Explaining Outcomes in Major System Change: A qualitative study of implementing centralised acute stroke services in two large metropolitan regions in England Naomi Fulop Professor of Health Care Organisation and Management, University College London Academy Health Annual Research Meeting, 26 th -28 th June 2016, Boston USA
Innovations in major system reconfiguration in England: a study of the effectiveness, acceptability and processes of implementation of two models of stroke care Naomi Fulop, Steve Morris, Angus Ramsay, Rachael Hunter, Simon Turner Tony Rudd, Charles Wolfe, Christopher McKevitt Pippa Tyrrell, Ruth Boaden, Catherine Perry Funded by the NIHR Health Services & Delivery Research programme (HS&DR) (Project number 10/1009/09). The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the HS&DR, NIHR, NHS or the Department of Health
Background • Major system change: reorganisation of delivery system to e.g. increase provision of evidence-based care and clinical outcomes (not just individual level) • e.g. service centralisation – some evidence on outcomes, but very little on how implemented • To open up ‘black box’ need to: (i) distinguish between implementation outcomes (e.g. adoption of, fidelity to the intervention) and intervention outcomes (Proctor et al, 2011) (ii) study relationships factors that influence implementation (e.g. nature of intervention and approaches to implementation) and intervention outcomes • i.e. combine ‘what works?’ with ‘how and why?’
Framework for major system change Drivers for change, governance & leadership of C1 decision-making [Best 2012; CFIR] Decision to change C2 Decision on which model to implement Context, approaches to facilitation [PARIHS, CFIR, KTA] C3 Intervention development/ Implementation selection approaches [PARIHS, CFIR, KTA] C4 Implementation outcomes C5 Adoption, spread, fidelity Intervention outcomes [Proctor 2011, Mendel 2008] Evidence based Clinical care outcomes Patient Cost experience effectiveness Including care provision, health outcomes, patient and carer experience [KTA, Mendel 2008] Fulop et al, 2016
Background to major system change in stroke National Stroke Strategy (2007) set out case for change: • 3rd biggest cause of death in UK • Outcomes for stroke in UK compared poorly with those internationally • Services not organised to enable evidence-based clinical practices to be provided • Greater Manchester and London led way in reconfiguring services to address these concerns
The changes After Before London Greater Manchester ‘A’ Suspected stroke Suspected stroke Suspected stroke >4 hrs ≤4 hrs 8 HASUs (all 24/7) Stroke unit/ward 3 HASUs Greater Manchester (x12) 11 DSCs (1x24/7, 2 in-hours) London (x30) 24 SUs Community Community Community rehabilitation services rehabilitation services rehabilitation services
What we know: why different models chosen? • Both systems aimed to increase provision of evidence- based clinical interventions; little evidence on which model best to achieve this • Both systems originally proposed to implement the more radical, 24 hour model • More radical transformation in London was enabled by ‘top down’ leaders’ political authority challenging the existing service delivery context and managing (some) stakeholders’ resistance to change
What we know: dealing with resistance London : ‘holding the line’ “Stroke was their [clinician representatives’] life, and they wanted to get the best for stroke […] but actually what got it through was being straight with them, trying to explain it to them, but in the end holding the line.” (Commissioner and Project Board Member) Greater Manchester: ‘consensus’ “The minute it felt like unanimity was being compromised on that clinical discussion on the 24 versus the 4 hour pathway, I think we were always going to be minded then to tilt towards holding unanimity and taking what might be a small step, but still the right step.” (Commissioner)
What we know: provision of clinical interventions • Looked at provision of evidence based clinical interventions: access to stroke unit, scans, aspirin, assessments (physio, swallow, nutrition) Care got better in all areas over time… London: more likely than elsewhere to provide interventions BUT Manchester: no different from elsewhere WHY? Fidelity to the model • HASUs significantly more likely to provide interventions • But different proportions of pts treated in HASU – 93% in London, 39% in Manchester Ramsay et al, 2015
What we know: clinical outcomes • Mortality in London reduced significantly more than in the rest of England (96 additional lives saved per year) – no equivalent effect in Greater Manchester • Length of stay in hospital reduced significantly more in Greater Manchester and London than in the rest of England (by 2 days in GM, 1.4 days in London)
What we need to know Both regions : National Stroke Strategy (2007); Greater Variation in provision of evidence based care C1 Decision to Manchester Network-led: change ‘Consensus’ [Turner 2016] C2 Decision 1x 24/7 HASU; Influence of Led by regional on which model 2x IH HASUs;10x DSCs model selected? authority: to implement 4h window for HASU ‘Holding the line’ [Turner 2016] C3 Facilitation? 8x 24/7 HASUs; 24x SUs Implementation Management? All eligible for HASU approaches C4 Facilitation? HASUs provide 39% treated Implementation Management? evidence based in HASU outcomes care; DSCs vary Influence of [Ramsay 2015] model selected? C5 Intervention outcomes HASUs provide Evidence based evidence based No more likely than care care elsewhere [Ramsay 2015] 93% treated More likely than in HASU London LoS= ▼ elsewhere Mortality = NSD Clinical LoS = ▼ outcomes Mortality = ▼ [Morris 2014] Fulop et al, 2016
Methods • Qualitative fieldwork at 2 levels: governance & service • Analysis of 653 documents • 125 semi-structured interviews • 4 stages of analysis: (i) narrative summaries & timelines of changes (Turner et al, 2016) (ii) service-level narrative summaries using constant comparative method (iii) Used i) and ii) to identify implementation tasks and how undertaken (iv) Applied our framework to cross-case analysis Fulop et al, 2016
Findings: service model selected Greater ‘I don’t understand who’s supposed to be going here and who’s supposed Manchester to be going there, and if I don’t, I bet other people don’t know’ (GM stroke physician) Complex ‘We need to have a definite time of onset […] and if that time exceeds Decision on the four hours then we won't be taking them to Simple which model Selective the Hyper Acute Stroke ‘we cannot give crews to implement Unit’ fragmented messages, you (GM ambulance representative) can’t say that you can get this type of care between 8 and 5 Monday to Friday Inclusive but not on the second Wednesday of the month because there’s a meeting, ‘it’s not about “can you do it crews don’t work that way’ right?” its about, “how (London ambulance consistently can you do it, and representative) will everybody get that care?”’ (London regional health authority London representative) Fulop et al, 2016
Findings: implementation approaches ‘If you phase it, it does create a Greater degree of confusion […] you change it, and then you change it, Manchester and then you may change it again ’ (GM ambulance representative) ‘the one thing that we really did push for was a ‘it probably wasn’t a “go live” date, not a “go written principle but Pilot, then live” date in one area actually what we do is phased and another in other hold consensus and try areas ’ (London ambulance No and deliver this through representative) accreditation unanimity’ (GM or financial commissioner representative) ‘Big bang’ Implementation levers approaches Platform to share Accreditation & learning financial levers ‘Having to meet all these standards […] Hands-on ‘so much learning came out of has provided a stick for facilitation it through this […] informing hospital management how the model should look […] to invest in stroke it was all very emergent’ (GM services’ (London stroke network representative) ‘We were there to remind them of what physician) they had signed up to […] of the quality standards that they needed to meet, but London always in a supportive manner’ (London network representative) Fulop et al, 2016
Recommend
More recommend
