Screening for depression in patients with cognitive impairment: a local audit

Depression is highly prevalent in cognitive impairment and is considered one of the most common comorbid conditions in dementia.1 Despite this, underlying depressive symptoms can easily be missed2 by health care providers, mainly due to lack of awareness and lack of routine screening for depressive symptoms. Here, the authors examine the results of their local audit in Oxleas NHS Foundation Trust screening for depression in those diagnosed with dementia and discuss what their recommendations are for improving this aspect of clinical care in the future.

Knapskog et al.3 found that depressive symptoms were common among patients referred to specialist memory services. Since both depression and dementia can cause memory impairment, it is essential to screen effectively to ascertain the most appropriate treatment plan. Hancock et al.4 found that those individuals without dementia who attended cognitive clinics were more likely to score above the threshold for depression on the PHQ-9 screening tool: 41% in non-demented patients versus 12% in those with dementia.

NICE dementia guidelines5 advise that there should be an assessment for depression at the time of diagnosis of dementia and at regular intervals subsequently. This, however, is not widely seen in routine clinical practice. Our local guidelines are in line with this and recommend screening for depression in all new referrals of patients with cognitive impairment. The local Trust guidelines also recommend having a clear discussion about diagnosis, treatment and treatment benefit with patients and family/carers at the time of initial and subsequent appointments. It is also recommended that this discussion is clearly documented on the Trust electronic patient record keeping system for future reference.

For the purpose of monitoring and evaluating adherence to NICE guidelines and local Trust recommendations an audit cycle was completed. There is no evidence of previous audits on this topic. The outcome from this audit may inform, and hopefully improve, screening and treatment for depression in cognitive impairment across memory services.


The audit process was completed over a two-year period from 2013 to 2015 at the Memory Service, Oxleas NHS Foundation Trust, UK. The memory service comprises a separate multidisciplinary team across each of three London boroughs.

A data collection tool (Box 1) was agreed so that information could be obtained from the electronic notes.

The cohort assessed for the initial audit cycle was patients referred to the memory service for investigation of cognitive problems between 1st January 2012 and 1st July 2012. The cohort was chosen by stratifying the referrals during this period so that every fourth referral chronologically was selected.

The first three face-to-face meetings with each patient were reviewed on the electronic notes. It was checked whether the patient had a pre-existing diagnosis of depression. It was then noted whether mood was assessed during the first three meetings with the patient, either informally or using any standardised screening tool (eg Geriatric Depression Scale [GDS],6 Cornell Scale for Depression in Dementia [CSDD]7). If a patient was on treatment for depression, then it was noted whether there was documentation of a discussion about this with the patient or carer. On some occasions there was not a statement specifically about mood but there were details of core symptoms relating to mood such as problems with sleep, appetite, anhedonia or concentration. Therefore we also included in the data collection tool space to document the inclusion of this.

A re-audit was completed 18 months after dissemination of the results of the original audit. The aim of the re-audit was to measure improvement in clinical practice since the initial audit. A list of all new referrals to the memory service was obtained from the informatics department during the period of February 2014 to July 2014. A computer-generated random list of every fourth person was again selected to be included in the re-audit. This aimed to reduce selection bias. Data was collected retrospectively from the Trust’s electronic record keeping system using the data collection tool (Box 1).


Audit cycle 1

There were 913 referrals to the memory teams across Oxleas NHS Foundation Trust from 1st January 2012 to 1st July 2012. After stratifying the cohort, there were 229 patients, of these 19 (8%) were not assessed either due to the patient declining assessment, the patient moving area, the death of the patient or the referral recorded in error. The patients predominately had an organic diagnosis with some diagnosed with an affective disorder. The most common ICD-10 diagnoses amongst the 210 patients assessed were dementia in Alzheimer’s disease with late onset, F00.1 (54 patients) and mixed dementia in Alzheimer’s, F00.2 (54 patients).

There was no existing diagnosis of depression in 85% of patients. Of these, 83% had their mood assessed when they presented at services (see Figure 1) with only 4% using a standardised screening tool. This was the GDS in all cases.

Of those who did have an existing diagnosis of depression, there was documentation of their mood at assessment in 87% of cases (see Figure 2) with 6% of these using the GDS or CSDD.

There were 34 patients on medication for depression. Of these, 47% had documentation in their notes of a discussion about treatment for depression with the patient or family/carer.

After the initial audit the following recommendations were made:
• Mood should be assessed in every assessment at the memory service.
• The use of standardised tools can be helpful in evaluating mood.
• If there is a pre-existing diagnosis of depression treatment should be reviewed regularly at every contact with the patient.
• Discussions need to be had with the patient and/or carer about the treatment and documentation must be clearly made in the patient records.

The results were disseminated among all teams during local multidisciplinary team (MDT) teaching sessions. Advice was given to use screening tools, either GDS or CSDD, which were provided in clinics, especially when the assessor was lacking confidence in the assessment of depression. Teaching and feedback sessions were useful in ensuring that all teams were aware of these guidelines.

Following this initial audit, it was decided that in the subsequent re-audit an attempt would be made to increase the sample size. It was also decided to include an additional question in the data collection tool: Were the core symptoms and biological symptoms used to assess depression?


The re-audit was carried out 18 months later. There were a total of 1280 new referrals received during the period of February 2014 to July 2014. After stratifying the cohort, there were 302 patients in the sample. The sample size was calculated in line with a recommendation from the original audit to increase sample size.

