The definition of medication error mainly refers to the definition provided by the US National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP). A medication error is defined as “any preventable event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer” [3]. Medication errors occur at each stage of the medication process, including prescribing, dispensing, administering, monitoring, and use.

Dispensing medication is the core function of pharmaceutical care and involves a complex combination of processes, technologies, and human interactions. Dispensing errors include, but are not limited to, dispensing the medicine for the wrong patient, incorrect medicine name, incorrect strength, and incorrect dosage [7].

Our hospital experiences 190,000 outpatient visits; 12,000 emergency visits; 7600 inpatient admissions; and 5100 surgeries monthly. To evaluate the impact of implementing medication-related technologies (ADCs, BCMAs, and SDCs) on patient medication safety, we conducted a before-and-after study at our 2202-bed academic medical center in Taiwan.

We analyzed medication error rates within the medication use process from January 1, 2017, to December 31, 2023, using data from the China Medical University Hospital Patient Safety Database (Multimedia Appendices 1 and 2). Our analysis specifically focused on medication errors that occurred during pharmacist medication dispensing. Power business intelligence was used for visual data analytics to conduct a frequency and cause analysis of these errors.

To address common errors in pharmacist medication dispensing, we used ADC, BCMA, and SDC as interventions. The study periods were defined as follows: stage 0 (preintervention, January to November 2017), stage 1 (post-ADC intervention, December 2017 to June 2018), stage 2 (post-BCMA intervention, July 2018 to October 2020), and stage 3 (post-SDC intervention, November 2020 to December 2023).

All medication orders, including those written or cosigned by physicians during the study period, were included in the analysis. The pharmacy, providing continuous services for both outpatients and inpatients, adheres to standardized protocols for medication preparation and dispensing. The primary outcome measure was the monthly rate of prescription dispensing errors before and after the implementation of medication-related technologies (ADC, BCMA, and SDC). This rate was calculated by dividing the number of reported prescription dispensing errors by the total number of prescriptions filled each month. Secondary outcome measures included the types of dispensing errors and their severity. The severity of medication errors was assessed by two pharmacists using the NCC MERP method. In case of disagreement, consensus was reached through discussion with a third senior pharmacist.

Descriptive statistics were used to summarize the dispensing error rates at each stage of the medication use process. The dispensing error rate was calculated by dividing the total number of dispensing errors by the total number of medications dispensed. We compared the incidence of dispensing errors before and after the implementation of medication-related technologies (ADCs, BCMAs, and SDCs). Furthermore, we stratified the analyses by the type and severity of dispensing errors.

Differences among the intervention periods were analyzed using chi-square or Fisher exact tests. The Bonferroni method was applied to adjust for multiple comparisons. All statistical analyses were conducted using R software (version 4.1.0, R Foundation for Statistical Computing). Statistical significance was defined as P<.05.

This study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of the Research Ethics Committee at China Medical University Hospital, Taichung, Taiwan (approval CMUH114-REC2-041). Following the 'Patient Safety Incident Reporting and Reward Operation Guidelines' of China Medical University Hospital, reporter information is automatically identified and transferred upon acceptance and establishment of a patient safety incident report, facilitating subsequent review and collaborative processes. This study utilized secondary use data, derived from a descriptive qualitative analysis of the patient safety notification database, to quantify error type categories related to healthcare information technologies. Notably, this study did not involve any patient or staff identification, and therefore, participant compensation disclosure is not applicable.

Table 1 presents the proportions of prescription dispensing errors by prevalence, type, and potential severity before and after the implementation of medication-related technologies. A total of 15,410,968 medication orders were dispensed in stage 0 (preintervention), followed by 10,721,238 in stage 1; 44,193,666 in stage 2; and 56,260,136 in stage 3. The total number of dispensing errors observed were 968, 406, 1556, and 773 in stages 0, 1, 2, and 3, respectively. The average number of dispensing errors per month decreased from 88 in stage 0 to 58 in stage 1, 56 in stage 2, and 20 in stage 3. Correspondingly, the dispensing error rates declined from 0.0063% in stage 0 to 0.0038% in stage 1, 0.0035% in stage 2, and 0.0014% in stage 3.

