Primary Care practices
Interested general practices in the North Thames DeNDRoN area will be identified in collaboration with the local Primary Care Research Networks: the Primary Care Research Network-Greater London (PCRN-GL) and the Primary Care Research Network-East of England (PCRN-EoE). Practices will be contacted by the Trial research team, by letter and awareness-raising through general practice educational meetings and by regular newsletters.
The inclusion criteria for practices in this study will be: 1) routine data collection from clinical encounters on electronic medical records and; 2) team commitment to participate in educational workshops held in the practice. (All staff working in the practice will be eligible to participate in the study.)
Practices that do not routinely capture clinical data in electronic records will be excluded.
Patients with dementia and their carers
The inclusion criteria for people with dementia in this study will be a diagnosis or suspected diagnosis of dementia of any type, with no lower age limit.
Exclusion criteria will be: 1) people who cannot speak English and for whom an interpreter cannot be located; 2) if the patient or carer are involved in concurrent research; 3) if the key professional feels that an approach to the person with dementia or their carer would be inappropriate where, for example, the dementia is very severe or that an approach may increase distress and; 4) any other important reason that the key professional may have for why the person with dementia or their carer should not be contacted.
Every effort will be made to include those who meet the inclusion criteria but may not adequately understand verbal explanations or written information given in English. Where applicable, an independent and qualified translator/interpreter will be sought and invited to assist.
We derived our primary outcome measure and the effect size from discussions with practitioners in the feasibility phase of the trial. The consensus was that the clinical tasks involved in providing good quality care required at least two encounters per year, and that the educational intervention would promote this effectively in a majority of those in the intervention arm. Our hypothesis is, therefore, that in the intervention arm, the proportion of patients receiving two dementia-specific management reviews per year will increase between groups of patients by 50%, i.e. 20% (control) versus 70% (intervention), after the introduction of the educational intervention. Data relating to dementia-specific patient reviews and consultations relevant to dementia management will be extracted from the practice records.
Concordance with guidelines.
We will transcribe and scrutinise manual and electronic records for the recording of actions considered to be best practice in the diagnosis and management of dementia in primary care by using the NICE/SCIE dementia guidelines . Appendix 1 shows the components of good practice that will be captured. The index consultation is the one where the suspicion of possible dementia is first recorded (Appendix 2).
Measurement of unmet needs in patients and carers.
Unmet needs will be captured using a questionnaire developed and validated by us in a previous trial .
Based on the study having 90% power to detect as significant, using a 5% 2-tailed significance level, a difference in the proportion of individuals with 2 or more dementia related GP visits of 50% (control: 20% versus Intervention: 70%) the required sample size, based on individual randomisation would be 23 per group - a total of 46 individuals. However, due to the use of cluster randomisation, the total required sample size needs to be inflated in order to take account of this clustering. The number of patients recruited per practice will also need to be inflated in order to maintain the sample sizes in the presence of attrition [19, 20]. With 20 practices (10 per arm), the power to detect the differences postulated would be maintained if the ICC were of the order 0.37 or less. Thus the effective sample size with 10 patients per cluster and 20 practices would need to be 200. If the expected attrition rate were 1/3 (33.3%) then 15 patients would need to be recruited per practice in order to maintain the sample size of 10 patients per practice.
All practices will be asked to identify patients with symptoms of dementia by using electronic searches of their clinical record system updated by manual checks of the resulting list by medical and nursing staff. Where there is no informal carer eligible for interview, advice regarding consent for examination of medical records will be sought from an appropriate consultee. No personalised information will be retained following the record examinations, which will be carried out in the surgery by research team members who have honorary contracts with the relevant PCT.
This trial will last 36 months allowing for a follow-up period of 12 months that will capture effects on clinical practice, carer satisfaction, met and unmet needs. Data will be gathered before and 12 months after the introduction of the educational interventions.
Participating practices will be randomised by an independent person to intervention or control arms using a computer generated randomisation programme. Practices randomly allocated to the intervention arm will be asked to participate in tailored learning activities consisting of an educational needs assessment followed by up to three face-to-face educational workshop sessions on dementia over a three-month period and will be given electronic resources which they can use during and after consultations with people with known or suspected dementia syndrome. Each educational workshop will be arranged at dates/times convenient to the practice team.
