Development and evaluation of a clinical dashboard for nephrology inpatients: a pilot study Authors: Nadieh Rouhani, Khadijeh Makhdoomi, Aynaz Nourani, Mohamad Jebraeily, Bahlol Rahimi Published: 2026 ABSTRACT IntroductionKidney disease imposes a significant burden on healthcare systems due to high complications, mortality, and costs. Effective management requires regular monitoring, which electronic information systems can support. Clinical dashboards visualize critical data and key performance indicators (KPIs) in real-time to aid decision-making. This study aimed to design, develop, and evaluate a clinical dashboard for inpatient nephrology care in Iran.MethodsThis multi-phase developmental study (2022–2023) was conducted in the Nephrology and Kidney Transplantation Department at Imam Khomeini Hospital, Urmia. A two-stage Delphi process with 25 nephrologists identified KPIs. Conceptual and logical models were designed, followed by dashboard development using Laravel, MySQL, and Android Studio (with barcode integration). Usability was evaluated by 10 experts using a Nielsen-based questionnaire, and system success was assessed by 40 clinical staff after a three-month pilot using the DeLone & McLean model.ResultsThe Delphi yielded 58 KPIs categorized into demographic, medical history, clinical/laboratory, and discharge groups. The highest consensus was on electrolytes (4.88 ± 0.33) and vital signs/blood pressure. Usability evaluation showed acceptable scores in most Nielsen principles (e.g., high in visibility and aesthetic design), though orientation was lower and subsequently improved. End-user evaluation indicated high system quality, information quality, user satisfaction, and perceived net benefits.ConclusionThe dashboard demonstrated acceptable usability and success in a pilot setting, with clear visualizations, barcode-enabled bedside access, and categorized KPIs, providing a platform with the potential to support real-time monitoring and clinical decision-making in nephrology.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12911-026-03629-4. Introduction The kidneys are vital organs responsible for maintaining electrolyte balance, filtering blood, and regulating blood pressure [1]. Damage to these organs can lead to severe health issues, including structural and functional disorders [2]. Major risk factors for kidney disease include hypertension, diabetes, genetic predisposition, aging, and environmental or regional determinants [3]. Kidney damage, regardless of its etiology, can substantially diminish quality of life and lead to multi-organ complications or mortality [4]. Globally, kidney diseases affect approximately 10% of the population and represent the twelfth leading cause of death [5]. In Iran, the prevalence of kidney diseases ranges from 6% to 15%, placing a considerable burden on the healthcare system through increased costs and elevated mortality rates [6]. Given the high and rising prevalence of kidney diseases, healthcare systems worldwide are facing mounting economic and operational pressures in managing this chronic condition [7, 8]. The management of hospitalized patients in nephrology departments demands continuous monitoring of vital signs, laboratory values (blood and urine), electrolyte levels, imaging studies, and physical examinations [9, 10]. These clinical data are essential for establishing treatment protocols, adjusting medication dosages, and planning ongoing care [11, 12]. However, the absence of efficient tools for real-time data analysis significantly hampers timely clinical decision-making and accurate tracking of disease progression. As data are generated continuously over extended periods, manual analysis becomes increasingly impractical, underlining the urgent need for advanced health information technologies [13–15]. To address these challenges, various information technologies, such as electronic health records [16, 17], decision support systems [18], and telemedicine systems [19, 20] have been developed to aid in the monitoring and management of kidney patients. These technologies have proven effective in improving disease management, supporting clinical decision-making, and enhancing patient outcomes. Among these tools, clinical dashboards have emerged as particularly effective for real-time patient management and performance tracking [21–24]. A clinical dashboard is a dynamic, interactive decision-support tool that presents key performance indicators (KPIs) in a visual format, typically by integrating data from multiple sources [25, 26]. Dashboards provide timely, accurate, and easily interpretable information, enabling clinicians to make informed decisions efficiently, particularly in fast-paced clinical environments [23, 26, 27]. Visual analytics, a core feature of healthcare dashboards, is widely used to interpret clinical, scientific, and administrative data [28]. The advantages of dashboards include their ability to manage large datasets, present information intuitively, and support evidence-based decision-making to improve care efficiency and quality [29]. Given the specific needs in the care of kidney patients and the capabilities of dashboards, several studies have focused on developing and evaluating clinical dashboards for managing this patient group [21, 23]. For example, Swee et al. emphasized that monitoring kidney patients largely depends on tracking laboratory results and vital signs. As a result, dashboards for kidney disease management