Skip Navigation



NDT Advance Access published online on March 6, 2008
Nephrology Dialysis Transplantation, doi:10.1093/ndt/gfn105
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/8/2653    most recent
gfn105v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Chilcot, J.
Right arrow Articles by Farrington, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chilcot, J.
Right arrow Articles by Farrington, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author [2008]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org


Screening for depression while patients dialyse: an evaluation

Joseph Chilcot1,2, David Wellsted2 and Ken Farrington1

1 Renal Unit, Lister Hospital Stevenage 2 Research Development Support Unit, University of Hertfordshire, UK

Correspondence and offprint requests to: J. Chilcot, Renal Unit Lister Hospital, Coreys Mill Lane, Stevenage, Hertfordshire, SG1 4AB, UK. Tel: +44-7988662216; E-mail: josephchilcot{at}nhs.net



   Abstract
Top
 Abstract
Introduction
 Methodology
 Materials
 Results
 Discussion
 References
 
Background. The lack of routine depression screening among the haemodialysis (HD) population may contribute to depression being under-recognised. While screening patients could be beneficial, the optimum screening procedure remains unclear. One method would be to screen HD patients while they receive their treatment. The purpose of this investigation was to determine whether the Beck Depression Inventory-II (BDI) could be administered while patients dialysed.

Methods. Forty HD patients completed the BDI while dialysing and again at a time when off-dialysis. Level of agreement analysis (Bland and Altman) was undertaken to determine if the assessment condition influenced BDI scoring. The off-dialysis assessment also involved a short clinical interview that was compared with the BDI assessment.

Results. There was a high level of agreement between the on- and off-dialysis assessments, but differences in response to the somatic items on the BDI scale were apparent between the conditions. The clinical interview revealed that 22% of the sample met the DSM-IV criteria for major depressive disorder. The optimal cut-off value for the BDI as determined by receiver operating characteristic curves was ≥16, with 88.9% sensitivity and 87.1% specificity.

Conclusion. The results indicate that the procedure of on-dialysis assessment using the BDI is a viable screening procedure. The practicality of employing this screening procedure may facilitate improved detection of depression in the dialysis population.

Keywords: agreement; beck depression inventory (BDI-II); depression; diagnosis; dialyse



   Introduction
Top
 Abstract
 Introduction
Methodology
 Materials
 Results
 Discussion
 References
 
Depression is a common psychopathological condition among renal dialysis patients [1–4], yet its detection and management are usually not recognised as being part of the routine care of this population [5]. An expanding literature base has shown the negative impact of depression upon patients with chronic illness. There are associations between depression and mortality among the end-stage renal disease (ESRD) population [1,6–8], and depression is a risk factor for withdrawal from haemodialysis (HD), even after controlling for other clinical factors and age [9]. While the effects of depression may further complicate the clinical course of individuals on dialysis, its prevalence remains contentious. The best estimates of prevalence lie between 20 and 30% [10,11], but the reported variation is vast partly due to the type of assessment tool employed [12,13]. There is a difference between clinical interviews that can be used to diagnose a depressive disorder, and screening tools that are used to assess the severity of depressive symptoms. Typically, estimates based upon some defined screening cut-off score lead to higher estimates than diagnostic schemes.

Furthermore, the estimates obtained by current screening and diagnostic tools are complicated by criterion contamination, that is, they include enquiry about a varied amount of somatic symptomatology which overlaps with physical symptoms directly attributable to ESRD.

While the issues surrounding prevalence and criterion contamination are of empirical interest, the procedure of assessment has received little attention. Screening the HD population may be a practical method of assessment compared to professional psychiatric evaluation [1,12,14], which is costly and ineffective when applied to the whole clinical population. On-dialysis screening may be a convenient and practical method of assessment, promoting regular evaluation. Drayer et al. [8] assessed depression (via clinical interviews and depression screening) while HD patients were dialysing. This procedure was adopted in an attempt to minimize the impact of fluctuating uraemic symptoms on mood and cognition [8,15] that would inevitably be present in patients during the interdialytic period. However, few studies have explicitly employed on-dialysis assessment [16] and little is known about the legitimacy of such a procedure. Does the context of actively dialysing influence depression screening? Our aim was to investigate the level of agreement between depression screening scores (BDI-II) obtained on-dialysis with scores obtained off-dialysis. We were particularly interested in understanding if on-dialysis assessment influenced somatic scoring on the BDI. Additionally, the accuracy of the BDI was also obtained in relation to a diagnostic standard for MDD.



