Stroke Aphasic Depression Questionnaire (SADQ)

Evidence Reviewed as of before: 22-08-2008
Author(s)*: Lisa Zeltzer, MSc OT
Editor(s): Nicol Korner-Bitensky, PhD OT; Elissa Sitcoff, BA BSc

Purpose

The Stroke Aphasic Depression Questionnaire (SADQ-21) was developed to detect depressed mood in clients with stroke and significant aphasia living in the community. It is a 21-item questionnaire developed based on observable behaviours thought to be associated with depressed mood. It is completed by the client’s caregiver on behalf of the client. A shortened version of the SADQ has been developed (SADQ-10), which is comprised of 10 questions that best differentiate those with high scores on depression questionnaires from those with low scores. A revised version of the scale has also been developed for clients in hospital to be completed by hospital staff (SADQ-H).

In-Depth Review

Purpose of the measure

The Stroke Aphasic Depression Questionnaire (SADQ-21) was developed to detect depressed mood in clients with stroke and significant aphasia living in the community. It is a 21-item questionnaire developed based on observable behaviours thought to be associated with depressed mood. It is completed by the client’s caregiver on behalf of the client. A shortened version of the SADQ has been developed (SADQ-10), which is comprised of 10 questions that best differentiate those with high scores on depression questionnaires from those with low scores. A revised version of the scale has also been developed for clients in hospital to be completed by hospital staff (SADQ-H).

Available versions

The original version of the SADQ and the shortened version (SADQ-10) were developed by Sutcliffe and Lincoln in 1998. The SADQ was revised by Sutcliffe, Lincoln, and Unsworth in 2000 so that the measure could be used with clients in the hospital (SADQ-H).

Features of the measure

Items:
SADQ-21
In this 21-item scale, caregivers must rate the frequency at which certain observable behaviours that are thought to be associated with depressed mood occur. Each item requires the responder to rate how often in the last week these behaviours occurred on a scale of ‘often’, ‘sometimes’, ‘rarely’, ‘never’. Examples of items include: “Did he/she take sleeping tablets; Did he/she refuse to eat meals?; Did he/she have weeping spells?”.

SADQ-10
The SADQ-10 was developed as an abbreviated version of the SADQ-21 in order to improve the validity of the SADQ. For clinical use, this version is recommended as the best version currently available for detecting depressed mood. The SADQ-10 is comprised of 10 items from the SADQ-21 that best differentiated between those with high scores on depression questionnaires from those with low scores.

SADQ-H
The SADQ-H was developed to be used specifically with clients in the hospital. The SADQ-H is comprised of the same 21 items as the SADQ-21, but some modifications were made to the wording of the items and the response categories. Some items were rephrased in order to make them easier to answer with respect to a patient with aphasia (e.g. ‘Did he/she indicate suffering from aches and pains?’ rather than ‘Did he/she complain of aches and pains?’). Items were also reworded to take disability into account (e.g. ‘Did he/she take care of his/her appearance to the extent of his/her physical ability?’ rather than asking if he/she take care of his/her appearance). Questions were made more specific (e.g. ‘Did his/her waking cause disturbance in sleep patterns?’ rather than ‘Does he/she wake early in the morning?’, emphasizing the behaviour as a problem, rather than merely early waking per se). Moreover, the tense of the questions was altered because the rater was considering patient behaviours over the previous week. The response categories were altered, from ‘often’, ‘sometimes’, ‘rarely’ and ‘never’ to ‘every day this week’, on ‘4-6 days this week’, on ‘1-4 days this week’ and ‘not at all this week’.

SADQ-H 10
The SADQ-H 10 was developed to be used specifically with clients in the hospital. It is comprised of the same 10 items found in the SADQ-10 but these items were first modified in the SADQ-H as described above.

Scoring:
For all versions of the SADQ a score of 0-3 is selected for each question. The total score is produced by adding the individual scores from each question.

SADQ-21
Scores range from 0-63 on the SADQ-21, with higher scores indicating higher levels of depression. No cut-off score has been established for this version of the SADQ.

SADQ-10
Scores range from 0-30 in this version. A cut-off score of 14 out of 30 on the SADQ-10 has been found to be optimal for detecting the presence of depression (Leeds et al., 2002).

SADQ-H
Same as the above for the SADQ-21. A cut-off score of 17/18 has been found optimal for detecting the presence of depression.
Note: The cut-off values indicate a lower figure that is the highest likely to be obtained by people without low mood and a higher figure that is the lowest likely to be obtained by people with low mood.

SADQ-H 10
Same as the above for SADQ-10. A cut-off score of 5/6 has been found optimal for detecting the presence of depression.
Note: The cut-off values indicate a lower figure that is the highest likely to be obtained by people without low mood and a higher figure that is the lowest likely to be obtained by people with low mood.
Time: It takes approximately 2 minutes to complete the SADQ-10 (and SADQ-H 10) and approximately 4 minutes to complete the full SADQ (and SADQ-H).

