liaison psychiatry

Interventions in Consultation/Liaison Psychiatry Part I: Patterns of Recommendations Frits J. Huyse, M.D., Ph.D., James J. Strain, M.D., and Jeffrey S. Hammer, M.D.

Abstract: In a previous study, a checklist and a schema for operafionalized interventions have been described [II. In this study, these operafionalized interventions have been used in clinical practice in 820 cases. In 287 cases (35%), recommendations focusing on the medical treatment, other than diagnostic action (41%) and medication (680/o),were provided. This emphasis on the intensify of medical treatment is an unreported finding. The further distribution of recommendations over the different intervention domains was: obtaining additional psychosocial information, 30%; psychosocial management on the ward, 61%, specifically ifs organization; discharge planning, 41%; and affercare management, 24%. Recommendations infrequently used included: monitoring of cognition and behavior, referrals to occupational therapy and alcohol- and drug-related facilities, and a detailed specification of postdischarge care. Since consulfee concordance with the consultant’s recommendations is important to treatment outcome, this method of a systematic recording of operafionalized recommendations is a first step toward enhancing the evaluation of treatment recommended and provided by consultation/liaison (C/L) psychiatrists. Moreover, these operationalized recommendations help advance the specification of protocols for psychiatric intervention studies.

Introduction Epidemiological reports of health care delivery by consultation/liaison (C/L) psychiatrists are limited with regard to the description of the components of patient management [l]. For example, it is difFrom the Free University Hospital (F.J.H.), Amsterdam, Holland; Division of Consultation Psvchiatrv and Behavioral Medicine (J.J.S.), The Mount Sinai School of Medicine, New York, New York; and VAMC West Los Angeles f1.S.H.). v ~I I. Los Angeles, California. Address reprint requests to: Frits J. Huyse, M.D., Ph.D., Consultation Liaison Psychiatry, P.C.D. 010.D2, Free University Hospital, De Boelelaan 1117, P.O. Box 7057, 1007 MB Amsterdam. the Netherlands. General Hospital Psychiatry 12, 213-220, 1990 0 1990 Elsevier Science Publishing Co., Inc. 655 Avenue of the Americas, New York, NY 10010

ficult to understand the meaning of psychotherapy and ward management [l]. Since only two followup visits by the psychiatrist are provided for two thirds of the patients seen [2,3], psychotherapy in a classical sense is not delivered. The accurate assessment of the psychiatrist’s effectiveness in the general hospital requires that their interventions be described both qualitatively and quantitatively [4]. First, such specification enhances the capacity of consultees in other medical specialties to understand the efforts of C/L psychiatrists [5]. Second, it allows the consultant to evaluate and enhance treatment programs [l]. Third, measurable interventions are a basic requisite for outcome research [4,6,7]. Mumford et al. [8] have reported that timelimited psychiatric and psychologic interventions using simple therapeutic techniques were most effective in a number of medical conditions. Furthermore, Smith et al. [9] have provided evidence that management strategies, suggested by consultants but performed by primary care physicians, can alter patient behavior. Smith et al. [9], Popkin et al. [lO,ll], and Callies et al. [12] underscore that the effectiveness of the C/L consultant depends on the dedicated participation of other members of the health care system to implement their interventions. They emphasize the importance of the specification of recommendations to enhance the collaboration (concordance) of the consultee and ward staff [13]. Based on the above findings, a checklist divided into domains for biopsychosocial management interventions was developed by the author (F.J.H.); this can be employed either by the psychiatric consultant or recommended to the ward staff [l]. In summary, the intervention schema provides a systematic focus for assessment, 213 ISSN 0163~8343/90/$3.50

F. J. Huyse et al.

ward, discharge, and aftercare management on four axes: biologic, psychologic, health care organization, and social support. The primary obiectives of this study are first to describe the pattern of recommendations provided by consultants and their distribution over the domains of the schema, and then to distinguish components of the interventions that are performed by the consultant or the ward staff. Specific relationships among the main demographic, psychiatric diagnoses, physical illness, and hospitalization variables and the recommendations (intervention profiles) will be described separately [ 141.

