Depression occurs in one third of patients with stroke 1,2

Transcription

1 Depression Screening in Stroke A Comparison of Alternative Measures With the Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (Major Depressive Episode) as Criterion Standard Alyna Turner, PhD; John Hambridge, Dip Clin Psych; Jennifer White, M Health Science (OT); Gregory Carter, FRANZCP, PhD; Kerrie Clover, PhD; Louise Nelson, BSc(Hons); Maree Hackett, PhD Background and Purpose Screening tools for depression and psychological distress commonly used in medical settings have not been well validated in stroke populations. We aimed to determine the accuracy of common screening tools for depression or distress in detecting caseness for a major depressive episode compared with a clinician-administered structured clinical interview for Diagnostic and Statistical Manual of Mental Disorders Fourth Edition as the gold standard. Methods Seventy-two participants 3 weeks poststroke underwent a diagnostic interview for major depressive episode and completed the Patient Health Questionnaire-2 and -9, Hospital Anxiety and Depression Scale, Beck Depression Inventory-II, Distress Thermometer, and Kessler-10. Internal consistency, sensitivity, specificity, likelihood ratios, and posttest probabilities were calculated. Each measure was validated against the gold standard using receiver operating characteristic curves with comparison of the area under the curve for all measures. Results Internal consistency ranged from acceptable to excellent for all measures (Cronbach ). Areas under the curve for the Patient Health Questionnaire-2, Patient Health Questionnaire-9, Hospital Anxiety and Depression Scale depression and total score, Beck Depression Inventory-II, and Kessler-10 ranged from 0.80 ( ) for the Kessler-10 to 0.89 ( ) for the Beck Depression Inventory-II with no significant differences between measures. The Distress Thermometer had an area under the curve of 0.73 ( ), significantly smaller than the Beck Depression Inventory-II (P 0.05). Conclusions Apart from the Distress Thermometer, selected scales performed adequately in a stroke population with no significant difference between measures. The Patient Health Questionnaire-2 would be the most useful single screen given free availability and the shortest number of items. (Stroke. 2012;43: ) Key Words: assessment of depression depression distress sensitivity/specificity stroke Depression occurs in one third of patients with stroke 1,2 and is associated with increased mortality, 3 poorer functional outcome, 4 and increased caregiver distress. 5 Stroke management guidelines recommend routine screening for depression. 6 9 Because guidelines also recommend screening for other forms of psychological distress such as anxiety, 9 a general distress measure may be useful provided depressive disorders are detected. Many depression and distress screening tools commonly used in medical settings have not been well validated in stroke populations. The Patient Heath Questionnaire-9 (PHQ-9), 10 Hospital Anxiety and Depression Scale (HADS), 11 Beck Depression Inventory II (BDI-II), 12 Distress Thermometer (DT) 13 and Kessler-10 (K-10) 14 are depression or distress measures commonly used in primary and/or specialist care settings. The PHQ-9 has been recommended for use in stroke 8,9 due to its brevity and strong psychometric properties. Along with the 2-item PHQ-2, it has been evaluated in patients with nonaphasic stroke with promising results. 15 Similarly, the HADS depression (HADS-D) and anxiety (HADS-A) subscales and total score (HADS-Total) have been validated for use with patients with stroke without communication problems Several other depression measures commonly used in stroke practice (eg, Geriatric Depression Scale) also have evidence supporting their use. 21 Evidence for diagnostic accuracy of the BDI-II, DT, and K-10 is lacking in stroke populations. Diagnostic accuracy of the original BDI has been evaluated, 18,22 24 but not the Received October 27, 2011; final revision received December 6, 2011; accepted December 22, From the Heart Research Centre (A.T.), Melbourne, Victoria, Australia; the School of Medical Practice and Population Health (A.T., G.C., K.C., L.N.), University of Newcastle, New South Wales, Australia; Hunter New England Local Health Network (J.H., J.W.), Newcastle, New South Wales, Australia; Newcastle Calvary Mater Hospital (G.C., K.C.), Newcastle, New South Wales, Australia; and The George Institute for Global Health (M.H.), Sydney, New South Wales, Australia. Correspondence to Alyna Turner, PhD, Senior Research Fellow, Heart Research Centre, Box 2137 Post Office, The Royal Melbourne Hospital, Victoria 3050, Australia. alyna.turner@heartresearchcentre.org 2012 American Heart Association, Inc. Stroke is available at DOI: /STROKEAHA

