Cardiovascular diseases (CVDs) remain the leading cause of morbidity and mortality worldwide1. According to the latest available data from the Global Burden of Disease study, there has been a significant increase in the number of patients with CVD and CVD-related deaths in recent years. There is no sign of this trend abating2. This increase has been attributed to several factors, including an ageing population, an increase in risk factors such as hypertension, diabetes, tobacco use, obesity, and sedentary lifestyles, as well as improvements in diagnostic methods, which allow these diseases to be identified at an earlier stage3–5. Demographic projections suggest that the number of new cases of CVD and CVD-related deaths will continue to increase in the coming decades, and will reach alarming levels if prevention and control strategies are not intensified6.

In 2023, 9.8% of the Spanish population had some form of CVD, with 52.6% of cases occurring in women and 47.4% in men. These diseases are the second leading cause of death in Spain, accounting for 26.5% of all deaths, behind tumours (26.6%) and ahead of respiratory diseases (10.8%)7.

In response to the need to adopt new perspectives in CVD care, the Spanish National Health System implemented the Cardiovascular Health Strategy8, which aims to minimise the impact of these diseases and to prevent their onset by taking a comprehensive, patient-centred approach to cardiovascular health. This approach encompasses all influencing factors, including social determinants, health education, training of healthcare personnel, improvements in early diagnosis, and the importance of prevention and rehabilitation.

Likewise, the “Strategy for Addressing Chronicity in the National Health System” has established a key strategic approach to reorient healthcare delivery9, aiming to shift the system's focus from disease-centred to people-centred, while considering the needs of both the population as a whole and each individual.

To achieve this, stratification and prediction models have been recommended as the basic tools for managing chronicity. The “Population Stratification Project in the Spanish NHS” has been developed within this framework to promote this change in approach and ensure more personalised, patient-centred care10. Adopting this stratification approach makes it possible to accurately identify individuals within the population who are at the highest risk. This facilitates the design of more effective interventions to improve chronic disease management, as well as the implementation of preventive measures and health promotion.

This approach is consistent with the Capacity-Motivation-Opportunity (CMO) model developed by the Spanish Society of Hospital Pharmacy (SEFH) and is promoted through the SEFH “Strategic Map for Outpatient Care” (MAPEX) project. It aims to improve care and health outcomes for patients attending hospital pharmacy (HP) outpatient clinics and has already been adapted to different patient profiles11,12. At the project's second consensus conference in 2023, 5 strategic lines were prioritised for the period 2024 to 2027. One of these lines was to develop and implement new adaptations of the CMO-MAPEX methodology for different types of patients. This involved collecting and analysing data to enable their parameterisation, utilisation, and clinical evaluation with the aim of obtaining information on adherence, quality of life, and health outcomes13.

Although various tools for stratifying pharmaceutical care (PC) have been developed for patient groups treated by HPs, none to date have focused on CVD14–16.

The aim of this study was to develop a stratification tool for PC for patients with CVD, and to establish interventions tailored to each established priority level.

The study was developed through a methodological process that combined expert consensus, a critical review of previous models, and empirical validation via a pilot study. The project ran from February 2024 to December 2024, covering all the stages required to define and implement the model.

The first stage comprised organisation and context analysis. To this end, a working group comprising 12 hospital pharmacists specialising in cardiovascular patient care was formed. Most of the pharmacists were members of the SEFH Cardio Group and had either accreditation or experience in advanced PC in the cardiovascular field, working either part-time or full-time in this area. The group conducted an exhaustive review of the literature and previous models in order to identify the specific needs of the cardiovascular population. Given this background, it was deemed relevant to include quantitative and qualitative variables that could be primarily extracted from the electronic medical record and supplemented by structured interviews with the patients in the HF.

In stage 2, the stratification tool itself was defined and designed. Accordingly, during online workshops, the expert group defined the variables to be incorporated into the tool based on previous SEFH models and specific bibliographic searches. They grouped these variables into four dimensions: demographic; socio-health and functional; clinical and service use; and treatment-related.

It was agreed that a score would be assigned to each variable on a scale of 1 to 4, with higher values indicating greater importance for achieving health outcomes. These weights were assigned based on expert assessment and the existing bibliographic support for each item, thus ensuring that the clinical relevance of each variable was given the appropriate weighting.

In stage 3, the pre-test study was conducted and the thresholds between priority levels were defined. To achieve this, a pilot study was conducted with CVD patients recruited from 6 representative hospitals. The pilot study aimed to apply the tool in clinical practice in order to determine the classification thresholds for 3 risk levels (priorities 1, 2, and 3) according to the Kaiser Permanente pyramid care model. Data were collected systematically using a standardised Excel tool, drawing on information available in the electronic medical records of each centre, supplemented by data obtained from structured patient interviews. The weighted mean scores were calculated for the total model and each dimension. A sequential exclusion analysis was then performed to assess the individual impact of each variable on risk classification. The findings of the pilot study enabled the precise definition of the thresholds for the 3 levels, which were adjusted to the clinical reality of the cardiovascular environment.

