Past Mental Health or Substance Use Treatment History and Psychiatric Differential Diagnosis in Consultation-Liaison Patients

Past Mental Health or Substance Use Treatment History and Psychiatric Differential Diagnosis in Consultation-Liaison Patients JAMES

R.

RICHARD

M.D. HALL, M.D.

RUNDELL,

c.w.

The purpose of this study was to determine whether frequencies of any current primary or secondary (organic) psychiatric diagnoses in consultation-liaison (C-L) patients are associated with patient report ofpast history of mental health or substance use disorder treatment. Clinical information recorded in 134 CoL patients was retrospectively analyzed with the chi-square test. two-tailed. or Fisher s Exact Test when the sample sizes were $ 5. The results revealed that the overall likelihood of the existence of a secondary psychiatric disorder diagnosis is elevated when there is no treatment history (P =0.(09). A primary psychiatric disorder is more likely to be diagnosed when there is treatment history (P = 0.009). Specific psychiatric diagnoses more likely to be present when there is no treatment history are organic mood disorder. depressed (P = 0.020) and adjustment disorder (P = 0.010). Specific diagnoses more likely to be present when there is treatment history are psychoactive substance-withdrawal syndrome (P = 0.014) and alcohol abuse or dependence (P = 0.028). The authors conclude that psychiatric treatment history can help CoL psychiatrists with differential diagnostic considerations. (Psychosomatics 1997; 38:262-268)

D

uring psychiatric consultations on medical-surgical patients, information about psychiatric predisposition is believed to be essential in developing a case formulation. 1-4 The absence of past mental health (MH) or substance use disorder (SUD) history increases many consultation-liaison (C-L) psychiatrists' Received February 27.1996; revised April 19. 1996; accepted May 24. 1996. From the Department of Psychiatry. Uniformed Services University of the Health Sciences (USUHS). Bethesda. Maryland; and the Psychiatric Programs. Florida Hospital Center for Psychiatry. Orlando. Florida. Address reprint requests to Dr. Rundell. Dept. of Psychiatry. USUHS. 4301 Jones Bridge Rd.. Bethesda. MD 20814-4799. Copyright © 1997 The Academy of Psychosomatic Medicine. 262

suspicions that a current psychiatric syndrome may have a toxic or general medical etiology.5-7 Unfortunately, confirming and ruling out past psychiatric diagnoses can be difficult and timeconsuming for clinicians, researchers, and patients. Instead, some CoL psychiatrists ask patients whether they have ever been "treated" before for MH or SUD problems. This approach is probably more reliable and certainly easier than attempting to reconstruct whether a patient ever met criteria for one or more psychiatric illnesses. However, some patients who have psychiatric disorders that have never been treated will be misidentified by this approach as lacking psychiatric predisposition. Similarly, patients may have been treated for a psychiatric

PSYCHOSOMATICS

Rundell and Hall

condition by a primary care physician and not readily identify that treatment as a MH intervention. There is a need for research to establish whether there are associations in C-L patients between current psychiatric diagnosis frequencies and psychiatric predisposition, as detected by patient report of past MH or SUD treatment. If specific diagnoses can be identified that are relatively more or less likely when there is MH or SUD treatment history, C-L psychiatrists will have another empirically based tool in their armamentariums to aid in the differential diagnosis of psychiatric syndromes in medical-surgical patients. For example, if a hospitalized medical-surgical patient has a depressive syndrome and reports no past history of psychiatric treatment, the clinician has a reason to intensify his or her search for a toxic or medical etiology, particularly because research has confirmed that a diagnosis of secondary (organic) mood disorder (depressed) is indeed more likely. This study has the following hypotheses: I) primary psychiatric disorders, as a group, are more likely to be diagnosed when C-L patients report a history of previous MH or SUD treatment than when they do not; 2) secondary (organic) psychiatric disorders, as a group, are more likely to be diagnosed when C-L patients report there is not a history of previous MH or SUD treatment than when they report there is; and 3) some specific psychiatric diagnoses among C-L patients are more highly associated with the presence or absence of reported history of MH or SUD treatment than other diagnoses. METHODS Between July 1992 and October 1992, 155 consecutive patients who received psychiatric consultations were clinically assessed, according to a standard C-L service-operating procedure. The following information was collected at the time of initial consultation and recorded for each patient by a single examiner (JR): age, gender, ethnicity, DSM-III-R8 psychiatric diagnosis, and whether the patient acknowledged there had ever been MH or SUD inpatient or VOLUME 38 • NUMBER 3 • MAY - JUNE 1997

