Background and Importance In recent years, pharmacy services have shifted towards centralised preparation of sterile drugs to ensure compatibility, stability, and sterility. Quality controls will identify preparation errors preventing them from reaching patients.

Aim and Objectives Analyse errors detected in the manual preparation of sterile drugs from stock during January 2022 to April 2023.

Material and Methods The preparation of sterile drugs requires a series of quality/safety controls to detect errors, preventing them from reaching patients. Following a manual work methodology, a pharmacy technician selects the medicines/materials, generates the labels and records quantities, batches and expiry dates on the processing sheet. Another technician performs a double safety check. Once prepared, the pharmacist records the conformity, after inspecting the preparation sheet together with one of the preparations of each batch.

The incidents noted in the elaboration sheets from January 2022 to April 2023 were reviewed. The errors were recorded in a Microsoft Excel file, being classified based on the type and severity, according to pharmacist criteria: minor (errors on the preparation sheet, labels or batches); and serious (errors in expiry date or dose on the label, wrong administration system; wrong drug/serum, excess/defect dose, presence of particles/air, unfinished packaging, and unprotection from light).

Results 88 errors were detected, affecting 4.4% of the batches produced. 44.3% were considered minor errors and 55.7% were considered serious. The most frequent error was the completion of the processing sheets (26%). Regarding labelling, the most detected errors were related to expiry date (15.9%), batch (11.36%), dose/name/colour (2.3% each) and label hiding the graduation of the syringe (1.13%). Other errors: 9.1% of non-complete final packaging; 5.68% excess doses, 6.81% defect doses; incorrect serum and infusion systems (3.4% each); unprotected from light (3.4%); presence of particles/air (2.3%) and duplicate batches (1.13%).

Conclusion and Relevance The error rate detected is lower than that reported in the literature. More than half of them were considered potentially serious if they had reached the patient.

According to our results and the literature, this methodology presents a low error detection, incorporating new technologies (comprehensive software, barcode verification, image capture, gravimetry) could enhance error detection and reduce preparation errors, ultimately leading to improved patient safety.

Conflict of Interest No conflict of interest.