2 Turner et al Depression Screening in Stroke 1001 BDI-II, 12 although a factor analytic study in a neurorehabilitation population indicated the measure provides a meaningful severity score of depression in patients with stroke. 25 The DT, a single-item distress measure commonly used in oncology, 13,26 is being used in a UK-based stroke service enhancement project. 27 The K-10, a distress screening scale used in several population-based surveys, 28,29 is being used by Australian general practitioners within a national government rebated mental healthcare plan. 30 Validation studies of the PHQ-9, PHQ-2, and HADS have several methodological limitations. Recruitment has occurred between 2 weeks and 6 months poststroke, whereas stroke guidelines recommend depression screening beyond the 6-month period. Furthermore, in the PHQ-9 validation study, 15 a clinician interview was only administered to patients endorsing symptoms of depression on the PHQ-9, and the study population was limited to patients 1 to 2 months after ischemic stroke with minimal aphasia or cognitive impairment. These findings therefore need to be replicated, preferably with a more heterogeneous group of patients with stroke. Stroke clinicians might manage patients for many years postevent and screening measures should be validated across all time points in the recovery pathway. The aim of the present study was to examine performance of these commonly used depression-specific or distress measures with regard to internal consistency, concurrent validity, and diagnostic accuracy in detecting a major depressive episode (MDE) compared with a clinician-administered structured diagnostic interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition as the gold standard. A heterogeneous sample of patients with stroke representing those seen in postacute inpatient and outpatient stroke services was examined. Methods Eligible participants had a confirmed stroke 3 weeks prior, were aged 18 years, and able to attend a local health site for assessment in the Hunter region of New South Wales, Australia. Exclusion criteria included inability to read or understand English or severe cognitive or physical impairment precluding participation. Patients with expressive dysphasia or dysarthria with adequate receptive and expressive communication strategies were included. Ethics approval was gained from the Hunter New England Human Research Ethics Committee (No. 08/08/20/5.02). Outpatients were recruited through the Community Stroke Team, Hunter Medical Research Institute Research Register, stroke clubs, and hospital discharge lists. Consecutive rehabilitation unit inpatients were also approached. Consenting participants were interviewed using the Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition MDE module by a trained clinical psychologist at the hospital or health site. Participants self-completed the HADS and BDI-II using pencil and paper and PHQ-9, DT, and K-10 by computer touchscreen. Outpatients completed their assessments during 1 appointment; inpatients had the option of having a break if fatigued. Measures The Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition MDE module was used as the gold standard (criterion standard). 31 A depression case was defined as a positive score on the Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition diagnosis of MDE. Selected questionnaires were administered for the depression-specific ( Depression : PHQ-2, PHQ-9, HADS-D, BDI-II) and multifactorial/general distress ( Distress : DT, K-10, and HADS-Total) measures of symptom burden. The 9-item PHQ-9 10 is based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition diagnostic symptoms of MDE. Responses are rated 0 to 3 in reference to the past fortnight (total score range 0 27) with higher scores indicating greater severity. The PHQ-2 is the sum of the first 2 items (range, 0 6). Previously recommended cutoffs for the PHQ-9 and PHQ-2 are 9 10 and 2, 32 respectively. The PHQ-9 also provides a preliminary diagnosis of major depressive syndrome using a scoring algorithm ( 5 items, including Items 1 and/or 2, are rated 2 unless suicidality item endorsed, which can be rated 1). The 14-item HADS consists of 2 7-item subscales assessing depression and anxiety. The HADS was developed for medical settings and omits somatic symptoms of depression. 