gms | German Medical Science

GMS Psycho-Social-Medicine

Corporate journal of German scientific societies in psychosocial medicine

ISSN 1860-5214

The psychosocial evaluation of medically-ill inpatients: accordance between mental disorders and self-rated psychosocial distress

Die Evaluation psychosozial belasteter Krankenhauspatienten: Übereinstimmung zwischen psychischer Diagnose und Selbstbeurteilung psychosozialer Belastungen

Research Article

  • corresponding author Kurt Fritzsche - Department of Psychosomatics and Psychotherapy, University Hospital Freiburg, Freiburg, Germany
  • Thorsten Burger - Department of Psychosomatics and Psychotherapy, University Hospital Freiburg, Freiburg, Germany
  • Armin Hartmann - Department of Psychosomatics and Psychotherapy, University Hospital Freiburg, Freiburg, Germany
  • Matthias Nübling - Department of Psychosomatics and Psychotherapy, University Hospital Freiburg, Freiburg, Germany
  • Claudia Spahn - Department of Psychosomatics and Psychotherapy, University Hospital Freiburg, Freiburg, Germany

GMS Psychosoc Med 2005;2:Doc11

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/journals/psm/2005-2/psm000020.shtml

Published: December 15, 2005

© 2005 Fritzsche et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Abstract

Background: Both psychometric questionnaires and structured psychodiagnostic interviews are used to identify medically-ill inpatients requiring psychotherapeutic treatment. The study examines the accordance between the diagnosis of a mental disorder (ICD-10) and self-rated psychosocial distress of unselected inpatients within the framework of a psychosomatic liaison service.

Methods: Of n=532 consecutive patients hospitalized in five departments of the University Hospital Freiburg, n=357 patients underwent a structured psychodiagnostic interview (Mini-DIPS) to obtain diagnoses according to ICD-10 F. Psychosocial distress (HADS), somatoform symptoms (SOMS-2 J), quality of life (EDLQ) and coping strategies (FQCI) were evaluated by self-rating questionnaires.

Results: A mental disorder requiring treatment was diagnosed in 44% of the patients. Predictors for the diagnosis of a mental disorder were the depression subscale of the HADS, the frequency of somatoform symptoms and depressive coping. The greatest accordance between mental disorders and screening instruments was found for the depression subscale of the HADS at a cut-off 8+. The area under the curve (AUC) was 0.75. With a specificity of 87% and a sensitivity of 53%, nearly half of all mental disorders requiring treatment were not identified by the HADS, especially in patients with neurotic, stress-related and somatoform disorders (F4), most of them cancer patients with adjustment disorders.

Conclusions: Case selection by the HADS is successfull in disorders, where depressive symptoms are prevalent. For F4-diagnoses a lower cut-off of 6+ is recommended. The importance of a clinical interview is underlined. Other screening procedures, specific for the general hospital population, are discussed.

Keywords: mental disorders, psychosocial distress, screening, psychosomatic consultation liaison service, general hospital

Zusammenfassung

Hintergrund: Zur Identifizierung behandlungsbedürftiger psychischer Störungen bei körperlich Kranken werden sowohl psychometrische Fragebögen als auch strukturierte klinische Interviews eingesetzt. Untersuchungen zur Übereinstimmung zwischen Screeninginstrumenten und diagnostischem Interview für unselegierte Patienten im Akutkrankenhaus im Rahmen des psychosomatischen Konsil- und Liaisondienstes liegen bisher noch nicht vor.

Methoden: Von n=532 Patienten, die auf fünf Stationen des Universitätsklinikums Freiburg stationär behandelt wurden, wurden n=357 konsekutiv mit einem strukturierten diagnostischen Interview (Mini-DIPS) zur Erfassung einer ICD-10 Diagnose untersucht. Parallel wurden auf Patientenseite Angst und Depression (HADS), somatoforme Symptome (SOMS-2 J), Lebensqualität (Alltagsleben) und Copingstrategien (FKV) mit Hilfe von Selbstbeurteilungsbögen erhoben.