outpatient treatment. This information was recorded on the C-L service's clinical information form. Psychiatric diagnoses were made, according to DSM-III-R criteria. 8 Information about past MH or SUD treatment was gathered, according to the standard C-L service-operating procedure, by asking each patient the following questions: 1. .. Have you ever been treated before now by a mental health professional as an outpatient for any reason?" 2. .. Have you ever been treated before in a mental health or psychiatric inpatient unit for any reason?" 3. .. Have you ever been treated for an alcohol or drug problem before?" Though further historical data was collected if a patient answered "yes" to any of these questions, only responses to these three standardly asked questions were used for data analysis. Data used in this retrospective study were collected from the records maintained in a computerized database by the C-L service. Data were not used if the patients had died or were discharged from the hospital before the followup set of information could be completed, or if they were unwilling or cognitively unable to cooperate. The independent variable for data analysis is patient report of any past inpatient or outpatient MH or SUD treatment and is not subcategorized, according to the form of treatment (e.g., group psychotherapy, medications). The data were analyzed with the chi-square test, two-tailed. Fisher's Exact Test was used when one or more sample sizes was less than or equal to five. The level of statistical significance was set at 0.05. RESULTS Of the 155 consecutively consulted patients seen by the C-L service, 10 (6.4%) were too cognitively impaired to obtain the clinical information later needed for this study. Eleven (7.1 %) died or were discharged from the hospital before the information could be completed. 263

Treatment History and Differential Diagnosis in C-L Patients

The records of 134 patients (85.9%) who had C-L service standard information completed were used for data analysis. The study group's mean age was 47.1 (range: 9-93). There were 67 men (50.0%) and 67 women (50.0%). There were 116 Caucasians (86.6%), 12 African Americans (9.0%), 5 Hispanics (3.7%), and 1 Asian-American (0.7%). The most frequent DSM-IU-R diagnoses in this group of patients were delirium (16.4%); current primary major depression, single episode (16.4%); alcohol abuse (10.4%); adjustment disorder (9.7%); uncomplicated alcohol withdrawal (8.2%); current primary major depression, recurrent (7.5%); and organic (secondary) mood disorder. depressed (7.5%). Frequencies of psychiatric diagnosis categories used for data analysis are listed in Tables 1 and 2. Seventy-six patients (56.7%) reported past MH or SUD histories; fifty-eight patients (43.3%) did not. Outpatient MH treatment history was reported by 63 (47.0%) patients, inpatient MH TABLE I.

treatment history by 39 (29.1 %) patients, and SUD treatment (inpatient or outpatient) by 35 (26.1 %) patients. Considerable overlap occurs within these three categories of treatment. For example. only three patients with inpatient MH treatment histories did not also report outpatient treatment histories. Only nine patients with SUD treatment histories did not also report inpatient or outpatient MH treatment histories. Table 1 shows the associations between primary and phenomenologically related secondary (organic) psychiatric disorders and the presence or absence of past MH or SUD treatment histories. The first comparison in Table 1 is between any primary psychiatric disorder diagnosis and any secondary (organic) psychiatric disorder diagnosis. Patients were excluded who had both primary and secondary psychiatric disorder diagnoses. The patients with only primary psychiatric diagnoses were twice as likely as the patients with only secondary psychiatric disorder diagnoses to have reported past

Primary vs. secondary (organic) psychiatric disorder frequency comparisons between patients without vs. with mental health or substance use disorder treatment history

Diagnosis Comparison

Without Mental Health or Substance Use Disorder Treatment History (n 58)

=

With Mental Health Or Substance Use Disorder Treatment History (n 76)

=

Any primary psychiatry disorder (n = 94)

36 (38.3%)

58 (61.7%)

Any secondary (organic) disorder (n = 20)

14 (70.0%)

6(30.0%)

14 (38.9%)

22 (61.6%)

Secondary (organic) mood disorder, depressed (n = 10)

8 (80.0%)

2 (20.0%)

2

X

p

6.371

0.009

.

