Psychosomatics 39:244-252, June 1998
© 1998 The Academy of Psychosomatic Medine
Consultation-Liaison Psychiatrists' Management of Depression
Graeme C. Smith, M.B.B.S., M.D., D.P.M., F.R.A.N.Z.C.P.,
David M. Clarke, M.B.B.S., M.P.M., F.R.A.C.G.P., F.R.A.N.Z.C.P.,
Dennis Handrinos, M.B.B.S., M.P.M., F.R.A.N.Z.C.P., and
Astrid Dunsis, M.B.B.S., F.R.A.N.Z.C.P., D.P.M., D.C.H.
Received May 30, 1997; revised August 25, 1997; accepted September 16, 1997. From the Consultation-Liaison Psychiatry Research Unit, Monash University, Department of Psychological Medicine, Melbourne, Australia. Address reprint requests to Dr. Smith, Monash University Department of Psychological Medicine, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia.
|
ABSTRACT
|
Prospective data on 1,360 consecutive inpatients referred to the consultation-liaison psychiatry service of 2 metropolitan general teaching hospitals and diagnosed as having a Depressive Illness Spectrum Disorder were collected by using the MICRO-CARES clinical database system. The distribution of DSM-III-R diagnoses was major depression (MD) 49%; dysthymia (DYS) 15%; organic or substance-induced mood disorder or depressive disorder not otherwise specified (ORG/NOS) 14%; and adjustment disorder with depressed mood (AD) 29%. Antidepressants were prescribed in 59% of the MD cases, 40% of the DYS cases, 36% of the ORG/NOS cases, and 17% of the AD cases. In confirmed MD, antidepressants were prescribed in 69%, and significantly more often in those who were older, female, had a prior history of physical illness, had a neoplasm or a disorder of the nervous or musculoskeletal systems, had higher Axis IV scores, or were referred because of pain or terminal illness. The patients with confirmed MD prescribed antidepressants had a longer length of stay and were referred later than those not prescribed antidepressants. The results illustrate the importance of all of the forms of depression in consultation-liaison psychiatry and the vigor with which all forms are treated.
Key Words: Consultation-Liaison Psychiatry • Depression
|
INTRODUCTION
|
Depression is common in those who are physically ill or who somatize, and this is the most common way in which depression presents in the community.1–5 Both syndromal and subsyndromal depressive symptoms are associated with significant impairment of function and well-being and increased mortality in the physically ill.3,6–14 However, there is little except consensus opinion to guide the management of this common form of physical/psychiatric comorbidity. As has been observed,15 most controlled trials of the treatment of depression have been performed on patients with major depression, without physical comorbidity, and in psychiatric settings. Those few that addressed depression in the physically ill or in primary care were described as inconclusive. Both the Agency for Health Care Policy and Research (AHCPR) and the American Psychiatric Association guidelines for the treatment of depression are evasive about physical/psychiatric comorbidity.15,16
There are many reasons for this lack of evidence-based guidelines. They include problems with the classification system,2,17 the difficulties of defining and measuring depression in the presence of physical illness,18,19 and the difficulties encountered in trials of antidepressant medication in this population. Reliable documentation of the interventions made by experienced consultation-liaison (C-L) psychiatrists when they diagnose depression is an important part of the process of developing guidelines.10,20 Here we report such data, using a well defined and supervised prospective clinical database for inpatients referred to the coordinated C-L psychiatry services of a group of teaching hospitals. This method will permit comparison with practice in other centers.21 In an article, Hamner et al.22 review the history of the development and application of such clinical databases: a prior study by themselves and another by McKegney et al.