Ergebnisse: Bei 44% der Patienten wurde eine behandlungsbedürftige psychische Störung diagnostiziert. Prädiktoren für die Diagnose einer psychischen Störung waren Depressivität im HADS, die Häufigkeit somatoformer Symptome und depressives Coping. Die größte Übereinstimmung zwischen psychischen Störungen und den Screeninginstrumenten findet sich für die Depressionsskala (HADS) bei einem Cut-Off von 8+. Die Fläche unter der Kurve (AUC) betrug 0,75. Mit einer Spezifität von 87% und einer Sensitivität von 53% wurden fast die Hälfte aller Patienten mit einer behandlungsbedürftigen psychischen Störung nicht durch den HADS erfasst. Dies betraf vor allem Patienten mit neurotischen Störungen, Anpassungsstörungen und somatoformen Störungen (F4). Die meisten dieser Patienten waren Tumorpatienten mit einer Anpassungsstörung.

Schlussfolgerungen: Zur Identifizierung von Patienten mit einer behandlungsbedürftigen psychischen Störung im Rahmen des psychosomatischen Konsil- und Liaisondienstes ist der HADS geeignet, vor allem wenn es sich um Symptome einer affektiven Störungen handelt. Bei F4-Diagnosen wird ein niedrigerer Cutoff von 6+ empfohlen. Die Bedeutung eines klinischen Interviews wird betont. Andere Screeningverfahren werden diskutiert.

Schlüsselwörter: psychische Störung, psychosoziale Belastungen, Screening, psychosomatischer Konsil- und Liaisondienst


Introduction

The overall prevalence of mental disorders in patients in a general hospital and in university hospitals fluctuates between 15% and 47% [1], [2], [3], [4], [5]. The marked difference in results is due to the use of various diagnostic classification systems (ICD, DSM) and different definition of the cases, such as clinical assessment, structured or diagnostic interview and questionnaires. There are different results to be found with the questionnaires due to different cut-off scores selected. A comparison of the screening questionnaires with structured or standardized clinical interview was undertaken mainly in patients with cancer [6], [7], [8], [9], [10], [11], [12], [13]. Other studies examined the usefulness of screening instruments to diagnose specific mental disorders like major depression [14]. However, to date there have been only a few studies which tested the screening results of questionnaires against the "gold standard" of a structured clinical interview in unselected patient groups in hospital [15], [16].

Integrated treatment of mental disorders in one health service facility as liaison service is the most effective way [17], [18]. The development of the consultation-liaison (CL) services in Europe and Germany was promoted by the "European Consultation Liaison Workgroup (ECLW)". Psychiatrists and psychosomatic physicians in 14 countries have united with the aim of improving psychosocial care in general hospitals by means of pan-European cooperation [19]. At the university hospital Freiburg a psychosomatic liaison service exists for the departments of neurology, dermatology, gastroenterology, hematology-oncology and radiation therapy.

The study presented is part of a broader study of need and use of psychotherapeutic interventions of medically-ill inpatients in the framework of the psychosomatic liaison service at the University Hospital of Freiburg. This first part of the study has three objectives:

1.
Assessement of self-rated psychosocial distress and frequency of mental disorders requiring treatment.
2.
Identification of predictors for the diagnosis of a mental disorder.
3.
Accordance between mental disorders and self-rated psychosocial distress of the patients.

Patients and methods

Sample

Within a period of 6 months, a total of 532 patients was admitted to the 7 wards in 5 departments participating in this study. 48 patients (9%) were admitted more than once during this period. These patients were only counted once. The sample size was n=392. Figure 1 [Fig. 1] shows the patient recruitment and reasons for exclusion.

The patients who were excluded from the study did not differ with respect to age and sex from the study population. Their illness was, however, significantly more severe (Karnofsky-Performance Status (KPS); p<0.001).

The study was performed within the framework of routine care by the psychosomatic liaison service. Each newly-admitted patient was visited by the research assistant within the first two days and informed about the study after a brief introductory conversation. Informed consent was given by all subjects. The study was approved by the Ethics-Review-Committee of the Freiburg University.

Diagnostic procedures and measurements

The diagnostic interview for mental disorders - the Mini-DIPS - [20] was conducted by research assistants, who were in their final semester and have since completed their doctorate. They were well trained in interview technique and were under continued supervision.

The Mini-DIPS serves to record the most important mental disorders in the psychotherapeutic area based on the criteria of DSM-IV and ICD-10. The interrater-reliability of the Mini-DIPS presents with a kappa coefficient between 0.84 and 1.0 [20]. The Mini-DIPS permits a sufficiently precise diagnosis of all mental disorders except organic mental disorders (F0) and personality disorders (F6). These diagnoses are only made with clinical assessment and in consultation with the study director.

