The number of people living with dementia is increasing as the population ages. The UK government has initiated a "Dementia strategy" to respond to this increase, and has set four main goals; 1) to identify those with dementia as early as possible, 2) to prolong their independence, 3) to reduce the need for hospital and nursing home admissions, and 4) to delay the time until death. So far lots of research has looked at what factors might cause or increase the risk of dementia, but less is known about how dementia affects people over time once they have dementia. One of the quickest and efficient ways to identify the actual course of dementia may be to use existing medical records from primary care. Primary care medical records are those kept by the patient's GP. Having the ability to look at the course of dementia over time in these records will help identify people who have a different course (for example a faster and more severe course) and also identify the factors that alter that course. This information would be useful to clinicians and indicate potential targets for treatment in the future to alter the course of someone's disease. To assess whether this is possible, around 1000 patients with dementia who have recently been seen by a dementia service (where patients with a diagnosis of dementia are seen normally every year) will be contacted. Consent to link their primary care medical records (medical records kept by their GP) to their dementia service medical records will be sought. The target for recruitment is 400 patients who will provide consent. Then the assessment of how well the potential markers of disease progression identified from primary care medical records relate to actual changes in disease course from their dementia service medical records can be determined.

The UK government has prioritised early recognition and treatment of dementia, with specific goals to prolong independence, delay nursing home and hospital admissions, and reduce mortality. It is recognised that this strategy in the UK health system necessitates primary care involvement in diagnosis and management of dementia. While advances have been made in identifying patients with dementia within primary care, there is little research on the course of the disease after diagnosis, even though information about disease course and the factors that alter the course are essential to patient management and wider health policy.

One potential major resource for studying these issues in primary care are Electronic Health Record (EHR) databases which contain routinely recorded information from consultations with a GP and other healthcare professional such as illnesses and treatment provided. They are a useful resource for studying how illnesses progress after diagnosis in primary care and have been used to investigate disease course, outcomes and patient care in a wide spectrum of diseases. However, to date, they have not been used to research the course of dementia after diagnosis. Existing evidence has established important dementia outcomes (e.g. nursing home placement, hospital admissions, mortality) are recorded within EHR and a systematic review that has recently been performed suggests other illnesses and symptoms (e.g. cognitive status, neuropsychiatric symptoms), may be recorded within primary care EHR and could serve as markers of disease progression.

To address this research gap the research will identify possible markers of disease progression from routinely collected in primary care medical records and compare them against assessments of disease progression from specialist dementia service medical records. The study will enable healthcare professionals to use routinely recorded information in primary care, without the need for intensive assessments, to identify individuals with dementia who are at risk of a faster disease progression. This will drive the future development and delivery of treatment to the individual patient with dementia based on their likely future course.

The CoMed study is a retrospective cohort design linking dementia service and primary care medical records. Potential participants' only involvement will be a request for consent to access and link these medical records.

Participants will be recruited from a secondary care dementia service.

All patients who have a confirmed diagnosis of dementia in the dementia service medical records and who have been assessed by the dementia service at more than one-time point at least 12 months apart will be identified (a 12-month gap is the minimum necessary to determine disease progression). The first 1000 eligible patients will be initially identified (expected number required to obtain sample size required of 400 patients), however, further searches can be carried out if the expected response and recruitment rate is not met. The list of patients to be invited will be screened by the patient's clinical care team in the participating dementia service to exclude those who it would be inappropriate to contact as it may cause undue distress or harm.

Eligible patients who have been seen by the dementia service within the previous 12 months will be identified by the dementia service clinical care team. These eligible patients will be sent by post the study information pack containing: an invitation letter, patient and personal consultee information leaflets, patient consent and personal consultee declaration forms, and a stamped addressed return envelope. If there is no reply to the initial invitation, then a reminder invitation and study information pack will be sent after a minimum of 2 weeks. A poster about the study will be displayed in the dementia service clinical areas and waiting areas, and a copy of the poster in leaflet form will be available for patients to take away and read if they wish.

A review of the response and recruitment rate will be undertaken at regular intervals to assess whether more patients need to be identified (based on a 40% recruitment rate). If the required number of patients is not achieved additional recruitment strategy will be initiated. An additional search of the dementia service medical records to identify further eligible patients and study packs will be posted out.

For those who are willing to take part in the study written informed consent will be obtained from patients with dementia (or personal consultee's advice for those not able to give consent) to access their dementia service medical records and their primary care medical records and for these to be linked and used for research purposes.

In those consenting to take part in the study, patient data will be retrieved from the dementia service records including detailed clinical assessments of cognitive status, behavioural and function assessments, and the time points of these assessments. The patient's primary care medical records will be requested from their GP practice. Details on how to electronically download the requested medical records and securely transfer them via the secure NHS email system will be included with the letter to GP practices.

Confidentiality and anonymity will be maintained for all participants. All information collected regarding participants will be kept confidential. Participants will be free to withdraw from the study and to request their medical records are not used up to a period of one month from the date consent is given. After this point, pseudoanonymised data will be used and it will not be possible to identify patients to remove their medical records from the analysis dataset. The presence or absence of potential markers of progression recorded in patients' primary care records will then be compared to change in scores from standard clinic assessments recorded at the specialist dementia clinic.

Inclusion Criteria:

• Adults aged 18 years and over
• Confirmed diagnosis of dementia in the dementia service medical records
• Assessment by the dementia service on two or more occasions at least 12 months apart (timescale required to determine disease progression)
• Recorded appointment with the dementia service in last 12 months

Exclusion Criteria:

• Lists of patients who are potentially eligible to take part will be screened by their respective clinical care teams to exclude those who may have significant issues (e.g. significant life event) where contact to participate may cause harm or additional distress or who have indicated that they do not wish to be part of any research or for their data to be shared including those expressed as an advance directive or statement.

Participants will be recruited from a secondary care dementia service. All patients who have a confirmed diagnosis of dementia in the dementia service medical records and who have been assessed by the dementia service at more than one-time point at least 12 months apart will be identified (timescale is necessary to determine disease progression). The first 1000 eligible patients will be initially identified (expected number required to obtain sample size requirement of 400 patients), however further searches can be carried out if the expected response and recruitment rate is not met.