Different HCPs have distinct responsibilities in the narcotic consumption card (NCC) process [12]. Physicians verify NCCs with their signatures before pharmacy approval and archiving, whereas nurses and pharmacists perform broader tasks, including medication administration and error investigation [12].

Experiences with the eNCC process were examined using a cross-sectional electronic survey, a common method in health care research to complement qualitative findings [29]. We reported our survey following the CROSS (Consensus-Based Checklist for Reporting of Survey Studies) guidelines [30].

A survey questionnaire (Multimedia Appendix 1) comprising 4 themes was developed by a multidisciplinary group of experts in pharmacy, medicine, and UX research, as well as those with experience in the EHR system. Instead of using standardized usability questionnaires (eg, System Usability Scale and Usability Metric for User Experience), we developed a study-specific questionnaire. While the System Usability Scale was used in an earlier study by our team [12], it did not differentiate between the workflows examined in the present study.

Separate versions were created for physicians and for nurses and pharmacists to reflect their distinct roles in the eNCC process [12]. The core themes addressed (1) usability, (2) context-dependent use, and (3) attitudes and perceptions. These themes were based on established usability dimensions [31]. In addition, the questionnaire included demographic items such as professional background, department, hospital, years of professional experience, current working unit, and experience with the EHR system. The questionnaire was implemented using the REDCap software (Vanderbilt University) [32,33].

Questionnaire items were designed to reflect the 4 predefined themes and ensure comprehensive coverage of usability aspects. All professional groups were presented with 10 general UX statements rated on a 5-point Likert scale (“strongly disagree”-“strongly agree”), 1 net promoter score (NPS) item (0‐10) [34], and 1 open-ended question. The open-ended item captured qualitative feedback on the eNCC process and its development. Additional role-specific modules were included for nurses and pharmacists. These comprised statements related to dosage forms (ie, solids, such as tablets, capsules, and transdermal patches, and liquids, such as injections, infusions, and liquid oral preparations), reflecting common hospital formulary. To assess perceived workload effects among nurses and pharmacists, additional items were included to estimate the time spent on investigations and error corrections after eNCC implementation and during the most recent month.

Content validity was ensured through iterative refinement with a research group experienced in surveys and eNCC workflows. Face validity was supported by piloting (n=8) the questionnaire with representatives from each HCP group using a discussion-based approach consistent with cognitive interviewing techniques [35]; feedback from these sessions informed minor revisions. Pilot responses were excluded from analysis as the pilot aimed to evaluate question clarity and questionnaire usability rather than generate study data.

The targeted population comprised physicians, registered nurses, practical nurses, and ward pharmacists across 84 hospital wards, as well as pharmacists at the HUS Helsinki University Hospital pharmacy. With ward staffing levels typically ranging from 30 to 70 HCPs, the estimated study population was 2520 to 5880 individuals. At the time of the survey, the eNCC was used across 14 specialized health care departments, all represented in the sample, and had been implemented for at least 6 months in every department.

This study was conducted at HUS Helsinki University Hospital, Finland, in May 2024. Survey invitations were distributed via head nurses and pharmacy supervisors, who received the link via email and forwarded it to all eligible HCPs in their units. The survey remained open for 3 weeks, during which 2 reminders were sent through the same intermediaries. This personalized approach to a defined population has been shown to improve online survey response rates [36].

Before analysis, participants who selected “other” as their professional group were combined with registered nurses. Work experience categories were reduced from 5 to 3.

Data were analyzed using both qualitative and quantitative methods. For statistical analysis, nonparametric tests were applied. The Kruskal-Wallis one-way ANOVA was used to assess UX differences across professional groups, work experience, and years of EHR system use, with pairwise comparisons performed when significant differences were found. The Wilcoxon matched-pair signed-rank test compared the eNCC process for different dosage forms (solid vs liquid). The NPS was calculated as the percentage of promoters (ratings of 9‐10) minus the percentage of detractors (ratings of 0‐6) [34]. All statistical analyses were performed using the original 5-point Likert scale. For graphical presentation, Likert-scale responses were aggregated into 3 categories to improve the readability and interpretability of the figures. Analyses were conducted using Microsoft Excel and SPSS Statistics (version 28.0; IBM Corp), whereas open-ended responses were analyzed thematically in Microsoft Excel.

