A regional initiative in Crete, Greece, has demonstrated both the potential and the significant constraints of building a digital cardiovascular disease (CVD) surveillance system within primary care. The Cardiovascular Disease Observatory in Primary Health Care (CVD-PHC Observatory) was designed to support multidisciplinary care by systematically recording morbidity, risk factors, and treatments among primary care attendees. Published in Frontiers in Cardiovascular Medicine, this community case study details the five-phase implementation of the project, which ran from early 2019 to late 2020, and offers practical lessons for future digital health infrastructure in southern Europe. The initiative was a collaboration between the School of Medicine at the University of Crete, the Cardiology Department of the University Hospital of Heraklion, and the Region of Crete.
The project aimed to address the fragmented nature of CVD surveillance in Greece, where data has historically been scattered across paper records, hospital discharge data, and isolated electronic prescribing systems. Crete, the largest island in Greece with approximately 624,000 permanent residents, presents unique challenges due to its mix of urban centers and extensive rural areas. The observatory sought to create a regional big-data infrastructure for the systematic recording of CVD among primary care attendees aged 18 and older. The design was informed by the changing risk profile of the Cretan population, which has seen a substantial rise in obesity, metabolic syndrome, and dyslipidaemia over the past three decades.
Implementation proceeded through five sequential phases, beginning with the co-design of a structured data collection instrument. This instrument was developed by a multidisciplinary working group of general practitioners, epidemiologists, and cardiologists. The final version covered ten domains, including cardiovascular diagnosis, sociodemographic characteristics, lifestyle behaviours, family history, personal medical history, anthropometric measures, laboratory results, cardiac biomarkers, imaging findings, and pharmacotherapy. The data collection form was built around constrained picklists and conditional fields to reduce entry errors, with specific attention to ICD-10 coding for diagnoses and ATC coding for medications.
The second phase involved the development of a stand-alone digital recording platform. Due to pragmatic constraints, including unreliable connectivity in rural sites and limited budget, the team selected a Windows desktop application using C++ with a Microsoft Access back end. This choice allowed for offline operability and minimal infrastructure requirements on the practice side. The platform featured two user roles: a recorder role with rights to enter new records and an administrator role with rights to edit, upload, and manage user accounts. Although the original vision included a relational big-data architecture, the delivered system was a structured registry tool rather than a large-scale data platform.
Training for the participating healthcare professionals occurred in the third phase, with ten general practitioners and their teams invited to act as sentinel sites across the four prefectures of Crete. A structured two-hour training session was delivered in July 2020, combining a live demonstration of the platform with a moderated focus group discussion. Sites that could not attend the central session were trained individually via videoconference. The fourth phase involved a three-month pilot deployment from August to October 2020, designed specifically to elicit usability and integration feedback rather than to generate a representative epidemiological dataset.
Feedback from the pilot was critical in shaping the final deployment. General practitioners reported that the platform was technically feasible and comprehensible, but they unanimously recommended that any future iteration must be integrated with existing electronic prescribing and medical record systems to avoid duplicate data entry. Other recommendations included embedding ICD-10 coding throughout the form, allowing for multiple concurrent diagnoses, and upgrading the platform to support file attachments. These insights highlighted the gap between the project’s initial big-data ambition and the technical and legislative realities of the Greek primary care landscape.
The fifth and final phase was a 12-month deployment from December 2021 to December 2022, during which an upgraded version of the platform was used across the same sentinel network. The upgrade addressed several pilot feedback items, including improved ICD-10 coverage and extended export formats. During this period, the ten sites entered a total of 600 patient records. The recorded population had a mean age of 65.1 years, with 54.2% being women. The mean body mass index was 28.9 kg/m², and 41.7% of patients had an entry diagnosis of hypertensive diseases. These data provided a first snapshot of the cardiovascular risk profile across urban and rural Crete.
Despite the technical success of the deployment, the long-term value of the platform was constrained by several factors. The lack of interoperability with national e-prescription and electronic health record systems remained a significant barrier, as the stand-alone design limited scalability. The burden of duplicate data entry was a major concern for sustainability in routine care. Additionally, the disruption of the COVID-19 pandemic absorbed primary care capacity and affected the continuity of the project. The data collected during the pilot phase remained accessible only for a short period, highlighting the need for institutional data stewardship agreements in future initiatives.
A notable limitation of the initiative was the absence of cardiovascular nurses in the participating sites. Although CVD nursing is a formally recognised specialisation in Greece, it is not currently differentiated in the primary care staffing model. Consequently, the deployment was delivered primarily by general practitioners, and the role of nurses in CVD surveillance was under-represented. The authors emphasise that future iterations of such observatories should be co-designed with the nursing workforce to ensure that the system supports, rather than duplicates, interdisciplinary workflows. This gap represents a key lesson for the meaningful inclusion of nursing in regional health observatories.
