A Synergistic Bridge Between Human–Computer Interaction and Data Management Within CDSS

Abstract

Clinical Decision Support Systems (CDSSs) have become indispensable in medical decision-making. The heterogeneity and vast volume of medical data require firm attention to data management and integration strategies. On the other hand, CDSS functionality must be enhanced through improved human–computer interaction (HCI) principles. This study investigates the bidirectional relationship between data management practices (specifically data entry management, data transformation, and data integration) and HCI principles within CDSSs. Through a novel framework and practical case studies, we demonstrate how high-quality data entry, driven by controlled workflows and automated technologies, is crucial for system usability and reliability. We explore the transformative positive impact of robust data management techniques, including standardization, normalization, and advanced integration solutions, on the HCI elements and overall system performance. Conversely, we illustrate how effective HCI design improves data quality by reducing cognitive load, minimizing errors, and fostering user engagement. The findings reveal a synergistic relationship between HCI and data science, providing actionable insights for designing intuitive and efficient CDSSs. This research bridges the gap between technical and human-centric approaches, advancing CDSS usability, decision accuracy, and clinician trust for better patient outcomes.

1. Introduction

Clinical decision-making systems are helping medical practitioners more and more throughout medical treatment [1]. The rapid evolution of healthcare systems, along with massive and diverse medical data, has highlighted the importance of data management. The systems need to control and steer the data flow [2]. Medical data must be managed in different phases, including data generation, fetching, tagging, and processing [3,4]. Integrating heterogeneous, low-quality data from diverse subsystems for medical decision support emphasizes the significance of robust data management practices [5]. Data entry and generation are significant aspects of data management that require profound investigation as human–computer interaction (HCI) elements because they form the foundation of user–system interactions in healthcare environments [6]. The medical data generated by user–system interactions and fetched data from other systems must be scrutinized as a data management field within Clinical Decision Support Systems. In other words, high-quality data provided for Clinical Decision Support Systems (CDSSs) functionality stems from medical data management and integration [7,8]. Accurate and timely data entry supports clinicians by ensuring access to consistent and reliable information during critical decision-making [9]. For instance, poorly managed data entry systems often lead to increased mental workload, user frustration, and potential medical errors [10,11].

Despite numerous studies exploring data integration and management within CDSSs, a significant gap persists in understanding the role of human–computer interaction in this process. Existing studies focus on optimizing data accuracy, consistency, and integration, while HCI research in CDSSs primarily addresses usability, user experience, and cognitive workload. However, the interplay between these two domains (how data management directly influences HCI elements and, consequently, CDSS functionality) remains unexplored. Investigating these HCI aspects can bridge the gap between system designers and end-users, facilitating smoother user adoption of CDSSs. The current study discusses why data entry management must be investigated as an HCI element and how it can impact CDSS functionality. Furthermore, it will illustrate how data integration improves HCI factors relevant to CDSS environments, introducing a beneficial and bilateral cycle between data management and HCI elements within CDSSs. The primary contribution of this article is establishing a scientific framework that connects HCI factors to data management strategies, offering a roadmap to reach more intuitive and effective CDSS environments.

To achieve our objectives, we employ a qualitative research approach, synthesizing existing studies on data entry management, standardization, and integration within CDSSs to construct a framework that integrates data management with HCI principles. To reinforce the proposed framework, we incorporate empirical findings from previous case studies that examine the real-world application of these components. These studies evaluate how specific data management strategies affect data quality, system usability, and decision-making accuracy. The metrics such as data completeness, error frequency, process efficiency, and alignment with active clinical workflows are benchmarks to assess the framework’s practical relevance. This dual approach (literature synthesis and empirical grounding) ensures that the framework is both conceptually robust and contextually applicable.

This paper is organized into sections that collectively address the critical role of data management and integration within CDSS environments, highlighting their impact on HCI elements. In the upcoming sections, at the first step, we delve into data entry management and control, examining existing features for its implementation. In the next section, we will articulate the significance of data standardization and normalization within CDSSs and their concurrent role in enhancing HCI. The subsequent section will address data integration within CDSSs, exploring its diverse forms. The discussion section expands on these foundations by examining real-world standardization and implementation challenges, human-centered design considerations, system limitations, and practical deployment strategies. Finally, we conclude how their collective effect elevates data quality and improves CDSS accuracy.

2. Data Entry Management and Control

After surgery, surgeons commonly note rare medical cases in the surgery reports utilizing professional expressions like “unicornuate uterus with a rudimentary horn”. Another surgeon might use varying terminology like “rudimentary horn anomaly” to describe the same medical condition. This inconsistency in medical terminology can pose challenges when researchers attempt to identify and analyze specific cases. If researchers aim to investigate the prevalence or characteristics of a particular rare condition, they may encounter difficulties in finding surgical reports that include the intended terms and expressions. These challenges can even impact the accuracy level of this research project and, consequently, the obtained results, highlighting the importance of medical data entry.