have emerged as a promising tool for improving care quality. Their findings revealed that the dashboard improved early diagnosis and identification of kidney disease, saving both time and costs for patients [21]. Similarly, Stevens et al. developed a dashboard that successfully facilitated disaster planning during the COVID-19 pandemic, enabling systematic monitoring of kidney patients [23]. Despite international advancements in clinical dashboards for kidney disease (primarily in outpatient or resource-rich settings), no context-specific dashboard has been reported for hospitalized nephrology patients in Iran or similar resource-constrained environments. This study addresses this gap by developing and pilot-evaluating the first dedicated inpatient nephrology dashboard in Iran, featuring: (1) a Delphi-derived comprehensive KPI framework with four categories (demographic, medical history, clinical/laboratory, and discharge), (2) barcode wristband integration for rapid bedside access, and (3) a hybrid approach combining UML modeling, expert usability evaluation, and end-user success assessment in a real-world setting. Methods Study design This multi-phase developmental and quantitative study was conducted between 2022 and 2023 in the Nephrology and Kidney Transplantation Department of Imam Khomeini Teaching and Therapeutic Hospital, affiliated with Urmia University of Medical Sciences. The study process consisted of four main stages, which began with the identification and refinement of key performance indicators (KPIs). A conceptual and logical model was designed as the basis for the clinical dashboard. The usability of the developed dashboard was evaluated, and finally, its success in the clinical setting was examined. Determination of key performance indicators To determine the KPIs for the nephrology clinical dashboard, a narrative review of relevant literature was conducted, including scientific studies, dashboard development reports [30–32], and international Kidney Disease Improving Global Outcomes (KDIGO) guidelines published between 2012 and 2022 [33, 34]. Key guidelines included KIGO 2012, 2021, and 2022 for CKD evaluation, blood pressure and management, and diabetes management [33–35]. As no specific national Iranian guidelines for inpatient nephrology KPIs were identified, international standards were adapted to the local context. The extracted information was compiled into a preliminary checklist of 56 items, including potential KPIs and supporting data elements, categorized into four main groups: demographic information, medical history, clinical and laboratory data, and discharge information (Supplementary File 1, Part 1). Before the formal Delphi rounds, a panel of five experienced nephrologists (mean 12 years of experience in CKD and dialysis management) reviewed these 56 literature-derived items in an interactive consensus session. Using predefined criteria, they distinguished critical KPIs, metrics reflecting clinical performance, that enable ongoing patient monitoring, guide therapeutic decisions, and support care quality evaluation from supporting data elements. This step minimized individual bias and ensured clinical relevance before the broader Delphi process. Given the patient-centered focus, KPIs were broadly defined as essential variables for individualized real-time monitoring, risk stratification, treatment planning, and bedside decision-making, including laboratory/vital sign metrics as well as key demographic and social factors directly influencing CKD etiology, progression, complication risk, and personalized care (e.g., age/gender differences, occupational kidney exposures, socioeconomic barriers). The 25 pre-selected KPIs were compiled into a Delphi questionnaire using a 5-point Likert scale (ranging from “very valuable” to “not valuable”). After validation, the questionnaire (Supplementary File 1, Part 2) was distributed electronically via the Iranian Society of Nephrology’s social and professional networks to collect responses from a geographically diverse community across eight provinces. Inclusion criteria were limited to active clinical practice in nephrology, and participation was voluntary to ensure diverse representation. Between rounds, participants received anonymized aggregated feedback, including mean scores, standard deviations, and de-identified comments. New suggestions from open-ended questions in the first round (n = 15) were reviewed for clinical relevance and duplication, and included in the second-round questionnaire for formal scoring and consensus evaluation. Sample size followed established Delphi recommendations for specialized panels of 20–30 experts. Of 29 initial volunteers, 25 nephrologists completed both rounds (86% overall response rate; 100% retention between rounds). In the first round, items with > 75% agreement were selected for inclusion in the dashboard; items with 50–75% agreement advanced to the second round, and items with < 50% agreement were excluded. The second round applied the same thresholds (Supplementary File 1, Part 3). These percentage-based consensus thresholds (> 75% inclusion, 50–75% second-round review, < 50% exclusion) were predefined according to standard practice in clinical and health informatics Delphi studies, providing precise and reproducible decision rules for specialized panels [36, 37]. New suggestions from open-ended questions in the first round (n = 15) were reviewed by the research team for relevance