   Methodology
Top
 Abstract
 Introduction
 Methodology
Materials
 Results
 Discussion
 References
 
Study population
HD patients from the renal service of the East and North Hertfordshire NHS Trust were approached, excluding patients with dementia or a history of major psychiatric disorders other than depression. Adult ESRD patients who had been receiving HD for >3 months were recruited. A standardised cognitive assessment was conducted [17] (The Mini Mental State Examination, MMSE), in which a score of <23 indicated a level of cognitive impairment, in which case patients were excluded. All patients were receiving high-flux haemodialysis or on-line haemodiafiltration three times weekly. None of the patients required hospitalisation between assessments.



   Materials
Top
 Abstract
 Introduction
 Methodology
 Materials
Results
 Discussion
 References
 
The severity of depression symptomatology was measured using the Beck Depression Inventory (BDI-II) [18]. The BDI-II consists of 21 items including a range of cognitive, affective and somatic symptoms elements. Each question is measured via a 4-point scale (0–3), that when summed range from 0 to 63. Typically a score of below 9 indicates a normal range, between 10 and 18 suggests mild depressive symptoms, between 19 and 29 moderate and 30 plus indicates severe depressive symptomatology. In a recent study of HD patients [10] the BDI was validated against a structured clinical interview for depression (SCID) suggesting a BDI score of ≥16 as an optimal cut-off for significant depression symptoms, revealing 91% sensitivity and 86% specificity. Eliminating the somatic elements of the BDI forms the Cognitive Depression Index (CDI) [19], which has been proposed as a better predictor of depression due to the reduction of confounding symptoms [19]. Sacks et al. [19] revealed significant associations between the CDI and BDI, revealing a standardised internal consistency of {alpha} = 0.74.

The Mini International Neuropsychiatry Interview [20] (MINI) was also administered, to assess whether patients were depressed following a diagnostic criteria (DSM-IV) for Major Depressive Disorder.

Design and procedure
The study employed a within subjects design, recruiting HD patients who satisfied the inclusion criteria. A research psychologist provided patients with an information sheet, and obtained consent. Patients were asked to complete the BDI and a demographic questionnaire during a stable dialysis session, commencing 30 minutes after initiation. The off-dialysis assessments were completed either before or after a treatment session, according to the patients’ preference. Patients were invited to the renal unit to complete the same questionnaires, with the addition of the interview component involving the MINI and MMSE. The research psychologist was trained to use the MINI by a consultant psychiatrist.

The on- and off-dialysis assessments took place on average 10.7 (±4.2) days apart, with the order of assessments counterbalanced to control for order effects. Assessment order was randomised, determined by a random number sequence, with the order allocated on the basis of the next available order as the patients were recruited. A local NHS research ethics committee approved the study protocol.

Analysis
The primary aim was to determine the extent to which the situation in which the patients were assessed (on-dialysis or not) influenced their responses. The default assumption was that the situation would have no discernable influence on the patients’ responses. A high level of agreement (LoA Bland and Altman [21]) was therefore expected between the two BDI assessment scores. A high level of agreement would indicate that the location of the assessment did not influence the patients’ responses to the assessment tool. The statistical methods were repeated in a sub-analysis involving the CDI, attaining the level of agreement over the two conditions to evaluate the relative importance of the somatic components of depression included in the BDI. ROC curves were used to determine suitable cut-offs for the BDI and CDI, measured against the diagnostic scheme. Chi-square analysis was employed to test associations between the diagnostic scheme, BDI and CDI. Jacknife re-sampling was undertaken to identify any observation that was having an undue influence on the statistical model.