Subjects and Method Sample Medical charts of 820 consecutive consultations performed by the C/L Psychiatry Service of the Free University Hospital, Amsterdam, the Netherlands, between January 1985 and January 1987 were examined. Suicide attempts by the ingestion of pills were excluded, if patients were directly discharged from the Intensive Care Unit of the Department of Medicine (n = 187). The psychiatric management of these patients is under the direct care of the psychiatric consultant, who carries the management of the patient into effect alone. Moreover, since these patients stay only for a very short time in the hospital, as such they constitute a distinct category. Children and neonates were also excluded from this sample, as the adult C/L service does not provide cons;lts for them. Finally, repeat consultations in the same patient later in the hospitalization or during a second admission were excluded. The C/L service employed a faculty of one fulltime C/L psychiatrist (F.J.H.), two PGY IV psychiatric consultation residents, and a full-time liaison clinical nurse specialist. The consultation procedures are comparable to those described in the U.S. literature [15,16]. During the study, the Free University Hospital (614 beds) had 33,218 admissions with a mean length of stay (LOS) of 11.5 days during the study. The consultation rate was 3.2% (2.6% if suicide attempts are excluded). The bed per capita size and the composition of the C/L service is comparable with other C/L services in Dutch university hospitals at that time [17,18]. However, as a result of a collaborative liaison psychiatry arrangement for the extensive psychiatric needs of the oral cavity 214

tumor patients, an overrepresentation of consultations from the otolaryngology (ENT) department occurred.

Instruments and Assessment Definition of Recommendations The C/L service employs an especially designed three-page consult form. It includes the following: (1) the formulation of the problem, the assessment, and the conclusion; (2) the description of the intervention including the recommendations for the wardstaff and the role of the psychiatric consultant; and (3) follow-up observations. Recommendations were defined as (1) all remarks found on the consultation form “recommendation page” and (2) statements identified on the follow-up note as “recommendations” and/or including the phrase “I would recommend.” All remarks by the consultant that he or she would perform or already performed the recommendation alone were listed separately. Recommendations may contain both the encouragement or the discouragement of action. Finally, recommendations were excluded if it was not specified whether an action should be taken or not. For instance, the consultant could have written the following: “One could consider the disposition of a patient to a psychiatric facility.” This consideration would not be included in the study.

Categories of Recommendations All recommendations were categorized according to the previously developed checklist [l]. If items could not be assigned, they were listed separately, screened, and if appropriate, additional categories were constructed.

Assessment of Recommendations Two studies were undertaken to evaluate the reliability of the assessment of recommendations from the consultant’s chart note. All consultations of the first 360 patients were individually screened for recommendations by the research assistant and the principle investigator. In 308 cases, 1650 recommendations other than medication and physical diagnostic action were observed. All recommendations that could not be agreed upon by both of the raters were excluded from the study (n = 61, inter-rater agreement 96%). A second cohort of cases admitted during 1 week (n = 13) was also

Interventions in C/L Psychiatry: Part I

examined for reliability. Both raters identified 54 recommendations, of which 50 were the same. Because of the high interrater agreement, the research assistant consulted the primary investigator in subsequent cases only when there was doubt.

Assessment of Patient and Hospitalization Variables Demographic, psychiatric, and consultation characteristics were collected on a structured data base form utilized for C/L psychiatry consultations [3,191.

Data Analysis The data have been analyzed on a mainframe com-

Table 1. Psychiatric diagnoses: DSM-III Axis I Percentage of all cases (n = 820) Organic mental disorders a. Axis-III-related b. Substance-induced c. Primary dementias

Substance use disorders Schizophrenia Affective disorders Anxiety disorders Somatoform disorders Adiustment disorders None/V-codes Diagnosis deferred

Others

37 57 25 19 11 4 8 2 3 13 16 2 4

puter with the use of SPSS-X.