Considering these factors, there is indeed an underestimate for these two diagnosis subgroups. Nevertheless, this was accepted because these two diagnosis groups did not play a role in the psychotherapeutic treatment of the liaison service. The interview is manualized. In contrast to standardized interviews, the interviewer took advantage of the opportunity to ask additional questions if something was ambiguous. Based on our experience, this increases the clinical usefulness of the procedure as well as its acceptance in practice and research.

The diagnosis was only made at the end of the entire interview and after in-depth discussions with the team. It took between 45 and 60 minutes to perform this procedure. The interviewer received accompanying diagnostic supervision. During the study, meetings for discussion and classification of the interviews were held every two weeks. The ICD-10 training of the interviewer consisted of sample interviews with inpatients and ratings of video-recorded conversations. Twelve selected cases from the ECLW Study [21] were used as proof of reliability of the diagnosis based on ICD-10. These cases dealt with typical case descriptions from consultation services that were to be classified according to ICD-10-F. The following cases arose: disorders resulting from use of alcohol, depressive episode, recurrent depressive disorder, social phobia, panic disorder, adjustment disorder, dissociative disorder, somatization disorder, hypochondrical disorder, psychological or behavioral factors associated with disorders or diseases classified elsewhere. For these 10 cases a consensual gold standard was defined.

The spectrum of these diagnoses basically coincides with the expected spectrum for consultation liaison services area for mental disorders. The four interviewers correctly identified 36 of 48 possible diagnoses. Some clinically justified deviations from gold standard were accepted e.g. agoraphobia vs. panic disorder with agoraphobia, depressive episode vs. adjustment disorder [21]. The kappa-values for the four interviewers were 0.54, 0.72, 0.81 and 0.81. In the ECLW-study some 76% of the raters had a kappa value of at least 0.70; 6% had a value between 0.40 and 0.70.

Cooper's [22] criteria were used to estimate the severity of the mental disorders. 0 means no abnormalities, 1 means mild symptoms, no treatment necessary, 2 to 4 means increasing severity of symptoms and impairment, need for treatment ranging from outpatient to inpatient. Only patients with a severity of ≥2 (obvious mental illness requiring treatment) are listed (see Table 3). The research assistants were blinded to the results of the psychometric questionnaires.

Psychometric questionnaires

Patients filled in the following psychometric instruments:

Hospital Anxiety and Depression Scale

The HADS [23] is a well-validated, routinely-used screening instrument for the detection of mental disorders in physically-ill patients. The HADS has been translated into more than 30 languages and extensively validated in many of these versions [24]. The German version (HADS-D) was validated in 6200 patients and control persons and found equivalent to the English original [25].

Freiburg Questionnaire Coping with Illness - FQCI

The FQCI [26] addresses ways of coping at the levels cognition, emotions and behavior. The possible responses are on a 5-point scale from 1=not at all to 5=very strong. 35 items were assigned to 5 scales extracted on the basis of a factor analysis. For our study, three scales were assessed: Depressive coping, e.g. "withdrawl from other people". Minimization and wishful thinking, e.g. "unwillingness to believe the event" and Active problem-orientated coping, e.g. "seeking information about disease and treatment"or "determined fighting against disease".

Screening for Somatoform Disorders - SOMS

The "Screening for somatoform Disorders (SOMS)" [27] is based on the somatization index (cut-off: 4 for male, 6 for female) of Escobar [28] and makes it possible in the form of self-rating to record 53 physical complaints from the DSM-IV which were present during the past 2 years (SOMS-2) or during the past 7 days (SOMS-7) and which were not attributable to an organic cause. After discussion with the authors of the instrument, the cutoff was increased from 6+ to 9+ symptoms, in order to avoid false positive cases confounding with physical disease.

The Every Day Life Questionnaire - EDLQ

The questionnaire consists of 42 items with a 5-point Likert scale [29]. It contains questions on physical, emotional, social and functional components of the quality of life, as well as satisfaction with life and medical treatment. For the study, the two scales Everyday life and Social life were assessed. The Everyday life scale includes the extent to which a person is able to work at a regular job, can take care of himself, can cope with daily problems and engages in leisure activities. The Social life scale asks about contacts with other people, the extent to which the person feels understood and supported and about the quality of the partner-relationship.