11 Responses are rated 0 to 3 in reference to the past week (subscale scores range 0 21) with a higher score indicating greater severity. The HADS total score (HADS-Total) is scored by adding the HADS-A and HADS-D subscales (range, 0 42). Clinical cutoffs have been provided for HADS-D ( 7), 11 whereas past studies suggest an ideal cutoff for the HADS-Total of 11 in patients with stroke. 17,18 The 21-item BDI-II was developed to detect changes in depressive symptoms in mental healthcare settings. 12 Responses are on a scale of 0 to 3 in reference to the past fortnight (total score range 0 63) with higher scores indicating greater severity. A clinical cutoff of 13 has been suggested. 12 The single-item DT is a visual analog scale recommended for use in oncology services. 13 It was found to be comparable to the HADS in detecting distress in patients with cancer. 24 Response options range from 0 ( no distress ) to 10 ( extreme distress ) in reference to the past week. The recommended cutoff for clinically significant emotional distress in oncology patients is The 10-item K-10 is a measure of nonspecific psychological distress 14 designed as a brief screen for serious mental illness. 29 It discriminates well between people with a diagnosed mental health disorder (anxiety, mood or nonaffective psychosis) and all other community respondents. 29 Responses are on a scale of 1 to 5 in reference to the past month (total score range 10 50) with higher scores indicating greater levels of distress. The recommended cutoff for clinically significant psychological distress is Statistical Analysis Group differences were examined using independent-samples t tests or 2 tests. Missing data were replaced by the rounded mean item value if 20% of items were missing. Internal consistency was assessed with Cronbach. Concurrent validity was examined as Pearson product-moment correlations between the continuous scale scores. For each scale, receiver operating characteristic curves were constructed and area under the curve (AUC) and CIs (binomial exact) calculated. Each receiver operating characteristic curve AUC was compared with chance (AUC 0.5) and with the receiver operating characteristic curves of the other scales with the method of DeLong et al 34 used to calculate SE and results expressed as a z-statistic. The sensitivity, specificity, likelihood ratio positive and negative, and posttest probability positive and negative were calculated for established and ideal cut points. PASW Statistics Version (SPSS Inc, an IBM Company, Chicago, IL), MedCalc Version (MedCalc Software, Mariakerke, Belgium), and Microsoft Excel 2007 were used. Results Recruitment strategies identified 124 potential participants. Nonparticipation reasons included refusal (n 28), distance (n 5), illness or cognitive impairment (n 5), uncontactable/ missed (n 10), and unable to read (n 3). One patient s data were lost due to technical issues. The final sample (n 72) were aged 25 to 91 years (mean, 66.7 years; SD, 13.1), 38

3 1002 Stroke April 2012 Table 1. Test Score Performance, Internal Consistency, and Concurrent Validity Scale Depression Distress No.* PHQ-2 PHQ-9 HADS-D BDI-II DT K-10 HADS-Total Total scores Mean (SD) (1.87) 7.24 (6.00) 5.26 (3.92) (12.09) 3.80 (2.97) (7.88) (7.72) MDE case Mean (SD) (2.17) (6.75) (4.60) (15.60) 5.85 (3.02) (10.40) (7.66) MDE noncase Mean (SD) (1.44) 5.90 (4.95) 4.22 (2.87) (8.11) 3.34 (2.78) (5.90) 9.76 (6.40) df t Internal consistency Concurrent validity 72 PHQ HADS-D BDI-II DT K HADS-Total PHQ indicates Patient Health Questionnaire; HADS-D, Hospital Anxiety and Depression Scale-Depression; BDI, Beck Depression Inventory; DT, Distress Thermometer; K-10, Kessler-10; MDE, major depressive episode. *n 71 for DT (n 58 for DT noncase). P P Cronbach. Pearson product-moment correlations. (52.8%) were male, and 61 (84.4%) were Australian-born. Twenty (28%) were inpatients and 52 (72%) outpatients. Fifty-one (70.8%) participants had experienced 1 stroke with 17 (23.6%) having 2 to 3 strokes and 4 (5.6%) having 4 to 6 strokes. Time since stroke was 3 weeks to 540 months (median, 