and duplication and included in the second-round questionnaire for formal scoring and consensus evaluation, in line with established Delphi methodology. Dashboard design and development Conceptual and logical models, including use case, activity, class, entity-relationship, data flow, and sequence diagrams (Figs. 1 and 2), were developed using Visual Paradigm v8 to provide a structured blueprint and facilitate communication between researchers and developers. These models were reviewed and finalized in dedicated meetings with the research team and clinical consultants. The core web-based dashboard was implemented using PHP and CSS within the Laravel framework (MVC architecture) for scalability, maintainability, and built-in security. The relational database was developed in MySQL for efficient data management. A complementary mobile barcode-scanning module was developed separately using Android Studio, IntelliJ IDEA, and Google ML Kit to enable rapid bedside patient identification by scanning wristband barcodes, providing secure, immediate access to patient profiles in the web dashboard. Together, this approach produced a comprehensive, user-friendly, and safe system for managing complex patient data in nephrology and kidney transplantation workflows. To address technical challenges in a resource-constrained setting with limited standardized interoperability, real-time data synchronization from the hospital information system (HIS) and laboratory information system (LIS) was implemented via custom pull-based RESTful APIs developed in Laravel. Barcode-based patient identification was integrated through Google ML Kit, ensuring accurate bedside access and minimizing identification errors. The dashboard was designed to be responsive across web and mobile devices, supporting both nurses (data entry) and physicians (monitoring). Complex interoperability standards such as HL7 or FHIR were deliberately not adopted due to local infrastructure limitations, prioritizing rapid deployment, usability, and maintainability. Security was maintained through Laravel’s authentication system, role-based access control distinguishing nurses and physicians, and encrypted barcode-mediated patient identification. The validated UML conceptual and logical models served as the foundation for prototyping before full-scale coding. Dashboard usability evaluation Ten usability experts evaluated the developed dashboard before deployment in the nephrology and kidney transplantation department to extract usability problems. The evaluation utilized a customized usability checklist based on Nielsen’s framework, explicitly designed for dashboards [38]. The checklist comprised 49 questions across 10 sections, each corresponding to Nielsen’s usability principles (Supplementary file 2). To analyze the questionnaires, we aggregated participants’ responses across all 10 categories and calculated the mean for each category. A mean score of ≥ 6.5 (out of 10) per heuristic was considered acceptable. Severity levels for each heuristic category (No problem, Minor problem, Major problem) were qualitatively assigned based on mean scores, adapted from Nielsen’s severity rating framework [39]. Dashboard success After resolving usability issues, the nephrology dashboard was deployed in the nephrology and kidney transplant department of Imam Khomeini Hospital. User onboarding consisted of brief hands-on demonstrations (approximately 30 min) provided by the research team to nurses and nephrologists. No formal training protocol or manual was developed, given the intuitive design and the exploratory nature of this pilot study. Nurses entered data into the system, and nephrologists regularly used the system to monitor patients’ status. Three months after the system was deployed, 40 nurses and nephrologists who used the dashboard completed the Delon and McLean System Success Questionnaire (Supplementary file 3). The questionnaire consisted of 24 questions on a four-point Likert scale (from “strongly agree” to “strongly disagree”) and covered six main dimensions: system quality, information quality, service quality, utilization, user satisfaction, and perceived net benefits. Finally, participants’ responses were analyzed using descriptive statistics, including calculation of frequencies, means, and standard deviation. Responses were complete with no missing data. Means ± SD were used for Likert items, following common practice in DeLone & McLean-based health informatics evaluations, to provide interpretable summaries for this pilot study (n = 40); confidence intervals were not reported due to the descriptive focus. Results Findings are presented separately for each phase of the study. Determination of key performance indicators To conduct the Delphi process, 29 Nephrologists volunteered following a call from the Iranian Society of Nephrology’s social and professional networks. The questionnaire, covering KPIs, was distributed electronically to enhance participant engagement. A total of 25 Nephrologists participated in the first round, and the same participants continued in the second round. Demographic information of the participants is shown in Table 1. Table 1Demographic information of participants in the determination of KPIsVariablesCategoriesFrequencyPercentageAge31–40520%41–501248%51–60832%GenderMale1768%Female832%Work experience6–10832%11–15936%16–20520%21–25312% Table 1Demographic information of participants in the determination of