Particular questions examined were those of the relationship between interventions recommended and the type and certainty of diagnosis, concordance with recommendations, and the factors that correlate with the use of antidepressants.
|
METHODS
|
Data were collected prospectively over a 4-year period (1990–1993) on 3,372 consecutive inpatient referrals to the integrated adult C-L psychiatry services of 2 metropolitan general teaching hospitals affiliated with Monash University, Monash Medical Centre, and Dandenong Hospital (the "referred cohort"). Ten experienced C-L psychiatrists (including the authors) and their 3rd- and 4th-year psychiatry trainees doing their 6-month C-L psychiatry rotation, working in a mixture of liaison and consultation mode, saw referrals from medical, surgical, and specialty units, including obstetrics and gynecology. The MICRO-CARES clinical database system is used routinely in clinical work.21,22 Data from it are used for reporting and for subsequent research purposes. Supervisors are responsible for seeing that the data entered reflect accurately the clinical process, particularly the diagnoses; the results thus reflect the practice of experienced C-L psychiatrists. Training and quality assurance practices are used to ensure reliability and were supervised by one of the authors (GCS). This involves checking for adherence to protocol and use of glossary definitions for all items, including interventions.21
Data collected that were relevant to this study were 1) demographic data; 2) reasons for referral and relevant problems as stated by the consultee (referring doctor) and by the consultant (psychiatrist)—up to 5 reasons/problems per patient; 3) DSM-III-R Axis I & II terminal diagnoses for the admission episode—up to 6 diagnoses per patient, "confirmed" (meets DSM-III-R criteria) or "rule-out" (considered likely but does not reach criteria because of insufficient data, confounding by physical factors, or subthreshold status); Axis IV, greatest of acute or enduring circumstances in last 12 months; Axis V, highest level for past 12 months; Karnofsky rating of physical functioning in past month; 4) ICD-9-CM (International Classification of Diseases, 9th Edition, Clinical Modification) diagnoses for the admission episode—up to 3 diagnoses; 5) interventions—drug and nondrug (note that the category "behavioral management" refers to nonpsychological interventions such as mobilization); and 6) hospital process—referring unit, length of stay in general hospital unit excluding any stay in a psychiatric unit, lagtime in referral, urgency of referral, administrative action, and discharge location.
From the referred cohort, we extracted a group of 1,360 patients with a depressive illness spectrum disorder (DISD),23 made up of 4 mutually exclusive subgroups: 1) "major depression," those patients with a DSM-III-R major depressive syndrome (major depression, bipolar disorder depressed or mixed or not otherwise specified (NOS), or cyclothymia); 2) "dysthymia," those with DSM-III-R dysthymia; 3) "organic/NOS," a collapsed group of those with DSM-III-R organic or substance-induced mood disorder and those with depressive disorder NOS (our practice has been to use this category when there is uncertainty as to the contribution of organic factors to what is nevertheless a depressive picture); and 4) "adjustment/depressed," those with DSM-III-R adjustment disorder with depressed mood or mixed emotional features. Each subgroup was divided into "confirmed" and "rule-out" groups on the basis of certainty of terminal (admission episode) diagnosis with respect to DSM-III-R criteria. The distribution of patients across subgroups is shown in Table 1.
Statistical Analyses
Comparison between the groups was performed by using analysis of variance (ANOVA) and t-tests for continuous variables and chi-squared tests for categorical variables, using the SPSS/PC+ package.24 The magnitude of effects was assessed by using the eta-squared25 and phi-squared26 coefficients for continuous and categorical variables, respectively. The larger the value of eta-square and phi-square, the greater the amount of variance in the dependent variable is accounted for by a particular independent variable. In the advent of a significant (P<0.01) ANOVA or chi-squared test, KnowledgeSEEKER27 was employed to form homogenous clusters28 and for exploratory decision-tree analysis, as previously applied to psychiatric data.29 Thus, in Table 2, the significance figures refer to the difference between clusters of the four subgroups.
|
RESULTS
|
Demographic data for those patients given a confirmed DISD diagnosis were as follows: mean age 55.1 (19.1 standard deviation), 60% female, 50% married, 93% white, 20% employed, 18% professionals, and 23% living alone. For the confirmed DISD group, a diagnosis from within this spectrum was made as the sole diagnosis in 73% of the cases, giving a comorbidity rate of 27% on Axis I & II. The most frequent comorbidities were personality disorder (9%), organic mental disorder (9%), psychoactive substance use disorder (7%), and somatoform and related disorders (6%). It should be noted that these are the comorbidity rates for confirmed diagnoses; the rates for rule-out diagnosis comorbidity were much higher, but they were not analyzed in the present study. Diagnostic data for the confirmed DISD cohort subgroups are shown in Table 2. There were no significant differences in distribution between sites. The dysthymia patients stand out as having more comorbidity with somatoform disorder and personality disorder. Levels of physical functioning in the month prior to admission (Karnofsky rating) were similar across the groups.