Data entry is still discussed as a significant bottleneck in electronic health record (EHR) platforms, and in many cases, it hinders optimizing the full potential of EHRs [12]. In other words, data entry difficulties push the medical staff away from interacting with the system, causing user attrition and frustration [13,14,15,16]. One of the significant concerns among medical system designers is whether data entry will be carried out correctly to ensure accurate data entry and enhance the quality of the recommendations [17]. Minor data entry errors may impact the treatment plan results and sometimes deceive the physicians [18]. Considering HCI principles is crucial for mitigating many user challenges [19]. Medical practitioners frequently express concerns about the prevalent error from a lack of appropriate user interaction with the medical systems [20]. High-quality and accurate data entry significantly improves the precision of CDSS data and better medical results [21]. In this section, two methods to control medical data entry will be introduced.

2.1. Data Range Control

In medical information systems, a reasonable data range must be defined for some data fields to impede data entry beyond acceptable limits. This restriction applies to augment the reliability of provided data for CDSSs, resulting in more accurate data analysis and decision-making [22]. The mentioned data range is mutable, and medical practitioners can determine the specific values and modify them whenever necessary. In some cases, the color of the data entry field will change to provide visual feedback to the user. As shown in Figure 1, if a value is entered outside the normal range (too low or too high), the field might turn red to alert the user about the problematic value [23]. This color change indicates potential data entry errors or data abnormalities.

Data entries that fall outside the expected range can significantly deviate between research findings and clinical decision-making. Studies have demonstrated that such anomalous data can undermine the validity and reliability of statistical analyses and the accuracy of clinical predictions and outcomes [24]. Hence, it is essential to identify and control out-of-range data entries for data reliability and subsequent data analyses.

Data entry errors usually occur in circumstances such as transcription mistakes, inaccuracies in lab measurements, or reference range misinterpretation. Studies emphasize the need for data quality control processes and validation assessment to detect and refine out-of-normal range data entries [25]. These controls promote more reliable data in healthcare systems.

2.2. Data Entry Obligations Controlling

The mandatory medical data fields are pivotal in guaranteeing data completeness and reliability. This obligation prevents data-missing phenomena, meaning blank data fields and incomplete medical records will be firmly diminished [26,27]. As depicted in Figure 2, the medical systems equipped by the data entry obligation technology are implemented in this way: they adhere to this compulsion seriously so that the medical data forms are arranged sequentially, and medical staff will not be able to complete the following forms unless the mandatory medical records have been filled within the current form.

The implementation of data entry compulsion promotes significant advantages within healthcare systems. First, it vouches for data completeness and inclusion of all necessary information to assist the physicians on the treatment path. Second, the mandatory data entry method allows medical practitioners to fill out specific data fields determined by the physicians. Hence, the risk of errors and data inconsistencies will be firmly reduced. Third, exploiting this method leads to a streamlined medical workflow. When healthcare providers have timely and proper data, they can make accurate medical decisions, ultimately improving patient outcomes [28].

2.3. Leveraging Automated Text Generation for Accurate Data Entry Management

Building upon data entry concepts discussed in the previous sections, we conducted another empirical study [29] to implement a novel technology to resolve data management concerns. The designed platform meets the needs of professionals, such as sonographers, radiologists, and surgeons, to emit medical reports in text format. At the same time, the data entry process can be entirely managed. First, the user fills in data fields within medical forms by forcing the system’s policy during the data entry. Then, the medical text (report) derived from data fields will be generated automatically. By exploiting this technology, the data entry process can be managed and controlled record by record, hindering missed or out-of-range data. This feature eliminates the reliance on manual report writing, which is prone to errors and inconsistencies, and produces standardized and reliable documentation.

One of the prominent features of this system is its data range control mechanism. For every data field, predefined acceptable ranges are established, and any value outside these ranges triggers immediate feedback. Additionally, through mandatory data entry requirements, the system ensures that users cannot save or submit records until all essential fields have been completed. This enforcement guarantees comprehensive data entry and helps capture accurate medical data. However, the system is not rigid; physicians maintain complete authority over clinical decisions. They can manually edit the generated medical text when necessary, ensuring that rare or exceptional cases are appropriately documented. Additionally, embedded text boxes within the system allow practitioners to input specific details that may not fit the predefined structured fields. These features provide flexibility while preserving data accuracy and completeness.

Another innovative aspect of the mentioned technology is its focus on user-centric design. The interface is intuitive and minimally disruptive to clinicians’ workflows. It allows the user to perform data entry without excessive cognitive load. Consequently, physicians can focus on patient care instead of text typing. Furthermore, the system’s policies regarding data field obligations and raised alarms have been designed with built-in flexibility. Medical professionals can modify these data entry rules without requiring specialized technical expertise, allowing them to adjust defined data constraints to match evolving clinical needs.