14 months; interquartile range, 3 36 months). Twenty-three (31.9%) were 3 weeks to 6 months poststroke; 21 (29.2%) had their stroke 6 months to 24 months prior, and 28 (38.9%) were 24 months poststroke. Twenty-one (29.2%) were currently taking antidepressant medication and 32 (44.5%) had a caregiver in the last 6 months. Thirteen patients (18%) were defined as MDE cases based on the Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition. For all scales, the score was significantly higher for cases than noncases (P 0.005; Table 1). Cronbach exceeded 0.70 for all scales indicating acceptable internal consistency. Strong correlations (P 0.001) occurred between measures (Table 1). AUC was 0.80 for all measures apart from the DT (AUC 0.73; Figures 1 and 2). Z-statistics indicated all AUCs were significantly 0.50 (P 0.005); that is, all scales detected MDE cases at greater than chance level. Pairwise comparisons of the AUC for all measures showed a significant difference between the BDI-II and the DT only (z 2.16, P 0.05). The 2 ideal cutoff options displayed in Table 2 are those meeting criteria for a good screening tool (sensitivity 0.80 and specificity 0.60) 21 and those with highest sum of sensitivity and specificity. The PHQ-9 categorical diagnosis of major depressive syndrome derived from the diagnostic algorithm had high specificity but poor sensitivity. The recommended stroke cutoff of 10 for HADS-Total had high sensitivity and adequate specificity; however, a higher cutoff of 14 increased specificity and had the highest sum of sensitivity and specificity. For the HADS-D and BDI-II, the ideal cutoff on both criteria were 5 and 11, respectively, lower than previously recommended. Similarly, the PHQ-9 ideal cutoffs were lower than the recommended cutoff with scores 6 meeting criteria for a good screening tool and scores 8 providing the highest sum of sensitivity and specificity. At no point was sensitivity 0.80 and specificity 0.60 reached concurrently for the PHQ-2, DT, and K-10. PHQ-2 scores 1 and K-10 scores 17 almost met these criteria. Highest sum of sensitivity and specificity was found at higher than previously recommended cutoffs for the PHQ-2 ( 3) and K-10 ( 25). For the DT, the highest sum of sensitivity and specificity was found at 1; however, this had poor sensitivity. The BDI-II ( 11) had the highest overall sum of sensitivity and specificity. Discussion We examined the performance of several depression and distress screening tools in a stroke population. Overall, this study of a heterogeneous group of patients with stroke able to complete a self-report measure found that, apart from the DT, selected measures performed comparably. Performance differences between longer and shorter tools were not of sufficient size to justify length of time lost in administration (opportunity cost). Study findings support previous literature indicating the PHQ-2, PHQ-9, and HADS are appropriate for use in patients with stroke without aphasia Unlike previous studies of

4 Turner et al Depression Screening in Stroke 1003 Figure 1. ROC curves for depression measures versus SCID cases of MDE. Area under the curve : PHQ ( ), z 4.76; PHQ ( ), z 4.93; HADS- D 0.87 ( ), z 7.27; BDI-II 0.89 ( ), z P for all. ROC indicates receiver operating characteristic; SCID, Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; MDE, major depressive episode; PHQ, Patient Health Questionnaire; HADS-D, Hospital Anxiety and Depression Scale-Depression; BDI, Beck Depression Inventory. these measures, we have included participants who varied in time since stroke and number of strokes, representing patients seen across a range of inpatient and outpatient services. Findings are therefore relevant for clinicians who see patients with stroke at both early and late stages of the recovery pathway. Specificity of the PHQ-2 was equivalent to that previously seen in the early postacute period. 15 For the PHQ-9, specificity was lower but acceptable. However, previously recommended PHQ-2 and PHQ-9 cutoffs resulted in lower sensitivity than ideal. Supporting previous literature, 16,17 ideal cut points seen for the HADS-D were lower than traditionally recommended cut points. This study provides diagnostic accuracy information for depression measures not previously evaluated in a stroke population. The AUC of the BDI-II (0.89) was the highest of all measures examined and was within the range seen for the BDI at different time points over the first 18 months poststroke (AUCs ). 