Table 4 Summary of measurement tools

GP behaviour

 Cognitive assessment using MMSE

CAT electronic search^†

MMSE results recorded in the patient file or free text indicates an MMSE has been undertaken.

 Depression assessment using validated scale

CAT electronic search^†

Geriatric Depression Scale (GDS) results recorded in the patient file or free text indicates GDS, Hamilton Rating Scale for Depression, Even Briefer Assessment Scale for Depression has been undertaken.

 Referral to CDAMS or specialist

CAT electronic search^†

Free text indicates that the patient has been referred to CDAMS, ACAS (Aged Care Assessment Service), or a geriatrician.

 Referral for CT scan

CAT electronic search^†

CT scan has been requested or free text indicates that a CT (head) scan has been requested/undertaken.

 Dementia diagnosis

CAT electronic search^†

Coded diagnosis of dementia or free text indicates that the patient has dementia or Alzheimer’s disease.

 Cognitive assessment using MMSE (all patients aged 70+ years)

CAT electronic search^†

MMSE results recorded in the patient file or free text indicates an MMSE has been undertaken.

 Reported suspicion of cognitive impairment (all patients aged 70+ years)

CAT electronic search^†

Coded diagnosis of cognitive impairment or free text indicates a suspicion of cognitive impairment (e.g., confusion, muddled, cognitive), or; MMSE undertaken in isolation of the 75+ Health Check (an indication of GP’s suspicion of cognitive impairment), or; MMSE undertaken as part of 75+ Health Check with score indicating cognitive impairment (i.e., a score between 10 and 24).

 Dementia diagnosis (all patients aged 70+ years)

CAT electronic search^†

Coded diagnosis of dementia or free text indicates that the patient has dementia or Alzheimer’s disease.

Proxy measures of GP behaviour

 Self-report of adherence to recommended behaviours (e.g., Cognitive assessment using MMSE)

Questionnaire* (1 item per behaviour)

Adapted from Eccles et al. [45]. Example item: ‘Thinking about the last 10 patients you saw who you suspected had cognitive impairment, how many of them did you assess for cognitive function using the Mini Mental State Examination (MMSE)?’

 Behavioural simulation to adhere to recommended behaviours (e.g., Cognitive assessment using MMSE)

Questionnaire (6 clinical vignettes)

Vignettes simulate clinical decision-making about detection, diagnosis and management of dementia. Vignettes include a range of clinical variables: sex, age (72 – 88 years), cognitive function (including changes to memory, personality, behaviour, cognition), depression, and other elements. These clinical variables were drawn from previously published vignettes [46–52] and from the experience of the clinical investigators. The vignettes, and response options, will be piloted with two to three GPs prior to being administered.

Hypothesised mediators of GP behaviour

 Intention to adhere to recommended behaviours (e.g., Cognitive assessment using MMSE)

Questionnaire* (3 items per behaviour)

Adapted from Eccles et al. [45], Francis et al. [53],and Foy et al. [54]. Items for Cognitive assessment using MMSE include ‘I would make it a high priority to use the MMSE to assess the cognitive function of these patients,’ ‘I plan to use the MMSE to assess the cognitive function of these patients,’ ‘I intend to use the MMSE to assess the cognitive function of these patients.’ Each item measured on a 7-point Likert scale ranging from strongly disagree to strongly agree (1 to 7). Scores are then averaged to create a behavioural intention score. Higher scores reflect greater intention to assess cognitive function using the MMSE.

 Behavioural constructs for primary outcomes

Questionnaire*

Adapted from [34]. There are 49 items in total. Example of the items used to measure various domains (noted in brackets) for the behaviour depression assessment using a validated scale include: ‘How much do you know about validated scales for assessing depression in these patients?’ (knowledge), ‘Using a validated scale to assess these patients for depression is sometimes stressful’ (emotion), ‘Lack of time may prevent me from using a validated scale to assess these patients for depression’ (environmental context and resources). Each item is measured on a 7-point Likert scale. All constructs are measured using three items (which are averaged to create a final score for the construct), except beliefs about consequences, which is measured using four items.