Intervention recommendations for the confirmed DISD group are shown in Table 2. The groups were differentiated on only one of the five investigative recommendations (ordering laboratory tests) and three of the six psychosocial intervention categories (psychological management, environmental manipulation, aftercare referrals). Psychotropic recommendations were made for 64% of the patients; 82% of these were to commence or continue a drug and 18% to discontinue one. The major depression group had the highest level of recommendations concerning antidepressants (including lithium), and with the organic/NOS group, the highest level for antipsychotics. Each group received low but similar levels of recommendations about anxiolytics and sedatives.
Table 3 shows the numbers of patients with either a confirmed or rule-out DISD diagnosis, for whom a recommendation to commence or continue an antidepressant were made. Antidepressants used were tricyclics 65%, tetracyclics 15%, selective serotonin reuptake inhibitors 12%, monoamine oxidase inhibitors 4%, and reversible inhibitors of monoamine oxidase 4%. There were no significant differences in rate of prescription between sites or referring services (grouped as medical vs. surgical). Although the confirmed major depression patients were the most likely to receive an antidepressant (in 69% of cases), 19% of confirmed adjustment/depressed and 10% of the rule-out adjustment/depressed patients were prescribed antidepressants.
The factors associated with the recommendation to commence or continue an antidepressant were explored in the confirmed major depression group (Table 4). Those receiving such a recommendation were older and more likely to be female, with a later referral and longer length of stay. Those variables with the greatest effect size (phi-square) were Axis III injury or poisoning, environmental manipulation recommended, and admission to psychiatry recommended, all of which were less likely, and history of physical illness in the last 12 months and higher Axis IV scores, both of which made prescription more likely. No combination of variables examined using KnowledgeSEEKER decision-tree analysis gave a classification accuracy that was more than 5% greater than that of a single variable such as referral for suicide risk. The degree of noncompliance with antidepressant recommendations (by consultees or patients) in the confirmed major depression group was 2%. A reaction to an antidepressant drug was considered likely in 13%.
|
DISCUSSION
|
Only 70% of patients given a DISD diagnosis were able to be so diagnosed at a confirmed level with respect to DSM-III-R criteria. The percentage was highest for the adjustment/depressed group and lowest for the organic/NOS group. It is likely that many factors contributed to this, including the limited time available to complete the assessment in many cases, and the relative inadequacy of the DSM-III-R classification in the field of physical/psychiatric comorbidity.10,17–19 The fact that the highest level of certainty of diagnosis was for the adjustment/depressed group is an indication of the fact that C-L psychiatrists find this less well-defined category to be a useful way of capturing the phenomena observed in this field.21 Thus, we included data about those patients given rule-out DISD diagnoses in the analyses of the use of antidepressants.
The rates of prevalence of comorbidity are low, especially for anxiety disorders. All diagnoses were systematically considered in the clinical process, but only confirmed comorbid diagnoses have been reported here; the prevalence of rule-out comorbid diagnoses was much higher.
The authors of the AHPCR guidelines made some suggestions about the treatment of depression in the medically ill.15 They recommended a 1- to 2-week observation period in nonsuicidal, nonpsychotic patients. Emphasis was placed on treating the medical condition first and the depression later, except for severe depression. Medications received relatively greater emphasis than psychotherapy.
In keeping with the AHCPR recommendations, for patients with a confirmed DISD diagnosis the C-L psychiatrists in this study made recommendations for additional laboratory tests in 1 in 20 cases, about additional medical-surgical consultations in 1 in 10 cases, and about increasing the vigor of medical treatment in 1 in 7 cases. To that degree they were not content with the extent to which any physical contribution to the depression had been assessed and was being treated. This is in keeping with the findings of Epstein et al.,20 who reported that expert C-L psychiatrists did not take for granted a statement that no organic etiology had been found.