Subscales:
None.

Equipment:
Only the test copy and a pencil are required to complete the SADQ.

Training of administrator:
None required.

Alternative forms of the SADQ-21

Please see the “items” section under ‘features of the measure’ for a detailed description of the alternative versions of the SADQ-21, namely the SADQ-10, SADQ-H, and SADQ-H 10.

Client suitability

Can be used with:

  • Clients with stroke and significant aphasia.

Should not be used with:

  • Individuals who may be depressed but who have not had a stroke or clients who do not have significant aphasia, that is, minimal ability to understand and response (Leeds, Meara, Hobson, 2002). For these clients, other depression measures exist that have more evidence to support their psychometric properties (e.g. Hamilton Depression Rating Scale (HDRS), Beck Depression Inventory (BDI), Geriatric Depression Scale (GDS), etc.).

In what languages is the measure available?

The SADQ-21 and SADQ-10 have been translated into French and the SADQ-H and SADQ-H 10 are available in Italian. These translations are available online at: http://www.nottingham.ac.uk/IWHO/Research/PublishedAssessments.aspx

Summary

What does the tool measure? To detect depressed mood in clients with stroke and significant aphasia.
What types of clients can the tool be used for? The SADQ is intended for use with clients with stroke having significant aphasia.
Is this a screening or assessment tool? Assessment or screening.
Time to administer From 2 to 4 minutes.
Versions SADQ-10; SADQ-H; SADQ-H 10
Other Languages French; Italian
Measurement Properties
Reliability Internal consistency:
– Two studies examined the internal consistency of the SADQ-21 and the SADQ-H and reported excellent internal consistency.
– One study examined the internal consistency of the SADQ-10 and reported excellent internal consistency.
– One study examined the internal consistency of the SADQ-H 10 and reported poor internal consistency.

Test-retest:
One study examined the test-retest reliability of the SADQ-21 and SADQ-10. Both had adequate test-retest reliability using Spearman rho correlation.

Inter-rater:
One study examined the inter-rater reliability of the SADQ-H between patients and their nurses using Cohen’s kappa and reported 5 items had excellent inter-rater reliability and the remaining 16 items had adequate inter-rater reliability.

Validity Construct:
Convergent:
– Two studies examined convergent validity of the SADQ-21 and SADQ-10 with the Hospital Anxiety and Depression Scale (HADS) and the Wakefield Depression Inventory (WDI). Both studies found a poor correlation between SADQ-21 and the HADS depression subscale (HADS-D) and an adequate correlation between the SADQ-21 and the WDI.
– One study reported an adequate correlation with the HADS anxiety subscale (HADS-A) and one reported a poor correlation with the HADS-A.
– One study reported a poor correlation between the SADQ-10 and the HADS-D and WDI, and an adequate correlation with the HADS-A, and one study reported an adequate correlation between the SADQ-10 and the HADS-D and an excellent correlation with the HADS-A and WDI.
– One study examined the convergent validity of the SADQ-H and SADQ-H 10 with the HADS and reported an adequate correlation with the HADS-D and a poor to adequate correlation with the HADS-A.
– One study examined the convergent validity of the SADQ-10 with the Geriatric Depression Scale-15 (GDS-15) and reported an adequate correlation.
Floor/Ceiling Effects No studies have examined the floor or ceiling effects of the SADQ.
Sensitivity/Specificity – One study reported that for the SADQ-10, a cut-off score of 14/30 produced a sensitivity of 70% and a specificity of 77% to detect depression.
– One study reported an optimum cut-off of 17/18 on the SADQ-H (sensitivity of 1.00; specificity of 0.81), and an optimum cut-off of 5/6 on the SADQ-H 10 (sensitivity of 1.00; specificity of 0.78) to detect depression.
Note: The cut-off values indicate a lower figure that is the highest likely to be obtained by people without low mood and a higher figure that is the lowest likely to be obtained by people with low mood.
Does the tool detect change in patients? At present no studies have examined the responsiveness of the SADQ.
Acceptability The SADQ should not be used with individuals who may be depressed but who have not had a stroke, or patients who do not have aphasia.
Feasibility Completion of all versions of the SADQ is quick and simple.
How to obtain the tool?

The SADQ is available free with permission from the authors and can be obtained from the following website:
http://www.nottingham.ac.uk/IWHO/Research/PublishedAssessments.aspx

Psychometric Properties

Overview

We conducted a literature search to identify all relevant publications examining the psychometric properties of the Stroke Aphasic Depression Questionnaire (SADQ-21) and its alternative versions in individuals with aphasia.

Floor/Ceiling Effects

No studies have examined the floor or ceiling effects of the SADQ.