Results Specific Characteristics of the Consultation Sample (CS) Sociodemographics. The characteristics of the CS cohort were the following: single or divorced (32%), widowed (15%), living with spouse or children (57%), alone (32%), and living in institutions (10%). With regard to financial support, 12% were employed, 21% had disability benefits, and 29% received retirement pay. Physical illness. Twenty-four percent of the patients seen have been admitted to the hospital for an oncologic disease, 13% for diseases of the nervous system, and 13% for injuries. Only 2% of the patients have been admitted for an infectious disease as a primary physical diagnosis; there were no AIDS patients within the consultation sample. Psychiatric history and assessment. Seventeen percent of the CS had a previous history of psychiatric admissions, and 12% had experienced only outpatient psychiatric treatment. The psychiatric diagnoses of the CS sample are described (Table 1). Substance use disorders were the most prevalent secondary diagnosis (6%). Axis II personality disorders were observed in 18% of the CS cohort. Patient-staff conflicts according to specific criteria occurred in 22% of the CS [20].

Recommendations to the consultee. The most prevalent categories (in 10% or more of the cases) from the checklist (Table 2), a detailed description of the interventions distributed over the different domains (Table 3), and the quantitative distribution over the main domains (Table 4) are presented. Each separate recommendation is presented both as frequency and as percentage of all cases for whom a specific recommendation was suggested. Recommendations carried into effect by the consultant. Arrangements for the disposition to psychiatric facilities (both hospital and nursing homes) were performed by the consultants themselves as they are the only persons with access to these facilities. Other recommendations carried into effect by the consultant are described (Table 5).

Discussion Although a difference with regard to ENT referrals is evident, the sample is representative of patients consulted on other C/L services [3]. The first use of the schema and checklist demonstrated that psychiatric interventions by C/L psychiatrists are mainly composed of elements representing the following different domains: biologicul management (enhancement of physical assessment, application of medication management, management of the patient’s medical status); psychosocial management

(enhancement of the psychosocial diagnoses, psy215

F. J. Huyse et al.

Table 2. Rank-ordered

recommendations”

Total Medication; neuroleptics Medication; hypnotics Medication; anxiolytics Hematological tests Organize the patient/staff interaction Obtain more information from PCP Arrangements for outpatient care Educate patient; cognitive training Nonmedical consultations Educate patient; provide information Educate or counsel the family Character management of the patient Determine timing of discharge Increase patient activities Redirect patient activities; activate Medication; antidepressants Make arrangements for visits Objects for orientation Inform PCP on postdischarge management Discharge to psychiatric hospital Medical diagnostic procedures Intensify medical treatment Obtain more information from family Consultations from medical specialists Others

Number of recommendations

% of aII cases (n = 820)

5092 604 360 313 290 188 176 172 172 171 165 157 136 136

98 (n = 804) 38 27 27 23 20 21 24 19 19 19 18 13 13

129 129 117 113 113 109 104 101 94 93 950

13 10 13 13 13 12 12 11 11 10 -

“Providedin 10% or more of all cases.

chological management of the patient, and/or his or her family in relation to the unit-staff); discharge management; and aftercare management. The assessment and treatment domains are more frequently used than those for discharge and aftercare (Table 4). Specifically, the most frequent recommendations involve medication, laboratory tests, and suggestions for the unit staff as to who should do what with regard to the biopsychosocial care of the patient. Of particular importance are recommendations concerning the flow of information-specifically the meaning and course of illness and treatment-

among patient, family, and unit staff. Furthermore, management of the patient’s behavior and its mobilization were often recommended. Although the literature seldom addressed the issue of the consultants’ role in the medical management of the patient, for example, treatment of the hypoxia secondary to pneumonia or decreased medical intervention in somatizers, it was an im216

portant focus of the interventions of this study. These medical recommendations were necessary because of the staff’s misunderstanding of the meaning of the patient’s behavior, and their lack of knowledge of psychopathology. This could result in the staff erroneously assigning a psychiatric

diagnosis and ignoring the etiologic role of organic factors in the psychiatric diagnosis as well as the production of a negative staff attitude toward the patient. This inappropriate interpretation, also seen in staff conflicts, could result in insufficient medical treatment [20]. Consequently, the consultant needs to encourage the staff to assume necessary medical treatment and consider whether psychologic and environmental recommendations are appropriate as well. On the other hand, the consultant might need to discourage the staff from pursuing unnecessary medical treatment or further investigations, for instance, in those patients with somatoform disorders [lo]. In all phases of the intervention-diagnostic, ward management, dis-

