D-PACT Findings: Qualitative Analysis of Dementia Support

This post explains how to apply AI-enabled qualitative research to post-diagnostic dementia support, using the D-PACT evaluation as a worked example. The primary keyword is qualitative analysis of dementia support. Researchers and primary care teams face high-volume qualitative data: realist interviews, case notes, observations and reflections collected in the D-PACT programme. According to the D-PACT research programme, published on LSE Research Online in August 2026, mixed-methods and realist qualitative approaches identified four impact areas for dementia support workers: living well, proactive care, reactive care and care transitions. This article shows concrete coding, synthesis, and reporting steps that speed insight generation while preserving methodological transparency.

Key Takeaways

According to the study "Development and evaluation of personalised post-diagnostic dementia support based in primary care: the D-PACT research programme" (LSE Research Online), embedding dementia support workers in primary care delivered measurable qualitative benefits in 2026; see the original report for full methods and data.

In Phase 2, the D-PACT research programme recruited 126 people with dementia and 121 carers from 13 practices, reaching 70% (126/180) of the recruitment target in 2026, according to the LSE report deposited on 25 August 2026.
Retention in the D-PACT intervention remained high at 79% in Phase 2, while only 22% (28/126) of participants completed one quantitative follow-up measure at 9–12 months, according to the D-PACT research programme.
The D-PACT economic analysis identified 23 cases where early intervention by dementia support workers likely prevented high-cost care escalation, and estimated a mean annual cost per dyad of £1, 222.48 at a caseload of 53 dyads, according to the LSE publication in 2026.
The D-PACT intervention was described as delivering "personalised, proactive and integrated dementia care, " a phrase used by the D-PACT research programme to characterise the model.

What happened: D-PACT study design and data

The D-PACT research programme used a mixed-methods realist approach to develop and evaluate a primary-care-based dementia support worker intervention.

According to the LSE publication, Phase 1 tested the intervention in 10 GP practices and met 51% (56/110) of an initial recruitment target during feasibility work in early stages, with COVID-19 forcing remote delivery changes in 2020 and later stages.

According to the D-PACT research programme, Phase 2 combined longitudinal realist interviews, dementia support worker case notes, observations, and quantitative measures collected at three time points to answer what worked, for whom and in what circumstances.

Findings snapshot

Date / Phase	Metric	Value	Implication
Phase 1 (feasibility)	Recruitment achieved	56 of 110 target (51%)	Full cluster RCT not feasible within funding, moved to mixed-methods
Phase 2 (2026 report)	Participants recruited	126 people with dementia; 121 carers (13 practices)	Reached 70% of target (126/180) for people with dementia
Follow-up participation	Response to Engagement & Independence questionnaire at 9–12 months	28 of 126 participants (22%)	Quantitative data completeness was low, limiting statistical conclusions
Economic finding	Estimated mean annual cost per dyad	£1, 222.48 at caseload of 53 dyads	Model suggests community care use increased while secondary care costs reduced
Service impact	Cases where escalation likely prevented	23 cases identified	Early intervention may reduce high-cost care events

Implications for primary care researchers and service designers

Embedding dementia support workers in primary care can improve social engagement, medication management and care transitions, according to the D-PACT research programme's qualitative longitudinal analysis reported in August 2026.

Design implication: Plan for low quantitative follow-up rates (the D-PACT study saw 22% completion on a key measure at 9–12 months) and rely on richer qualitative longitudinal interviews to assess lived experience.
Operational implication: The D-PACT publication links high retention (79%) to supervision and peer support for dementia support workers, implying training and ongoing reflective practice are critical.
Economic implication: The D-PACT realist-informed analysis identified 23 cases where escalation was likely prevented, suggesting early targeted support may shift costs toward primary and community services.

How Evidano helps with AI-enabled qualitative analysis of dementia support

Evidano definition and fit

Evidano is an AI-powered qualitative data analysis platform that helps researchers analyze interviews, open-ended surveys, and documents.

Evidano ingests interview transcripts, case notes and observation logs and applies thematic, frequency and cross-segment analyses to speed realist and longitudinal synthesis while preserving audit trails.

For example, teams following the D-PACT approach can upload interview transcripts and case notes to Evidano, run codebook development, and produce comparison matrices by site, time point, or carer status; see Evidano features at Evidano features.

Problem: high-volume, low-completion quantitative data

Answer: Use theme-first synthesis to centre lived experience when numeric follow-up is sparse.

Evidano automates extraction of recurring themes and links quotes to cases and time points, which is useful when, as the D-PACT research programme reported in 2026, only 22% of participants completed a key quantitative follow-up measure.

Problem: preserving realist explanatory mechanisms across sources

Answer: Use cross-source linking and memoing to map contexts, mechanisms and outcomes.

Evidano supports hierarchical codes and case-level linking so researchers can trace how dementia support worker actions produce outcomes across living well, proactive care, reactive care and transitions identified by the D-PACT team.

Problem: transcription and multilingual notes

Answer: Use integrated transcription and custom dictionaries to maintain accuracy.

Evidano provides transcription and translation tools with custom dictionaries and PII redaction to ensure consistent coding of clinical terms, names and local service labels; see Evidano speech-to-text.

FAQ: qualitative analysis of dementia support

How can qualitative data show impact when quantitative outcomes are incomplete?

Answer: Qualitative longitudinal data can document proximal outcomes and mechanisms even when numeric follow-up is low.

The D-PACT research programme used realist interviews, case notes and observations to identify four impact areas despite low quantitative completion, according to the LSE report deposited on 25 August 2026.

Researchers should triangulate interviews, practitioner reflections and records to build causal explanations rather than relying solely on underpowered quantitative comparisons.

What minimal sample and data types are needed for realist qualitative synthesis in dementia studies?

Answer: A realist synthesis needs purposive sampling across contexts, longitudinal interviews, and rich practitioner documentation.

The D-PACT evaluation combined realist interviews at two time points, dementia support worker case notes, observations and medical records across 13 practices, which provided the contextual variation needed to infer what worked and why in 2026.

Can AI tools bias thematic coding in sensitive health research?

Answer: AI tools can reproduce biases unless models are tuned and human oversight is retained.

Evidano uses human-in-the-loop workflows where machine suggestions are reviewed by researchers, an approach that aligns with the D-PACT programme's emphasis on expert-by-experience input and peer researcher involvement reported in the LSE publication.

What cost metrics should teams collect alongside qualitative work?

Answer: Collect resource use, escalation events, and caseload-driven per-dyad costs.

The D-PACT economic analysis reported a mean annual cost per dyad of £1, 222.48 at a caseload of 53 dyads and identified 23 prevented escalation cases, showing the value of linking qualitative accounts to resource-use data.

Conclusion & Next Steps

The D-PACT research programme demonstrates that high-value insight about dementia support can come from rigorous, AI-augmented qualitative synthesis when quantitative follow-up is limited, according to the LSE report deposited on 25 August 2026.

Teams designing or evaluating primary-care-based dementia support should plan for intensive qualitative collection, supervision mechanisms for support workers, and economic linkage to identify prevented escalations.

If you want to accelerate thematic synthesis, maintain traceability, and link qualitative findings to costs and case notes, explore how Evidano supports these workflows and tooling.

Try Evidano for free: Try Evidano for free.

Topics

qualitative analysis of dementia support
AI qualitative research dementia
dementia support evaluation

Keep reading

Browse all articles