5.301

0.032

.

0.064

1.000

0.190

1.000

(Excludes 20 patients with both primary and secondary disorders) Primary major depressive disorder (n = 36)

Any primary anxiety disorder (n = /0)

4 (40.0%)

6 (60.0%)

Any secondary anxiety disorder (n = II)

4 (45.4%)

6 (54.5%)

Any primary psychotic disorder (n = 7)

3 (42.9%)

4(57.1%)

Secondary halucinosis or delusional disorder (n = 8)

3 (37.5%)

5 (62.5%)

Note: Statistic used: chi-square test, two-tailed; Fisher's Exact Test when I or more statistical cells is less than or equal to 5. Data are n (%). Numbers add to 100% across rows, not down columns. Statistically significant at P < 0.05.

264

PSYCHOSOMATICS

Rundell and Hall

TABLE 2.

Psychiatric disorder frequency comparisons between patients without vs. with mental health or substance use disorder treatment history Without Mental Health or Substance Use Disorder Treatment History (n =58)

Diagnosis Comparison

= 16)

With Mental Health Substance Use Disorder Treatment History (n =76)

Xl 2.474

0.116

1.409

0.235

5.934

0.020'

0.619

0.531

0.001

1.000

0.082

0.774

0.074

0.785

1.251

0.263

6.001

0.014

5.192

0.028

1.236

0.266

6.636

0.010'

4 (25.0%)

12 (75.0%)

= 118)

54 (45.8%)

64 (54.2%)

7 (31.8%)

15 (68.2%)

= 112)

51 (45.5%)

61 (54.5%)

Secondary (organic mood disorder. depressed (n = 10)

8 (80.0%)

2 (20.0%)

Without secondary mood disorder, depressed (n = 124)

50 (40.3%)

74 (59.7%)

Dementia (n

Without dementia (n Delirium (n

= 22)

Without delirium (n

Secondary (organic) anxiety disorder (n = II)

6 (54.5%)

5 (45.5%)

Without secondary anxiety disorder (n = 123)

52 (42.3%)

71 (57.7%)

Schizophrenia or delusional disorder (n = 7)

3 (42.9%)

4(57.1%)

55 (43.3%)

72 (56.7%)

14 (41.2%)

20 (58.8%)

44 (44.0%)

56 (56.0%)

Panic disorder or generalized anxiety disorder (n = 15)

6 (40.0%)

9 (60.0%)

Without panic or generalized anxiety disorder (n = 119)

52 (43.7%)

67 (56.3%)

Without schizophrenia or delusional disorder (n = 127) Primary major depression (n

= 34)

Without primary major depression (n = 1(0)

Psychiatric factor affecting physical condition or somatoform disorder (n = 11) Without either category (n

= 123)

Psychoactive substance withdrawal syndrome (n = 18) Without psychoactive substance withdrawal (n = 116)

3 (27.2%)

8 (72.7%)

55 (44.7%)

68 (55.3%)

3 (16.7%)

15 (83.3%)

55 (47.4%)

61 (52.6%)

Alcohol abuse or alcohol dependence (n =20)

4(20.0%)

16 (80.0%)

Without alcohol abuse or dependence (n = 114)

54 (47.4%)

60 (52.6%)

6(31.6%)

13 (68.4%)

52 (45.2%)

63 (54.8%)

Drug abuse or drug dependence (n = 19) Without drug abuse or dependence (n = 115) Adjustment disorder (n

= 13)

Without adjustment disorder (n

= 121)

10 (76.9%)

3(23.1%)

48 (39.7%)

73 (6O.3%)

p

,

,

Note: Statistic used: chi-square test, two-tailed; Fisher's Exact Test when I or more statistical cells is less than or equal to 5. Data are n (%). Numbers add to 100% across rows, not down columns. Statistically significant at P < 0.05.

