When Randomised Controlled Trials are not possible-design issues from a national evaluation
|
Shipman, Cathy, King's College, London, England, Addington-Hall, Julia, King's College, London, England, Richardson, Alison, King's College, London, England, Burt, Jenni, King's College, London, England, Ream, Emma, King's College, London, England |
Background: In April 2001 the Department of Health in England launched a national educational and support programme for community nurses in palliative care. The aim was to improve palliative care provision and enable more patients to die at home should they wish. The programmes are provided through 34 Cancer Networks across England. Each network varies considerably.
Objective: To identify the range of programmes being provided and evaluate their impact on community nurses, family doctors and patients.
Methods: We are using both formative (process) and summative (outcome) approaches. Nationally we have categorised the wide range of educational models being used and tracked milestones, obstacles and achievements. Because of limited resources, we used a before and after design on 8 randomly selected networks. In phase one, postal questionnaires have been sent to 5000 different respondents. We are also undertaking a qualitative study using focus groups and telephone interviews.
Results: We have completed phase 1 with questionnaire responses from 56.4% community nurses, 60.1% of family doctors and 43.1% of bereaved carers. Fewer nurses are confident with agitation and confusion (41%) and breathlessness (48%) than assessing pain (73%). Focus groups and telephone interviews have produced exploratory data on the experience of attending and providing the education.
Discussion: Before and after designs incorporating mixed methods produce exploratory and generalisable results. There are however many learning experiences from research within the community.
|