KPIsVariablesCategoriesFrequencyPercentageAge31–40520%41–501248%51–60832%GenderMale1768%Female832%Work experience6–10832%11–15936%16–20520%21–25312% Demographic information of participants in the determination of KPIs As shown in Table 1, most participants were aged 41–50 (48%), male (68%), and had 11–15 years of professional experience in nephrology (36%). This study used the Delphi method to identify KPIs in two rounds. Before the Delphi phase, five nephrologists with an average of 12 years of experience in managing kidney and dialysis patients first assessed 56 items extracted from the literature review (Supplementary File 1: Part 1). They labeled 25 items as KPIs and 31 items as data elements. Then, the KPIs were compiled into a questionnaire and provided to the participants. In the first round, 25 KPIs were assessed (Supplementary File 1: Part 2). Eighteen KPIs with a level of agreement of 75% or higher were identified and selected for inclusion in the dashboard design. One item (Picture Archiving and Communication System (PACS) Code) was excluded because it received less than 50% agreement among panelists. The other seven items (including insurance, occupation, socio-economic status, pathology (Biopsy results), therapeutic interventions, Pathology results follow-up, and Subsequent patient visits), with levels of agreement between 50% and 75%, advanced to the second round. Additionally, 15 open proposals submitted by nephrologists in the first round were included in the second round and achieved the required score. The final key performance indicators are presented in Table 2. Table 2Final key performance indicators (KPIs) selected for the dashboard#KPIs(Mean ± SD) Demographic 1Age(4.24 ± 0.60)2Gender(4.34 ± 0.76)3Insurance(4.24 ± 0.60)4BMI (Body Mass Index)(4.28 ± 0.84)5Occupation(4.08 ± 0.49)6Socio-economic status(3.76 ± 0.72) Medical History 7Diabetes(4.56 ± 0.51)8Hypertension(4.68 ± 0.48)9Dialysis(4.72 ± 0.46)10Cardiovascular Diseases(4.68 ± 0.48)11Kidney Diseases(3.74 ± 1.19)12History of anemia related to CKD(3.96 ± 0.74)13Duration of CKD(4.44 ± 0.51)14Stage of CKD(3.74 ± 1.13)15History of electrolyte imbalance(3.76 ± 0.66)16History of vascular access complications(4.44 ± 0.65)17History of cardiovascular events (MI (Myocardial infarction), HF (Heart Failure))(4.32 ± 0.48) Clinical and Laboratory 18Pathology (Biopsy results)(4.52 ± 0.51)19Therapeutic Interventions(3.76 ± 0.72)20eGFR (Estimated Glomerular Filtration Rate) (4.72 ± 0.46) 21Parathyroid Hormone (PTH)(4.32 ± 0.51)22Ferritin & Iron Studies(3.68 ± 0.69)23HbA1c(3.84 ± 1.03)24Lipid Profile (Chol, LDL, HDL, TG)(4.32 ± 0.37)25Inflammatory Markers (CRP, ESR)(3.96 ± 0.68)26Specific Viral Markers (HBV, HCV, HIV)(3.76 ± 0.52)27Vital Signs (BP, HR, Temp, RR)(4.72 ± 0.46)28Hematology (CBC, Hb, Hct)(4.64 ± 0.49)29Electrolytes (Na, K, Ca, P, Mg) (4.88 ± 0.33) 30Biochemical Tests (Creatinine, BUN, Liver enzymes, Albumin)(4.32 ± 0.48)31Serology (Viral markers, antibodies)(4.12 ± 0.60)32Urinalysis (Protein, Microalbumin, Creatinine, Specific gravity, etc.)(4.68 ± 0.48)33CSF (cerebrospinal fluid) Analysis(4.32 ± 0.48)34Synovial Fluid Analysis(3.92 ± 0.70)35Pleural Fluid Analysis(3.52 ± 0.71)36Ascitic Fluid Analysis(4.36 ± 0.49)37Blood Gases (ABG)(3.92 ± 0.76) Discharge 38Adherence to dialysis schedule(4.48 ± 0.51)39Patient-reported outcomes(4.36 ± 0.49)40Nutritional Recommendations(4.36 ± 0.57)41Drug Treatment Recommendations(4.64 ± 0.41)42Therapeutic Interventions Follow-up(4.56 ± 0.51)43Pathology Results Follow-up(4.24 ± 0.52)44Subsequent Patient Visits(3.52 ± 0.74) Table 2Final key performance indicators (KPIs) selected for the dashboard#KPIs(Mean ± SD) Demographic 1Age(4.24 ± 0.60)2Gender(4.34 ± 0.76)3Insurance(4.24 ± 0.60)4BMI (Body Mass Index)(4.28 ± 0.84)5Occupation(4.08 ± 0.49)6Socio-economic status(3.76 ± 0.72) Medical History 7Diabetes(4.56 ± 0.51)8Hypertension(4.68 ± 0.48)9Dialysis(4.72 ± 0.46)10Cardiovascular Diseases(4.68 ± 0.48)11Kidney Diseases(3.74 ± 1.19)12History of anemia related to CKD(3.96 ± 0.74)13Duration of CKD(4.44 ± 0.51)14Stage of CKD(3.74 ± 1.13)15History of electrolyte imbalance(3.76 ± 0.66)16History of vascular access complications(4.44 ± 0.65)17History of cardiovascular events (MI (Myocardial infarction), HF (Heart Failure))(4.32 ± 0.48) Clinical and Laboratory 18Pathology (Biopsy results)(4.52 ± 0.51)19Therapeutic Interventions(3.76 ± 0.72)20eGFR (Estimated Glomerular Filtration Rate) (4.72 ± 0.46) 21Parathyroid Hormone (PTH)(4.32 ± 0.51)22Ferritin & Iron Studies(3.68 ± 0.69)23HbA1c(3.84 ± 1.03)24Lipid Profile (Chol, LDL, HDL, TG)(4.32 ± 0.37)25Inflammatory Markers (CRP, ESR)(3.96 ± 0.68)26Specific Viral Markers (HBV, HCV, HIV)(3.76 ± 0.52)27Vital Signs (BP, HR, Temp, RR)(4.72 ± 0.46)28Hematology (CBC, Hb, Hct)(4.64 ± 0.49)29Electrolytes (Na, K, Ca, P, Mg) (4.88 ± 0.33) 30Biochemical Tests (Creatinine, BUN, Liver enzymes, Albumin)(4.32 ± 0.48)31Serology (Viral markers, antibodies)(4.12 ± 0.60)32Urinalysis (Protein, Microalbumin, Creatinine, Specific gravity, etc.)