   Results
Top
 Abstract
 Introduction
 Methodology
 Materials
 Results
Discussion
 References
 
Forty-three adult HD patients agreed to participate, corresponding to a 78% consent rate. Two patients dropped out of the study prior to follow-up, and one patient was excluded due to a significant counselling intervention being provided between assessments. Forty patients completed the study of which 24 were male and 16 were female with a mean age of 53.2 (±14.2) years. A summary of demographic and clinical data is presented in Table 1. Severe extra-renal co-morbidity [22] was prevalent in 20% of the sample with an additional 40% having intermediate co-morbidity. There were no significant gender differences with regards to age (P = 0.062), working status (P = 0.775), marital status (P = 0.328) and co-morbidity (P = 0.932).


View this table:
[in this window]
[in a new window]

  		Table 1 Demographic and clinical data for the sample, comparing depressed and non-depressed

 
On- versus off-dialysis BDI assessment
Table 2 shows the mean BDI and CDI scores on- and off-dialysis, for the whole sample and for subsets defined by the presence of depression based upon the diagnostic assessment. For depressed patients, the mean BDI and CDI scores did not differ significantly across the assessment conditions. However the non-depressed scored significantly higher on the somatic component of the BDI while on-dialysis, evident by significantly higher BDI scores but similar CDI scores. Across all patients, this effect was maintained (effect size = 0.41 and power = 0.80). The data was examined using a jacknife estimation, which demonstrated that no observations were having an undue influence on the model.


View this table:
[in this window]
[in a new window]

  		Table 2 Mean BDI and CDI scores for patients assessed on- and off-dialysis, defined by those who meet the diagnostic criteria for MDD using the MINI. The depressed patients scored consistently across the assessment conditions. The non-depressed varied in their somatic scoring, evident by greater on dialysis BDI scores (includes somatic items), but consistent CDI (somatic items removed) scoring across conditions. The difference between on- and off-dialysis scores was maintained when the sample was analysed collectively

 
The level of agreement for repeated use of an interval measure (e.g. the BDI) was assessed using Bland–Altman's level of agreement (LoA). LoA plots for the BDI and CDI are displayed in Figures 1 and 2. The mean difference between the BDI scores on- and off-dialysis was 1.8 (±4.29) indicating a bias for higher scores on dialysis, and the 95% confidence range of agreement was –6.8 to +10.4. Similarly, the mean difference between CDI scores was 0.95 (±3.26), biased to on-dialysis with a confidence range of –5.6 to +7.5. Both plots depict good levels of agreement across the two conditions.


Figure 1
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 The level of agreement between the BDI assessments made on- and off-dialysis. For each patient, the difference between their two BDI scores was plotted against their average score following Altman and Bland's method. Shown are the limits of agreement (mean ± 2SD) between which the vast majority of points lie.

 

Figure 2
View larger version (16K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 The level of agreement between the two CDI assessments across the dialysis conditions. Construction of the plot involved the same procedure as in Figure 1, excluding the somatic elements of the BDI to form the CDI.

 
Accuracy of the BDI and CDI
Of the 40 HD patients assessed, 22.5% met the criteria for MDD based upon the MINI. In order to assess the accuracy of the BDI in our sample of patients, off-dialysis scores were compared with the MINI. A ROC curve showed high predictive accuracy for the BDI (area under the curve 0.961, P = 0.001) versus the MINI. Optimal cut-off values were determined by the location in which the number of accumulating false positive values exceeded the number of false negatives. The optimal cut-off detected for the BDI was ≥16. Employing this cut-off revealed an 88.8% positive predictive value, and an 87.0% negative predictive value, with 88.9% sensitivity (95% CI 75.1 and 95.5) and 87.1% specificity (95% CI 72.9 and 94.4). Chi-square analysis revealed significant associations between the MINI (depressed/non-depressed) and BDI classification ≥16 (P = 0.001) (Table 3). The mean BDI score for patients who did not meet the MINI criteria for depression but had a BDI ≥16 (discrepant cases) was 17.7 (range = 3), compared to 25.7 (range = 20) for those meeting both the MINI and BDI criteria. This difference was significant (P < 0.05).


