Monitorización de errores de medicación en dispensación individualizada mediante el método "carro centinela": comparación entre dos sistemas de distribución en dosis unitaria

Thumbnail Image
Publication date
Reading date
Journal Title
Journal ISSN
Volume Title
Medication errors (ME) in hospitals pose a significant clinical problem involving a potential risk to the patient, as well as a source of possible claims and compensations. ME reduction in the Prescription and Transcription phases, especially after the introduction of electronic prescribing systems, has been the subject of numerous studies. However, the generation of ME during Dispensing including preparing and distributing medication up to its final Administration, and primary responsibility of Pharmacy Services, has been scarcely analyzed. Aims: To analyze the impact on the quality of the process of unit dose (UD) dispensing as a result of its progressive automation, by comparing two systems of distribution and monitoring with the Sentinel Cart (SC) method. Methods: Design and Scope of the study: Observational, prospective and randomized sampling, conducted in two study periods: before (semi-manual (SM) system) and, after full automation ((FA) system) of the dispensing process; Methodology: Monitoring for 24 h every week of a randomly selected medication cart (SC), for the determination and registration of the following Variables: Dependent(s) variable(s): Medication Errors (ME) Claims (C) and Returns (R) of medication from the wards. Independent variables: Administrative and demographic, Clinical (ward specialty and type of medication, type and cause of ME, cause of C, cause of R)) and operational (technical time and cost). Statistical analysis: Descriptive statistics, comparison of means (T test or analysis of variance) for quantitative variables and proportions comparison (z test or Poisson test) for qualitative variables. Relative Risk (RR) between both periods for main dependent and independent variables. Multivariate analysis of clinical and operational parameters that could be related to the generation of ME (mixed effects analysis). Cost analysis. Results: 130 carts (19,973 lines of medication) for 19 UH during the first Period (SM system), and 52 carts (11,782 lines, 21 UH) in the second (FA system) were monitored. A statistically significant decrease in the rate of ME between SM and FA systems: 1.73% vs 1.39%; RR [IC 95%]: 0.81 [0.67-0.97], as well as a stastically significant increase in the rates of D: 4.9% vs 6.17%; RR [C 95%]: 1.26 [1.15-1.38] were found. No statistically significant difference were found in the rates of C between the two systems. Among the types and causes of ME, absence of medication (32.2%) and human error (49.1%) showed the highest prevalence, respectively. Non dispensed medication (24.7%) and patient discharge (31.1%) were the most frequent causes for C and R, respectively. In addition, multivariate analysis (logistic regression with mixed effects analysis) identified a decreased risk of ME during the second period (FA system), associated with the crossed random effect of the variables Ward Specialty and type of Medication and Odds Ratio 0.16 [0.12-0.22]. A statistically significant decrease of 17,5 min (51,8%) per 100 lines of medication, between SM and FA dispensing systems, was observed. This resulted in an estimated savings of € 3,23 per 100 lines of dispensed medication and annual net accumulated of 56200,5€ (42258,8 € including amortization of equipment). Conclusions: The implementation of a new system for medication preparing and dispensing, based on its progressive automation and monitoring by the SC method, improves system efficiency (less ME) and effectiveness (lower technicians time consuming and operating costs). Key words: Medication Errors, Dispensing, Unit Dose, Quality Control.
Bibliographic reference