Evaluation of complex integrated care programmes: the approach in North West London

Abstract

BACKGROUND: Several local attempts to introduce integrated care in the English National Health Service have been tried, with limited success. The Northwest London Integrated Care Pilot attempts to improve the quality of care of the elderly and people with diabetes by providing a novel integration process across primary, secondary and social care organisations. It involves predictive risk modelling, care planning, multidisciplinary management of complex cases and an information technology tool to support information sharing. This paper sets out the evaluation approach adopted to measure its effect. STUDY DESIGN: We present a mixed methods evaluation methodology. It includes a quantitative approach measuring changes in service utilization, costs, clinical outcomes and quality of care using routine primary and secondary data sources. It also contains a qualitative component, involving observations, interviews and focus groups with patients and professionals, to understand participant experiences and to understand the pilot within the national policy context. THEORY AND DISCUSSION: This study considers the complexity of evaluating a large, multi-organisational intervention in a changing healthcare economy. We locate the evaluation within the theory of evaluation of complex interventions. We present the specific challenges faced by evaluating an intervention of this sort, and the responses made to mitigate against them. CONCLUSIONS: We hope this broad, dynamic and responsive evaluation will allow us to clarify the contribution of the pilot, and provide a potential model for evaluation of other similar interventions. Because of the priority given to the integrated agenda by governments internationally, the need to develop and improve strong evaluation methodologies remains strikingly important.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 32 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.