Since there are no cutoffs for FQCI or EDLQ, a limit value for each scale calculated from the mean and one standard deviation was defined for the selection of distressed patients.

"Severity of the physical disease" Scale

The factor "Severity of the physical disease" was formed by factor analysis from Karnofsky-Performance Status (KPS) [30], the Charlson Morbidity Index (CMI) [31], the level of care required and the Physical stress scale of the Every Day Life Questionnaire [29]. The scores of this index were calculated as a weighted sum of the single scores, according to their factor loadings. The Karnofsky Index (0.815) and the level of care required (0.751) have the greatest impact on the factor.

A structured diagnostic interview could be performed in n=357 patients (91%). The return quota for the psychometric instruments was 75% for HADS, 74% for FQCI, 67% for SOMS and 71% for EDLQ. Complete data include age, sex, physical disease and severity of the physical disease. A multivariate model with the variables age, sex, severity of physical disease, mental disorders, somatic diagnoses and department/ward was tested and fitted to the data in order to determine the predictive factors with respect to filling out all four psychometric instruments. Logistic regression shows that the probability of filling out all four psychometric instruments decreases with higher age (p<0.001), the presence of a mental disorder (p<0.001) and in patients in the hematological-oncological ward (p<0.001). Model parameters: Chi2=49.9; df=6; p<0.001; Hosmer-Lemeshow's C=5.2; Nagelkerkes R2=0.18.

Statistical procedures

The data were analyzed using SPSS 10.0. The most important statistical analytical procedures on the bivariate level are contingency tables, Chi2-Tests and Fisher's exact test, if necessary and the comparison of means by single and multiple analyses of variance. Multiple linear regression, multiple logistical regression and factor analyses were used for multivariate analyses.

Receiver operating characteristics (ROC) analysis [32], [33] was used to evaluate the diagnostic accuracy of the HADS subscales. The ROC-curve expresses the sensitivity and specificity for each score and represents an index of the overall ability of the instrument to discriminate between cases and non-cases [33].


Results

Sociodemographic data

The mean age of the sample was 54.5 years (SD 16.46) with a range from 18-84 years. The number of women (44%) and men (56%) was approximately equal. More than half the patients (54%) had a low educational level. 34% of the patients were retired, 15% unemployed and 14% on sickleave or disabled. More than half the patients were skilled and unskilled workers and 69% were married or living with a partner.

Somatic diagnoses

Due to the focus of the liaison activities, cancer diseases were most frequent (46%), followed by skin diseases (16%) and neurological diseases (11%) (see Table 1 [Tab. 1]).

Screening for psychosocial distress

Table 2 [Tab. 2]shows the means and the percentage of patients with increased values above the cut-off for the four psychometric instruments.

Mental disorders

A psychodiagnostic interview was conducted with n=357 patients. Of these, n=142 (39.8%) had no mental disorder, n=57 (16%) had a diagnosis not requiring treatment (Cooper severity grade 1 ), n=158 (44.3%) had a diagnosis requiring treatment (Cooper severity grade 2 to 4). The distribution by ICD-10 classification is shown in Table 3 [Tab. 3].

Adjustment disorder (F43) was the most frequent diagnosis (n = 67; 18.8%). Most of the patient with a F43 diagnosis were cancer patients.

Accordance between mental disorder and self-rated psychosocial distress

Table 4 [Tab. 4] shows a comparison of mean values of the psychometric instruments with and without mental disorder by ICD-10, Cooper severity grade 2 to 4. The strongest relationships between a mental disorder and the scales of the self-rating instruments are found for Anxiety, Depression, Somatization and Depressive coping.

If, instead of means, the increased values above the cut-off with and without ICD-10 diagnosis are compared, the strongest relationships are again found in the subscales Anxiety and Depression of the HADS, Somatization and Depressive coping (no data shown).