View this table:
[in this window]
[in a new window]

  		Table 3 Comparison of BDI and CDI diagnostic accuracy, compared with the diagnostic scheme (MINI). The data show that the BDI has greater positive and negative predictive value, compared with the CDI, although both compare well with the MINI

 
The analysis was replicated using the off-dialysis CDI assessment. The ROC curve again revealed high predictive accuracy for the CDI (area under curve 0.941, P = 0.001). The optimal cut-off detected for the CDI was ≥10. Employing a CDI ≥10 led to 77.8% sensitivity (95% CI 62.3 and 88.1) and 80.6% specificity (95% CI 65.4 and 90.1) with 77.7% positive predictive value, 80.6% negative predictive value. These findings suggest that the CDI, excluding somatic enquiry, has reduced sensitivity and specificity compared to the BDI. Further, as compared to the MINI, using a BDI ≥16 yielded four false positives and one false negative, whereas a CDI ≥10 had six false positives and two false negatives.

BDI and CDI classification on- and off-dialysis
After employing the determined cut-offs described above, 32.5% of patients had a BDI ≥16 when assessed on-dialysis, compared to 30.0% off-dialysis. The CDI assessment was consistent across the conditions, with 32.5% of patients scoring ≥10. Chi-square analysis confirmed that the associations between the two sets of BDI (P = 0.001) and CDI (P = 0.001) classifications were significant (Table 4).


View this table:
[in this window]
[in a new window]

  		Table 4 Comparison of patients classified as having depressive symptomology on- and off-dialysis.

 
Timing of the off-dialysis assessment
The off-dialysis BDI and MINI assessments were timed according to patient convenience. Twenty-three patients conducted their off-dialysis assessment before a treatment session, and 17 after. In order to determine if the timing of the off-dialysis assessment influenced our findings, we compared those who were assessed pre- and post-dialysis. The number of patients who satisfied the MINI criterion when assessed before a dialysis session was 4, compared to 5 when assessed after a dialysis session. This difference was non-significant (P = 0.374). The mean off-dialysis BDI score, when completed before dialysis, was 10.8 (±8.6) compared to 11.5 (±10.6) after dialysis. This difference was non-significant (P = 0.923), suggesting that the timing of the off-dialysis assessment had no influence upon the BDI scores. As no differences were apparent, we suggest that the above findings were not influenced by the timing of off-dialysis assessments (pre-/post-treatment).

Differences between the depressed and non-depressed
All the clinical and demographic variables were investigated in relation to depression, and only target haemoglobin (<11g/dl) and EPO dose (IU) were shown to differ significantly between depressed and non-depressed patients (Table 1). A cross tabulation between depressed and non-depressed patients by target haemoglobin demonstrated that patients with low haemoglobin were more likely to be depressed (P = 0.01), (OR 6.6, 95% CI 1.2 and 29). Weekly EPO dose (IU) was significantly higher in the depressed group (P = 0.027). EPO dose was shown in a regression model to be predictive of depression (MINI) (P = 0.03) accounting for 22% of the variance. There was no relationship between depression and co-morbidity.



   Discussion
Top
 Abstract
 Introduction
 Methodology
 Materials
 Results
 Discussion
References
 
The primary aim of this pilot investigation was to examine the legitimacy of on-dialysis depression screening, using the BDI. We set out to understand if on-dialysis screening influenced scoring, particularly somatic enquiry. To summarise our findings, we suggest that if screening HD patients is desired, on-dialysis assessment compares suitably with off-dialysis assessment, providing a convenient procedure that could promote regular evaluation. Compared to off-dialysis screening, on-dialysis screening may be important for three main reasons: (1) increased on-dialysis scoring suggests that the BDI will capture more cases of depression and although increasing false positives, will reduce false negatives; (2) on-dialysis screening is convenient and may allow regular assessment at a time when patients are easily accessible; and (3) clinical interviews cannot be conducted on-dialysis, limiting the identification of depression. Taken together, this suggests a regimen of on-dialysis screening supplemented with diagnostic interviews for those patients identified as being at risk.