The most common diagnoses in the sample were mixed dementia (17%), dementia in Alzheimer’s disease-late onset (15%), and multi-infarct dementia (12%).

Only 9% of the cohort had an existing diagnosis of depression. Mood was assessed in 93% of these patients (see Figure 3) with a screening tool used in 25% of cases.

Of the 91% with no existing diagnosis of depression, 88% had documentation of a mood assessment (see Figure 4) with 19% using screening tools.

Biological symptoms of depression were documented in 69% of all patients. Discussion about treatment was documented in 82% of cases. Of the patients receiving treatment for depression, 96% were on antidepressant medication.


The initial audit revealed a significant proportion of patients were lacking documentation of a mood assessment in their notes (17% of patients with no pre-existing depression and 13% of those with pre-existing depression). This suggests that a large number of people were either not assessed for depression or, if assessed, no documentation to that effect was made. This raised the concern that there were missed opportunities to provide treatment for these patients and also that this would complicate the effective management of their cognitive impairment.

There was an improvement following the re-audit (12% of patients with no pre-existing depression and 7% of those with pre-existing depression were lacking documentation of mood assessment). However, it is clear that, despite improvement, new measures are needed to ensure greater compliance with the recommendation.

Standardised tools can be helpful in screening for mood disorders. In the initial audit, a standardised tool was used in only 4% of those with no pre-existing depression and 6% of those with pre-existing depression. This increased to 19% and 25% respectively in the re-audit. There has been an improvement but there is scope for further optimisation. The greater use of standardised tools could make the assessment of mood more reliable across the board. Furthermore, the more frequently the tools are used the more proficient people become at using them and the quicker the assessments are completed.

There was a large deficiency in the documentation of treatment discussions. The discussion gives the opportunity to address concerns about the medications, discuss alternative treatment options, clarify the benefits and side-effects of chosen treatment and improve the patient-clinician therapeutic relationship. Screening for core and biological symptoms of depression was not audited in the initial audit so this was added to the audit tool to enhance the re-audit. Although core symptoms were documented in the majority of cases, there remains further scope for improvement to increase the accurate recognition of patients with depression.

The results of the re-audit showed some improvement in screening for depression, however, there was room for further improvement. Further recommendations were made to achieve sustained improvement.


1. Screening for depression: All patients referred to the memory service should be screened for underlying depression.
2. Core symptoms: Initial assessment should include core and biological symptoms of depression when screening for depression.
3. Clinical tools: A culture of using validated and standardised tools to screen for depression in cognitive impairment should be prompted; Geriatric Depression Scale is advised. These tools should be used by all dementia health care providers including doctors, nursing staff, occupational therapists and psychologists.
4. Accessibility of depression screening tools: Depression screening tools should be made widely and easily accessible to health care providers within memory services involved in routine clinical care of dementia patients. These have been printed and placed in clinical areas and will be available on the intranet.
5. Discuss treatment options: Once depression is identified treatment options should be discussed with patients, and families/carers.
6. Psychological treatment: Consider and offer psychological treatment to those eligible for treatment of depression in addition to pharmacological treatment options.
7. Improve treatment: Improve the number of patients with depression receiving treatment for the condition once diagnosed.
8. Improve documentation/standard template for documentation: Use a standard template for assessments to allow essential areas to be included.

The following is an example of a template of headings to be included in clinical documentation:
• Screening for depression (include initial assessment of mood and score of clinical tools used to screen for depression).
• Discussion of diagnosis (if identified).
• Discussion of treatment options.
• Progression of treatment and possible side-effects.
• Patient/family/carer views and satisfaction.

A further audit is advised to assess the effectiveness of the implementation of these recommendations.

Declaration of interests

No conflicts of interest were declared.

Dr Wood is Higher Trainee in Old Age and General Adult Psychiatry, Tees, Esk and Wear Valleys NHS Foundation Trust, Harrogate District Hospital; Dr Nabi is Higher Trainee in Old Age and General Adult Psychiatry, South London and Maudsley NHS Foundation Trust, Lewisham Hospital, and Dr Adebekun is Consultant Psychiatrist in Old Age Psychiatry, Oxleas NHS Foundation Trust Bexleyheath Centre, Bexleyheath.


1. Ismail Z, Elbayoumi H, Fischer CE, et al. Prevalence of Depression in Patients with Mild Cognitive Impairment: A Systematic Review and Meta-analysis. JAMA Psychiatry 2017;74(1):58–67.
2. Rodda J, Walker Z, Carter J. Depression in older adults. BMJ 2011;343:d5219.
3. Knapskog AB, Barca ML, Engedal K. Prevalence of depression among memory clinic patients as measured by the Cornell Scale of Depression in Dementia. Aging Ment Health 2014;18(5):579–87.
4. Hancock P, Larner AJ. Clinical Utility of Patient Health Questionnaire-9 (PHQ-9) in memory clinics. Int J Psychiatry Clin Pract 2009;13:188–91.
5. National Institute for Health and Care Excellence (NICE). Dementia: supporting people with dementia and their carers in health and social care. London: NICE, 2006. (accessed October 2018).
6. Yesavage JA, Sheikh JI. Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version. Clinical Gerontologist 1986;5:165–73.
7. Alexopoulos GS, Abrams RC, Young RC, et al. Cornell Scale for Depression in Dementia. Biol Psychiatry 1988;23:271–84.

Screening for depression in patients with cognitive impairment: a local audit.

Add yours ↓
Web design and marketing agency Leamington Spa