Accordance between mental disorders and HADS

The subscale anxiety has the greatest accordance with a mental disorder at a cut-off 7+ (Chi2=23.776; df=1; p<0.001; Phi=0.299). The percentage of correctly-classified cases was 65%. The specificity at a cut-off 7+ attained a value of 0.75 with sensitivity of 0.54.The subscale depression has the greatest accordance with a mental disorder at a cut-off 8+ (Chi2=52.121; df=1; p<0.001; Phi=0.443). The percentage of correctly-classified cases was then 74%. Specificity at a cut-off 8+ attained a value of 0.87 with sensitivity of 0.53.

Looking at the total HADS score, there were no considerable changes: we found the greatest accordance at a cut-off of 14+. The percentage of correctly-classified cases is 73%. Specificity was 0.79 and sensitivity 0.63.

A more balanced ratio between specificity and sensitivity we found at a cutoff score of 6+, with a specificity of 0.67 and a sensitivity of 0.72 for the subscale depression.

When ROC-analysis is performed, the AUC of the subscale depression surpasses the subscale anxiety, while the AUC of the total HADS-score expresses almost the same values as the subscale depression (see Figure 2 [Fig. 2]).

Specificity and sensitivity are also dependent on ICD-10 categories (see Table 5 [Tab. 5]). They were high for mood disorders (F3) and low for substance abuse (F1) and neurotic, stress-related and somatoform disorders (F4).

Comparison of the HADS subscales anxiety and depression with the ICD-10 diagnosis groups for anxiety (ICD-10: F 40, 41, 43.22) vs. others and depression (ICD-10: F 32-34, 43.20, 43.21, 43.22) vs. others shows only a moderately significant relationship for depression (F=6.711; df=1; p<0.05; Eta=0.247) and no significant relationship for anxiety (F=0.266; df=1; p=n.s.; Eta=0.051). Comparison of the SOMS scale with the ICD-10 diagnosis group for somatoform disorders (ICD-10: F 45) vs. others also shows only a moderately significant relationship (F=7.939; df=1; p<0.01; Eta=0.286).

Predictors for mental disorder

Predictors for the presence of a mental disorder were the subscale depression of the HADS (p<0.001), the frequency of somatoform symptoms (SOMS) (p<0.001) and depressive coping with illness (FQCI) (p<0.05) (Model parameter: Chi2=63.2; df=3; p<0.001; Nagelkerke R2=0.33).


Discussion

HADS as screening instrument

The subscale depression of the HADS at a cut-off 8+ attained the best accordance (74%) with the presence of a mental disorder in general. The specificity was 87%, but the sensitivity with 53% was low. The best balance between specificity and sensitivity was found at a cutoff score of 6+. A review [34] about the validity of the Hospital Anxiety and Depression Scale identified 12 studies of noncancer medical patients and 10 studies of cancer patients. For noncancer patients the mean optimal cut-off score for the HADS depression subscale was approximately 8+, with mean specificity 0.79 and mean sensitivity 0.83. For cancer patients the mean optimal cut-off score of the depression subscale was also approximately 8+ with mean specificity 0.83 and lower mean sensitivity of 0.66.

The reported values of specificity and sensitivity for the HADS score vary depending on the homogeneity of the sample, the prevalence of disorders in the sample, structured or clinical interview and the diagnosis included. In some of the previous studies, stricter criteria for the presence of a mental disorder requiring treatment were applied. For example, adjustment disorders were not recorded in the CIDI or SCID. But it is important, especially in patients with physical diseases, to select a broader approach in order to identify as far as possible all patients who require psychotherapeutic intervention. Sensitivity and specificity are sample-dependent. The sensitivity increases markedly in Chapter F3 and is considerably lower in Chapter F4 and F1. Patients with mood disorders and increased emotional distress were better identified than patients with substance abuse, adjustment disorders or somatoform disorders. Because of the small sample size of patients with diagnosis of Chapter F1 and F3 the results can not be generalized. Examination of other factors like sociodemographic, somatic and psychosocial variables, produced no significant results. Another reason for the false negative results could be patients who denied their mental distress but who were identified as cases in the diagnostic interview.

Other than the HADS, only the scales "depressive coping" in the FQCI and the SOMS attained satisfactory accordance with the presence of a mental disorder. The subscale depression of the HADS, the scale "depressive coping" and the SOMS were also significant independent predictors for the diagnosis of a mental disorder. The FQCI is rejected as a screening instrument, since there is no standardized cut-off. The SOMS is a screening instrument for somatoform disorders. Some of the patients are unable to follow the instructions to mark only those complaints for which no organic cause has been found. Therefore there is a danger of confounding by somatic symptoms of underlying physical disease.