Item-level missing data were possible because individual responses were not mandatory, but REDCap prevented partially completed questionnaire submissions because respondents had to reach and confirm the final page before submitting the questionnaire. Missing data were not imputed and were handled using pairwise deletion such that each analysis included all available observations for the variables involved. Nonresponse error could not be evaluated as the target population was not fully identifiable and the response rate could not be determined precisely.

This study was conducted in accordance with Finnish regulations and the principles of the Declaration of Helsinki. According to the Finnish National Board on Research Integrity guidelines [37], no separate ethical review was required because participants were surveyed in their professional roles and no sensitive information was collected. Participation was voluntary, and electronic informed consent was obtained before respondents accessed the survey. All responses were anonymous and accessible only to the research team. HUS Helsinki University Hospital granted research approval (47/2024).

This study confirmed that HCPs generally viewed the eNCC process positively, reinforcing earlier findings that it supports reliable real-time documentation of controlled substances and improves efficiency, particularly for solid dosage forms [12]. Among respondents, 80.9% (484/598) opposed returning to paper-based workflows, indicating broad acceptance despite some persistent challenges. However, the NPS was 18, reflecting a polarized user base with both strong promoters and detractors [34]. Because the NPS classifies ratings of 9 and 10 as promoters and ratings of 0 to 6 as detractors and excludes ratings of 7 and 8 from the final scoring, it can be considered a rather strict metric and is therefore best used as a complementary measure rather than a comprehensive indicator on its own [38]. The moderately positive NPS score of 18, despite generally favorable Likert-scale responses, may reflect the stringent scoring methodology of the NPS, where moderately satisfied users (scores of 7‐8) do not contribute to the overall score. Open-ended responses highlighted problematic waste entries, slower workflows, and error handling as key concerns, which can be mitigated through system refinements and robust training, both recognized as critical for successful EHR implementations [39,40].

Perceptions varied by dosage form: processes for liquids were rated as less efficient and more cumbersome than those for solids, and the eNCC was reported to reduce investigations and error corrections more effectively for solids compared to liquids. These findings align with prior observations that injection workflows present particular usability challenges [12]. Encouragingly, the frequency of investigations and error corrections declined markedly from implementation to the survey period, suggesting that system familiarity and process stabilization improve performance, consistent with broader evidence from public sector information system implementations [41]. Pharmacists consistently performed more investigations than nurses, underscoring their central role in controlled substance surveillance and their essential contribution within hospital-based health care teams [42].

Differences emerged between professional groups. Physicians expressed less confidence in the eNCC’s surveillance benefits compared with nurses and pharmacists. This likely reflects their limited interaction with the system, primarily verifying entries rather than performing documentation and error resolution. Previous research shows that physicians’ EHR satisfaction depends on workflow alignment and information accessibility, whereas poor usability is linked to stress and burnout [21,43]. In our study, physicians’ less positive attitudes may have also stemmed from difficulties locating information and limited awareness of other professionals’ responsibilities within the highly role-specific eNCC process. Interview data from our earlier study revealed that physicians may find reports and verification of entries challenging to use [12]. Indeed, prior to the implementation of the eNCC, their role in the controlled substance oversight process comprised manually signing paper cards. The paper-based process is easier to manage without familiarity with the EHR system workflows. However, each entry in the eNCC is based on medication orders and administration in the EHR system [12]. Improved traceability may therefore be less apparent to physicians whose involvement is limited to the verification phase of eNCC surveillance.

Conversely, nurses and pharmacists, who use the eNCC most frequently, reported greater trust in its surveillance benefits. Their familiarity with barcode scanning–based workflows likely contributed to this confidence [12]. Ward pharmacists were the most common professionals who investigated errors, and pharmacists at the hospital pharmacy approved each eNCC before archiving, making them more aware of issues related to the eNCC. Interestingly, practical nurses reported more positive UX than registered nurses. This may reflect their more limited involvement in workflow details related to the eNCC. Practical nurses reported conducting investigations and error corrections only rarely, which may have reduced their exposure to the challenges associated with eNCC. However, other factors such as differences in use frequency, task complexity, or user expectations may also have contributed to the observed differences. These findings underscore the importance of considering role-specific workflows and access rights in interaction design to support multi-professional work and minimize disruptions during digital transitions [44,45].