There was high agreement between on- and off-dialysis BDI assessments, as determined by Bland–Altman plots. Importantly our findings were not influenced by the timing of the off-dialysis assessment, as some patients were assessed before a dialysis session, others after. On-dialysis assessments yielded significantly higher BDI somatic scoring, compared with the off-dialysis scores. While statistically significant, the practical implications of such a difference are relatively small. Critically the BDI and CDI scores did not reliably differ across the assessment conditions for patients who met the MINI criteria for MDD. Evidently the assessment condition did not influence the screening response for depressed individuals. The non-depressed scored consistently on the CDI but differed on the BDI, scoring reliably higher while dialysing. The heightened BDI scores reflected increased somatic scoring while dialysing. This effect was maintained across all patients. Given our data, on-dialysis assessment increases somatic not cognitive scoring, in contrast to suggestions that on-dialysis assessment may control the impact of fluctuating uraemic symptoms upon mood and cognition [8,15]. In addition, those who met the diagnostic criteria for depression were comparable to the non-depressed with regard to co-morbidity, age and gender, suggesting that these confounding factors did not have an impact upon the stated outcomes. Depressed patients did have, however, reduced haemoglobin and were on a higher weekly EPO dose although study power is an issue.

Although the aim of this study was to examine the procedure rather than the tool, we were able to demonstrate a good diagnostic accuracy for the BDI compared with a short clinical interview when using a cut-off of ≥16, supporting the findings of Watnick et al. [10].

In addition, our data suggest that when compared to a diagnostic scheme (MINI) the BDI (≥16), inclusive of all somatic enquiries, had increased sensitivity and specificity over the CDI (≥10) (which excludes somatic enquiry), favouring an inclusive approach to depression screening. Furthermore, the BDI (≥16) had fewer false positive and false negatives cases than the CDI (≥10). An inclusive approach has also been recommended by Drayer et al., who found that depressed dialysis patients report more somatic symptoms than the non-depressed [8], and that physical symptoms relate more closely to depression than medical co-morbidity [23].

We have established the viability of on-dialysis assessment and described some of its potential benefits but there are some limitations to consider. While we have focused on the procedure of assessment, little data exist relating to the psychopathology of HD patients, especially in the UK and Europe. Further investigation of the BDI in dialysis populations in the UK and elsewhere is required before firm conclusions can be drawn on the appropriateness of the methodology. For the moment we suggest that on-dialysis screening is viable, and that the BDI is a suitable screening tool. A further potential limitation of our study is that all patients recruited were receiving high-flux dialysis or on-line haemodiafiltration that is not the only standard in the UK. However we think that this is unlikely to affect the generalisability of our findings. Lastly, due to the limited sample size of our study we cannot exclude the possibility that the observed effects are due to confounding factors that cannot be fully evaluated due to limited power (i.e. differences between the depressed and non-depressed).

In response to increasing evidence highlighting the negative impact of depression upon the dialysis population, regular depression screening may be justified. Screening patients while actively dialysing may facilitate regular and practical evaluation, increase identification of the condition, and allow monitoring of the effects of treatment interventions.



   Acknowledgments
 
This project was kindly supported by a British Renal Society–Kidney Research UK Fellowship. We wish to thank all the patients who participated and staff members who assisted in the conduct of this research. Particular gratitude goes to Prof. N. Fineberg for her training input, and Maria Da Silva-Gane for her advice and support.

Conflict of interest statement. We confirm that the results presented in this paper have not been published previously in whole or part, except in abstract format.



   References
Top
 Abstract
 Introduction
 Methodology
 Materials
 Results
 Discussion
 References
 