Interventions in C/L Psychiatry: Part I

Table 3. Recommendations and checklist ]2]

organized

according

to the domains

of the intervention

schema

Number of recommendations

Percentage of cases

93 86 204 106 39 36

10 10 23 12 5 4

27

3

176 16

21 2

94

11

604 129 313 360 63 21 20 28

38 10 27 27 6 2 2 3

Biologic other Restraints Increase patient’s daily activities Decrease patient’s daily activities Continue substance abuse (nicotine) Discontinue substance abuse (alcohol) Decrease medical treatment Abstain medical treatment in terminal illness Intensify medical treatment Refer to physiotherapy

63 129 33 11 8 7 10 101 48

7 13 4 1 1 1 1 11 6

Psychologic Educate patient; cognitive training Educate patient; provide information Ignore behavioral disturbances Approach to character and/or coping

176 165 13 123

19 19 2 13

188 37 126 41

20 4 13 4

Diagnostic Biologic Consultations from medical specialists Lab tests; routine specific: vitamins, hormones Medical diagnostic procedures Obtain old medical records Decrease or stop biologic diagnostic procedures Psychosocial Monitor patients behavior and/or psychologic functions Health care system Obtain more information from general practitioner/P0 social work, interpreter Support system Obtain more information from family Ward management Biologic medication Neuroleptics Antidepressants Anxiolytics Hypnotics Vitamins Methadon Analgetics Other medication

Health care system Organize the patient/staff interaction Organize the family/staff interaction Provide objects for orientation (Do not) change wards Consultation from Psycho Geriatric community service Social work Occupational therapy Others

20 58 38 33 217

F. J. Huyse et al.

Table 3. bnfinued~ Number of recommendations Page consultant in case of emergency Day service Off-office hour service Weekend dispensation No Yes Keep the patient on the ward

Percentage of cases

76 73

9 9

13, 8 4

2 1 1

117 157

13 18

76 34 22 113 36 18

9 4 2 13 4 2

Support system Make arrangements for visits Educate and/or counsel family Discharge management Planning of term with Ctee Accelerate Delay Inform general practitioner on postdischarge management Incorporate psychiatric conclusion in letter of discharge Send an extra copy of the letter of discharge to Discharge to Home Psychiatric hospital Psychiatric nursing home Other

68 57 52 4

After care by General practitioner Alcohol and/or drug facility Mental health outpatient C/L outpatient service Recommendations concerning the approach to be taken by PCP Other Total number of recommendations

Table 4. Recommendations: Percentage of cases per main domain Biologic Diagnostic Management: Management: Psychosocial Diagnostic Ward Discharge Aftercare

medication other

41 68 35 30 61 41 24

charge planning, and aftercare-the organization of the health system is an important aspect of the intervention. It concerns the obtainment of in218

36 15 107 50 46 67

4 2 12 6 6 7

5092

98

formation (diagnostic), the distribution of tasks between members of the ward staff (ward management), and the continuation of care (discharge and aftercare management). However, aftercare providers are infrequently informed on the specific approach to be taken with regard to the psychiatric comorbidity (see below). The consultants se2domofferedrecommendations for the following: (1) monitoring of behavior and cognition (n = 27, 3%); (2) consultations for occupational therapy (n = 38, 5%); (3) referrals to substance use disorder facilities (n = 15, 2%); and (4) specified and written recommendations for the content of the postdischarge care (n = 46, 6%) (as mentioned in Section III 2.2 and 2.3 of the checklist [l]). Recommendations that were provided in less

Interventions in C/L Psychiatry: Part I

Table 5. Recommendations performed

most frequently by consultant Number of recommendations Percentage”

Obtain more information from PCP Inform PCP on postdischarge management Obtain more information from family Educate or counsel the

family

86

49

41

39

32

35

30

21

“Percentage of this specific recommendation, taken into effect by the consultant himself.

