Abstract Background Lewy body dementia, comprising both dementia with Lewy bodies and Parkinson’s disease dementia, is the second commonest cause of neurodegenerative dementia. Existing evidence suggests that it is underdiagnosed and without a consistent approach to management. Objectives To improve the diagnosis and management of Lewy body dementia by (1) understanding current diagnostic practice for dementia with Lewy bodies and Parkinson’s disease dementia; (2) identifying barriers to and facilitators of diagnosis and management; (3) developing evidence-based assessment toolkits to improve diagnosis of dementia with Lewy bodies and Parkinson’s disease dementia; (4) producing a management toolkit to facilitate management; and (5) undertaking a pilot cluster randomised clinical trial. Design Work package 1 assessed clinical diagnostic rates from case notes for dementia with Lewy bodies and Parkinson’s disease dementia before and after (work package 1 repeated) introduction of an assessment toolkit. In work package 2, we developed a management toolkit for Lewy body dementia. In work package 3, we developed assessment toolkits for dementia with Lewy bodies and Parkinson’s disease dementia and piloted these and the management toolkit in a clinical service. In work package 4, we undertook a pilot study of 23 services in nine NHS trusts that were cluster randomised to receiving and using the management toolkit or standard care. Work package 5 comprised a series of qualitative studies, examining barriers to and facilitators of diagnosis and management. Setting Secondary care memory assessment and movement disorder services in England. Interventions Assessment toolkits for Lewy body dementia consisted of questions for diagnostic symptoms, and management toolkits comprised 161 guidance statements grouped under five symptom domains. Review methods The systematic reviews of pharmacological and non-pharmacological management were based on published literature, with meta-analysis when possible, following a search of several electronic databases and the grey literature using terms related to Lewy body dementia, without restriction on time or language. Participants Participants aged ≥ 50 years diagnosed with dementia with Lewy bodies or Parkinson’s disease dementia and, for work package 1 and work package 1 repeated, non-dementia with Lewy bodies and non-Parkinson’s disease dementia controls. The qualitative studies included people with Lewy body dementia, carers and professionals. Main outcome measures For work packages 1 and 1 repeated, diagnostic rates for dementia with Lewy bodies and Parkinson’s disease dementia as a proportion of all dementia or Parkinson’s disease. For work packages 2 and 3, the production of diagnostic and management toolkits. For work package 4, feasibility of undertaking a cluster randomised trial of the toolkits, measured by number of participants recruited and use of the toolkits, assessed qualitatively. Results Work package 1 – 4.6% of dementia cases in secondary care received a dementia with Lewy bodies diagnosis (with significant differences in diagnostic rates between services) and 9.7% of those with Parkinson’s disease had a diagnosis of Parkinson’s disease dementia. There was evidence of delays in diagnosis for both dementia with Lewy bodies and Parkinson’s disease dementia compared with control patients, and the costs of dementia with Lewy bodies and Parkinson’s disease dementia were also greater than those for matched controls (p < 0.01 for both). Work package 2 – we produced 252 statements regarding Lewy body dementia management and, following a Delphi process, 161 statements were included in a management toolkit. Work package 3 – piloting indicated that separate assessment toolkits for use in memory clinic and movement disorder services were preferred, but a single toolkit for Lewy body dementia management was suitable. Work package 4 – we were able to recruit Lewy body dementia patients to target and recruited 131 patients within 6 months (target n = 120), of whom > 80% were retained in the study at 6 months. Work package 5 – barriers to diagnosis and management of Lewy body dementia were complex. Managing Lewy body dementia often requires input from a range of specialties and, therefore, care pathways may be fragmented. Positive attitudes to diagnosing Lewy body dementia, working with a team with expertise in Lewy body dementia and opportunities for cross-specialty discussion of patients with complex needs facilitated diagnosis and management. The toolkits were generally well received, particularly the management toolkit. Implementation, however, varied, reflecting differences in attitudes, skills, time and local leadership. Work package 1 repeated – following introduction of the assessment toolkit, we found that 9.7% of dementia cases had dementia with Lewy bodies (a significant increase from baseline; p = 0.0019), but Parkinson’s disease dementia rates were similar (8.2%) to baseline. Limitations We included only two geographical regions and evidence informing the management toolkit was limited. Work package 4 was a pilot study and, therefore, we did not set out to assess the extent to which use of the management toolkit altered outcomes at the individual patient level. We noted implementation of the toolkits was variable. The increase in diagnostic rates in dementia with Lewy bodies following introduction of the assessment toolkits cannot be necessarily causally attributed to them. Conclusions Dementia with Lewy bodies and Parkinson’s disease dementia were diagnosed in secondary care NHS services, with a lower frequency (around half) than that expected from known prevalence rates. The introduction of assessment toolkits for dementia with Lewy bodies and Parkinson’s disease dementia was associated with increased diagnostic rates of dementia with Lewy bodies, but not Parkinson’s disease dementia. Qualitative studies indicated inherent complexities of the disease itself, with treatment requiring input from different specialties and the potential for fragmented services, a workforce with variable training and confidence in Lewy body dementia, and negative attitudes towards diagnosis. The cluster randomised pilot trial demonstrated that patients could be successfully recruited, and provided preliminary evidence that the toolkits could be implemented in clinical services. Future work The evidence base informing the management of Lewy body dementia is limited, especially for non-pharmacological interventions. More well-designed randomised controlled trials for both cognitive and non-cognitive symptoms are needed. Trial registration Current Controlled Trials ISRCTN11083027. Funding This project was funded by the National Institute for Health Research (NIHR) Programme Grants for Applied Research programme and will be published in full in Programme Grants for Applied Research; Vol. 9, No. 7. See the NIHR Journals Library website for further project information.