An experienced general practitioner with a background in postgraduateeducation will facilitate the small group workshops with the practice team. Tailoring of the education programme is carried out in a three step process: 1) in the first workshop, an educational needs analysis is undertaken using a standard checklist to identify aspects of dementia care which the practice perceives as problematic for them; 2) a prescription to meet the practice team's educational needs is then written to address shortcomings and sorting out the appropriate written educational resources, and; 3) the best forms of learning are then identified with a delivery and discussion with the team of the selected educational resource materials in the light of their perceived needs.
The 'normal care' arm will be given a summary of the NICE/SCIE dementia guidelines  and offered workshop training and electronic tools and resources at the end of the study. General practitioners, practice nurses and any other staff involved will be invited to give feedback on the educational resources.
Practices will be offered financial reimbursement to cover the use of locums and data collection costs (e.g. recruitment of patients and carers). Payments will be based on a sliding scale according to the number of partners in the practice, up to a maximum allocation.
People with dementia will be identified by practices and their lead clinician will check whether they fulfil the inclusion criteria. Where the lead clinician believes that the individual should not be approached (e.g. because they are receiving palliative care) they will be excluded. Before seeking consent from patients to participate in the study, practitioners will be asked for their opinion about the capacity of the person with dementia to give informed consent, using the Mental Capacity Act  as the framework for their judgement.
For patients judged as having capacity, the following information will be posted to the person with dementia: 1) a covering letter explaining the involvement of the practice and signed by the lead clinician; 2) a participant information sheet and; 3) a response letter and pre paid envelope to be returned to the research team. A researcher will arrange to see those patients with dementia and their carers who express an interest in participating in the study. This encounter can take place at the patient's home or in the practice, as they wish. Its purpose is to seek consent from the carer and consent/assent from the person with dementia for their participation in the study and allowing access to patient records.
For those judged by the lead clinician as lacking capacity to give informed consent, a consultee as outlined in the Mental Capacity Act  will be identified and consulted about possible involvement of the person with dementia in the trial. The relevant clinician will write to the consultee providing full information about the trial and asking whether or not they consent on behalf of the patient to enrolment.
Every effort will be made to identify a consultee for those judged to lack capacity to give informed consent. In the event that a consultee cannot be identified, the person with dementia will be excluded from participation in the trial. Figure 2 shows the process of identifying participants and how these will be approached, recruited and consented/consulted about EVIDEM-ED.
The process for obtaining participant informed consent or assent and guardian informed consent will be in accordance with the REC guidance , the Mental Capacity Act  and Good Clinical Practice (GCP) . The investigator or their nominee and the participant or other legally authorised representative shall both sign and date the Informed Consent Form before the person can participate in the study. The participant will receive a copy of the signed and dated forms and the original will be retained in the Trial Master File. A third copy will be filed in the participant's medical notes and a signed and dated note made in the notes that informed consent was obtained for the trial.
With the informed consent of the person with dementia or their consultee, members of the research team will examine medical records in the surgery, using the themes shown in Appendix 1 to guide data extraction. No personalised information will be recorded or retained by the researcher and each case will be allocated a unique study number for the purposes of recording data. All other information will be stored in accordance with the Data Protection Act.
We will assess the effect of the interventions at the practice level because the data will be cluster based and analyses will be performed on an intention to treat basis. Analyses of all quantitative responses will be performed with a general linear model with the arm and time as fixed effects and practice identity as a random effect.
We will analyse differences in detection rates by using binary logistic regression to include the cluster effect. These will be calculated before and after the intervention, excluding cases previously diagnosed in another practice. Concordance scores for diagnosis and management before and after the intervention will not directly be comparable as they comprise counts of actions taken over two different lengths of time. We will use Vickers' method and examine differences in baseline concordance scores across the arms of the study and then repeat the analysis for scores after the intervention. This analysis will also incorporate the cluster effect.