than 10% of the cases (n = 950) were distributed over all domains. The psychiatrists’ infrequent suggestion to monitor patients’ behavior results in a missed opportunity for the unit staff to make important observations that are essential for diagnosis and management of psychiatric illness. A quarter of the organic mental disorders is substance induced (9%), 11% of the cases have primary substance use disorders and 6% a secondary diagnosis; the referral rate to outpatient alcohol- and drug-related facilities is 2%. This low percentage reflects the tenuous integration of referrals to substance abuse disorder institutions from general hospitals in the Netherlands. Of equal importance is the fact that hospitalized patients often refuse referrals to outpatient facilities, which are not a part of ongoing hospital services. Consequently, patients with serious medical illness related to substance use disorders often receive inadequate postdischarge medical and psychiatric outpatient treatment that may ultimately result in further medical deterioration and enhanced use of medical resources. Moreover, the low referral rate by consultants might reflect upon their own feelings of impotence to influence the course of alcoholism in any setting. In fact, a simple technique like a limited nurse interview combined with a booklet with proper information on the outcome of drinking behavior after 16 months has been reported to be effective [21]. Finally, specific treatment plans to enhance the continuity of care after discharge or to prevent relapse are seldom made. This is similar to findings

of an English study [22], demonstrating the lack of preparation of the elderly for discharge from the hospital. The importance of specific recommendations for postdischarge care in the outpatient setting was also demonstrated by Smiths intervention study of somatizers [9]. In conclusion, this study offers a model for the organization of operationalized interventions that has been empirically tested. This is a first step needed for the evaluation of the profile and fate of the consultants’ recommendations. This will allow for an enhancement of consulting techniques and the development of more organized C/L health service delivery. The finding that the organization of the health care system was an important focus of attention for our consultants underscores the need to distribute tasks among health care providers in patients with medical and psychiatric comorbidity. Subsequently, the organization of tasks of several health care providers will result in a real operationalized biopsychosocial intervention.

Consequences

for CZinical Pracfice

and Research

The majority of the patients seen by the psychiatric consultant have both serious psychiatric and physical illnesses. They are often confused, depressed, and/or anxious and consequently more dependent on the support of others. Supportive psychotherapy, medication, and physical diagnostic action are only some of the many possibilities of intervention [l]. The full range of services would also contain the appropriate use of unit staff, family, and primary care physician resources. Research protocols mandate careful specification of the intervention as well as whether or not they are complied with in order to evaluate the effectiveness of psychiatric recommendations in the medically ill. Such specification of the intervention in behavioral terms was lacking in the important cost offset study of Levitan and Kornfeld [23], and it confounded the interpretation of the results. The current study demonstrated the applicability of the checklist in the practice of C/L psychiatry. Yet, it is suggested that such a checklist can only serve as a reference guide: The recommendations have to be formulated according to the specific clinical situation. Although multivariate analyses are needed, the results of the study indicate that lacking systematic review of the possible interventions provided by consultants will result in missed opportunities for treatment. A more systematic and operationalized

219

F. J. Huyse et al.

approach to biopsychosocial interventions will enhance the effectivity of the consultants as (1) they will consider the full range of possible interventions, (2) be able to formulate more standardized approaches toward frequently encountered problems, and (3) enhance the range of psychiatric interventions in the medically ill. It encourages liaison psychiatrists to conduct an ongoing review of all intervention possibilities with members of the unit staff. Such a screening assessment approach encourages the enactment of those psychiatric recommendations deemed essential for optimal care of patients with psychiatric and medical comorbidities. At the same time, such an approach promotes the systematic evaluation of the effects of the completed recommendations on the quality and outcome of care. In keeping with the precepts of the marketplace model, the outcome of these integrated psychosocial/psychiatric interventions can be shared with the directors of medical and surgical residencies [5]. For every department interested in the services provided by the C/L service, a careful description of the collaborative effort can be provided. The current study highlighted the need for the delineation of specified and measurable interventions. The article that follows focuses on whether the unit staff implement the specific recommendations (see Part II [24]). -