To evaluate the effectiveness of this technology, a comparative examination was conducted involving 10 physicians across 150 surgeries (including three different surgery types). The study analyzed 4200 patient records, comparing traditional manual text entry with the automated text generation system. Quantitative metrics were measured, including data completeness, out-of-range error frequency, and accuracy. Statistical analyses revealed that the automated approach eliminated missing data (0%) compared to 6.67% blank entries in the manual text-based method (χ² = 280.82, df = 1, p < 0.001). Additionally, the system prevented out-of-range data entries altogether, whereas 1.55% of entries in the manual approach fell outside the acceptable range (χ² = 73.86, df = 1, p < 0.001). These findings confirm that automated text generation can be a supportive tool rather than a restrictive mechanism, reinforcing clinical decision-making by providing accurate and complete data.

This study compared missed and out-of-range data in two text-based (traditional) and automated text-generating approaches, demonstrating that recent technology significantly reduces data corruption, improving accuracy and reliability. Beyond preventing documentation errors, the system is an intelligent assistant, providing physicians with complete, structured, high-quality patient data to boost medical decision-making. Rather than enforcing a rigid treatment framework, it enhances decision-making clarity, minimizing the risk of errors due to incomplete or inconsistent records. Physicians retain complete authority, leveraging the system’s structured data entry as a protective mechanism against potential medical errors rather than as a limitation on their professional expertise.

3. Data Standardization and Normalization

Data standardization and normalization are critical processes in managing the complexity of medical data within Clinical Decision Support Systems. The principal mission of these processes is to govern the consistency attribute regarding data derived from different sources, such as laboratory systems, radiology reports, the patient’s medical history, and wearable devices [30]. They assist the system in interpreting and analyzing medical data properly. Considering these procedures, coherent results will be acquired in the same medical circumstances, even in different systems and platforms [31].

The CDSS environments can provide physicians with a more consistent and reliable experience by standardizing and normalizing medical data. Exploiting the mentioned procedures, the CDSSs will be ameliorated in several HCI dimensions:

• Improved user interface: Standardized data formats and consistent terminology make user interfaces more straightforward to learn and navigate [32];
• Reduced cognitive load: Normalized data structures simplify data entry and retrieval processes, reducing physicians’ cognitive burden [33,34];
• Enhanced decision-making: Consistent data representation across different platforms and systems allows for more accurate and reliable clinical decision-making [24];
• Increased system interoperability: Standardized data formats facilitate seamless integration with other healthcare systems, improving data sharing and collaboration [30].

Ultimately, by investing in data standardization and normalization, healthcare organizations can empower clinicians with a more efficient, effective, and user-friendly CDSS.

Data Transformation Pipeline

The raw data from medical environments must be in a standard available format to feed the data into CDSSs. Standardization and normalization procedures would enable data processing in such a context [35]. To be more precise, this is a data transformation pipeline consisting of several operations on transforming the raw, heterogeneous, and often incomplete medical data into structured and standardized formats usable by CDSSs [36]. This process addresses inconsistencies and significant issues relevant to data gaps, which can compromise decision accuracy.

From an HCI perspective, this pipeline is deeply tied to usability and user interaction with CDSSs. The systems’ usability level depends on their potential to present clean, standardized, and interpretable data to end-users [37,38]. This pipeline helps physicians receive reliable and actionable information (without manually interpreting or adjusting inconsistent data inputs) by resolving data gaps and transforming data into a structured and consistent format. In this manner, cognitive load, as one of the most prominent HCI elements, will be mitigated, and the user can focus on clinical decision-making rather than system navigation or data troubleshooting [39]. The functionality of the data transformation pipeline consists of several stages [40,41,42], as shown in Figure 3. These stages are described as the following:

➢

Data collection:

▪

Sources: Data are gathered from multiple sources, such as electronic medical records (EMRs), laboratory systems, radiology databases, wearable health devices, and patient-reported outcomes;

▪

Challenge: Data often arrives in different formats, with missing fields or irregular entries.

➢

Data cleaning:

▪

Operations: Handles erroneous, incomplete, or irrelevant data through imputation (filling in missing values), outlier removal, and error correction;

▪

Focus on gaps: Missing data due to human error, system inefficiencies, or disconnected subsystems are resolved here. Techniques like predictive imputation, default values, or physician-driven corrections are applied.

➢

Data standardization:

▪

Purpose: Converts diverse terminologies, units, and coding schemas into a unified framework—for example, ICD-10 codes and SNOMED CT terminology;

▪

Impact on gaps: Standardization ensures that gaps caused by mismatched terminologies are bridged, making datasets interoperable.

➢

Data normalization:

▪

Process: Scales data into consistent ranges (e.g., blood pressure in mmHg) or formats (e.g., dates in “YYYY-MM-DD”);

▪

Addressing gaps: Normalization identifies outliers and irregularities that may hint at gaps in understanding or reporting.