(4.68 ± 0.48)33CSF (cerebrospinal fluid) Analysis(4.32 ± 0.48)34Synovial Fluid Analysis(3.92 ± 0.70)35Pleural Fluid Analysis(3.52 ± 0.71)36Ascitic Fluid Analysis(4.36 ± 0.49)37Blood Gases (ABG)(3.92 ± 0.76) Discharge 38Adherence to dialysis schedule(4.48 ± 0.51)39Patient-reported outcomes(4.36 ± 0.49)40Nutritional Recommendations(4.36 ± 0.57)41Drug Treatment Recommendations(4.64 ± 0.41)42Therapeutic Interventions Follow-up(4.56 ± 0.51)43Pathology Results Follow-up(4.24 ± 0.52)44Subsequent Patient Visits(3.52 ± 0.74) Final key performance indicators (KPIs) selected for the dashboard Based on Table 2, items related to Electrolytes (Na, K, Ca, P, Mg) showed the highest consensus. In contrast, follow-up and advanced fluid analyses demonstrated lower consensus, reflecting differences in perceived clinical relevance. Design and development At this stage, conceptual models (including use cases, activity diagrams, and class diagrams) and logical models (including ERDs, DFDs, and sequence diagrams) were designed. Subsequently, a validation session was held with the research team and two experienced specialists who have master’s degrees in software engineering. After incorporating the necessary revisions, the finalized conceptual models were utilized in the software development process. For illustration, Fig. 1 presents a sequence diagram, and Fig. 2 depicts an example of an activity diagram. This systematic approach ensures clarity and alignment between the design and development phases. Fig. 1Sequence diagram for retrieving a patient’s report Sequence diagram for retrieving a patient’s report Fig. 2Activity diagram for entering new data for a patient who is already registered in the dashboard Activity diagram for entering new data for a patient who is already registered in the dashboard After the conceptual and logical models were validated, they were transferred to a team of four specialists- web developers, graphic designers, database engineers, and mobile programmers- for software development. The dashboard was developed over three months, based on the approved model. The design emphasized simplicity, clarity, and a user-friendly approach. Colors were selected based on established clinical dashboard guidelines: green for normal/safe values and trends, amber/yellow for warning or moderate deviations, and red for critical alerts and abnormal values, in order to enhance rapid visual interpretation and align with usability principles. Figure 3 shows the patient profile login page, accessible via username and password or by scanning the wristband barcode. Wristband barcode technology was designed to connect the dashboard profile to the patient, making it easy to access and view each patient’s information, and ensuring that the correct data is available for each patient on the dashboard. Fig. 3Patient profile login page (with barcode scanning option) Patient profile login page (with barcode scanning option) Figures 4 and 5 illustrate the dashboard’s longitudinal visualization of trends in biochemical data and blood gas status over time. Additional Screenshots related to the dashboard are provided in Supplementary File 4. Fig. 4Patients’ biochemical data changes Patients’ biochemical data changes Fig. 5The patient’s blood gas status The patient’s blood gas status Usability evaluation Ten usability experts evaluated the clinical dashboard for nephrology inpatients. Most of them were younger, in the 21–40 age group (80%). The gender composition also showed that about 60% were male and 40% were female. In terms of educational background, half of the participants held a PhD in medical informatics. In contrast, the remaining participants had either a PhD in health information management (20%) or a Master’s degree in medical informatics (30%). In terms of work experience, most experts had shorter experience, ranging from 1 to 5 years (60%), while another group reported more than 6 years of experience. The usability evaluation of the dashboard is summarized in Table 3. Table 3Usability evaluation of the clinical dashboard for nephrology inpatients#Evaluation categoryNumber of QuestionsFrequency (Mean) YesFrequency (Mean) NoFrequency (Mean) N/ASeverity Level*1Visibility of System Status656 (9.33)3 (0.50)1 (0.17)No problem2Match between the System and the Real World543 (8.60)7 (1.40)0 (0.00)Minor problem3User Control and Freedom539 (7.80)11 (2.00)0 (0.00)Minor problem4Consistency and Standards641 (6.83)11 (2.67)3 (0.50)Minor problem5Recognition rather than Recall426 (6.50)12 (3.00)2 (0.50)Minor problem6Flexibility and Efficiency of Use759 (8.42)9 (1.28)2 (0.28)No problem7Aesthetic and Minimalist Design752 (7.43)18 (2.57)1 (0.14)No problem8Spatial Organization329 (9.67)1 (0.33)0 (0.00)No problem9Information Coding217 (8.50)3 (1.50)0 (0.00)No problem10Orientation419 (4.75)11 (2.75)10 (2.50)Major problem*Severity level: Qualitative classification based on mean score per category – No problem (≥ 8.0: excellent compliance); Minor problem (6.5–8.0: low-priority issues); Major problem (≤ 6.5: significant issue) Table 3Usability evaluation of the clinical dashboard for nephrology inpatients#Evaluation categoryNumber of QuestionsFrequency (Mean) YesFrequency (Mean) NoFrequency (Mean) N/ASeverity Level*1Visibility of System Status656 (9.33)3 (0.50)1 (0.17)No problem2Match between the System and the Real World543 (8.60)7 (1.40)0 (0.00)Minor problem3User Control and Freedom539 (7.80)11 (2.00)0 (0.00)Minor problem4Consistency and Standards641 (6.83)11 (2.67)3 (0.50)Minor problem5Recognition rather than Recall426 (6.50)12 (3.00)2 (0.50)Minor problem6Flexibility and Efficiency of Use759 (8.42)9 (1.28)2 (0.28)No problem7Aesthetic and Minimalist Design752 (7.43)18 (2.57)1 (0.14)No problem8Spatial Organization329 (9.67)1 (0.33)0 (0.00)No problem9Information Coding217 (8.50)3 (1.50)0 (0.00)No problem10Orientation419 (4.75)11 (2.75)10 (2.50)Major problem*Severity level: Qualitative classification based on