The study concludes that while a region-wide CVD surveillance instrument can be co-designed and deployed under modest funding, sustainable next iterations must address the gap between big-data ambition and available resources. Interoperability with national systems is central to any future success, as is the integration of the full primary care team, including CVD nursing, from the first phase of design. The initiative offers transferable assets for digital CVD surveillance and highlights the importance of modern web-based architectures and long-term data stewardship for future regional health observatories in southern Europe.
The initiative was grounded in a specific epidemiological shift within the Cretan population. While the island historically maintained a low cardiovascular risk profile, a pattern documented in the Seven Countries Study and confirmed by later regional cohorts, this advantage has eroded over the past three decades. The authors note a substantial rise in obesity, metabolic syndrome, and dyslipidaemia, particularly among rural and ageing demographics. A cross-sectional study conducted by the same group in primary care patients aged 40 and older identified a metabolic syndrome prevalence of approximately 74%. Furthermore, a substantial proportion of these individuals presented with a 10-year Systematic COronary Risk Evaluation (SCORE) risk exceeding 10%, underscoring the urgent need for systematic surveillance and targeted action within the community.
The selection of the ten sentinel sites was purposive rather than random, reflecting the exploratory nature of the initiative and the small sample size. Sites were chosen from areas of elevated cardiovascular burden, as identified by the Regional Health Authority of Crete and prior epidemiological work. The network comprised seven health centres, two satellite rural practices, and one private general practice, distributed across all four prefectures. This distribution was intended to capture the diverse healthcare access challenges posed by Crete’s mix of urban centres and extensive rural and mountainous areas, where connectivity and continuity of care are well-documented challenges.
The structured data collection instrument was designed to capture a comprehensive clinical picture, moving beyond basic demographics to include specific cardiac biomarkers. The final version covered domains such as B-type natriuretic peptide, troponins, creatine kinase-MB, soluble urokinase plasminogen activator receptor, C-reactive protein, and lipoprotein-associated phospholipase A2 assay. It also included electrocardiography and imaging findings, such as left ventricular hypertrophy and ejection fraction. This level of detail was selected based on routine availability in Greek primary care records and informed by European guidelines on cardiovascular risk assessment, ensuring that the data collected was both clinically relevant and feasible for general practitioners to record during routine consultations.
During the 12-month deployment, the recorded population exhibited specific clinical characteristics that highlight the burden of disease in this setting. In addition to the mean age of 65.1 years and mean body mass index of 28.9 kg/m², the data revealed that 75.8% of patients had dyslipidaemia as a cardiovascular comorbidity. Family history of cardiovascular disease was present in 70.2% of the recorded population, while 49.9% were current or former smokers. These aggregate figures provide a snapshot of the risk factor profile in primary care attendees with confirmed cardiovascular diagnoses, although the authors clarify that a detailed epidemiological analysis of this dataset is beyond the scope of this report and will be addressed in separate publications.
The authors explicitly distinguish this work from a clinical trial or a nursing intervention study. Because cardiovascular nurses were not embedded in the participating sites, the deployment was delivered by general practitioners, and the manuscript reports lessons learned for future interdisciplinary efforts rather than evaluating the efficacy of a specific nursing-led intervention. The study is a reflective community case study, and its findings are limited to the specific organisational and regulatory features of Greek primary care. The qualitative feedback from general practitioners was synthesised rather than analysed through formal qualitative methods, and patient invitation and consent rates were not systematically recorded, preventing denominator-based participation assessments.
Looking forward, the authors propose that future iterations of such observatories must address the technical and workforce gaps identified in this pilot. They suggest a formally defined split of data capture responsibilities, where nurses routinely record anthropometry, vital signs, and lifestyle behaviours, while physicians handle diagnosis coding and pharmacotherapy. This approach would require co-design with the nursing workforce to ensure the digital platform supports, rather than duplicates, interdisciplinary workflows. The authors argue that small regional initiatives like this one can generate valuable implementation knowledge, mapping closely onto established usability and interoperability frameworks, thereby informing larger national or regional digital health strategies in southern Europe.
Subject of Research: Design and implementation of a digital cardiovascular disease observatory in primary care in Crete, Greece
Article Title: Building a cardiovascular disease observatory in primary care in Crete, Greece: design, implementation, and lessons learned for interdisciplinary digital surveillance
Article References: Anastasaki, M., Anastasiou, F., Parthenakis, F., & Lionis, C. (2026). Building a cardiovascular disease observatory in primary care in Crete, Greece: design, implementation, and lessons learned for interdisciplinary digital surveillance. Frontiers in Cardiovascular Medicine, 13, Article 1906642. https://doi.org/10.3389/fcvm.2026.1906642
Image Credits: AI Generated
DOI: 10.3389/fcvm.2026.1906642
Keywords: cardiovascular disease, primary care, digital health, surveillance, Crete, Greece, interoperability, nursing, epidemiology, health informatics, Building, cardiovascular
News Source: Ophelia Keating. (November 12, 2026). Crete’s CVD Observatory: Digital Surveillance Lessons. Scienmag.