➢

Data integration:

▪

Objective: Combines data from multiple sources into a single, unified repository;

▪

Role in gaps: Integration resolves gaps caused by data fragmentation across subsystems, creating a comprehensive view for decision support.

➢

Data analysis and enrichment:

▪

Techniques: Applies advanced analytics, predictive modeling, and enrichment using external datasets (e.g., population health statistics);

▪

Gap insights: Identifies systemic gaps by analyzing patterns and generating recommendations for future improvements.

➢

Data presentation:

▪

Focus: Presents clean, consistent, and actionable data to users via dashboards or alerts in the CDSS;

▪

HCI element: Assists in making the output comprehensible and prevents user fatigue and system override.

The data after passing this route will be ready for decision proceedings, and the probable data gaps will be detected in several aspects:

➢

Incomplete data: Missing entries in key fields, such as medication history or lab results, can severely degrade decision-making processes. Research shows that healthcare systems often face challenges with data incompleteness, impacting clinical outcomes [43];

➢

Inconsistent data: Variations in data representation, like “hypertension” vs. “HTN”, introduce inconsistencies that hinder practical analysis. Some methods can handle inconsistencies in prediction models [44];

➢

Fragmented data: Data spread across isolated systems without integration leads to care fragmentation, which is associated with adverse outcomes and higher costs [45];

➢

Outdated data: Relying on outdated information affects decision-making accuracy and relevance [46].

Since this pipeline plays a profound role in recuperating the different data gaps, it diminishes CDSS problems in some dimensions:

❖

Clinical risk: Missing or erroneous data in a CDSS can directly impact patient outcomes by compromising the quality of diagnoses and treatment plans [47]. For instance, incomplete lab results, missing imaging studies, or gaps in medication history can prevent healthcare providers from identifying critical clinical patterns that may delay lifesaving interventions. Moreover, inaccurate or incomplete records can lead to physicians’ inability to identify contraindications, allergies, and comorbidities and provide poor treatment options. These risks underlie the importance of having a robust integrated data pipeline that will ensure the CDSS systems present the clinician with a complete and accurate patient profile for decision-making [24];

❖

User dissatisfaction: Users’ trust in a CDSS is considerably breached if incomplete or unreliable data come up with questionable recommendations [48]. Generally, physicians depend on timely, accurate, and complete data for precise medical decisions. Suppose either data gaps or inconsistencies provide a CDSS output. In that case, healthcare providers may avoid investing more time in manually verifying information or searching for missing details across systems. This can also raise the cognitive load, break workflows, lower efficiency, and lead to frustration; this diminishes the system’s perceived value. These experiences, over time, decrease user confidence in CDSS usage, limiting its adoption and effectiveness within clinical environments [49];

❖

Algorithm bias: Machine learning algorithms in CDSSs require complete and representative data for reliable predictions. When datasets contain the mentioned data gaps (due to incomplete records, underrepresented populations, or inconsistent data integration), bias can be introduced to the model. In some cases, if some demographic groups or clinical variables are underrepresented, the system may generate predictions that advantage one subset of patients at the expense of others, thereby undermining fairness and equity in care [50]. Additionally, missing data degrade model stability and decrease its predictive accuracy, leading to unreliable recommendations that can hinder clinical decision-making. These biases can only be resolved by robust data integration approaches and efficient handling of incomplete records for CDSS tools to provide fair and consistent outcomes [51,52].

The data as the primary feed for CDSSs must be presented coherently and consistently, although they may be gathered from various resources [24]. The prerequisite for this data consistency is the data transformation process. The mentioned transformation pipeline significantly boosts the data integration procedures within CDSSs. In this circumstance, the system’s reliability level and, consequently, system usability will be elevated [48].

4. Data Integration Within Clinical Decision Support Systems

Data integration combines data from multiple sources into a unified and coherent format. Within CDSSs, data integration is crucial in furnishing relevant, timely, and accurate information to assist healthcare providers in making purposeful decisions [53]. Medical data often originate from separate systems, such as electronic health records, laboratory systems, radiology reports, and real-time monitoring devices. Integrating these diverse data enables a comprehensive view of the patient and impedes ignoring critical patient data [54].

Since fragmented, inconsistent, or missing data can significantly compromise decision-making accuracy, the importance of data integration in CDSSs has been increasingly highlighted. In other words, enhancing data integration assists medical practitioners by enabling seamless communication between different systems and improving the precision of medical recommendations [55].