mean score per category – No problem (≥ 8.0: excellent compliance); Minor problem (6.5–8.0: low-priority issues); Major problem (≤ 6.5: significant issue) Usability evaluation of the clinical dashboard for nephrology inpatients *Severity level: Qualitative classification based on mean score per category – No problem (≥ 8.0: excellent compliance); Minor problem (6.5–8.0: low-priority issues); Major problem (≤ 6.5: significant issue) Overall, the usability evaluation showed strong performance in Visibility of System Status, Spatial Organization, Information Coding, and Aesthetic and Minimalist Design, reflecting effective information conveyance and a straightforward, uncluttered interface. Moderate scores were observed in Match between System and Real World, User Control and Freedom, Consistency and Standards, and Recognition rather than Recall, indicating minor opportunities for improvement. For each heuristic, a mean score out of 10 was calculated based on the proportion of “Yes” responses across all experts. A score of > 6.5 was considered acceptable, consistent with previous studies and Nielsen’s severity framework, where ≤ 5.0 indicates major problems, 6.5–8.0 minor issues, and ≥ 8.0 no problems. The Orientation category received the lowest scores and was rated as a major usability problem. Experts specifically cited confusion in the screen hierarchy, difficulty transitioning between KPI categories (e.g., demographic to clinical/laboratory sections), and a lack of contextual awareness due to missing navigational aids, such as breadcrumbs or location indicators. To address this major usability problem, we implemented specific navigation improvements, including persistent breadcrumbs, a fixed sidebar menu with highlighted active sections, quick-jump tabs across KPI categories, and prominent “Back to Overview” buttons on each screen. These changes were verified by the usability experts before pilot deployment and likely contributed to the subsequent high end-user satisfaction. Before clinical deployment, we addressed this issue by adding persistent breadcrumbs, a fixed sidebar menu with highlighted active sections, quick-jump tabs, and prominent “Back to Overview” buttons. Post-redesign expert checks confirmed substantial improvement in navigation clarity and reduced disorientation, likely contributing to the subsequent high end-user ratings in System Quality and Use. Severity levels were assigned based on Nielsen’s standard framework (no problem: mean ≥ 8.0; minor problem: 6.5-8.0; major problem: ≤6.5) and academic backgrounds (primarily medical informatics and health information management). The consistent agreement on orientation as the primary major issue supports the reliability of the findings. Although the dashboard incorporated local clinical expertise and benefited from end-user feedback during the three-month pilot, a limitation of the user-centered design process was the absence of direct methods such as interviews, clinical observations, or co-design workshops with frontline nurses and physicians. Future iterations could incorporate these approaches to further align features with daily workflows. Evaluation of system success A total of 40 staff members from the nephrology department participated in evaluating the system’s success. The gender distribution was nearly balanced, with 52.5% female and 47.5% male participants. Most participants were between 20 and 40 years old (67.5%), while the remainder were aged 40–60 years. Professionally, the group included physicians (20%), residents (20%), interns (35%), and nurses (25%). Regarding work experience, the majority (55%) had one to five years of experience, while the rest reported between 5 and 25 years. The evaluation results (Table 4) indicated that the system’s quality was at a high level, with an average score of (4.2 ± 0.61). This demonstrates that the dashboard is user-friendly and largely meets the users’ expectations in terms of functionality. The quality of information was also assessed with an average of (4.0 ± 0.53), which indicates that the data provided by the system is accurate and practical and can play an effective role in clinical decision-making. In the field of service quality, an average score of (3.8 ± 0.69) was obtained. Although users assessed the technical support as adequate overall, they indicated that the speed and continuity of response could be improved to some extent. The “use” dimension, with an average of (4.1 ± 0.62), showed that this system helps to increase efficiency in performing tasks and improve the overall performance of employees. The highest level of satisfaction was related to the dimension of “user satisfaction”, which achieved an average score of (4.3 ± 0.53). This score reflects the overall user satisfaction with the system. Finally, the dimension of “Perceived Net Benefits”, with an average of (4.2 ± 0.64), showed that the dashboard has a favorable impact on improving patient care and management. Table 4Success evaluation results about the clinical dashboard for nephrology inpatients#Evaluation categoryNumber of QuestionsMeanSDInterpretation1System Quality44.20.61High usability2Information Quality44.00.53Accurate and useful information3Service Quality43.80.69Adequate technical support4Use44.10.62Enhanced task efficiency and job performance5User Satisfaction34.30.53High satisfaction with the system6Perceived Net Benefits54.10.64Positive impact on patient care, communication, and decision-making