22 The K-10 performed equivalently to the depression-specific measures and correlates strongly with the PHQ-2, PHQ-9, and the BDI-II. However, the DT demonstrated the lowest AUC of all measures. It did not demonstrate a sensitivity specificity combination that met with guidelines for a good depression screening tool 21 and had the weakest correlations with the other scales. Although factor analysis of the K-10 has identified a depression specific factor, 35 the DT addresses global distress in a single item, potentially capturing a range of nondepressive states of emotional distress (reducing sensitivity). Alternatively, the item s presentation format may have impacted the results: patients with stroke have been found to be less likely Figure 2. ROC curves for distress measures versus SCID cases of MDE. Area under the curve : DT 0.73 ( ), z 2.99; K ( ), z 3.94; HADS-Total 0.85 ( ), z P for all. ROC indicates receiver operating characteristic; SCID, Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; MDE, major depressive episode; DT, Distress Thermometer; K-10, Kessler-10; HADS, Hospital Anxiety and Depression Scale. to correctly complete visual analog scales than healthy control subjects, 36 and visual analog mood scales have been found to lack sensitivity when used with patients with stroke. 22 Further work is required to determine the role for the DT in stroke care. Study strengths included a sample that covered a long period after stroke; inclusion of patients with multiple strokes; use of a strong criterion standard; examination of multiple instruments contemporaneously; and use of an appropriate range of accuracy statistics and receiver operating characteristic comparisons. Study limitations included small sample size (leading to low numbers of depression cases) resulting in low precision of the accuracy statistics; no formal assessment of cognitive and functional status; nonrandomization of test order potentially resulting in test fatigue; and having the same assessor administer the Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition and the screening questionnaires, potentially contributing to observed agreement. A higher rate of major depression was found in our sample (18.1%) than seen in previous studies (12.5% 15.8%) It may be that those with depression symptoms were more likely to volunteer. Comparison with published data 37 also suggests the present sample may have participants of a younger age and a higher ratio of patients having multiple strokes than seen in the region s stroke population. It may not be possible to generalize our results to other populations and replication studies are needed. Several clinical implications arise from our results. Findings support recommendations advocating use of the PHQ-2, PHQ-9, and HADS in patients with stroke able to complete a

5 1004 Stroke April 2012 Table 2. Sensitivity, Specificity, Likelihood Ratios, and Posttest Probabilities at Recommended and Ideal Cut Points Test Test / Negative/ Likelihood Ratios Posttest Probabilities Categorical SCID SCID Negative Sensitivity Specificity Negative PHQ-9 diagnosis 4/13 54/ ( ) 0.92 ( ) 3.63 ( ) 0.76 ( ) 0.44 ( ) 0.86 ( ) Depression Cut point scales PHQ /13 37/ ( ) 0.63 ( ) 2.06 ( ) 0.37 ( ) 0.31 ( ) 0.93 ( ) 2* 9/13 49/ ( ) 0.83 ( ) 4.08 ( ) 0.37 ( ) 0.47 ( ) 0.93 ( ) 3 8/13 54/ ( ) 0.92 ( ) 7.26 ( ) 0.42 ( ) 0.62 ( ) 0.92 ( ) PHQ /13 37/ ( ) 0.63 ( ) 2.27 ( ) 0.25 ( ) 0.33 ( ) 0.95 ( ) 8 10/13 44/ ( ) 0.75 ( ) 3.03 ( ) 0.31 ( ) 0.40 ( ) 0.94 ( ) 9* 9/13 46/ ( ) 0.78 ( ) 3.14 ( ) 0.39 ( ) 0.41 ( ) 0.92 ( ) HADS-D 5 12/13 40/ ( ) 0.68 ( ) 2.87 ( ) 0.11 ( ) 0.39 ( ) 0.98 ( ) 7* 8/13 49/ ( ) 0.83 ( ) 3.63 ( ) 0.46 ( ) 0.44 ( ) 0.91 ( ) BDI-II 11 12/13 42/ ( ) 0.71 ( ) 3.20 ( ) 0.11 ( ) 0.41 ( ) 0.98 ( ) 13* 11/13 44/ ( ) 0.75 ( ) 3.33 ( ) 0.21 ( ) 0.42 ( ) 0.96 ( ) Distress scales DT 1 13/13 19/ ( ) 0.33 ( ) 1.49 ( ) ( ) 1.00 ( ) 3* 9/13 33/ ( ) 0.57 ( ) 1.61 ( ) 0.54 ( ) 0.27 ( ) 0.89 ( ) K /13 35/ ( ) 0.59 ( ) 2.08 ( ) 0.26 ( ) 0.31 ( ) 0.95 ( ) 19* 10/13 41/ ( ) 0.69 ( ) 2.52 ( ) 0.33 ( ) 0.36 ( ) 0.93 ( ) 25 7/13 56/ ( ) 0.95 ( ) ( ) 0.49 ( ) 0.70 ( ) 0.90 ( ) HADS-Total 10* 12/13 37/ ( ) 0.63 ( ) 2.48 ( ) 0.12 ( ) 0.35 ( ) 0.97 ( ) 14 11/13 44/ ( ) 0.75 ( ) 3.33 ( ) 0.21 ( ) 0.42 ( ) 0.96 ( ) SCID indicates Structured Diagnostic Interview for the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; PHQ, Patient