Representativeness

The study population did not differ essentially with respect to age and gender from the patients at the participating departments in the years 1998 and 1999. The distribution of the somatic main diagnoses is also typical for the departments in question and includes a large part of the relevant medical diagnoses of the university hospital.

The other sociodemographic data also agree largely with the findings recorded in studies in general hospitals and university clinics in Germany [35], [36], [37], [38], [39]. The quota of patients refusing to participate was low (9%). Excluded patients and refusers showed significantly higher severity of disease. Age and gender were comparable.

Comparison of patients with complete data sets in the self-rating instruments with those of patients without questionnaire data showed higher age, presence of a mental disorder and hospitalization in a hematological-oncological ward to be significant factors of influence in a multivariate analysis. The patients who did not participate in the questionnaire study were thus not only more physically limited, but also under greater mental stress. These results correspond to previous studies [1], [15], [39].

The study was designed as a consecutive survey. Each patient had the same chance of being assigned to the sample. This avoided overrepresentation of patients with longer hospitalization times and corresponding characteristics, which would be the case in a purely cross-sectional selection.


Conclusions

The HADS is a valid screening instrument for mental disorders in which depressive symptoms are prevalent. The low sensitivity for patients with substance abuse and for neurotic, stress-related and somatoform disorders leads to the following considerations: most of patients with F4-diagnoses are cancer patients. As shown in other studies [34] the sensitivity of the HADS in cancer patients is low. A lower cut-off, like 6+, results in higher sensitivity. For somatoform disorders or substance abuse a lower cut-off point or a specific instrument is also necessary. The diagnostic weakness of the HADS in somatoform disorders cannot be offset by the SOMS. Our study also brings very little evidence in this regard, since the number of patients with somatoform disorders (n=9) is very small. Agreement between F45 diagnoses and the SOMS is significant (p<0.01), but not overwhelming (Eta=0.286).

Based on the data available from this study with its low number of cases, we can only recommend the SOMS as a screening instrument with reservations, and then preferably in combination with HADS or GHQ.

For patients who deny their psychosocial distress although physicians and nurses see need for psychotherapeutic support, a clinical interview is necessary to recognize a psychiatric morbidity. In this respect, appropriately-trained physicians and nurses are the best screening tools for psychosocial distress in medically-ill patients [40].

Other approaches to case-finding are the Patient Health Questionnaire (PHQ), developed and validated within the Prime-MD (an interview for mental disorders in primary care) [41] and the INTERMED [42]. The PHQ which has been translated into German, was favourably rated in comparison studies with respect to statistical parameters and ROC-curves [43], [44], [45]. The INTERMED has been developed by a European team to determine the need for treatment. INTERMED describes the patient's care needs in biological terms, (chronicity and severity of illness), psychological terms (coping and psychosocial distress), social terms (e.g. family disruption) and the healthcare needs (e.g. complexity of care). The goal of this instrument is to detect those patients that receive the most medical and nursing care during their stay but not to detect patients suffering mental disorders per se. There is a need for further research on differential usability of specialised screening procedures (PHQ, INTERMED), traditional psychometric testing (eg. HADS) and psychodiagnostic interview.