Psychological management and social support were mainstays of treatment, more so for the adjustment/depressed patients than the others, but as much for major depressive illness as for dysthymia. This reflects the long history of support for a psychosocial understanding within C-L psychiatry, which has resisted the biomedical model.
Nevertheless, the pattern of antidepressant medication recommendations was in keeping with the guidelines and other literature.15,16,30,31 Forty-two percent of those in whom the psychiatrists considered a DISD likely received a recommendation to commence or to continue an antidepressant. This rate ranged from 69% for confirmed major depression to 10% for rule-out adjustment/depressed. The C-L psychiatrists were prepared to prescribe antidepressants in all subgroups, even in 28% of those for whom the diagnosis was made at a rule-out level only.
The question of what determines whether a patient diagnosed as having depression and being prescribed an antidepressant was explored for the confirmed major depression subgroup. It was not possible to produce a significant cluster of variables that would characterize such patients, but a number of significant individual correlates were identified. Older age and female gender made prescription of antidepressants more likely, as did a prior history of physical illness or a current pain disorder or terminal illness. The patients prescribed antidepressants had been functioning as well as those not prescribed (they were no more likely to have had a psychiatric illness in the past 12 months or to have higher Axis V ratings), despite being likely to have had more severe stressors (Axis IV). The C-L psychiatrists were using antidepressants with vigor in this cohort, especially in the presence of a neoplasm or disorder of the central nervous or musculoskeletal system. However, they were less likely to prescribe antidepressants in those whom they considered required more medical or surgical consultations. The patients prescribed antidepressants were referred later and had a longer length of stay; cause and effect cannot be distinguished in this study but warrant further exploration. Referring staff were accepting of the use of antidepressants; the noncompliance rate was only 2%. This has been achieved partly because the completion of the database form requires a return to the referring service to check compliance, and also by application of techniques shown to enhance compliance.21 Psychiatrists do not usually write orders.
Within the spectrum of nondrug interventions, psychological management stands out as being commonly used in those receiving antidepressants, indicative of a broad approach to management.
The incidence of adverse reactions to antidepressants was relatively low, 14% in this cohort compared with 32% reported by Popkin et al.32 This difference may reflect the fact that the newer agents have an improved side-effect profile. The nature of such side effects and reactions is reviewed by Series.31
The limitations of this study include the fact that, although considerable effort was put into the training and ongoing surveillance of staff in the use of the database and glossary of definitions, it remains a study of clinical practice. Application of DSM-III-R criteria was checked, but reliability and validity ultimately depend on the clinician's practice. We have argued previously for the proposition that data collected in the course of clinical work have a different type of validity that complements that produced by use of structured interviews.21
The extent to which the results have been influenced by the referral patterns of these hospitals can only be assessed by repeating the study in other centers and controlling for this variable. However, there were no differences between sites or grouped referring services in distribution of diagnoses within the DISD spectrum. We made only a limited exploration of variation in practice by psychiatrists by establishing that there were no significant differences between the general markers of this, site and grouped referring services. A larger study would be needed to control for the linking of psychiatrist with referring service.
It would be interesting to dissect out practice with respect to DISD diagnoses and treatments already in place when the patient was referred. This would require a different methodology.
The results may not be generalizable to the outpatient setting; comparison with the AHCPR guidelines is therefore of limited validity.
|
CONCLUSION
|
C-L psychiatrists in these hospitals were dealing with DISD in almost half of their inpatient referrals, though for one-third of these the diagnosis could not be made in DSM-III-R terms with certainty. Only half of the diagnoses made were in the major depression domain. Considerable use was made of the category Adjustment Disorder with depressed mood or mixed emotional features; it had the highest certainty of diagnosis. They used a broad spectrum of biopsychosocial interventions, which included the use of antidepressants in 42%; this was not confined to cases of confirmed major depression, and the practice included the full range of physical illness and type and degree of certainty of diagnosis. The prevalence of comorbidity on Axis I and II—particularly for somatoform disorders, psychoactive substance use disorder, and personality disorder—complicates the work. There is a dearth of data that would serve as a research basis for guidelines about how to deal with physical/psychiatric comorbidity of this type; we must rely on guidelines produced by consensus review,33 to which the findings reported here contribute.