Health Questionnaire; HADS-D, Hospital Anxiety and Depression Scale-Depression; BDI, Beck Depression Inventory; DT, Distress Thermometer; K-10, Kessler-10. *Previously recommended cutoff. Cutoff best meeting criteria for a good screening tool 21 (none available for DT). Highest sum of specificity and sensitivity. screening tool. 8,9,21,38 However, the PHQ-9 diagnostic algorithm (providing a categorical diagnosis of major depressive syndrome) had high specificity but unacceptably low sensitivity and cannot be recommended. In a stroke population, it can be argued that screening should optimize detection of major depression cases by using assessments with maximum sensitivity because the clearly outlined negative effects of depression outweigh the consequences of conducting further assessments on patients with no depression. The 21-item BDI-II demonstrated best performance with regard to AUC and sensitivity and specificity at selected cut points; however, it was not significantly superior to the other measures except for the DT. These relatively small differences would not outweigh the time-saving benefits of using shorter measures in routine screening. The performance of the K-10 indicated that, although described as a psychological distress measure, it has good accuracy as a MDE screener in this population. This is important because many patients with stroke in Australia will be managed by general practitioners who are increasingly using this scale in clinical practice. Although results support potential use of these measures for depression screening and case-finding in stroke services, future research should use larger and representative stroke samples to enable subgroup analyses (eg, for age, gender, disability, and ethnicity) and to confirm ideal cut points. Further work is required to determine the accuracy of screening instruments to detect important related constructs in patients with stroke, for example, anxiety and global distress. Appropriate measures for patients with communication and cognitive impairments should be developed and validated. Finally, the impact of routine depression screening and feedback on clinician behavior and patient outcomes using appropriate screening measures needs further examination in intervention studies. Acknowledgments We thank Megan Alston and the Community Stroke Team; Dr Michael Pollack and the Hunter Stroke Service; and the Hunter Medical Research Institute Research Register for assistance with participant recruitment; and Dr Alison Beck for data collection. Sources of Funding Funding provided by a bequest from Jennie Thomas through Hunter Medical Research Institute. None. Disclosures

6 Turner et al Depression Screening in Stroke 1005 References 1. Hackett ML, Yapa C, Parag V, Anderson CS. Frequency of depression after stroke: a systematic review of observational studies. Stroke. 2005; 36: Ayerbe L, Ayis S, Rudd AG, Heuschmann PU, Wolfe CD. Natural history, predictors, and associations of depression 5 years after stroke: the South London Stroke Register. Stroke. 2011;42: Everson SA, Roberts RE, Goldberg DE, Kaplan GA. Depressive symptoms and increased risk of stroke mortality over a 29-year period. Arch Intern Med. 1998;158: Schmid AA, Kroenke K, Hendrie HC, Bakas T, Sutherland JM, Williams LS. Poststroke depression and treatment effects on functional outcomes. Neurology. 2011;76: Cameron JI, Cheung AM, Streiner DL, Coyte PC, Stewart DE. Stroke survivor depressive symptoms are associated with family caregiver depression during the first 2 years poststroke. Stroke. 2011;42: Duncan PW, Zorowitz R, Bates B, Choi JY, Glasberg JJ, Graham GD, et al. Management of adult stroke rehabilitation care: a clinical practice guideline. Stroke. 2005;36:e National Stroke Foundation. Clinical Guidelines for Stroke Management. Melbourne, Australia: National Stroke Foundation; Miller EL, Murray L, Richards L, Zorowitz RD, Bakas T, Clark P, Billinger SA. Comprehensive overview of nursing and interdisciplinary rehabilitation care of the stroke patient: a scientific statement from the American Heart Association. Stroke. 