Table 4Success evaluation results about the clinical dashboard for nephrology inpatients#Evaluation categoryNumber of QuestionsMeanSDInterpretation1System Quality44.20.61High usability2Information Quality44.00.53Accurate and useful information3Service Quality43.80.69Adequate technical support4Use44.10.62Enhanced task efficiency and job performance5User Satisfaction34.30.53High satisfaction with the system6Perceived Net Benefits54.10.64Positive impact on patient care, communication, and decision-making Success evaluation results about the clinical dashboard for nephrology inpatients Overall, these results demonstrate that the system successfully meets usability and information needs, supports clinical workflows, and provides tangible benefits in healthcare delivery. Discussion This research was conducted in four main stages: identifying key performance indicators (KPIs), designing conceptual and logical models, assessing usability, and finally implementing and evaluating the dashboard’s success. In the first stage of this study, due to the difficulty of identifying data elements from key performance indicators, a literature review was conducted, after which all relevant items were extracted. Then, an initial filter was performed on the extracted items, and the key performance indicators were identified and entered into the Delphi study. The indicators were classified into four main groups, including patient demographic characteristics, medical history, clinical and laboratory data, and discharge information. Among these, nephrologists identified indicators related to electrolytes and glomerular function as the most critical factors in patient management. Comparing the results of the first stage of this study with those of previous studies showed that classifying performance indicators into four categories can provide a comprehensive framework for monitoring and managing nephrology patients. In Dehghani et al.’s study [40], indicators were examined only in three axes: structure, process, and outcome, while some discharge and demographic indicators were overlooked. In the study by Ntabanganyimana et al. [41], the focus was solely on clinical and laboratory indicators. Compared with these studies, the broader classification in the present study provides a more comprehensive perspective for clinical and management decision-making. In the second stage, conceptual and logical modeling of the dashboard was carried out with a focus on the visual display of data and its interactivity. Pre-development modeling can provide a suitable foundation for system design by examining aspects such as roles, user activities, and specifying necessary data and data flow. This approach aligns with the findings of Yazdanian et al. [42], who emphasized the importance of creating a conceptual model to facilitate a common understanding among researchers and developers. The study by Berry et al. also highlighted the importance of developing a conceptual model before creating dashboards [43]. Similarly, Gray et al. considered conceptual and logical modeling as an essential step in the development of information systems [44]. While the UML diagrams served as a valuable design foundation and ensured traceability from requirements to implementation, clinical validation of user needs was primarily derived from the three-month pilot deployment and end-user evaluation rather than the models alone. After developing the dashboard, its usability was assessed by 10 usability experts using a questionnaire based on Nielsen’s principles before deployment in the clinical department. The results showed that, overall, the system’s usability is at a desirable level, although some minor challenges were observed in the “orientation” dimension and were corrected before deploying the dashboard. Many studies have highlighted the importance of assessing software system usability before use [45, 46]. However, there has always been disagreement about whether this assessment should be carried out by end users or by usability experts. In this study, following the deployment of the dashboard in the nephrology department, an attempt was made to assess its usability by experts before deployment and to address any usability issues raised by system end-users during the pilot deployment period. Therefore, an expert assessment was initially conducted using Nielsen principles to extract the main issues and review them from the experts’ perspective. This tool enabled the identification and prioritization of usability weaknesses. In a systematic review [46], tools such as the System Usability Scale (SUS), Technology Acceptance Model (TAM), and Health Information Technology Usability Evaluation Scale (Health-ITUES) were mentioned for evaluating dashboard systems and other information systems, but most of these tools rely on end-user perspectives and do not fully reflect the experts’ opinions on the nature and severity of usability issues. Therefore, in this study, medical informatics and health information management experts conducted an assessment to identify and resolve the main usability issues before presenting the dashboard to end users. The detailed identification and resolution of orientation challenges, such as unclear screen relationships addressed through breadcrumbs and sidebar navigation, offer practical lessons for designing intuitive clinical dashboards in time-pressured nephrology settings. The developed clinical dashboard was deployed as a pilot for three