4.1. Bridging Data Integration with Human–Computer Interaction

Data integration significantly enhances the user experience and system usability by presenting the required data to physicians in a unified and merged manner [56]. The integration function can be executed in two aspects: data integration and interface integration. Although these aspects are intertwined, as illustrated in Figure 4, each plays a disparate role in enhancing the system’s usability and user experience [56,57]. When physicians can make medical decisions based on the data derived from various data sources, they can trust the system’s veracity and comprehensiveness. On the other hand, if physicians interact with a seamless user interface instead of separate platforms, the necessary information will be considerably more accessible, leading to increased user satisfaction.

According to the undeniable importance of this approach in both data and interface integration, nowadays, system designers in the medical context attempt to unify the available systems by exploiting recent cloud technologies. Some dedicated services like iPaaS (Integration Platforms as a Service) provide a range of integration capabilities, including data integration, application integration, and API (Application Programming Interface) management [58,59]. The integration process impacts HCI within CDSSs in the following ways:

▪

Improved data presentation: Integrated and standardized data allow for precise, user-friendly interfaces that present medical information efficiently. Physicians can access patient records, test results, and diagnostic recommendations without navigating fragmented data sources [60];

▪

Enhanced trust and engagement: Since physicians receive reliable, comprehensive, and accurate data during interaction with the system, they can trust CDSSs. When clinicians acquire consistently desired results based on integrated data, they are more likely to engage with and rely on the system for decision-making [61].

As a result, data integration within CDSS, by providing a unified platform for data and user interaction and ensuring more reliable data collection, empowers physicians to achieve more accurate medical decisions and, ultimately, better patient outcomes.

4.2. Categorizing Data Integration to Enhance Usability and Functionality of Clinical Decision Support Systems

In the CDSS context, effective data integration is critical for enhancing clinical workflows, supporting evidence-based decision-making, and improving patient outcomes. To address the diverse needs of healthcare professionals, categorizing data integration based on content types (such as EHRs, genomic data, and medical imaging) provides a structured framework to evaluate their unique contributions to system functionality and usability [62]. The depicted categorization in

Table 1. Different data integration types within CDSSs.

Data Integration Type	Description	References
Electronic health record integration	Incorporating patient data from EHRs into CDSSs	[63]
Genomic data integration	Incorporating genomic information into CDSSs	[64]
Medical imaging integration	Incorporating imaging data (e.g., X-rays, MRIs) into CDSSs	[65]
Laboratory data integration	Incorporating lab results into CDSSs	[66]
Pharmacy data integration	Incorporating medication data into CDSSs	[67]
Wearable device data integration	Incorporating data from patient wearables (e.g., heart rate monitors) into CDSSs	[68]
Social determinants of health (SDOH) data integration	Incorporating socioeconomic and environmental data into CDSSs	[69]
Research data integration	Incorporating the latest clinical research findings into CDSSs	[70]
Patient-reported data integration	Incorporating data directly reported by patients (e.g., symptoms, outcomes) into CDSSs	[71]
Administrative data integration	Incorporating administrative information (e.g., billing, scheduling) into CDSSs	[72]

illustrates that integration strategies empower HCI principles such as reducing cognitive load, improving data accessibility, and supplying interface consistency.

Each data integration type alone has the potential to impact HCI elements and, consequently, decision-making capabilities. In this regard, EHR integration elevates accessibility to comprehensive patient data and reduces manual data entry, although well-designed workflows must be considered to prevent cognitive load [63]. Genomic data integration introduces personalized care options but demands advanced visualization tools to manage its inherent complexity [64]. Medical imaging integration enhances diagnostic precision by linking imaging and textual data, requiring intuitive interfaces for seamless navigation [65]. Since lab results have become prerequisites for disease diagnosis in recent years, laboratory data integration boosts decision-making speed and accuracy through real-time lab results [66]. On the other hand, pharmacy data integration improves medication safety but should strive to prevent alert fatigue through thoughtful interface design [67]. Wearable device data integration facilitates real-time monitoring, generating accurate and non-intrusive alerts to assist physicians and patients [68]. SDOH data integration combines non-medical factors (e.g., income, housing) with health data to understand broader health influences. Utilizing this data integration, the explainability and individuality aspects of the system will be enhanced, leading to improved health outcomes and equitable care [69]. Research data integration supports evidence-based activities and concentrates on streamlined filtering among medical data [70]. Patient-reported data integration promotes patient-centered care but requires user-friendly input and easy-to-use interfaces to gather essential medical data [71]. Although administrative data integration does not directly influence medical decision-making, it can mitigate the work burden by facilitating the administrative processes, leading to more concentration on the main medical processes [72].

As demonstrated above, data integration in different shapes ameliorates HCI factors, causing better CDSS functionality and more accurate medical decisions.

4.3. Empirical Case Studies About Data Integration Impact on Decision Accuracy and Human–Computer Interaction in Clinical Decision Support Systems

Several empirical studies have investigated the relationship between data integration, data accuracy, and HCI factors within medical settings.