Embedding controlled substance surveillance within the EHR offers clear advantages. All administrations are recorded in the medication administration record, enabling reconciliation with dispensing data and improving detection of discrepancies [7]. Structured and standardized documentation further enhances data quality and reuse [46], whereas remote access to integrated records supports workflow efficiency [47]. The eNCC consolidates medication orders and administration of controlled substances in a single location [12], which may explain its generally positive reception among HCPs. However, usability depends on aligning documentation requirements with clinical workflows; excessive clicks, fragmented access, or missing data can undermine efficiency. Future development should prioritize these aspects to maximize safety and user satisfaction. Furthermore, although the eNCC was specifically developed for a particular EHR system and within the framework of Finnish legislation, the experiences gained from its implementation may provide valuable insights for similar initiatives in other health care organizations.

Our study has several limitations that warrant consideration. First, the response rate (599/5880, 10%‐599/2520, 24% depending on the denominator) was lower than typical benchmarks for online surveys; this may be partly explained by the fact that the invitations were sent to supervisors to forward the survey links instead of directly sending invitations to all HCP end users. The survey link was not personalized, which in principle allowed a single respondent to submit multiple responses. However, the heterogeneity of the responses and their distribution across professional groups suggest that this was unlikely. The respondents represented all professional groups, with proportions comparable to those of the hospital staff overall. Respondents were drawn from all 14 departments that had implemented the eNCC, supporting the representativeness of the study population. Nevertheless, voluntary surveys may be subject to self-selection bias [48], whereby respondents may have constituted users who were more engaged with the eNCC or more confident in using digital tools than nonrespondents. However, the critical comments expressed in the open-ended responses suggest that participation was not restricted exclusively to highly satisfied users. Second, disentangling eNCC-specific tasks from broader EHR-based medication administration is challenging as several steps are largely similar irrespective of NCC modality [12]. Nonetheless, the survey design, grounded in prior observational work, enabled isolation of key differences by dosage form (solids vs liquids), offering meaningful UX insights. Third, our methodological approach relied on perceived usability and workload, which correlate only modestly with actual EHR-related fatigue [49]. Furthermore, retrospective evaluations may have been influenced by peak and end experiences rather than the cumulative number of implementation-related issues encountered [50,51]. However, these findings build on our earlier observational study [12], and further observations in the expanded deployment context would strengthen triangulation.

Instead of relying on standardized usability questionnaires [16,52-55] or unvalidated commercial instruments [56], we developed a study-specific survey tool. This approach was informed by our previous work [12], which provided an in-depth understanding of the medication workflows associated with the eNCC process. This knowledge allowed us to explicitly account for differences between workflows for solid and liquid dosage forms. Importantly, we recognized that the implementation of the eNCC could have divergent effects, potentially either accelerating or decelerating the documentation of controlled substances. Similarly, the frequency of errors in controlled substance surveillance could plausibly increase or decrease following eNCC implementation. As we used a study-specific questionnaire, direct comparisons with studies using standardized usability measures are not possible. However, the aim was to compare workflows rather than evaluate usability.

Future studies should build on our findings in several ways. Repeating observational work would be valuable as the number of wards using the eNCC has increased substantially since our initial study [12], enabling richer analyses of workflows in diverse contexts. Our survey focused solely on the eNCC process, and without a concurrent paper-based control group, we cannot directly compare UX across modalities. However, a remarkable portion of HCPs (484/598, 80.9%) were unwilling to return to a paper-based process. Therefore, future research should focus on the usability, workload, and error detection of the eNCC while considering role-based access rights and responsibilities. In addition, prospective research on the applicability of the eNCC process in different care settings, such as opioid substitution treatment units, could enhance surveillance practices.

The eNCC process was generally viewed positively by HCPs. It was preferred over traditional paper-based workflows despite some concerns such as problematic waste entries and challenges in error handling. The electronic process was perceived as easier to manage for solid dosage forms than for liquids. Addressing the usability concerns will be essential for optimizing the process efficiency and robustness as well as user satisfaction in the health care setting. Furthermore, the research highlights the importance of understanding the varying levels of interaction with the eNCC among different professional roles. Ward pharmacists played a central role in conducting more investigations and corrections than nurses responsible for the administration of controlled substances, emphasizing their critical role in controlled substance surveillance within health care teams.