|
ACKNOWLEDGMENTS
|
The authors thank Paul Low for data management; the psychiatry trainees and consultants who faithfully contributed to the database; and Dean McKenzie, who performed the statistical analyses. This work was supported by Upjohn Australia Pty., Ltd., and the Buckland Foundation.
|
REFERENCES
|
-
Mayou R, Hawton K: Psychiatric disorder in the general hospital. Br J Psychiatry 1986; 149:172–190[Abstract/Free Full Text]
-
Clarke DM, Minas IH, Stuart GW: The prevalence of psychiatric morbidity in general hospital inpatients. Aust N Z J Psychiatry 1991; 25:322–329[Medline]
-
Wells KB, Golding JM, Burnam MA: Psychiatric disorder and physical functioning in a sample of the Los Angeles population. Am J Psychiatry 1988; 145:712–717[Abstract/Free Full Text]
-
Wells KR, Golding JM, Burnam MA: Chronic medical conditions in a sample of the general population with anxiety, affective and substance use disorders. Am J Psychiatry 1989; 146:1440–1446[Abstract/Free Full Text]
-
Wells KR, Rogers W, Burnam A, et al: How the medical comorbidity of depressed patients differs across health care settings: results from the Medical Outcomes Studies. Am J Psychiatry 1991; 148:1688–1696[Abstract/Free Full Text]
-
Wells KB, Stewart A, Hays RD, et al: The functioning and well-being of depressed patients: results from the Medical Outcomes Study. JAMA 1989; 262:914–919[Abstract/Free Full Text]
-
Coulehan JL, Shulberg HC, Block MR, et al: Depressive symptomatology and medical co-morbidity in a primary care clinic. Int J Psychiatry Medicine 1990; 20:335–347[Medline]
-
VonKorff M, Ormel J, Katon W, et al: Disability and depression among high utilizers of health care. Arch Gen Psychiatry 1992; 49:91–100[Abstract/Free Full Text]
-
Rovner BS, German PS, Brant LJ, et al: Depression and mortality in nursing homes. JAMA 1991; 265:993–996[Abstract/Free Full Text]
-
Zinbarg RE, Barlow DH, Liebowitz M, et al: The DSM-IV field trial for mixed anxiety-depression. Am J Psychiatry 1994; 151:1153–1162[Abstract/Free Full Text]
-
Sherbourne CD, Wells KB, Hays RD, et al: Subthreshold depression and depressive disorder: clinical characteristics of general medical and mental health specialty outpatients. Am J Psychiatry 1994; 151:1777–1784[Abstract/Free Full Text]
-
Alexopoulos GS, Vrontou C, Kakuma T, et al: Disability in geriatric depression. Am J Psychiatry 1996; 153:877–885[Abstract/Free Full Text]
-
Bingefors K, Isacson D, Von Knorring L, et al: Antidepressant-treated patients in ambulatory care: mortality during a nine-year period after first treatment. Br J Psychiatry 1996; 169:647–654[Abstract/Free Full Text]
-
Judd LL, Paulus MP, Wells KB, et al: Socioeconomic burden of subsyndromal depressive symptoms and major depression in a sample of the general population. Am J Psychiatry 1996; 153:1411–1417[Abstract/Free Full Text]
-
Kathol R, Katon W, Smith GR, et al: Guidelines for the diagnosis and treatment of depression for primary care physicians: implications for consultation-liaison psychiatrists. Psychosomatics 1994; 35:1–12[Free Full Text]
-
American Psychiatric Association: American Psychiatric Association Practice Guideline for Major Depressive Disorder in Adults. Am J Psychiatry 1993; 150(suppl 4):1–26
-
Stoudemire GA, Strain JJ, Hales RE: DSM-IV issues for consultation psychiatry. Psychosomatics 1989; 30:239–244[Free Full Text]
-
House A: Mood disorders in the physically ill—problems of definition and measurement (invited review). J Psychosom Res 1988; 32:345–353[Medline]
