The Hotspotters Project targeted adults with complex care needs characterized by problems across multiple life domains and correspondingly high acute care use. Screening identified 153 eligible patients out of 1,403 screened, and only 30 patients were ultimately enrolled. Lower-than-expected acute care use in the sampled population contributed to the small eligible pool and limited enrollment, which constrained the ability to draw conclusions about intervention effectiveness or perform a cost-effectiveness analysis.
This was a mixed-methods process evaluation structured around the RE-AIM framework (Reach, Effectiveness, Adoption, Implementation, Maintenance). Quantitative data sources included validated questionnaires and routine care records; descriptive statistics were produced using SPSS. Qualitative data were collected through semi-structured interviews with patients and care providers and analyzed using inductive thematic content analysis (Atlas.ti). The mixed-method approach was intended to allow qualitative insights to explain or contextualize quantitative findings.
Quantitative results showed that fewer patients than anticipated met the study inclusion criteria, and the low enrollment (n=30) reduced statistical power to evaluate health outcomes or healthcare costs. Because enrollment and sample size were insufficient, the investigators report that cost-effectiveness could not be assessed. The study documents that the discrepancy between expected and observed acute care use affected the trial reach and generalizability.
Interview data provided contextual explanation for trial outcomes. Care providers valued the structured intake process for offering a comprehensive overview of patients’ multidomain problems. Patients and staff reported that the intake helped surface complex needs and facilitated clearer care planning. Qualitative findings also highlighted perceived burdens associated with the inclusion process and study participation, which may have contributed to lower enrollment and reduced patient motivation to participate.
Despite the complex inclusion procedures and limited trial enrollment, many participating practices adopted elements of the intervention into routine care. The project was associated with improved collaboration with social work services, an operational benefit noted by providers. These signals of local adoption and implementation suggest potential for maintenance of certain intervention components even if the trial could not demonstrate effectiveness or cost savings at scale.
Care providers identified several implementation barriers that limited trial reach and feasibility in the Dutch primary care context. Key recommendations emerging from the process evaluation were:
The authors emphasize that adapting the target population and recruitment approach to the local context can strengthen the rationale for implementation, increase patient motivation to enroll, and enhance the ability to evaluate effectiveness in future interventions.
The study received ethical approval from the Ethics committee of the Leiden University Medical Center (METC-LDD, P21.123). The clinical trial registration identifier is NCT05878054. Participants did not consent to public data sharing; the participants’ data are not available publicly. Funding sources declared include the Netherlands Organization for Health Research and Development (ZonMW) and several regional foundations and insurers.
The Hotspotters process evaluation demonstrates that implementation successes—improved collaboration with social work, positive experiences with a structured intake, and integration of intervention elements into routine care—can coexist with trial limitations when reach and enrollment fall short. The main actionable lessons are to recalibrate patient definitions and recruitment strategies to the local context and to reduce patient and provider burden to improve feasibility of future pragmatic trials targeting complex-care populations. Because the enrolled sample was small, claims about effectiveness and cost-effectiveness remain untested in this dataset.