FMs, which related mostly to systemic antibacterials (21.6%, n = 8), psycholeptics (13.5%, n = 5) and antithrombotic agents (10.8%, n = 4). FMs were divided in eleven categories, from which “drug interaction” (36.8%; n = 14), “dose adjustment” (21.1%, n = 8) and “food-drug interaction” (7.9%, n = 3) were the most frequent. The PC of the detected FMs varied between 28 and 294, and 42.1% (n = 16) of them presented PC above 100. Median of the indicators G (6 – min: 3; max: 9), D (7 – min: 3; max 7) and priority coefficient (72 – min: 28; max: 294) indicate that FM had generally moderate gravity, low prevalence and low detection. For the majority of FMs (72.7%, n = 28), the chosen conduct was ‘not to accept’ and the established contingency measure included a sentinel event notification. Conclusion: The use of FMEA enabled the identification, classification, and prioritization of risks of the clinical medication process in the UTIR. This study indicates the need to implement measures that increase safety in the clinical practice of the study Intensive Care Unit.