-
Cavanaugh SVA: Depression in the medically ill: critical issues in diagnostic assessment. Psychosomatics 1995; 36:48–59[Abstract/Free Full Text]
-
Epstein SA, Gonzales JJ, St. Onge J, et al: Practice patterns in the diagnosis and treatment of anxiety and depression in the medically ill. Psychosomatics 1996; 37:356–367[Abstract/Free Full Text]
-
Smith GC, Clarke DM, Herrman HE: Establishing a consultation-liaison psychiatry clinical database in an Australian general hospital. Gen Hosp Psychiatry 1993; 15:243–253[Medline]
-
Hammer JS, Strain JJ, Lyerly M: An optical scan/statistical package for clinical data management in C/L psychiatry. Gen Hosp Psychiatry 1993; 15:95–101[Medline]
-
Clarke DM, McKenzie DP, Smith GC: The recognition of depression in patients referred to consultation-liaison psychiatry. J Psychosom Res 1995; 39:327–334[Medline]
-
Norusis MJ, SPSS Inc.: SPSS/PC+ 4.0 base manual for the IBM PC/XT/AT and PS/2. Chicago, IL, SPSS Inc., 1990
-
Fisher RA: Statistical Methods for Research Workers, 14th Edition. New York, Hafner Press, 1970
-
Liebtrau AM: Measures of Association. Beverly Hills, CA, Sage, 1983
-
Angoss Software International: KnowledgeSEEKER User's Guide. Toronto, Ontario, Canada, Angoss Software International, 1993
-
Smith GC, Clarke DM, Handrinos D: Recognising drug and alcohol problems in patients referred to consultation-liaison psychiatry. Med J Aust 1995; 163:307–312[Medline]
-
McKenzie DP, McGorry PD, Wallace CS, et al: Constructing a minimal diagnostic decision tree. Methods Int Med 1993; 32:161–166
-
Snyder S, Strain JJ, Wolf D: Differentiating major depression from adjustment disorder with depressed mood in the medical setting. Gen Hosp Psychiatry 1990; 12:159–165[Medline]
-
Series HG: Drug treatment of depression in medically ill patients. J Psychosom Res 1992; 36:1–16[Medline]
-
Popkin MK, Callies AL, MacKenzie TM: The outcome of antidepressant use in the medically ill. Arch Gen Psychiatry 1985; 42:1160–1163[Abstract/Free Full Text]
-
Bass C, Benjamin S: The management of chronic somatisation. Br J Psychiatry 1993; 162:472–480[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
G. C. Smith, D. M. Clarke, D. Handrinos, and D. P. McKenzie
Consultation-Liaison Psychiatrists' Use of Antidepressants in the Physically Ill
Psychosomatics,
June 1, 2002;
43(3):
221 - 227.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. C. Smith, D. M. Clarke, D. Handrinos, and T. Trauer
Adverse Reactions to Antidepressants in Consultation-Liaison Psychiatry Inpatients
Psychosomatics,
June 1, 2002;
43(3):
228 - 233.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. C. Smith, D. M. Clarke, D. Handrinos, A. Dunsis, and D. P. McKenzie
Consultation-Liaison Psychiatrists' Management of Somatoform Disorders
Psychosomatics,
December 1, 2000;
41(6):
481 - 489.
[Abstract]
[Full Text]
|
|
|
|
|
|
|
D. M. Clarke, A. J. Mackinnon, G. C. Smith, D. P. McKenzie, and H. E. Herrman
Dimensions of Psychopathology in the Medically Ill: A Latent Trait Analysis
Psychosomatics,
October 1, 2000;
41(5):
418 - 425.
[Abstract]
[Full Text]
|
|
|
|
|
|
|
D. Handrinos, D. McKenzie, and G. C. Smith
Timing of Referral to a Consultation-Liaison Psychiatry Unit
Psychosomatics,
August 1, 1998;
39(4):
311 - 317.
[Abstract]
[Full Text]
|
|
|
Get information about faster international access.
Privacy Policy
Copyright © 1998
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