If outcome data are missing, we will assume they are "missing at random" (MAR). This means that the probability of missing data can be predicted by variables measured at baseline. In this case, an analysis which adjusts for the baseline predictors of 'missingness' (at least baseline response and treatment) will give an unbiased estimated of the treatment effect, making multiple imputation unnecessary. Multiple imputation will be used only if important baseline predictors are missing. Methods will then be employed which take account of the clustered nature of the data
The trial will be led by the Principal Investigator, Prof. Steve Iliffe and managed by the Programme Manager Ms. Jane Wilcock. The trial co-ordinating centre is the Department of Primary Care and Population Health at University College London. A Trial Management Committee will meet every two months to review the progress of the trial and its members will be recruited from; 1) the research teams in the EVIDEM programme and; 2) patient and public representatives working with the EVIDEM programme. A Trial Steering Committee (TSC) with a majority of members independent of the EVIDEM programme will meet six monthly to review the study. Because there were no adverse outcomes in the earlier trial of educational interventions, we propose not to have a data management committee but to give the chair of the TSC powers to convene a data management sub-group should there be any need to consider adverse events.
Stopping rules and discontinuation
Participants can chose to leave the trial at any point. The trial will be terminated if it is shown to have a negative impact on routine medical care.
Duration of the trial
The Trial will run for 36 months depending on the speed of recruitment of practices. Practices randomly allocated to receive training will receive this within three months period, with data collection at baseline and again at 12 months following baseline assessment.
Practices will receive regular updates on the trial's progress via newsletters and other media and will also be invited to annual summer schools and seminars run by the EVIDEM programme. This will include receiving study findings at the end of the Trial. Practices allocated to the control arm will receive training after the data collection period. The people with dementia and carers who participate will be offered an option to receive a lay version of the study outcomes and be invited to a lay conference at the end of the study period.
User and public involvement
All user and carer representatives from the NICE/SCIE guideline development group have agreed to join the Reference Group that meets annually to provide critical evaluation of the EVIDEM programme by which the EVIDEM-ED study is overseen. They will bring their experience of the guideline development process and in particular their awareness of the specific gaps in evidence and the methodological problems of dementia research in the community. Members of the Reference group also include representatives from the Alzheimer's Society, the Council for Palliative Care, Age Concern England, the Association of London Government, Dementias and Neurodegenerative Diseases Research Network (DeNDRoN), national co-ordinating centre and the Mental Health Trust.
We do not anticipate any unfavourable and unintended signs, symptoms, syndromes or illness to be caused to patients by general practitioners' participation in the trial. However, it is possible that an unknown serious condition, treatment or behaviour could come to the attention of the researchers.
Participants will be asked to contact the study site immediately in the event of any serious adverse event. All adverse events will be recorded and closely monitored until resolution or stabilisation, or until it has been shown that the study involvement is not the cause. The Chief Investigator shall be informed immediately of any serious adverse events and shall determine seriousness and causality in conjunction with any unexpected outcome of trial participation, together with the chair of the TSC.
In the event of a hitherto unknown severe or serious condition or situation becoming apparent to the research team, this will be discussed with the Chief Investigator who will notify the appropriate service as required. The Chief Investigator shall be responsible for all adverse event reporting. Any participant who experiences an adverse event may be withdrawn from the study at the discretion of the Investigator, but in consultation with the patient, their carer (where appropriate) and their general practitioner.
This study has particular ethical implications. Gaining consent from people with dementia raises complex issues and full research governance processes will be followed to gain meaningful informed consent, or where necessary, consult with those who are able to act as decision-makers. We have obtained ethical permissions from the NHS ethical review system and relevant NHS governance departments. We shall follow local adult protection procedures in each of the localities of the research and data collection.
Ethics committee and regulatory approvals
The trial will not be initiated before the protocol, informed consent forms and participant and GP information sheets have received approval/favourable opinion from the Research Ethics Committee (REC) and the respective National Health Service (NHS) Research and Development (R&D) department. Should a protocol amendment be made that requires REC approval, the changes in the protocol will not be instituted until the amendment and revised informed consent forms and participant and GP information sheets (if appropriate) have been reviewed and received approval/favourable opinion from the REC and R&D departments. A protocol amendment intended to eliminate an apparent immediate hazard to participants may be implemented immediately providing that the REC are notified as soon as possible and an approval is requested. Minor protocol amendments only for logistical or administrative changes may be implemented immediately and the REC will be informed.