A modified Delphi survey to build expert consensus on the structure and content of an enhanced care pathway for cognitive changes after stroke in the UK

Abstract

BackgroundEnhancing long-term support for post-stroke cognitive impairment is a top research priority. Addressing current gaps in UK post-stroke cognitive care provision requires a pragmatic and scalable intervention that can be integrated within the existing stroke care pathway. This study aimed to develop consensus on an initial set of core features for a UK-based monitoring and psychoeducational intervention for cognitive changes after stroke.MethodsAn expert panel of UK healthcare professionals and researchers participated in an online modified Delphi survey. Candidate intervention features were identified from clinical guidelines, existing literature, research team/collaborator expertise, and PPI group lived experience. Survey participants indicated whether they agreed/disagreed/had no opinion about including each candidate feature in the intervention and free-text responses were invited. We analysed responses for consensus (≥ 75% agreement) using descriptive statistics, with items not reaching consensus carried into subsequent rounds. Template analysis was used to identify similarities/differences in viewpoints for items that did not reach consensus.ResultsThree survey rounds were completed by 36, 29 and 26 participants, respectively. Participants agreed reviews should include a stroke-specific cognitive screen (97% agree) and assessment of other psychological changes (low mood, anxiety, fatigue: 94%, 90%, 89% agree, respectively). They agreed stroke survivors should be offered at least one review, regardless of their cognitive profile in hospital. They agreed on the importance of various cognition-focused psychoeducation topics, and formal (100% agree) and informal (79% agree) training for those conducting reviews. Consensus was not reached on the review mode (in person/remote options: 67% agree), offering reviews one-year post-discharge to patients without acute cognitive impairments (68% disagree), or including a dementia screen (63% disagree) and/or neuropsychological assessment battery (58% disagree). However, there were similarities in participant viewpoints. For example, participants emphasised the importance of onwards referral where clinically indicated. ConclusionsThe UK-based post-stroke monitoring and psychoeducation intervention was originally conceptualised as a cognitive care pathway, but expert participants agreed on the importance of simultaneously addressing related psychological changes (e.g. low mood, anxiety). There was clear consensus on a minimum set of intervention features. Recommendations outlined here may usefully inform local service improvements.