  1. Kimmel PL, Weihs K, Peterson RA. Survival in hemodialysis patients: the role of depression. J Am Soc Nephrol (1993) 4:12–27.[Abstract]
  2. Finkelstein FO, Finkelstein SH. Depression in chronic dialysis patients: assessment and treatment. Nephrol Dial Transplant (2000) 15:1911–1913.[Free Full Text]
  3. Levenson JL, Glocheski S. Psychological factors affecting end-stage renal disease. A review. Psychosomatics (1991) 32:382–389.[Abstract/Free Full Text]
  4. Kimmel PL. Depression in patients with chronic renal disease: what we know and what we need to know. J Psychosom Res (2002) 53:951–956.[CrossRef][Web of Science][Medline]
  5. Wang PL, Watnick SG. Depression: a common but underrecognized condition associated with end-stage renal disease. Semin Dial (2004) 17:237–241.[CrossRef][Web of Science][Medline]
  6. Lopes AA, Bragg J, Young E, et al. Depression as a predictor of mortality and hospitalization among hemodialysis patients in the United States and Europe. Kidney Int (2002) 62:199–207.[CrossRef][Web of Science][Medline]
  7. Kimmel PL, Peterson RA, Weihs KL, et al. Multiple measurements of depression predict mortality in a longitudinal study of chronic hemodialysis outpatients. Kidney Int (2000) 57:2093–2098.[CrossRef][Web of Science][Medline]
  8. Drayer RA, Piraino B, Reynolds CF III, et al. Characteristics of depression in hemodialysis patients: symptoms, quality of life and mortality risk. Gen Hosp Psychiatry (2006) 28:306–312.[CrossRef][Web of Science][Medline]
  9. McDale-Montez EA, Christensen AJ, Cvengros JA, et al. The role of depression symptoms in dialysis withdrawal. Health Psychol (2006) 25:198–204.[CrossRef][Web of Science][Medline]
  10. Watnick S, Wang PL, Demadura T, et al. Validation of 2 depression screening tools in dialysis patients. Am J Kidney Dis (2005) 46:919–924.[CrossRef][Web of Science][Medline]
  11. Kimmel PL. Psychosocial factors in dialysis patients. Kidney Int (2001) 59:1599–1613.[CrossRef][Web of Science][Medline]
  12. Smith MD, Hong BA, Robson AM. Diagnosis of depression in patients with end-stage renal disease. Comparative analysis. Am J Med (1985) 79:160–166.[CrossRef][Web of Science][Medline]
  13. Kimmel PL, Peterson RA. Depression in end-stage renal disease patients treated with hemodialysis: tools, correlates, outcomes, and needs. Semin Dial (2005) 18:91–97.[CrossRef][Web of Science][Medline]
  14. Kaveh K, Kimmel PL. Compliance in hemodialysis patients: multidimensional measures in search of a gold standard. Am J Kidney Dis (2001) 37:244–266.[Web of Science][Medline]
  15. Kurella M, Chertow GM, Luan J, et al. Cognitive impairment in chronic kidney disease. J Am Geriatr Soc (2004) 52:1863–1869.[CrossRef][Web of Science][Medline]
  16. Craven JL, Rodin GM, Johnson L, et al. The diagnosis of major depression in renal dialysis patients. Psychosom Med (1987) 49:482–492.[Abstract/Free Full Text]
  17. Folstein MF, Folstein SE, McHugh PR. ‘Mini-mental state’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res (1975) 12:189–198.[CrossRef][Web of Science][Medline]
  18. Beck AT, Steer RA, Brown GK. Beck Depression Inventory—2nd Edition Manual (1996) San Antonio: The Psychological Corporation.
  19. Sacks CR, Peterson RA, Kimmel PL. Perception of illness and depression in chronic renal disease. Am J Kidney Dis (1990) 15:31–39.[Web of Science][Medline]
  20. Sheehan DV, Lecrubier Y, Sheehan KH, et al. The Mini-international neuropsychiatric interview (MINI): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry (1998) 59(Suppl_20):22–33.[Medline]
  21. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet (1986) 1:307–310.[CrossRef][Web of Science][Medline]
  22. Chandna SM, Schulz J, Lawrence C, et al. Is there a rationale for rationing chronic dialysis? A hospital based cohort study of factors affecting survival and morbidity. BMJ (1999) 318:217–223.[Abstract/Free Full Text]
  23. Drayer RA, Mulsant BH, Lenze EJ, et al. Somatic symptoms of depression in elderly patients with medical comorbidities. Int J Geriatr Psychiatry (2005) 20:973–982.[CrossRef][Web of Science][Medline]
Received for publication: 24. 9.07
Accepted in revised form: 5. 2.08


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
23/8/2653    most recent
gfn105v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Chilcot, J.
Right arrow Articles by Farrington, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Chilcot, J.
Right arrow Articles by Farrington, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?