Liyuan et al. [73] implemented a data integration project embedding one of the best practice databases into a CDSS. Their study demonstrated that real-time access to evidence-based clinical guidelines (stemming from data integration) significantly improved diagnostic precision and reduced time-to-diagnosis. This integration allowed physicians to retrieve relevant information without disrupting their workflow, elevating decision accuracy. Some other studies concerning the user interface discuss that focusing merely on data retrieval without addressing user interface design limits the full potential of the medical systems [73,74]. They underscore that systems with poorly integrated interfaces often increase cognitive load, causing frustration among clinicians.

Another study [75] presented a successful implementation of data integration within a Medical Decision Support System (MDSS) framework designed to enhance end-user confidence through ubiquitous IOT (Internet of Things) devices. This study demonstrates to what extent data integrating from diverse data stems from IOT devices leads to significant improvement in system performance.

To illustrate the practical benefits of data integration within medical information systems, we draw on our prior empirical research [76], which investigated the impact of integrating a personnel attendance subsystem from a Management Information System (MIS) into a Hospital Information System (HIS), forming an integrated HMIS. This study aimed to assess how such integration affects the accuracy of performance evaluation reports and, by extension, decision-making reliability. The methodology involved a comparative analysis of reception staff performance before and after integration, focusing on a key metric in medical environments called “reception rate”.

This expression is applied to count the number of admitted patients in hospitals and is calculated as below:

Reception rate = number of admitted patients ÷ number of useful working hours

In 2019 (pre-integration), the reception rate (RR_19) was calculated using patient admissions data from the HIS and working hours data from the MIS, which were separate systems at the time. In 2020 (post-integration), the reception rate (RR_20) was derived entirely from the HMIS, unifying both variables under one system. The sample comprised 21 reception personnel (15 females, six males, aged 25–37), consistent across both years to control for demographic factors (e.g., age, education). Data were extracted from system reports: in 2019, the MIS provided 152,878.75 total working hours (average 16,803.75 ± 2488 SD), and the HIS reported 151,539 admitted patients (average 5673.95 ± 1658.39 SD); in 2020, the HMIS recorded 156,793.22 working hours (average 17,520.82 ± 1308.82 SD) and 156,392 admitted patients (average 7447.25 ± 1759.84 SD).

Statistical analysis employed the Mann–Whitney U Test due to non-normal data distribution (Shapiro–Wilk Sig. < 0.05). The results showed a significant difference between RR_19 (0.3468 ± 0.1153) and RR_20 (0.4298 ± 0.10909), with a Mann–Whitney U = 14,900, Z = −2.866, and p = 0.004 (<0.05), rejecting the null hypothesis of no difference. Notably, admitted patients increased by 31.24% in 2020, while working hours rose by only 4.2%, suggesting RR_20’s higher accuracy stemmed from capturing “active” working hours (time logged into the HMIS) rather than total presence, unlike RR_19, which relied on less precise MIS data (e.g., arrival/exit times without activity context).

These findings highlight those manual errors and untracked inactive hours (e.g., breaks) skewed RR_19, underestimating actual performance before integration. After integration, the HMIS enabled seamless cross-checking of patient records against active hours, reducing errors, streamlining workflows, and enhancing decision-making precision. For instance, managers could now distinguish productive time, improving staff evaluations and resource allocation. This aligns with broader CDSS goals: integrating subsystems like attendance tracking with clinical data mitigates data discrepancies, directly supporting accurate reporting and informed medical decisions, as seen in reduced cognitive load for users interpreting unified outputs.

The elevated accuracy level from the mentioned integration cases enhanced the system’s reliability and user acceptance, increasing user interaction. Hence, the system is improved in some identified HCI aspects, such as user satisfaction [77], ease of use [78], and data entry issues [19]. Therefore, these empirical studies revealed how data integration in medical settings can positively impact HCI status.

5. Discussion

The current study has delved into a critical but often overlooked aspect of healthcare technology: the intersection of data management and human–computer interaction in the CDSS environment. By examining data entry management, standardization practices, and the integration of diverse medical subsystems, this research investigates an underexplored area in healthcare technology: bridging data science and HCI.

While other studies have emphasized the technical aspects of data management and integration, a few have explored how these considerations directly impact the user interaction with the systems. Data entry issues focusing on error reduction have often been discussed in this context. In contrast, the mental efforts stemming from improper data presentation as one HCI element have rarely been addressed [79,80]. Hence, the mentioned area is considered a research gap. Although the automated text generator technology described in the current study significantly reduces manual error, it also alleviates physicians’ frustration and cognitive load when typing repetitive and complex data. Systems equipped with these features make the system smarter, facilitate user interaction, and prevent interoperability challenges arising from terminology diversity [29,81].

To indicate the consequential relationship between data management and HCI,

Table 2. Consequential effects of data management on HCI elements and CDSS functionality.

