The person, interactions and environment programme to improve care of people with dementia in hospital: a multisite study

Background: Improving care of people with dementia on acute hospital wards is a policy priority. Person-centred care is a marker of care quality; delivering such care is a goal for service improvement.

Objectives: PIE (Person: Interaction; Environment) comprises an observation tool and systematic approach to implement and embed a person-centred approach in routine care for hospitalised patients with dementia. The study aims were to: evaluate PIE as a method to improve the care of older people with dementia on acute hospital wards; and develop insight into what person-centred care might look like in practice in this setting.

Methods: We performed a longitudinal comparative case study design in ten purposively selected wards in five Trusts in three English regions; alongside an embedded process evaluation. Data was collected from multiple sources: staff, patients, relatives, organisational aggregate information and documents. Mixed methods were employed: ethnographic observation; interviews and questionnaires; patient case studies (patient observation and conversations ‘in the moment’, interviews with relatives and case records), patient and ward aggregate data. Data was synthesised to create individual case studies of PIE implementation and outcomes in context of ward structure, organisation, patient profile and process of care delivery. Cross case comparison facilitated a descriptive and explanatory account of PIE implementation in context, the pattern of variation, what shaped it and the consequences flowing from it. Quantitative data was analysed using simple descriptive statistics. Qualitative data analysis employed grounded theory methods.

Results: The study furthered understanding of dimensions of care quality for older people with dementia on acute hospital wards and the environmental, organisational and cultural factors that shaped delivery. Only two wards fully implemented PIE, sustaining and embedding change over 18 months. The remaining wards either did not install PIE (‘non-implementers’); or were ‘partial implementers’. The interaction between micro-level contextual factors (aspects of leadership (drivers, facilitators, team, networks), fit with strategic initiatives and salience with valued goals) and miso and macro level organisational factors, were the main barriers to PIE adoption. Where implemented, evidence suggests that the programme directly affected improvement in ward practice with positive impact on the experience of patients and caregivers, although the heterogeneity of need and severity of impairment meant that some of the more visible changes did not affect everyone equally.

Limitations: Although PIE has potential to improve the care of people with dementia when implemented, findings are indicative only: data on clinical outcomes was not systematically collected; and PIE was not adopted on most study wards.

Research implications: Further research is required to identify more precisely the skill-mix and resources necessary to provide person-focused care to hospitalised people with dementia, across the spectrum of need, including those with moderate and severe impairment. Implementing innovations to change practices in complex organisations requires more in-depth understanding of contextual factors that impact the capacity of organisations to absorb and embed new practices.