In stage 0 (preintervention), the 3 most frequent types of dispensing errors were wrong drug (459/968, 47.4%), wrong dose (335/968, 34.6%), and wrong technique (55/968, 5.7%). While the specific ranks varied slightly across stages, wrong drug, wrong dose, and wrong patient consistently ranked among the top 3 types of dispensing errors. According to the NCC MERP classification, the severity of harm (97%-98.8%) was categorized as “A” (no error) in all stages. The remaining severity of harm primarily fell into categories “B-D” (error, no harm). In stage 2, two cases were categorized as “C” (dose omission and wrong dose), and one case was categorized as “D” (wrong drug), all with no harm. In stage 3, only one case of “C” (wrong dose) was observed, again with no harm.

Figure 1 shows a significant decrease in the rate of reported prescription dispensing errors across the observation periods. Compared with the preintervention stage 0 (traditional manual medication picking), where the average dispensing error rate was 0.0063%, a notable reduction was observed in subsequent stages: stage 1 (post-ADC intervention) at 0.0038% (39.68% decrease), stage 2 (post-BCMA intervention) at 0.0035% (44.44% decrease), and stage 3 (post-SDC intervention) at 0.0014% (77.78% decrease).

The study revealed that prescription dispensing error rates significantly decreased after the introduction of medication-related technologies, including ADC, BCMA, and SDC. Specifically, error rates reduced by 39.68%, 44.44%, and 77.78% from stage 0 (preintervention) to stage 1 (post-ADC intervention), stage 2 (post-BCMA intervention), and stage 3 (post-SDC intervention), respectively. The average number of dispensing errors per month also significantly decreased from 88 in stage 0 to 58 in stage 1, to 56 in stage 2, and to 20 in stage 3. The most frequent error type in stage 0, wrong drug, was significantly reduced by 51.15%, 56.85%, and 81.26% in stages 1 to 3, respectively. Similarly, the wrong dose, the second most frequent error type, was significantly reduced by 42.93%, 42.85%, and 82.34% across the stages. Compared with stage 0, the severity of harm, as categorized by category A (no error) and category B (error, no harm), was significantly decreased in all subsequent stages. These findings demonstrate the effectiveness of medication-related technologies in reducing dispensing errors in a hospital pharmacy and improving medication safety.

In stage 1 (post-ADC intervention), the dispensing error rate decreased significantly from 0.0063% in stage 0 (preintervention) to 0.0038% (a reduction rate of 39.71%). This reduction was observed across all error types, including wrong drug, wrong dose, and the total number of dispensing errors. Notably, the reduction rate of wrong drug was substantial, decreasing from 0.0030% to 0.0015% (a 51.15% reduction), and the reduction rate of wrong dose also decreased significantly from 0.0022% to 0.0012% (a 42.93% reduction). These findings are consistent with previous studies that have demonstrated the effectiveness of ADCs in reducing dispensing errors [22,25-27]. The primary impact of ADC implementation was observed in a significant reduction of wrong drug and wrong dose. Previous research by Shah et al [26] highlighted that while ADCs can eliminate certain error types, they may also introduce new error types from a pharmacist’s perspective [26]. Therefore, pharmacists play a crucial role in ensuring the safe and effective use of ADCs by implementing appropriate policies, procedures, and quality assurance programs to address safety, accuracy, security, and patient confidentiality [28].

In stage 2 (post-BCMA intervention), the dispensing error rate decreased significantly by 43.95% compared with stage 0 (0.0063% vs 0.0035%). These findings were consistent across all error types, including wrong drug, wrong dose, monitoring error, wrong technique, dose omission, and the total number of dispensing errors. Notably, the reduction rate of the wrong drug was substantial, decreasing from 0.0030% to 0.0013% (56.85% reduction). The reduction rates of wrong dose, dose omission, and monitoring errors were also significant, decreasing from 0.0022% to 0.0012% (42.85% reduction), from 0.00015% to 0.00005% (66.67% reduction), and from 0.000058% to 0.000005% (92.25% reduction), respectively. The wrong technique error rate also decreased significantly from 0.0004% to 0.0001% (67.03% reduction). Compared with stage 1, only the wrong technique error rate showed a statistically significant difference. These findings are consistent with previous studies [29-31] and demonstrate the effectiveness of BCMA in reducing dispensing errors, particularly in preventing wrong drug, wrong dose, dose omission, monitoring error, and wrong technique.