Data Management Aspects	Impacted HCI Elements	Improvements in CDSS Functionality	References
Data range control (validity)	Error feedback, system reliability	Minimizes invalid data entries, enhancing algorithmic precision	[22,23,29]
Data entry obligations (completeness)	Structured interaction flow, task efficiency, usability	Ensures complete dataset availability, reducing query failures	[26,29,82]
Data consistency (terminology)	Lower interpretive ambiguity, interface clarity	Enhanced reasoning algorithms, improving natural language processing	[32,81,83,84]
Data standardization (interoperability)	Predictable interface behavior, workflow integration	System interoperability enables cross-system data synchronization	[30,31,85,86,87]
Data normalization (units/ranges)	Visual consistency, reliable data interpretation, minimized user confusion	Optimizing automated anomaly detection, accurate alerts	[88,89,90]
Data integration	Data accessibility, seamless multi-source data flow	Consolidates data streams, accelerating real-time analytics	[53,54,55,91]
Data presentation	Information visualization, decision clarity	Enhances graphical rendering, faster decision-making, improving predictive accuracy	[92,93,94]

presents a framework illustrating how data management aspects enhance HCI elements, subsequently improving CDSS functionality. This table is introduced to clarify the cascading effects: effective data management practices directly improve HCI components, which in turn elevate the technical performance of CDSSs. The table addresses the research gap identified earlier by systematically linking these domains, offering a structured perspective on mutual reinforcement. This approach aligns with the study’s objective of establishing a scientific framework that connects data management strategies to HCI factors, ultimately fostering more intuitive and effective CDSS environments.

5.1. Real-World Standardization Challenges

Despite recent advancements in data management technologies, several constraints remain in CDSS design. The variety in medical jargon is a severe limitation in this context. Inconsistent terminology complicates data entry and retrieval, impacting research accuracy and clinical decision-making [83,95]. Although the range control and obligatory data fields have been proposed as potential solutions, these modules require consistent updating with recent medical modifications [29]. On the other hand, the standard discrepancies among different countries present significant constraints on data formats. This limitation hinders standardization and normalization through a structured data transformation pipeline [81]. Moreover, healthcare institutions often use region-specific coding systems and documentation styles, making data interoperability a persistent challenge. To address these discrepancies, adopting widely recognized interoperability standards, such as HL7 FHIR (Fast Healthcare Interoperability Resources) and OMOP (Observational Medical Outcomes Partnership), can provide a structured approach to harmonizing medical data across institutions and regions [96,97].

Additionally, AI-driven terminology mapping and natural language processing (NLP) techniques can aid in translating disparate terminologies into unified medical vocabulary, reducing inconsistencies across different systems [84,98]. A phased, modular approach (starting with critical datasets like lab results and expanding via adaptable adapters) offers flexibility, allowing institutions to adopt standards incrementally without requiring extensive infrastructure upgrades [99]. This strategy mitigates resistance from regions with entrenched coding practices, enhancing data reliability and supporting cross-border research and care coordination [100].

5.2. Implementation Challenges

Implementing these data management solutions in CDSSs entails challenges, notably data privacy and interoperability with legacy systems. Data integration across systems with varying security policies raises privacy concerns, as sensitive patient information may be exposed during aggregation from diverse sources with inconsistent credential protocols (e.g., single sign-on vs. multi-factor authentication) [101]. Robust encryption (e.g., AES-256) and standardized access controls are essential to safeguard privacy [102]. Moreover, integrating with legacy systems (often built on outdated data architecture) poses additional challenges, necessitating the implementation of robust encryption protocols and standardized interfaces to ensure secure and seamless interoperability [103,104].

On the other hand, wearable devices and medical Internet of Things (IoT) solutions generate vast amounts of real-time data, such as vital signs, activity levels, and sleep patterns. While these integrations can enhance continuous patient monitoring and early warning systems, they bring significant challenges. One of these challenges is high-volume data, which can complicate CDSS functionality and make it harder for physicians to reach rational solutions. Furthermore, the lack of standardization in IOT data formats and device protocols entangles seamless integration into existing CDSSs [105]. In data integration, another challenge arises with alert sensitivity and the specified rules that trigger them. Specifically, medical alerts containing similar concerns originating from separate subsystems before the integration process negatively impact the system’s usability and lead to alert abandonment [106,107].

5.3. Human-Centered Design Considerations

Human-centered design (HCD) optimizes the data management–HCI bridge by prioritizing clinician needs. For data entry, real-time text previews and auto-suggestions reduce cognitive load and errors, allowing physicians to focus on patient care rather than manual input [108]. For standardization, unified dashboards displaying normalized data (e.g., vital trends in consistent units) improve comprehension and are designed with clinician input to highlight actionable insights [109]. For integration, customizable alert settings (adjustable to workflow preferences) mitigate fatigue, reflecting HCD’s emphasis on adaptability [110]. These designs, grounded in principles of empathy and simplicity, enhance usability and foster user trust. This enhancement is achieved through iterative usability testing with clinicians, ensuring alignment with the demands of clinical practice [111]. Such approaches elevate CDSS adoption by addressing human factors often neglected in technical solutions.