References 1. Huyse FJ, Hengeveld MW, Strain JJ, Hammer JS, Zwaan T: Interventions in consultation-liaison psychiatry: The development of a schema and checklist for operationalized interventions. Gen Hosp Psychiatry 10:88-101, 1988 Too 2. Sensky T: The general hospital psychiatrist: many tasks and too few roles? Br J Psychiatry 148:151-158, 1986 3. Hengeveld MW, Rooymans HGM, Vecht-van den Bergh R: Psychiatric consultations in a Dutch university hospital: A report on 1814 referrals compared with a literature review. Gen Hosp Psychiatry 6:271279, 1984 4. Pincus HA: Making the case for consultation-liaison psychiatry: Issues in cost-effectiveness analysis. Gen Hosp Psychiatry 6:173-179, 1984 5. Guggenheim FG: A marketplace model of consultation psychiatry in the general hospital. Am J Psychiatry 135:1380-1383, 1978 6. Lyons JS, Hammer JS, Wise T, Strain JJ: Consultation liaison psychiatry and costeffectiveness research: A review of methods. Gen Hosp Psychiatry 7:302-308, 1985 7. Cohen-Cole SA, Pincus HA, Stoudemire A, Fiester S, Houpt J: Recent research developments in con220

8.

9.

10.

11.

sultation liaison psychiatry. Gen Hasp Psychiatry 8:316-330, 1986 Mumford E, Schlesinger HJ, et al.: A new look at evidence about reduced cost of medical utihsation following mental health treatment. Am J Psychiatry 141:1145-1158, 1984 Smith GR, Monson RA, Ray DC: Psychiatric consultation in somatization disorder, a randomized controlled study. N Engl J Med 314:1407-1413, 1986 Popkin MK, Mackenzie TB, Hall RCW, Garrard J: Physicians’ concordance with consultants’ recommendations for psychotropic medication. Arch Gen Psychiatry 36:386-389, 1979 Popkin MK, Mackenzie TB, Callies AL: Consultees’ concordance with consultants’ recommendations for diagnostic action. J Nerv Ment Dis 168:9-12, 1980

12. Callies AL, Popkin MK, Mackenzie TB, Mitchell J: Consultees’ representations of consultants’ psychiatric diagnoses. Am J Psychiatry 137:1250-1253,198O 13. Popkin MK, Mackenzie TB, Callies AL: Consultation-liaison outcome evaluation system. Arch Gen Psychiatry 40:215-219, 1983 14. Huyse FJ: Interventions in C/L psychiatry. Patterns of recommendations: Related factors. In preparation. 15. Lipowski ZJ: Review of consultation psychiatry and psychosomatic medicine I and II. Psychosom Med 29:153-171, 201-224, 1967 16. Hackett TP, Cassem NH: Handbook of General Hospital Psychiatry. St. Louis, The CV Mosby Company, 1978. 17. Huyse FJ, Hengeveld MW: The development of C/L psychiatry in the Netherlands: Its social psychiatric heritage. Gen Hosp Psychiatry 11:9-15, 1988. 18. Hengeveld MW: Consultatieve Psychiatric in Nederland: 1984. Internal publication, Netherlands Consortium for Consultation/Liaison Psychiatry, 1987 19. Taintor Z, Spikes J, Gise LH, Strain JJ: Recording psychiatric consultations. A preliminary Report. Gen Hosp Psychiatry 1:139-149, 1979 20. Hengeveld MW, Rooymans HGM, Hermans J: Assessment of patient-staff and intrastaff problems in psychiatric consultations. Gen Hosp Psychiatry 9:2530, 1987 21. Chick J, Lloyd G, Crombie E: Counseling problem drinkers on medical wards: A controlled study. Br Med J 290:965-967, 1985 22. Victor CR, Vetter NJ: Preparing the elderly for discharge from hospital: A neglected aspect of patient care? Age Ageing 17:155-163, 1988 23. Levitan SJ, Kornfeld DS: Clinical and cost benefits of liaison psychiatry. Am J Psychiatry 138:790-793 1981 24. Huyse FJ, Strain JJ, Hammer JS: Interventions in consultation/liaison (C/L) psychiatry. Part II. Concordance. Gen Hosp Psychiatry 12:221-231, 1990 25. Querido A: Inleiding tot een Integrale Geneeskunde. De Tijdstroom, 1955 26. Kaufman MR: The role of a psychiatrist in a general hospital. Psychiatr Q 27:367-381, 1953 27. Strain JJ, Strain JS: Liaison psychiatry. In JG Howells (ed), Modern Perspectives in Clinical Psychiatry. New York, BrunneriMazel, 1988, pp. 76-101