Reliability

Internal consistency:
Sutcliffe and Lincoln (1998) examined the internal consistency of the SADQ-21 and SADQ-10 in 70 clients with aphasia who had been discharged from the hospital. Internal consistency was determined by performing item-total and split-half analyses. The SADQ-21 had a Cronbach’s alpha of 0.82 (excellent) and a split-half correlation of 0.79 (adequate). The SADQ-10 had excellent internal consistency, with a Cronbach’s alpha of 0.80 and a split-half reliability of r = 0.81.

Lincoln, Sutcliffe, and Unsworth (2000) examined the internal consistency of the SADQ in 50 in-patients with stroke and the hospital version of the SADQ (SADQ-H) in 30 in-patients with stroke. The primary nurse was asked to complete the SADQ in respect of the patient, for the week preceding the assessment interview. The SADQ had an excellent Cronbach’s alpha of 0.80 and an adequate split-half reliability of 0.70. The SADQ-H had an excellent Cronbach’s alpha of 0.82 and an adequate split-half reliability of 0.74.

Bennett, Thomas, Austen, Morris and Lincoln (2006) examined the internal consistency of the hospital version of the SADQ (SADQ-H) and the shortened version (SADQ-H 10) in 100 patients with stroke recruited from two acute hospitals. The SADQ-H had excellent internal consistency (alpha = 0.84), and the SADQ-H 10 had poor internal consistency (alpha = 0.68).

Test-retest:
Sutcliffe and Lincoln (1998) examined the test-retest reliability of the SADQ-21 and SADQ-10 in 17 clients with aphasia who had been discharged from hospital using Spearman correlations. The carers of clients (spouses or nursing home staff) were contacted by post and asked to complete and return the SADQ. Carers were contacted four weeks later and asked to complete and return the SADQ again. The SADQ-21 and SADQ-10 were found to have adequate test-retest reliability over a four-week interval (r = 0.72 and r = 0.69, respectively).

Inter-rater:
Lincoln et al. (2000) examined the inter-rater reliability of the SADQ-H in 30 in-patients with stroke. Cohen’s Kappas were calculated to assess the agreement between the SADQ as completed by the nurses, and the SADQ as completed by the patients themselves for each item. Using internal statistical categories, it was found that 5 items showed an excellent agreement and 16 items showed adequate agreement between patient and nurse (Cohen’s Kappa ranging from 0.40 to 0.90).

Validity

Content:
Items comprising the SADQ were chosen based on observable behaviours thought to be associated with depressed mood and included items derived from other questionnaire measures of depression. Ratings were compared with questionnaire responses in clients without aphasia.

Criterion:
Concurrent:
Not yet examined.

Predictive:
Not yet examined.

Construct:
Convergent/Discriminant:
Sutcliffe and Lincoln (1998) examined the convergent validity of the SADQ and SADQ-10 with the Hospital Anxiety Depression Scale (HADS – Zigmond & Snaith, 1983) and the Wakefield Depression Inventory (Snaith, Ahmed, Mehta, & Hamilton, 1971) in 70 clients with aphasia who had been discharged from the hospital. The relationship between the measures was examined using Spearman’s rank correlations. The SADQ had a poor correlation with the HADS depression subscale (r = 0.22), an adequate correlation with the anxiety subscale (r = 0.42) and with the Wakefield Depression Inventory (r = 0.52). The SADQ-10 had an adequate correlation with the HADS depression subscale (r = 0.32), an excellent correlation with the anxiety subscale (r = 0.63) and with the Wakefield Depression Inventory (r = 0.67).

Bennett et al. (2006) examined the convergent validity of the SADQ-H and SADQ-H 10 with the Hospital Anxiety Depression Scale (HADS – Zigmond & Snaith, 1983) in 100 patients with stroke recruited from two acute hospitals. The relationship between the measures was examined using Spearman’s rank correlations. The SADQ-H and SADQ-H 10 had an adequate correlation with the HADS depression subscale (r = 0.52 and 0.53, respectively), a poor to adequate correlation with the HADS anxiety subscale (r = 0.23 and 0.33, respectively), and an adequate correlation with the total HADS (r = 0.45 and 0.51, respectively).

Lincoln, Sutcliffe, and Unsworth (2000) examined the convergent validity of the SADQ and the SADQ-10 with the Hospital Anxiety and Depression Scale (HADS – Zigmond & Snaith, 1983) and the Wakefield Depression Inventory (Snaith et al., 1971) in 50 in-patients with stroke. The SADQ had a poor correlation with both the HADS depression (r = 0.12) and anxiety (r = 0.30) subscales, and an adequate correlation with the Wakefield Depression Inventory (r = 0.31). The SADQ-10 had an adequate correlation with the HADS anxiety subscale (r = 0.35) and a poor correlation with the HADS depression subscale (r = 0.05). The SADQ-10 had a poor correlation with the Wakefield Depression Inventory (r = 0.29).