In the final stage, the final adjustments were made, and the actions were defined and agreed upon, along with the frequency of stratification and restratification for each established priority level.

The risk stratification tool comprises a total of 20 variables grouped into 4 dimensions: 4 demographic, 2 socio-health, 8 clinical, and 6 pharmacotherapeutic. Table 1 shows the variables included and their specific weights. A key feature of the tool is that at least two-thirds of the data are already available prior to the patient's face-to-face consultation. This design was chosen to significantly optimise the pre-stratification process and intervention planning within the usual workflow, and to make implementation easier.

The pilot study included a total of 212 patients from 6 centres (Table 2). During the study, cut-off points were established for stratifying patients into 3 priority levels, according to the theoretical distribution of the Kaiser Permanente model (60–30–10). The cut-off points were defined as follows: priority 1, at least 23 points (10.4% of the sample); priority 2, 17 to 22 points (29.2%); and priority 3, less than 17 points (60.4%).

The distribution obtained showed good alignment with the expected pyramid model, allowing resources to be prioritised for patients at the highest risk while ensuring that the remaining patients received efficient care (Fig. 1).

Figure 1.

Cut-off points and priority levels established in the stratification tool for patients with cardiovascular disease.

Finally, specific interventions for PC were also established according to stratification level (Tables 3–5).

Priority 1: patients at this level require intensive follow-up, including face-to-face visits every 3–6months, a comprehensive review of pharmacotherapeutic objectives, and close coordination with the multidisciplinary team.

Priority 2: interventions focus on intermediate follow-up, with annual reviews and actions aimed at consolidating therapeutic adherence and adjusting treatments according to clinical needs.

Priority 3: lower-risk patients receive standard follow-up, focusing on health education, prevention, and general monitoring of their clinical situation.

Finally, the defined actions were divided into the following: pharmacotherapeutic follow-up interventions; patient education, training, and follow-up; and coordination with the healthcare team. It was also agreed that these interventions would be cumulative. Thus, priority 1 patients also receive the interventions planned for levels 2 and 3, while priority 2 patients also receive those for priority 3. This approach ensures comprehensive care tailored to the needs of each patient.

This study presents a stratification tool for PC—developed from the perspective of HP—to enable the early and accurate identification of patients at higher risk of poor outcomes and with greater care needs in the CVD setting. Thus, it facilitates personalised pharmacotherapeutic follow-up, allowing interventions to be individualised and reassessment intervals to be set according to the patient's risk profile. By integrating extensive data into the electronic medical record, including information from structured interviews, the model enables a multidimensional assessment that covers demographic, socio-health, clinical, and pharmacotherapeutic variables.

This comprehensive approach optimises resource allocation by focusing interventions on the patients who need them most. It also establishes the foundations for educational interventions that empower patients and improve the control of key factors, such as hypertension, cholesterol levels, and smoking. The tool is thus aligned with the paradigm shift proposed in the national strategies of Spain7,9, which seek to reorient healthcare from a disease-centred model to a patient-centred one.

The tool differs from other traditional and technological approaches that have attempted to address the complexity of cardiovascular risk. For example, tools such as the Medication Risk Complexity Index17 or the LACE index18 focus on one-dimensional aspects—either the complexity of the therapeutic regimen or the prediction of hospital readmissions—without systematically including the clinical, sociodemographic, and behavioural variables that are crucial in the clinical management of patients with CVD.

This difference is particularly relevant in light of the approaches of scientific societies such as the European Society of Cardiology. Its guidelines on cardiovascular disease prevention emphasise the importance of risk stratification to tailor preventive and therapeutic interventions to the individual characteristics of each patient19–21. Similarly, the American College of Cardiology/American Heart Association guideline on the primary prevention of CVD22 highlights the need for comprehensive cardiovascular risk assessment to guide personalised strategies that address both clinical and behavioural factors. These organisations concur that personalising interventions is essential to reducing the incidence of adverse events, thereby validating the implementation of models such as the one outlined in this study.

At the national level in Spain, the SEFH has also promoted initiatives within the MAPEX project—with approaches similar to the one proposed—which have been shown to improve patient health outcomes compared to more traditional approaches23–25.

One of the main advantages of this model is its ability to balance technical sophistication with clinical practicality. In contrast, although emerging models based on artificial intelligence and machine learning achieve high levels of accuracy, they often require complex infrastructures and large volumes of data, limiting their applicability in resource-constrained environments. On the other hand, the approach presented here relies on information already integrated into clinical systems, allowing for immediate and scalable implementation without sacrificing accuracy in risk prediction. One of the model's main strengths is its balance between complexity and operational feasibility, positioning it as a well-founded alternative for transforming clinical practice in the prevention and management of CVD26–28.

Other distinguishing features include the integration of educational interventions and the promotion of interdisciplinary collaboration. Although some models focus solely on number-based risk prediction, the proposed stratification tool encourages coordination between pharmacists, physicians, and other healthcare professionals, facilitating the exchange of information that enriches clinical decision-making. This collaborative approach has been consistently recommended by international organisations, which emphasise that a comprehensive and multidisciplinary care strategy is essential to address patient heterogeneity and improve health outcomes.