5.4. Implementation Constraints

While this study offers a novel framework bridging data management and HCI to enhance CDSS functionality, several limitations must be acknowledged. First, the proposed data processing strategies (such as automated data validation, normalization, and integration) may introduce computational overhead. These mechanisms require real-time performance, particularly in high-volume environments like emergency departments or intensive care units, which may challenge system responsiveness and scalability [112].

Second, the framework’s adaptability to diverse healthcare environments is not yet fully validated. Healthcare settings vary significantly in digital maturity, infrastructure availability, and staff training levels, which can influence the adoption and effectiveness of this framework. For instance, institutions with limited IT support or reliance on paper-based records may face more significant barriers in implementing standardized data structures and advanced interface designs [113].

Moreover, despite advancements, constraints persist in inconsistent medical jargon, complicating data entry and retrieval and impacting accuracy and decision-making [114]. Range control and obligatory fields need regular updates to reflect evolving standards [115]; however, cross-national format discrepancies hinder structured data transformation [116]. Although HL7 FHIR, OMOP, and NLP offer solutions [117,118], regional coding and resource disparities remain challenges, necessitating tailored strategies to ensure broad applicability [119].

5.5. Practical Implementation Guidelines

A structured, evidence-informed approach is recommended to operationalize the proposed framework across real-world healthcare environments. First, automated text generation tools should be deployed alongside structured clinician training programs, ensuring adoption and alignment with clinical documentation standards. These tools have demonstrated the potential to reduce manual errors and alleviate cognitive load, particularly in high-frequency data-entry tasks [115]. Second, interoperability should be pursued incrementally through the phased implementation of widely accepted standards such as HL7 FHIR. To address compatibility challenges with legacy systems, interfacing solutions (such as middleware or adapters) can be employed, with an initial focus on high-impact clinical domains like diagnostics and laboratory systems [117]. Third, human-centered design principles should guide interface development, emphasizing iterative usability testing involving clinicians. Tailored dashboards, alert configuration, and workflow-specific adaptations should be continuously refined through participatory design methods to ensure alignment with end-user needs and clinical efficiency [111,120]. Finally, pilot studies in controlled clinical environments are recommended before full-scale deployment. These pilots can help identify context-specific barriers and validate the framework’s effectiveness in enhancing CDSS usability, integration, and performance across diverse institutional settings [121].

Despite the mentioned constraints and obstacles to bridge between data management and HCI concepts, their mutual relationship cannot be denied. The proper data management and integration streamline the user interaction with the system through consistent data presentation and providing reliable medical data. On the other hand, enhancing the system usability and HCl aspects impedes data entry errors and missed data, leading to the supply of suitable data for medical decisions and empowering data management aspects. As depicted in Figure 5, this synergistic and bilateral effect can lead to a positive loop: improved HCI empowers data management by ensuring high-quality data, and reciprocally, effective data management enhances HCI by providing a seamless and user-friendly experience.

This study offers several scalable solutions like ‘automated text generation’ to mitigate data entry challenges, resulting in reduced user workload. Moreover, by integrating data management techniques with HCI principles, as illustrated in Table 2, this research highlights how effective data practices can enhance user satisfaction and elevate CDSS performance. This way, the system usability in different HCI aspects will be elevated, bringing tangible benefits for clinical outcomes.

Future research should focus on advanced NLP tools to address terminology inconsistencies and provide real-time harmonization of medical terms. Developing globally standardized data formats is essential for seamless interoperability across systems and regions. Enhanced training protocols will empower healthcare providers to effectively utilize complex, integrated systems. Advancing these areas will drive innovation and ameliorate CDSS functionality and patient outcomes.

6. Conclusions

This study has successfully bridged the gap between prominent human–computer interaction concepts and foundational data science principles, including data management, transformation, and integration, within the CDSS context. By delving into practical scenarios, we have demonstrated the mutual impact of data entry management and data integration on HCI elements such as usability, cognitive load, and user satisfaction (and vice versa). These analyses highlight the symbiotic relationship between data science and HCI, emphasizing how thoughtful system design can enhance data quality and user interaction.

Our findings underscore that prioritizing HCI principles significantly improves data accuracy and consistency, which is critical for effective medical decision-making. High-quality data stemming from data flow control and seamless integration enhances CDSS reliability and functionality, leading to the desired patient outcomes. By introducing this scientific framework, we contribute to a deeper comprehension of how integrating HCI into data practices can elevate the performance and adoption of advanced healthcare technologies.

Consistent attention to this interplay between data-driven and human-centered approaches will be crucial to advance healthcare systems further.