VOLUME 38 • NUMBER 3 • MAY - JUNE 1997

265

Treatment History and Differential Diagnosis in C-L Patients

treatment histories (X 2 = 6.371, P = 0.(09). Conversely, the patients with only secondary psychiatric diagnoses were about twice as likely as the patients with primary psychiatric diagnoses to not have reported past treatment histories 2 (X = 6.371, P = 0.(09). The patients with current primary major depression were three times more likely than the patients with organic mood disorder (depressed) to have reported past treatment histories (X 2 = 5.301, P = 0.032). The patients with organic mood disorder (depressed) were more than twice as likely than the patients with primary major depression to not have reported past treatment histories (X 2 =5.30 I, P = 0.032). The patients with primary and secondary anxiety disorders and the patients with primary and secondary psychotic disorders were similarly likely to have reported past treatment histories. Table 2 summarizes specific psychiatric disorder category frequency comparisons between the patients without (N = 58) vs. with (N =76) past MH or SUD treatment histories. Patients in 4 of the 12 psychiatric diagnostic categories were statistically more or less likely to receive that diagnosis when they reported a history of past MH or SUD treatment. Organic mood disorder, depressed (X 2 = 5.934, P = 2 0.020) and adjustment disorder (X =6.636, P = 0.010) are each more likely in the C-L patients who reported no past treatment history. Psychoactive substance withdrawal syndrome (X 2 = 6.001, P =0.014) and alcohol abuse or dependence (X 2 = 5.192, P = 0.028) are each more likely when patients did report a history of past MH or SUD treatment. The other eight diagnostic categories were similarly likely, regardless of whether the patients reported a past history of MH or SUD treatment. DISCUSSION The analysis in Table I confirms the hypothesis that patient report of past MH or SUD treatment history would significantly increase the probability of diagnosing a primary psychiatric disorder and decrease the probability of diagnosing a secondary psychiatric disorder 266

among the patients in this retrospective study. This phenomenon appears to be largely accounted for by statistical differences between the primary and secondary mood disorder categories. Primary and secondary anxiety and psychotic disorders are similarly likely whether there was report of past treatment or not, though smaller sample sizes limit interpretability for those two categories. In C-L training settings, residents and fellows are often taught that a depressive syndrome may be more likely secondary than primary when there is no past history of primary major depression, when there is no family history of primary major depression, and when there is an identifiable temporally related toxic or medical potential etiology.5.6 The data from Table I support at least the first aspect of this differential diagnostic triad, though the finding is only partially discriminating. Patients without past treatment histories may still be diagnosed with primary depression, and patients with past treatment histories are still subject to developing secondary (organic) mood disorder, depressed. Nevertheless, information about past MH and SUD treatment history in depressed patients can help the C-L psychiatrist rank-order differential diagnostic possibilities in the context of other examination findings. The likelihood of diagnosing psychoactive substance-withdrawal syndrome and alcohol abuse or dependence (Table 2) in this C-L setting is more likely when a patient reported past treatment history. The likelihood of diagnosing organic mood disorder (depressed) and adjustment disorder (Table 2) is more likely when a patient reported no past treatment history. For these four diagnoses, information about past MH and SUD treatment histories may help clinicians with differential diagnostic considerations. Relative frequencies of the other eight diagnostic categories statistically analyzed in this study are not significantly associated with history of MH or SUD treatment. This latter finding underscores the fact that other factors also determine a medical-surgical patient's clinical picture. There are several limitations and potential sources of bias in this study. First, the fact that PSYCHOSOMATICS