References

1.
Mayou R, Hawton K. Psychiatric disorder in the general hospital. Br J Psych. 1986;149:172-90.
2.
Mayou R, Hawton K, Feldman, E, Ardern M. Psychiatric problems among medical admissions. Int J Psychiatry Med. 1991;21:71-84.
3.
Levenson JL, Hamer RM, Rossiter LF. Relation of psychopathology in general medical inpatients to use and cost of services. Am J Psych. 1990;147:1498-503.
4.
Cavanaugh SA. The prevalence of emotional and cognitive dysfunction in a general medical population. Using the MMSE, GHQ and BDI. Gen Hosp Psychiatry. 1983;5:15-24.
5.
Bell G, Reinstein DZ, Rajiyah G, Rosser R. Psychiatric screening of admissions to an accident and emergency ward. Br J Psychiatry. 1991;158:554-57.
6.
Jenkins PL, May VE, Hughes LE. Psychological morbidity associated with local recurrence of breast cancer. Int J Psychiatry Med. 1991;21:149-55.
7.
Razavi D, Delvaux N, Bredart A, Paesmans M, Debusscher L, Bron D, Stryckmans P. Screening for psychiatric disorders in a lymphoma out-patient population. Eur J Cancer Prev. 1992;28A:1869-71.
8.
Razavi D, Delvaux N, Farvaques C, Robaye E. Screening for adjustment disorders and major depressive disorders in cancer in-patients. Br J Psych. 1990;156:79-83.
9.
Ibbotson T, Maguire P, Selby P, Priestman T, Wallace L. Screening for anxiety and depression in cancer patients: the effects of disease and treatment. Eur J Cancer Prev. 1994;30A:37-40.
10.
Grassi L, Malacarne P, Maestri A, Ramelli E. Depression, psychosocial variables and occurence of life events among patients with cancer. J Affect Disorders. 1997;44:21-30.
11.
Kugaya A, Akechi T, Okuyama T, Okamura H, Uchitomi Y. Screening for psychological distress in Japanese cancer patients. JPN J Clin Oncol .1998;28:333-8.
12.
Constantini M, Musso M, Viterbori P, Bonci F, Del Mastro L, Garrone O, Venturini M, Morasso G. Detecting psychological distress in cancer patients. Validity of the Italian version of the Hospital Anxiety and Depression Scale. Support Care Cancer. 1999;7(3):121-7.
13.
Reuter K, Härter M. Screening for mental disorders in cancer patients - discriminant validity of HADS and GHQ-12 assessed by standardized clinical interview. Int J Methods Psychiatr Res. 2001;10:86-96.
14.
Silverstone PH. Poor efficacy of the hospital anxiety and depression scale in the diagnosis of major depressive disorder in both medical and psychiatric patients. J Psychosom Res. 1994;5:441-50.
15.
Clarke DM, Smith GC, Herrmann HE. A comparative study of screening instruments for mental disorders in general hospital patients. Int J Psychiatry Med. 1993;23:323-37.
16.
Abiodum OA. A validity study of the Hospital Anxiety and Depression Scale in general hospital units and a community sample in Nigeria. Br J Psychiatry. 1994;165:669-72.
17.
Katon WJ, Gonzales J. A review of randomized trials of psychiatric CL studies in primary care. Psychosomatics. 1994;35:268-78.
18.
Creed F, Marks B. Liaison psychiatry in general practice. A comparison of the liaison attachment scheme and shifted outpatient clinic models. C Cell Gen Practa. 1989;39:514-7.
19.
Huyse FJ, Herzog Th, Lobo A, Malt U, Opmeer B, Stein B, de Jonge P, van Dijck R, Creed F, Crespo MD, Cardoso G, Guimares-Lopes R, Mayou R, van Moffaert M, Rigatelli M, Sakkas P, Tienary P. Consultation-Liaison psychiatric service delivery. Results from a European study. Gen Hosp Psych. 2001;23:124-32.
20.
Margraf J. Diagnostisches Kurzinterview bei psychischen Störungen [Diagnostic interview for mental disorders] (Mini-DIPS). Berlin, Heidelberg, New York: Springer; 1994.
21.
Malt UF, Huyse FJ, Herzog T, Lobo A, Rijssenbeek A, European Consultation Liaison Workgroup (ECLW): The European Consultation Liaison Workgroup (ECLW). Collaborative Study. III. Training of reliability and assessing of ICD-10 psychiatric diagnoses in the general hospital setting - an investigation of 220 consultants from 14 European countries. J Psychosom Res. 1996;41:451-63.
22.
Cooper B. Probleme der Falldefinition und Fallfindung.Problems of case definition and case finding. Nervenarzt. 1978;49:437-44.
23.
Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67:361-70.