This study by Lincoln et al. (2000) also examined the convergent validity of the SADQ-H with the Wakefield Depression Inventory (Snaith et al., 1971) in 30 in-patients with stroke and found that the measures were more highly correlated with the SADQ-H version than with the SADQ or SADQ-10 (r = 0.58).

Leeds, Meara, and Hobson (2002) examined the convergent validity of the SADQ-10 with the Geriatric Depression Scale-15 (GDS-15 – Sheik & Yesavage, 1986) in 65 non-aphasic patients with stroke undergoing rehabilitation. Bivariate analysis revealed an adequate correlation between the SADQ-10 and the GDS-15 (r = 0.40).

Known groups:
Not yet examined.

Sensitivity and Specificity:
Leeds et al. (2002) administered the SADQ-10 to 65 non-aphasic clients with stroke undergoing rehabilitation. Clients self-completed the Geriatric Depression Scale-15 (GDS-15 – Sheik & Yesavage, 1986) and a research nurse completed the SADQ-10 for each client. Using a cut-off point of ≥5 on the Geriatric Depression Scale-15 as the criterion for depression, a cut-off score of 14 out of 30 on the SADQ-10 produced a sensitivity of 70% and a specificity of 77% to detect depression.

Bennett et al. (2006) conducted receiver operating characteristic (ROC) curve analyses on the data from 100 patients with stroke to establish the appropriate cut-off scores for the SADQ-H and SADQ-H 10 in relation to depression and anxiety as identified on the Hospital Anxiety and Depression Scale (HADS – Zigmond & Snaith, 1983). The criterion for low mood was a score greater than 7 on the depression or anxiety subscales of the HADS. This criterion was based on cut-offs for mild mood disturbance found in a previous study in patients with stroke (O’Rourke MacHale, Signorini, & Dennis, 1998). Cut-off scores for depression were identified as an optimum cut-off of 17/18 on the SADQ-H (sensitivity of 1.00; specificity of 0.81), and an optimum cut-off of 5/6 on the SADQ-H 10 (sensitivity of 1.00; specificity of 0.78). Cut-off scores for anxiety were identified as an optimum cut-off of 9/10 on the SADQ-H (sensitivity of 0.75; specificity of 0.40) and an optimum cut-off of 4/5 on the SADQ-H 10 (sensitivity of 0.75; specificity of 0.50).
Note: The cut-off values indicate a lower figure that is the highest likely to be obtained by people without low mood and a higher figure that is the lowest likely to be obtained by people with low mood.

Responsiveness

Not examined.

References

  • Benaim, C., Cailly, B., Perennou, D., Pelissier, J. (2004). Validation of the aphasic depression rating scale. Stroke, 35, 1692.
  • Bennett, H. E., Thomas, S. A., Austen, R., Morris, A. M. S., & Lincoln, N. B. (2006). Validation of screening measures for assessing mood in stroke patients. British Journal of Clinical Psychology, 45, 367-376.
  • Leeds, L., Meara, R. J., & Hobson, J. P. (2004). The utility of the Stroke Aphasia Depression Questionnaire (SADQ) in a stroke rehabilitation unit. Clinical Rehabilitation, 18, 228-231.
  • Lincoln, N. B., Sutcliffe, L. M., & Unsworth, G. (2000). Validation of the Stroke Aphasic Depression Questionnaire (SADQ) for use with patients in hospital. Clinical Neuropsychological Assessment, 1, 88-96.
  • O’Rourke, S., MacHale, S., Signorini, D., & Dennis, M. (1998). Detecting psychiatric morbidity after stroke: Comparison of the GHQ and the HAD scale. Stroke, 29, 980-985.
  • Sheik, J. A., Yesavage, J. A. (1986). Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version. Clin Gerontol, 37, 819-820.
  • Snaith R.P., Ahmed S.M., Mehta S., Hamilton M. (1971). Assessment of the severity of primary depressive illness: the Wakefield Self Assessment Depression Inventory. Psychol Med, 1, 143-149.
  • Sutcliffe, L. M., Lincoln, N. B. (1998). The assessment of depression in aphasic stroke patients: The development of the Stroke Aphasic Depression Questionnaire. Clin Rehabil, 12, 506-513.
  • Zigmond, A. S., & Snaith, R. P. (1983). Hospital Anxiety and Depression Scale. Acta Psychiatrica Scandinavica, 67, 361-370.

See the measure

How to obtain the SADQ?

The SADQ is available free with permission from the authors and can be obtained from the following website: http://www.nottingham.ac.uk/medicine/about/rehabilitationageing/publishedassessments.aspx

Table of contents
Survey