Consistent with other studies, our analysis showed that BCMA implementation was associated with a reduction in dispensing errors, especially for medications with high error rates or those prone to confusion. However, the effectiveness of BCMA relies heavily on proper implementation, including comprehensive education and training on standard operating procedures. For example, simultaneous scanning of multiple medications can lead to errors, and issues with barcode quality or accuracy can also hinder the system’s effectiveness. In addition, damaged and unreadable drug barcodes, as well as medications labeled with incorrect barcodes, can also cause problems.

In stage 3 (post-SDC intervention), the dispensing error rate decreased significantly by 78.13% compared with stage 0 (0.0063% vs 0.0014%). These findings were consistent across all error types, including wrong drug, wrong dose, wrong patient, dose omission, monitoring error, wrong technique, and the total number of errors. Notably, the reduction rate of wrong drug errors was substantial, decreasing from 0.0030% to 0.0006% (81.26% reduction). The reduction rates of wrong dose and wrong patient errors were also significant, decreasing from 0.0022% to 0.0004% (82.34% reduction) and from 0.0003% to 0.0001% (59.56% reduction), respectively. Monitoring errors were completely eliminated in stage 3. The wrong technique and dose omission error rates also decreased significantly. Compared with stage 1 and stage 2, significant reductions were observed in multiple error types in stage 3, including wrong drug, wrong dose, wrong patient, and others. The implementation of SDC demonstrated significant reductions in wrong drug, wrong dose, wrong patient, dose omission, monitoring error, and wrong technique. These findings are consistent with a previous study by Teo et al [24], which observed that technology-assisted medication picking systems, such as an LED-guided system with a lockable drawer, significantly reduced near-miss medication errors in a tertiary referral hospital in Singapore.

Our observations highlight the potential of the SDC system, designed to address the unique dispensing needs of Taiwan, to significantly reduce medication errors. By incorporating visual, auditory, and perceptual cues, the SDC system can guide pharmacists through the dispensing process, minimizing errors that may occur due to factors such as infrequent medication use, similar packaging, or distractions. This is particularly important for new pharmacists who are still familiarizing themselves with medication locations and workflow.

This study has several limitations. First, variations in medication error definitions across the literature make comparisons challenging. We attempted to standardize error classification according to the NCC MERP guidelines and calculated error rates based on the total number of dispensing errors and medication orders per prescription. Second, the staggered implementation of ADC, BCMA, and SDC (December 2017, July 2018, and November 2020, respectively) may have introduced a time-related bias. To mitigate this, we implemented standardized staff training on system operation and troubleshooting and provided on-site technical support. Furthermore, direct observation of drug dispensing was not conducted, making it difficult to confirm the absence of management errors despite random checks of standard dispensing procedures.

Automated systems may face challenges with irregularly shaped or infrequently used medications, requiring backup systems during downtime and maintenance [32]. Concurrent manual workflows during automation can also introduce potential sources of error. Based on our findings, optimizing the benefits of dispensing and management technologies requires adequate resource allocation for postimplementation support, including education, training, and technical assistance. The occurrence of system-related errors highlights the need for continuous monitoring and proactive problem-solving. For example, the system should track the movement of drugs after delivery, including any necessary withdrawals by pharmacists and nursing staff. Clear procedures for responding to system failures must be communicated to all employees. To maximize the benefits of these technologies, continuous evaluation, system refinement, and process optimization are crucial.

This study demonstrated that the implementation of 3 medication-related technologies (ADC, BCMA, and SDC) significantly reduced dispensing errors compared with traditional manual medication picking. Overall, the study observed a significant reduction in dispensing errors across all stages. Except for wrong dosage form, wrong strength, wrong time, and other minor error types, all major error categories (including wrong drug, wrong dose, and wrong patient) showed statistically significant reductions (P<.001). The severity of harm also decreased significantly at each stage (P<.001). Significant differences in dispensing error rates were observed between all stages (P<.001). These findings underscore the potential of medication-related technologies to enhance medication safety and improve the efficiency of medication dispensing processes in hospital settings. Notably, the study observed significant reductions in critical error types such as wrong drug and wrong dose, which can have serious consequences for patient safety. This study shows that medication-related technology reduces dispensing errors and improves medication safety in hospitals. As hospitals work to enhance efficiency and safety, more targeted research is needed to assess the benefits and risks of these technologies. The findings highlight that technology can reduce errors and enhance medication management by improving dispensing and prescribing, optimizing clinical pharmacist resources, and allowing staff to focus more on patient care.