Rundell and Hall

a patient denies past MH or SUD treatment history does not necessarily equate to lack of psychiatric predisposition. Patient reports of past treatment history may underestimate actual psychiatric predisposition by not accounting for previously unmanifested genetic predisposition, health insurance limitations on obtaining treatment, and the presence of psychiatric illnesses that heretofore have gone untreated. 9 In addition, MH treatment may have been provided without patients' having perceived it, especially by primary care physicians. However, for busy clinicians, asking patients or families about whether a patient has ever been treated for a psychiatric or SUD is more straightforward and reliable than trying to make judgments about whether specific criteria for psychiatric disorders were ever met in the past. Second, because this is a retrospective record review, patient reports about treatment history, or its absence, cannot be confirmed. Some patients may have no memory of or an unwillingness to report previous MH or SUD treatment. Third, psychiatric diagnoses were based on DSM-III-R criteria, not on results of a standard psychiatric diagnostic interview. Usage of a standard interview instrument, though difficult in a busy hospital environment, would have provided a firmer basis for the psychiatric diagnostic categories. Fourth, comparisons are made between calculations involving different statistical-analysis cell sizes, limiting the interpretability of findings across categories. Fifth, the 21 patients excluded from this study because of unwillingness or cognitive inability to answer the relevant questions may have different past MH or SUD treatment histories than our final sample patients. The results of this study may not extend to them. Sixth, because the examiner for all patients also later developed the hypotheses for this study, observer bias is a possibility. This C-L service database was established with an eye toward future research applications. Though this study's hypotheses were not yet conceptualized and objectivity was attempted, the examiner's experience and beliefs about the meaning of psychiatric predisposition in the clinical VOLUME 38. NUMBER 3 • MAY - JUNE 1997

evaluation could have resulted in unconscious bias. Finally, it was not possible to separate inpatient, outpatient, and SUD treatment because there was so much overlap. Nor were specific types of MH treatments for specific types of disorders categorized. It may be possible that a specific history of inpatient MH treatment is more highly associated with current psychiatric diagnosis than outpatient MH treatment alone. It may also be possible that the nature and duration of prior treatments are related in more specific ways to current clinical status than the methodology of this study can identify. These limitations and potential sources of bias suggest areas of focus for future research. CONCLUSION The absence of a past psychiatric treatment history does not confer immunity from current primary psychiatric disorders in C-L patients. Similarly, the presence of a past psychiatric treatment history should not diminish a C-L psychiatrist's vigor in pursuing potential toxic and medical etiologies for a patient's psychiatric syndromes. However, asking patients about history of any previous MH or SUD treatment provides helpful data for the C-L psychiatrist who frequently faces difficult differential diagnosis and treatment decisions. Information about past psychiatric treatment, particularly for the depressed patient, can guide further history-gathering, suggest lines of questioning during patient examinations, provide data to rank-order differential diagnostic considerations, and help clinicians and patients make more informed treatment and management decisions.

References I. Murray GB: Confusion. delirium. and dementia. in Massachusetts General Hospital Handbook of General Hospital Psychiatry. edited by Hackett TP. Cassem NH. Littleton. MA. PSG Publishing. 1987. p. 99 2. Cohen-Cole SA: Consultation psychiatry: a practical guide. in Consultation-Liaison Psychiatry and Behavioral Medicine, edited by Houpt JL. Brodie HKH. Philadelphia. PA. JB Lippincott, 1986, p. 319 3. Cohen-Cole SA. Brown FW, McDaniel JS: Assessment

267

Treatment History and Differential Diagnosis in C-L Patients

of depression and grief reactions in the medically ill. in Psychiatric Care of the Medical Patient. edited by Stoudemire A. Fogel BS. New York. Oxford University Press. 1993. p. 60 4. Carroll BJ: Problems with diagnostic criteria for depression. J Clin Psychiatry 1984; 45:14-18 5. Wise MG. Rundell JR: Concise Guide to Consultation Psychiatry. 2nd Edition. Washington. DC. American Psychiatric Press Inc.• 1994. p. 57 6. Rouchell A. Tierney J. Pounds R: Depression. in The American Psychiatric Press Textbook of ConsultationLiaison Psychiatry. edited by Rundell JR. Wise MG. Washington DC. American Psychiatric Press. Inc.•

268

1996.p.319 7. McDaniel JS. Johnson KM: Mania. in The American Psychiatric Press Textbook of Consultation-Liaison Psychiatry. edited by Rundell JR. Wise MG. Washington DC. American Psychiatric Press. Inc.. 1996. pp. 348-356 8. American Psychiatric Association: Diagnostic and Statistical Manual for Mental Disorders. 3rd Edition. Revised. Washington. DC. American Psychiatric Association. 1987. pp. 22-23 9. Nurnberger JI. Goldin LR. Gershon ES: Genetics of psychiatric disorders. in The Medical Basis of Psychiatry. edited by Winokur G. Clayton P. Philadelphia. PA. WB Saunders. 1986. pp. 48~88

PSYCHOSOMATICS