2010;41: Intercollegiate Stroke Working Party. National Clinical Guidelines for Stroke, III Edition. London: Royal College of Physicians; Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16: Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand. 1983;67: Beck A, Steer R, Brown G. Manual for the Beck Depression Inventory-II. San Antonio, TX: Psychological Corporation; Holland JC, Andersen B, Breitbart WS, Compas B, Dudley MM, Fleishman S, et al. NCCN Clinical Practice Guidelines in Oncology: Distress Management. Version ; Available at: Accessed July 4, Kessler RC, Barker PR, Colpe LJ, Epstein JF, Gfroerer JC, Hiripi E, et al. Screening for serious mental illness in the general population. Arch Gen Psychiatry. 2003;60: Williams LS, Brizendine EJ, Plue L, Bakas T, Tu W, Hendrie H, et al. Performance of the PHQ-9 as a screening tool for depression after stroke. Stroke. 2005;36: O Rourke S, MacHale S, Signorini D, Dennis M. Detecting psychiatric morbidity after stroke: comparison of the GHQ and the HAD scale. Stroke. 1998;29: Sagen U, Vik TG, Moum T, Morland T, Finset A, Dammen T. Screening for anxiety and depression after stroke: comparison of the Hospital Anxiety and Depression Scale and the Montgomery and Asberg Depression Rating Scale. J Psychosom Res. 2009;67: Aben I, Verhey F, Lousberg R, Lodder J, Honig A. Validity of the Beck Depression Inventory, Hospital Anxiety and Depression scale, SCL-90, and Hamilton Depression Rating Scale as screening instruments for depression in stroke patients. Psychosomatics. 2002;43: Healey AK, Kneebone II, Carroll M, Anderson SJ. A preliminary investigation of the reliability and validity of the brief assessment schedule depression cards and the Beck Depression Inventory-Fast Screen to screen for depression in older stroke survivors. Int J Geriatr Psychiatry. 2008;23: Johnson G, Burvill PW, Anderson CS, Jamrozik K, Stewart-Wynne EG, Chakera TM. Screening instruments for depression and anxiety following stroke: experience in the Perth community stroke study. Acta Psychiatr Scand. 1995;91: Bennett HE, Lincoln NB. Potential screening measure for depression and anxiety after stroke. Int J Ther Rehabil. 2006;13: Berg A, Lonnqvist J, Palomaki H, Kaste M. Assessment of depression after stroke: a comparison of different screening instruments. Stroke. 2009;40: House A, Dennis M, Hawton K, Warlow C. Methods of identifying mood disorders in stroke patients: experience in the Oxfordshire Community Stroke Project. Age Ageing. 1989;18: Lincoln NB, Nicholl CR, Flannaghan T, Leonard M, Van der Gucht E. The validity of questionnaire measures for assessing depression after stroke. Clin Rehabil. 2003;17: Siegert RJ, Walkey FH, Turner-Stokes L. An examination of the factor structure of the Beck Depression Inventory-II in a neurorehabilitation inpatient sample. J Int Neuropsychol Soc. 2009;15: Mitchell AJ. Short screening tools for cancer-related distress: a review and diagnostic validity meta-analysis. J Natl Compr Canc Netw. 2010;8: Williams L, Thomson A, Lowden L. MCN for stroke distress management system. Int J Stroke. 2010;5(suppl 3): Australian Bureau of Statistics. National Survey of Mental Health and Wellbeing: Summary of Results. Cat no ; Australian Bureau of Statistics: Canberra, Australia. 29. Kessler RC, Andrews G, Colpe LJ, Hiripi E, Mroczek DK, Normand SL, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med. 2002;32: Pirkis J, Harris M, Hall W, Ftanou M. Evaluation of the better access to psychiatrists, psychologists and general practitioners through the Medicare Benefits Schedule initiative: Summative evaluation Available at: E31B4C0752BAE99FCA25784C0078BFCD/$File/sum.pdf. Accessed July 4, First MB, Spitzer RL, Gibbon Miriam, Williams JBW. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. New York: Biometrics Research, New York State Psychiatric Institute; Kroenke K, Spitzer RL, Williams JB. The Patient Health Questionnaire-2: validity of a two-item depression screener. Med Care. 2003;41: Australian Bureau of Statistics. Information paper: Use of the Kessler Psychological Distress Scale in ABS Health Surveys, Australia, Cat no ; Australian Bureau of Statistics: Canberra, Australia. 34. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44: Mackinnon A, Jorm AF, Hickie IB. A national depression index for Australia. Med J Aust. 2004;181:S Price CIM, Curless RH, Rodgers H. Can stroke patients use visual analogue scales? Stroke. 1999;30: Marsden DL, Spratt NJ, Walker R, Barker D, Attia J, Pollack M, et al. Trends in stroke attack rates and case fatality in the Hunter region, Australia Cerebrovasc Dis. 2010;30: Kneebone I; Baker J, O Malley H. Screening for depression after stroke: developing protocols for the occupational therapist. BJOT. 2010;73: