The study was conducted in the catchment area of the Hospital Clinic de Barcelona (HCB), within the Integrated Health District of Barcelona-Esquerra (AISBE, approximately 520 k citizens). The current research adopted a convergent mixed methods design [19-21], integrating quantitative and qualitative data to iteratively refine the hybrid care intervention. Quantitative data were used to characterize the study population, support patient profiling, and test the technology, whereas qualitative data provided contextual and experiential insights to inform the co-design process.

Table 1 describes the main characteristics of the study design for the qualitative component of the research. The information for qualitative analyses relied on three main sources: (1) longitudinal experiential data collected by the nurse case manager during routine interactions with patients and caregivers (including chat-based communication and follow-up contacts), (2) continuous feedback from the different supporting teams described in Table 1 that was compiled and synthesized in the weekly sessions carried out by the core research team across the PDSA cycles, and (3) a structured focus group conducted at the end of PDSA-3 to consolidate the strategy for deployment and sustainable adoption of the hybrid care approach in real-life clinical care. Qualitative data were synthesized iteratively by the research team and integrated with quantitative findings to inform decision-making and refinement of the hybrid care intervention components.

Each PDSA cycle simultaneously addressed the following four components, briefly described in the table: (1) structure of the nurse-led hybrid care intervention, (2) patients’ risk profiling and personalization of care, (3) advanced digital support, and (4) household indoor air quality (IAQ) assessment using low-cost sensors (LCS). Details on integration between outcomes from qualitative analyses and quantitative assessments for the 4 blocks reported in Table 1, within the relevant PDSA cycles, are described in detail in the results section.

The study cohort included patients visited in the context of two distinct clinical programs: (1) a community-based program for patients with chronic obstructive respiratory disorders, and (2) patients with severe asthma [2].

The first step for the enrollment of patients in the community-based program was elaborating the dataset of potential candidates using registry information from the Catalan Health Surveillance System (CHSS) [22], allowing the identification of the corresponding primary care physicians. The primary care physicians’ engagement in the study facilitated initial contact with the patients, exploring their willingness to participate and keeping them within their health care team during the follow-up period. An initial phone visit by the nurse case manager, on behalf of the primary care team, was used to introduce the project and set the first face-to-face visit to the patient’s home.

Inclusion and exclusion criteria for the community-based program were: (1) adults below 85 years, (2) residents in AISBE, and (3) with a diagnosis of chronic obstructive pulmonary disease (COPD), asthma, or bronchiectasis, and showing high multimorbidity burden [23]. Patients with dementia or those who were unable to independently perform activities of daily life were excluded from the study. Also, patients already enrolled in other chronic care programs such as home care support for advanced chronic complex patients, or palliative care, were not included in the study cohort.

Approximately 20% of the cohort was targeted for inclusion of patients with severe asthma [23] recruited from the outpatient Severe Asthma Unit of the Pulmonology Department of Hospital Clinic de Barcelona (HCB) following similar steps. Further details on the inclusion and exclusion criteria were listed in the reported study protocol [17].

As described in Table 1, the core research team (AGL, RGC, ICF, and JRT) implemented the 2-year (2024‐2025) co-design process to gain clinical applicability of the hybrid care intervention. To achieve this, 4 sequential PDSA co-design cycles, each lasting 6 months, were conducted by the research group with active continuous contributions of: (1) the patients and caregivers from the study cohort, (2) four clinical teams of health professionals: 2 primary care centers from AISBE, and 2 teams of respiratory specialists from HCB, one of them providing support to the community-based program (AISBE) and the second running the severe asthma unit at HCB, (3) two external technical support teams, and (4) stakeholders from the Digitalization for the Sustainability of the Healthcare System (DS3 group) pertaining to the single regional public payer in Catalonia (CatSalut), as described in detail in Appendix 1 in Multimedia Appendix 1.

The first PDSA cycle (PDSA-1, January to June 2024) had two main objectives: (1) to address patients’ profiling, and (2) to assess feasibility, beyond clinical research context, of the 4 components indicated in Table 1, in particular the technological tools used in the follow-up of the cohort: digital platform [12], heart rate variability (HRV) monitoring, respiratory oscillometry testing, and household IAQ monitoring with LCS.

PDSA-2, from July to December 2024, addressed the following aspects: (1) completion of patients’ profiling, (2) definition of clinical aspects of the hybrid care intervention with an integrated care approach, (3) elaboration of strategies for patients’ risk assessment and personalization of care, and (4) refinement of different aspects of the technological support.

The third PDSA cycle (PDSA-3, from January to September 2025) evaluated a consolidated version of the hybrid care intervention in the study cohort, introducing some refinements.

The patients’ and caregivers’ experience and outcomes regarding the characteristics of the hybrid care intervention were collected by the nurse case manager during the follow-up of the study cohort. The outputs from the co-design cycles were collectively evaluated by a dedicated focus group session with 14 leading professionals, mixing clinical staff (n=9), IT developers (n=1), data scientists (n=1), and managers (n=3) to consolidate findings and reach a final consensus on the structure of the integrated hybrid care intervention and the implementation strategy, aiming at fostering sustainable adoption in routine practice.

Contextual data were collected to characterize the study population across multiple dimensions, including sociodemographic, clinical, and functional characteristics, unhealthy lifestyle habits, multimorbidity burden [23,24], disabilities, social frailty, use of health care resources, and digital literacy. Adjusted morbidity groups (AMG) scoring reflects both the number and complexity of morbidity. The index is extensively used in Spain and other European regions, as reported in detail in [25]. Such comprehensive assessment of the study cohort aimed at identifying the candidates’ profiles for a personalized hybrid care intervention.

All data for patient profiling were gathered from 3 sources. At baseline, the nurse case manager instructed the patients, administered the different questionnaires, and performed lung function testing through 2 home visits of 1 hour duration each. Clinical information and therapies were gathered from the hospital care electronic records, with access to primary care information. A second, midterm assessment of the cohort was carried out, and a third home visit was planned at the end of the follow-up, as reported in [17]. We used the CHSS [22] to extract historical data on resource use, multimorbidity [23,24], and sociodemographic data. The CHSS centralizes data from primary and specialized care across the entire health care system.

Clinical characterization: at baseline, the patients were clinically characterized using a comprehensive set of variables, including sociodemographic data, primary respiratory diagnosis, comorbidities, lung function using forced spirometry and oscillometry, and health care resource use in the 12 months preceding enrollment (primary care visits, emergency department visits, hospital admissions, and total health care expenditure). Functional capacity was assessed using standardized patient-reported outcomes evaluating physical and mental health as well as respiratory symptoms and overall health status [26-34]. Additional variables included lifestyle risk factors (smoking, alcohol, physical activity, passive smoking), as well as housing characteristics and living conditions described in [17].

Digital literacy profiling: digital literacy was assessed with a purpose-designed questionnaire comprising different items evaluating knowledge, usage, and confidence in digital skills relevant to the hybrid care model. Results were compared with a parallel classification into 5 levels of digital skills (none, low, medium with no improvement capacity, medium with improvement capacity, high), based on the engagement and support needs observed by the nurse.

Respiratory diagnoses profiling: to ensure diagnostic accuracy, the single respiratory diagnosis validated by the research team examining the patients’ electronic health care records was compared with the respiratory diagnoses obtained from the CHSS, including diagnostic information across all health care tiers.

Continuous variables are reported as mean (SD) or median (IQR, defined as the 25th and 75th percentiles), as appropriate; categorical variables are presented as n (%). Between-group comparisons for continuous data used the student’s t test when distributional assumptions were met and the Wilcoxon rank-sum test otherwise. Categorical data were compared with the Fisher exact test. All tests were 2-sided, and P<.05 was considered statistically significant. All the analyses have been conducted using R (version 4.1.1; R Core Team) .

Ethical approval for the process evaluation was granted by the Ethical Committee for Human Research at the HCB on June 29, 2023 (HCB/2023/0126) and registered at ClinicalTrials.gov (NCT06421402). The patients participating in the research were required to provide and sign a written informed consent indicating the purpose of the study, the nature, use, and management of their data. Participants did not receive any compensation for their participation.

The Results section is organized into 2 main subsections. Section A presents the characteristics of the study cohort involved in the co-design process. Section B summarizes the key findings and lessons learned from the first 3 PDSA cycles, which informed the iterative development of the hybrid care intervention for routine clinical practice, and describes the refinements identified and incorporated during the fourth PDSA cycle to optimize the final intervention.

During the recruitment period, 205 patients were enrolled in the study: 152 (74%) from the community program (AISBE) and 53 (26%) from the severe asthma group (Figure 1). Of the 205 patients enrolled, 22 (10.7%) were excluded from analysis (19 withdrawals, 3 deaths), resulting in an analytic cohort of 183 participants.

Table 2 summarizes baseline clinical characteristics for (1) the overall analytic cohort (n=183), (2) community-managed patients in the AISBE program (n=135; 74%), and (3) patients followed at HCB’s Severe Asthma Unit (n=48; 26%).

Compared with patients from the severe asthma unit, the community-managed patients with chronic obstructive respiratory diseases were older (72.4 vs 57.4 y; P<.001), more often male (51.9% [70/183] vs 18.8% [9/183]; P<.001), and had a higher comorbidity burden (adjusted morbidity groups [AMG] 83.8 vs 63.7; P=.001). Lung function indicated greater airflow limitation in AISBE (forced expiratory volume in 1 second (FEV₁) % predicted 57.3% [135/183] vs 82.6% [48/183]; P<.001; FEV₁/FVC 48.5 vs 83.4; P=.01). No differences in the hospitalization rate during the previous year were observed between the 2 groups (all-cause hospitalization 35.8% [48/183] vs 33.3% [16/183], P=.86; unplanned 27.6% [37/183] vs 22.9% [11/183], P=.57). Health care use over the prior year was broadly similar between groups. We found no significant differences in health care expenditure per patient between the AISBE group (€5133 [US $5544]) and the severe asthma group (€8101 [US $8749]) (P=.11). US $ values were calculated using the average 2024 exchange rate of €1=US $1.08.

The World Health Organization Disability Assessment Schedule (12-item WHODAS questionnaire) [34] reflected graded levels of disability in 6 domain-specific scores. The mean WHODAS score was 1.88 (SD 0.77, IQR 1.5‐2.4), consistent with mild disability, as expected due to the entry criteria. Within the specific disability domains, a substantial interindividual dispersion was observed. Furthermore, a significant number of patients reported socioeconomic barriers, such as limited financial resources (<16,000€/y [US $17,300/y]; 74/160, 46.3%; 23/183, 12.5% missing), or living alone (45/183, 24.6%), highlighting potential sources of inequity that could influence patients’ capacity for effective self-management. (Appendices 2 and 3 in Multimedia Appendix 1).

Textboxes 1-4 summarize the clinical and operational characteristics of the hybrid care intervention at the end of the third PDSA cycle, detailing how the model articulates the roles of the main actors and the management procedures. These textboxes illustrates how the model is operationalized across care settings, distinguishing between the community-based AISBE program and the specialized Severe Asthma Unit at HCB.

Personalization of the hybrid care intervention (Table 3) was approached from three complementary perspectives: (1) clinical profiling of candidates, (2) contextual factors modulating care (eg, socio-economic, disability, living conditions); and (3) assessment of digital literacy.

Candidate profiling: baseline and midterm assessments, together with longitudinal monitoring and the contacts between the nurse case manager and participants over the first 18 months of follow-up, enabled the identification of six clinical groups: two within the severe asthma program (A1-A2) and 4 within the community-based AISBE program (C1-C4). These groups, and their corresponding hybrid care implications, were derived through an iterative profiling process conducted across the co-design cycles, integrating baseline characteristics, midterm assessment, and real-world observations over the follow-up period. The classification includes key dimensions: clinical severity of the respiratory disease, multimorbidity burden, functional status, socio-economic factors, and digital abilities, to inform the personalization of the hybrid care intervention. The characteristics of each group and their corresponding implications for hybrid care delivery are detailed in Table 3. Further information is displayed in Appendices 2‐3 in Multimedia Appendix 1.

These groups are intended to guide personalization of hybrid care rather than represent fixed clinical categories. Patient profiling and digital modality assignment represent complementary dimensions of personalization: the former defines care needs, while the latter reflects the level of digital support used to deliver the intervention.

Digital literacy: patients’ digital skills played a relevant role in the personalization of the hybrid care intervention. The level of digital skills directly influenced the type and intensity of digital support provided, ranging from phone-based management for patients with low or no digital skills or unwilling to use digital tools, to app-based communication and monitoring in those with higher or trainable digital capabilities. This allocation was dynamically adjusted by the nurse case manager according to patient progression and engagement. Briefly, 109 patients (59.6%) showed either high digital skills (n=51; 27.9%) or medium level scoring with significant improvements after training (n=58; 31.7%). The remaining patients presented significant limitations in the use of digital tools ranging from: (1) complete inability (n=16; 8.7%), or (2) poor digital skills (n=37; 20.2%), to (3) medium-level digital skills with no improvement after training (n=21; 11.5%) (Appendix 5 in Multimedia Appendix 1). Overall, 41 (22.4%) participants did not use the app, either due to clinical recommendation or personal preference. Improvements in digital literacy observed during follow-up appear to result from a combination of tailored training provided by the nurse case manager and progressive exposure to digital tools within the intervention, rather than from a single predefined training phase. At the end of PDSA-3, the distribution of the three different digital support modalities (M) was as follows: M1- phone-based follow-up (28% of the cohort), M2- app-based management with chat and patient-reported outcomes measures (PROMS; 21% of the cohort), and M3- app-based management including chat, PROMS, and HRV sensor (51% of the cohort). Allocation to each modality was personalized by the nurse case manager based on clinical characteristics, digital literacy, and patient engagement capacity and willingness.

Key lessons learned for scale-up of the hybrid-care intervention. The co-design process demonstrated applicability and stakeholders’ acceptance of the overall approach, contributing to the identification of some barriers partly limiting the digitally enabled management of the cohort. These findings provided emergent solutions for those limitations, as reported in Table 4. A major lesson learned was that further progress in the operationalization of the intervention must be made through its stepwise deployment in a real-world setting. Because of the direct relationships between patients and specialized professionals, the Severe Asthma program was selected as a priority scenario for deployment during the last PDSA cycle (PDSA-4).

Furthermore, several well-known consolidated frameworks guiding implementation processes were evaluated, aiming at fostering clinical adoption of the hybrid-care intervention. A pragmatic approach to the Integrated Theory-based Framework for Intervention Tailoring Strategies (ItFits-Toolkit) [41,42] was proposed to provide guidance. The selection of this implementation strategy was driven by two strategic priorities: (1) high potential for pragmatic application in the real-world setting covering the entire implementation cycle, and (2) plans for future use of these implementation tools at the regional level. This prospective framework is intended to bridge the gap between current operational results and systemic integration.