24.
Herrmann C. International experiences with the Hospital Anxiety and Depression Scale - a review of validation data and clinical results. J Psychosom Res. 1997;42:17-41.
25.
Herrmann C, Buss U, Snaith RP. HADS-D. Hospital Anxiety and Depression Scale - Deutsche Version. Bern: Huber; 1995.
26.
Muthny FA. FVK - Freiburger Fragebogen zur Krankheitsverarbeitung [FQCI - Freiburg Questionnaire of Coping with Illness]. Weinheim: Beltz Test; 1989.
27.
Rief W, Hiller W, Heuser J. SOMS, das Screening für somatoforme Störungen: Manual zum Fragebogen. Bern, Göttingen, Toronto, Seattle: Huber; 1997.
28.
Escobar JI, Waitzkin H, Silver RC, Gara M, Holman A. Abridged somatization: a study in primary care. Psychosom Med. 1998;60:466-72.
29.
Bullinger M, Kirchberg I, v Steinbüchel N. Der Fragebogen Alltagsleben - ein Verfahren zur Erfassung der gesundheitsbezogenen Lebensqualität [Every Day Life Questionnaire - An instrument for the assessment of health-related quality of life]. Z f Med Psychol. 1993;2:121-31.
30.
Karnofsky DA, Burchenal JH. The clinical evaluation of chemotherapeutic agents in cancer. In MacLeod CM, editor. Evaluation of chemotherapeutic agents. New York: Columbia University Press; 1949. p. 191-205.
31.
Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373-83.
32.
Beck JR, Shultz EK. The use of ROC curves in test performance evaluation. Arch Pathol Lab Med. 1986;110:13-20.
33.
Zweig MH, Campbell G. Receiver-Operating Characteristics (ROC) Plots. A Fundamental Evaluation Tool in Clinical Medicine. Clin Chem. 1993;39:561-77.
34.
Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the Hospital Anxiety and Depression Scale. An updated literature review. J Psychosom Res. 2002;52:69-77.
35.
Herzog T, Stein B. Psychotherapeutisch-psychosomatische Konsiliar-Liaisondienste. Entwicklungen, empirische Befunde, Perspektiven für die Praxis und Forschung. [Consultation/Liaison services by Psychotherapie and Psychosomatics. Empirical findings, open questions, ongoing research]. Psychol Med. 1994;5:10-7.
36.
Künsebeck H W, Lempa W, Freyberger H. Häufigkeit psychischer Störungen bei nicht-psychiatrischen Klinikpatienten [Frequency of mental disorders of nonpsychiatric inpatients]. DMW. 1984;109:1438-42.
37.
Herrmann C, Brand-Driehost S, Kaminsky B, Leibing E, Staats H, Rüger U. Diagnostic groups and depressed mood as predictors of 22-month mortality in medical inpatients. Psychosom Med.1998;60:570-7.
38.
Meyer T, Klemme H, Herrmann Ch.Depression but not anxiety is a significant predictor of physicians'assessments of medical status in physically ill patients. Psychotherapy Psychosomatics. 2000;69:147-54.
39.
Arolt V, Gehrmann A, John U, Dilling H. Psychiatrischer Konsiliardienst an einem Universitätsklinikum. Eine emprische Untersuchung zur Leistungscharakteristik [Consultation liaison psychiatric services at the University Hospital]. Nervenarzt. 1995; 66:347-54.
40.
Strain J. Liaison psychiatry. In Wiese MG, Rundel JR, editors. Textbook of consultation-liaison psychiatry. Washington DC: American Psychiatric Publishing; 2001. p. 33-48.
41.
Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD. JAMA. 1999;282:1737-44.
42.
De Jonge P, Huyse FJ, Stiefel FC, Slaets JPJ, Gans ROB. INTERMED - a clinical instrument for biopsychosocial assessment. Psychosomatics. 2001;42:106-9.
43.
Löwe B, Spitzer RL, Gräfe K, Kroenke K, Quenter A, Zipfel S, Buchholz C, Witte S, Herzog W. Comparative validity of three screening questionnaires for DSM-IV depressive disorders and physicians' diagnoses. J Affect Dis. 2004;78:131-40
44.
Löwe B, Gräfe K, Zipfel S, Witte S, Loerch B, Herzog W. Diagnosing ICD-10 Depressive Episodes. Superior criterion validity of the Patient Health Questionnaire. Psychother Psychosom. 2004;73:386-90
45.
Löwe B, Gräfe K, Zipfel S, Spitzer RL, Herrmann-Lingen C, Witte S, Herzog W. Detecting panic disorder in medical and psychosomatic outpatients. Comparative validation of the Hospital Anxiety and Depression Scale, the Patient Health Questionnaire, a screening question, and physicians' diagnosis. J Psychosom Res. 2003;55:515-9