Stratification is only one pillar of the CMO methodology that has been proposed and tested in other diseases. Therefore, this process must be complemented by aligning short-, medium-, and long-term pharmacotherapeutic objectives, and by providing longitudinal patient follow-up, thereby enabling real-time or timely responses to their needs. This paradigm, which combines accuracy, operability, and ease of implementation, is emerging as a tool with great potential to guide future PC and cardiovascular prevention strategies, while adapting to the demands of a health system that is increasingly oriented towards personalised care.

This study is limited by its reliance on qualitative data obtained through structured interviews, which could introduce bias in the assessment of adherence and other patient behaviour. Furthermore, the limited sample size of the pilot study means that the results cannot be generalised to more heterogeneous populations. However, standardising the data collection instruments and using pre-existing data from medical records has helped minimise these limitations and promote the tool's reproducibility.

Although the model has been designed such that approximately two-thirds of the required data are already recorded in the electronic medical record—which, based on our experience, makes completing the data easier—the time required to complete the remaining data was not measured systematically. Future research should focus on developing automated processes that reduce the healthcare burden.

On the other hand, the main strength of the model lies in its multidimensional approach and its ability to integrate objective and subjective variables, enabling a comprehensive assessment of cardiovascular risk. The potential for pre-stratifying patients using information already available in clinical systems optimises the allocation of resources, ensuring that patients with the greatest need receive intensive and personalised interventions. In addition, the ease with which the model can be integrated into clinical practice—without requiring excessively complex technological infrastructures—underscores its operational viability across settings with varying resource levels, which is particularly important in the current context.

To consolidate the validity of the model and broaden its applicability, multicentre prospective studies incorporating a greater diversity of populations and healthcare centres are required. In this context, the integration of technological tools could further optimise how variables are weighted, enabling the detection of risk patterns that might be overlooked by traditional methods. Likewise, developing automated follow-up protocols that facilitate real-time restratification could allow interventions to be adjusted dynamically according to the patient's clinical evolution.

In addition to its operational viability in specialised care settings, the model has high scalability potential in other healthcare settings, particularly primary care, where the early detection and long-term management of cardiovascular risk are crucial.

Another promising line of research could be extending the model to healthcare settings other than specialised care, enabling the implementation of comprehensive, inter-level PC strategies across various therapeutic contexts. Finally, the impact of educational and collaborative interventions on adherence and clinical outcomes should be systematically evaluated, and the reduction in adverse events and savings in healthcare resources quantified, thereby supporting the added value of the stratification methodology adopted.

In conclusion, the proposed stratification tool represents a significant innovation in PC, as it enables personalised monitoring and optimal resource allocation. Its coordinated implementation across hospital pharmacy services would generate a uniform set of comparable indicators which, when integrated into institutional scorecards, would provide added strategic value: identifying gaps in care, guiding resource allocation, and designing common training programmes based on real needs. The multidimensional approach of the model, underpinned by the recommendations of scientific societies, promotes the adoption of educational and collaborative interventions among healthcare professionals. Applied in the context of the CMO methodology, this tool could—after future multicentre validation and automation—contribute to transforming clinical practice into more multidimensional, accurate, and efficient PC that is fully aligned with national health strategies in Spain.

This study introduces an innovative risk stratification tool for patients with cardiovascular risk in the field of pharmaceutical care. The tool comprises 20 variables grouped into 4 dimensions (demographic, socio-health and functional, clinical and service use, and treatment-related), enabling patients to be classified into 3 priority levels. This classification facilitates the adoption of interventions tailored to the intensity of patient needs, thereby optimising the allocation of healthcare resources.

The results of the pilot study showed that the clinical and service use dimensions accounted for more than 80% of the total score, which underscores the importance of targeting interventions in these critical areas. Specific strategies were established based on the defined cut-off points. Priority 1 patients receive intensive pharmaceutical care, including more frequent follow-ups, detailed reviews of their therapeutic regimen, and close multidisciplinary coordination. Priority 2 and 3 patients are managed with less intensive interventions, adapted to the complexity of their clinical situation and optimising the use of available resources.

This contribution to the literature highlights the importance of having a stratification tool that can objectively identify the risk level of each patient and enable interventions to be tailored to the individual. In addition, the data provide a practical basis for strategic planning and the implementation of healthcare policies aimed at improving the efficiency and quality of pharmaceutical care in the cardiovascular context.

En representación de MAPEX-SEFH y grupo Cardio de la Sociedad Española de Farmacia Hospitalaria (por orden alfabético de apellidos): Cristina Díez Vallejo; Sara Guijarro Herrera; Sara Ibáñez García; Aránzazu Linares Alarcón; José Antonio Martín Conde; María José Mauriz Montero; Nuria Martínez Casanova; Rebeca Pelegrín Cruz.

All authors contributed to developing the original idea and designing the study. Ramón Morillo-Verdugo was responsible for writing the manuscript. The final version was reviewed and approved for publication by all the authors.