months in the nephrology department of Imam Khomeini Hospital, and 40 nurses and nephrologists used it in patient management. At the end of the three months, users rated the system’s success using a questionnaire based on the DeLone and McLean model. Based on the results of this evaluation, the quality of the system, information, and services, as well as the usability of the system and the level of user satisfaction with it, and finally, the Net benefits of the system were evaluated at an appropriate level, which indicated the success of the system during the pilot deployment period. Similar studies have also considered the success of the system [47]. It seems that this type of deployment and evaluation of information systems’ success can play an effective role in planning the permanent deployment of systems and their integration into the organization’s work processes. Overall, the main innovation of this study lies in its hybrid approach: a two-stage Delphi method was used to determine performance indicators, and conceptual and logical models were then used to design the dashboard. This process resulted in a final dashboard that was purposeful, efficient, and aligned with clinical needs. Integrating demographic, medical history, clinical, laboratory, and discharge data in this system provides physicians with a comprehensive view of the patient’s condition and may facilitate more informed clinical decisions. In addition, using an identification code (barcode) for quick access to each patient’s file enables longitudinal data to be viewed at the point of care. It facilitates faster decision-making at the patient’s bedside. Also, usability assessment first by experts and then by users can help provide a user-friendly and appropriate system. Ultimately, this dashboard offers a promising platform for managing nephrology patients and may facilitate data-driven decision-making in this area. Research limitations This study had several limitations that the researchers felt necessary to state. First, the implementation and evaluation were conducted at only one center, Imam Khomeini Hospital, the only specialized center for kidney and dialysis patients in West Azerbaijan province, which may limit the generalizability of the results to other hospitals or regions. Additionally, while the Delphi-driven KPI selection and staged evaluation methodology are replicable across settings, future multi-center studies are recommended to validate generalizability across diverse clinical practices, resources, and health information infrastructures. Second, the evaluation period was relatively short, and the system’s success was assessed only 3 months after deployment, which may not fully reflect its long-term effects on clinical processes and user satisfaction. Additionally, the brief pilot duration raises the possibility of the Hawthorne effect, whereby users may report higher satisfaction due to awareness of being studied; longer-term evaluations are therefore essential to confirm sustained usability and benefits. Third, the dashboard was designed explicitly for the nephrology and kidney transplantation department, and therefore, the results are not directly generalized to other medical disciplines or clinical settings. Also, the system’s performance and efficiency may be affected by local technical infrastructure, including network stability, hardware, and software environments, which may limit its transferability to other settings. Additionally, the absence of established interoperability standards such as HL7 FHIR or InterSystems IRIS limits potential integration with other hospital systems. Local infrastructure constraints necessitated this choice, but it represents a technical limitation for scalability and future multi-site deployment. Finally, due to the pilot nature and resource constraints, no objective measures of clinical outcomes (e.g., decision time, guideline adherence, length of stay, or readmission rates) were assessed. Additionally, the Delphi process relied on voluntary recruitment through professional networks, with no predefined formal inclusion/exclusion criteria beyond active nephrology practice, which may introduce minor selection bias despite achieving diversity in geography and experience. Conclusion This study aimed to design, develop, and evaluate the usability and success of a clinical dashboard for the nephrology inpatients. In this study, key performance indicators (KPIs) were categorized into four main categories: demographic information, medical history, clinical and laboratory history, and discharge information. In addition, a usability assessment using a custom-made Nielsen questionnaire indicated that the dashboard achieved acceptable usability scores from the experts’ perspective, although there were minor navigation challenges. Additionally, a three-month pilot deployment of the dashboard in the nephrology department and feedback from end users indicated positive perceptions of system success, laying a foundation with the potential to enhance patient monitoring and decision-making. However, no objective clinical outcomes were measured in this pilot study. The results of this study can serve as a basis for designing and developing similar dashboards in other medical specialties. It is suggested that future research investigate the practical impact of using this dashboard on improving patient outcomes and reducing clinicians’ workload. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4