Focus: Health Information Systems (HIS), including Pharmacy Information Systems (PhIS), offer the potential to enhance healthcare quality through improved data accessibility across administrative, inventory, clinical, and medication processes [1], [2], [3], [4]. However, this potential is often undermined by various socio-technical challenges within HIS, which can introduce new errors and lead to unintended consequences (UC) in healthcare delivery [1], [2], [5], [6], [7].

• Modification 1 (Focusing on PhIS UC): Most reported UC related to PhIS (UC-PhIS) arise during its adoption, particularly when pharmacists interact with the system or Computerized Provider Order Entry (CPOE). For example, PhIS has changed existing medication management by introducing novel issues, such as errors stemming from users' poor understanding of system machine rules [6].
• Modification 2 (Focusing on reliability and complexity): While a new dispensing system can help pharmacists safely dispense drugs, technical issues can reduce the system's reliability and its ability to support current tasks, thereby complicating established work processes [8].
• Modification 3 (Focusing on error identification): The incidence of latent errors and their unknown causes is rising [9]. The root cause of HIS-related errors often resembles a black box, and the use of workarounds further complicates efforts to understand this situation [10], [11].
• Modification 4 (Focusing on complexity): Understanding UC-Health Information Technology (HIT) is complex because it involves the intricate interaction of people, processes, and technology across all aspects of healthcare [12], [13].
• Modification 5 (Focusing on existing literature): Although the HIS literature has extensively covered basic UC concepts and related frameworks—including system interactions, usability, human factors (cognitive), classification and effects of UC-CPOE, UC-Clinical Decision Support (CDS), UC-Electronic Health Records (EHRs), HIT errors, and UC-workflow-HIT—studies specifically addressing UC in the context of PhIS remain limited.
• Modification 6 (Focusing on the study's proposal): Therefore, this study proposes an evaluation approach for UC-PhIS to facilitate a structured and comprehensive understanding of the causes and consequences of these unintended outcomes.
• Modification 7 (Focusing on benefits): Recognizing UC-HIS enables healthcare providers to better understand contributing factors and identify mitigation mechanisms. Similarly, developers can design more effective and efficient systems [1], [6], [14], [15].
• Modification 8 (Focusing on trade-offs): Evaluators can anticipate future PhIS challenges in improving healthcare quality by converting these UCs into predictable trade-offs, thereby reducing the risk of medication errors [14].
• Modification 9 (Focusing on the proposed approach): This study proposes an evaluation approach based on two established, complementary frameworks that guide the identification of UCs, their dominant factors, effects, and mitigation mechanisms from the user perspective in two pharmacy-based PhIS settings. This proposed approach is intended to guide the understanding of complex socio-technical interactions that trigger UC-HIS for system improvement and risk mitigation.

Section snippets

Theoretical Background

The concept of Unintended Consequences (UC) is often linked with Clinical Decision Support (CDS), which aims to reduce errors by providing targeted warnings, recommendations, and constraints [16], [17]. However, instead of maximizing their potential, these systems frequently generate new UCs and prove to be inherently error-prone [18].

The critical finding that human-technology interaction is a dominant factor in generating UC [19], [20] has pioneered new scholarly approaches. These include employing a human factors perspective for assessing UC-HIT [19], [21], evaluating HIT usability in relation to human cognitive load and the risk of automation bias [22], [23], and promoting the principles of safe HIT use [24].

Methods and Results

Methods

We employed a rigorous qualitative, explanatory case study evaluation to identify the contributing factors to Unintended Consequences of Pharmacy Information Systems (UC-PhIS) within two Malaysian ambulatory pharmacy dispensing services: Hospital T and Clinic R. Both facilities had fully transitioned to and utilized the Outpatient Pharmacy module in PhIS since 2018. Crucially, while the doctors continued to prescribe medications manually (i.e., on paper), pharmacists were responsible for accurately entering these orders into PhIS. Author NR systematically collected data in both settings from mid-November to mid-December 2020 through a combination of direct observations, in-depth interviews, and thorough document analysis.

Results

The implementation of PhIS at the two settings followed a structured, staged approach. Prior to the full rollout, all participating pharmacists underwent standardized basic PhIS training. Following this, the pharmacy unit and the PhIS implementation team successfully provided the necessary infrastructure to integrate PhIS hardware into the existing service environment. However, a critical limitation was encountered: PhIS could not be seamlessly integrated with patient health records because all wards and clinics still relied on traditional paper-based documentation. This integration failure inadvertently created an unobserved, yet significant, information transfer gap in the clinical workflow.

Discussion: Frameworks for Analyzing UC-PhIS

The combination of the Integrated Sociotechnical Analysis (ISTA) and Human-Organizational-Technology (HOPT) components proved highly effective for a comprehensive evaluation of Unintended Consequences of Pharmacy Information Systems (UC-PhIS).

• The ISTA component was instrumental in facilitating the systematic identification of UC-PhIS within the complex interplay of the sociotechnical systems involved in pharmacy practice. This framework allowed for a detailed mapping of how people, processes, and technology interact to generate unintended outcomes.
• Conversely, the HOPT component provided the critical lens necessary to move beyond mere identification, enabling us to focus strategically on the most effective aspects of mitigating and preventing future UC-PhIS occurrences.

The most critical and influential factors were identified through a meticulous analysis of the cause-and-effect relationship for each reported UC-PhIS using the ISTA framework. This rigorous analysis was then supplemented by identifying those UCs that served as primary triggers or dominant precursors leading directly to the occurrence of other subsequent UCs. The evidence illustrating this analytical progression is clearly shown in Appendix C (Cause-and-Effect Relationship Analysis) and Appendix D (Identification of Dominant Precursors), respectively.

Conclusion: Key Findings and Significance

The proposed evaluation approach, which judiciously combined the Integrated Sociotechnical Analysis (ISTA) and the Human-Organizational-Technology (HOPT-fit) frameworks, provided a robust and comprehensive methodology for assessing system impact. This approach successfully assisted us in identifying a significant total of 28 distinct Unintended Consequences (UCs) directly related to PhIS use across the two studied pharmacies.

The evaluation allowed for a detailed analysis of:

• Complex Socio-technical Factors and Effects: The ISTA framework was crucial for meticulously mapping the interaction between people, processes, and technology, thereby elucidating the factors contributing to each UC and detailing their immediate effects on the workflow and patient safety.
• Dominant UCs, Misfits, and Mitigation: The HOPT-fit framework enabled a deeper focus on identifying the dominant UCs (those that trigger others), the resulting misfits (e.g., between the system and the user's task), and the appropriate mitigation mechanisms designed to address these fundamental imbalances.

The primary socio-technical factors driving UC-PhIS were clearly identified as:

• System Rigidity and Complexity: The system's inflexible design and excessive operational complexity often forced users into inefficient workarounds.
• Users' Lack of Skills and Understanding: Deficiencies were noted in users' technical skills and, more critically, in their conceptual understanding of the system's intended purpose and the flow of information.
• Paper Persistence: The continued reliance on paper documentation created a problematic parallel workflow that directly undermined the system's goal of seamless electronic information transfer.
• Unclear Tasks and Role Shift: Ambiguities regarding new electronic tasks and an associated unplanned shift in professional roles and responsibilities within the dispensing workflows contributed significantly to confusion and errors.

In essence, the study demonstrates that UC-PhIS are rooted not just in technical flaws but in the deep-seated friction between a rigid digital system and the existing, often messy, reality of human and organizational workflows.

Declaration of Competing Interest : The authors declare that they have no known competing financial interests or personal.