verbal control of delusions

BEHAVIOR THERAPy 21, 461-479, 1990

Verbal Control of Delusions

C. F. LOWE

P. D. J. CHADWICK

Department of Psychology University College of North Wales

Bangor, UK

The verbal descriptions or rules people formulate to describe both themselves and their interactions with others can profoundly influence their subsequent behavior and rule formation. Delusions may be thought of as being one class of rules. Two single- case experiments, each comprising a multiple-baseline design across three behaviors (delusions), are reported. Each experiment incorporated a number of measures of delusional thinking and an intervention designed to challenge these delusions and to pro- vide the clients concerned with alternative rules with which to organize their behavior. In both experiments the intervention proved highly effective in reducing the degree of conviction in the delusional beliefs, and there was some generalization of the effects across beliefs. A number of secondary measures helped to identify associated clinical benefits, and there was evidence that both clients used the alternative rules to regulate their behavior during a 6-month follow-up phase.

Recent years have seen a growing awareness within behavior analysis and other areas of psychology of the importance of language or rules in the regulation of behavior. The process of learning to describe our environment alters our relation to it, because in so doing we organize and structure our subsequent interactions (Lowe, 1979; 1983; Lowe, Horne, & Higson, 1987; Skinner, 1969; Vygotsky, 1962; 1978; Zettle & Hayes, 1982). In recognition of the potency of the regulatory function of language, a number of researchers have come to understand many clinical disorders as being maintained, at least in part, by the ways in which clients describe both themselves and their interactions with others (e.g., Abramson, Seligman, & Teasdale, 1978; Beck, 1967; Hayes, 1987; Kelly, 1955). It seems plausible that delusional beliefs may also be maintained, to some extent, in this way. Strauss (1969) has argued that delusional behavior is on a continuum function with normal behavior, the difference between them being quantitative rather than qualitative. Delusions, like other strongly held beliefs, influence the way events are interpreted, and are similarly resistant to change (e.g., Maher, 1988). Indeed, because of this apparent continuity with other beliefs, the definition of what actually constitutes a delusion is somewhat problematic (see Garety, 1985).

461 0005-7894/90/0461-047951.00/0 Copyright 1990 by Association for Advancement of Behavior Therapy

All rights of reproduction in any form reserved.

462 LOWE AND CHADWICK

This study was conducted in 1988 while P.D.J. Chadwick was in receipt of a studentship funded by Clwyd Health Authority. Thanks are due to Mabel Tannahill for her help in setting up the research programme, to Peter Higson for help and guidance throughout and to many other staff of Clwyd Health Authority, all of whom contributed to the success of this project. Finally we should like to express our gratitude to the two individuals who participated as the main subjects of this study. Correspondence concerning this article should be addressed to C. F. Lowe, Depart- ment of Psychology, University College of North Wales, Bangor, Gwynedd LL57 2DG, United Kingdom.

Attempts at modifying delusions in people diagnosed as schizophrenic have been scarce, and methodological limitations abound. Basic requirements have frequently been overlooked, such as baseline data to describe any ongoing variability, and long-term follow-up data to determine the permanence of any change; a study by Alford (1986) remains one of the few to have met both criteria. Also, treatments have often combined a number of different interventions with little or no attempt at component analysis. Chadwick and Lowe (1990) used a multiple-baseline design across 6 people diagnosed as schizophrenic according to the criteria given in the Diagnostic and Statistical Manual o f Mental Disorders III (American Psychiatric Association, 1980), in order to evaluate the impact of two interventions on a number of dimensions of delusional behavior. The interventions were: first, a structured verbal challenge, and second, a reality-testing intervention during which the client and researcher collaborated to devise a simple test of the rules (see Beck, 1967). The theoretical rationale for the verbal challenge came from the literature on verbal self-regulation within normal and clinical populations (Lowe, Horne, & Higson, 1987). In accordance with numerous clinical approaches to changing covert and overt verbal behavior, the interventions were designed: (1) to give the client information about the ways in which rules or beliefs can govern behavior, (2) to generate alternatives to the maladaptive rules, and (3) to assist the client to monitor and evaluate these alternative rules. Of the 6 clients, 2 rejected their beliefs completely and 3 others reported significant reductions in the degree of belief conviction. These effects were maintained over a 6-month follow-up period, and there was evidence to suggest that the intervention had enabled 5 of the clients to effectively regulate their delusional thinking.

A multiple-baseline design can also be conducted across behaviors (Kazdin, 1982), and is appropriate in the cases of clients who hold more than one delusional belief. Our present study comprises two single-case experiments, each employing a multiple-baseline design across 3 beliefs. In both experiments, after a minimum of 5 weeks at baseline, a verbal challenge intervention was separately directed at each delusional belief in turn, with an intervention of at least four weeks between each intervention. A number of dimensions of delusional experience were measured throughout, and long-term follow-up data were collected. Because the verbal challenge intervention was applied to each belief separately, the study offered a unique opportunity to observe the rela- tionships between each client's three delusions and, in particular, how reduction of conviction in one affected the other two beliefs.

VERBAL CONTROL OF DELUSIONS 4 6 3

GENERAL METHOD Subject Selection

As in our earlier study (Chadwick & Lowe, 1990) participation was on a voluntary basis; clients who had held a delusional belief for the previous 2 or more years were asked whether they were willing to meet with a researcher who wished to discuss their beliefs. Two clients took part, both of whom were outpatients on a stable drug regime. Medication was held constant during the study. Both clients were diagnosed as schizophrenic according to the Diag- nostic and Statistical Manual of Mental Disorders (DSM-III) criteria (Amer- ican Psychiatric Association, 1980); these diagnoses were made independently by P.D.J. Chadwick and a clinical psychologist. Each subject presented with 3 delusional beliefs; there was no other delusional thinking.

Measures

Following Brett4ones, Garety and Hemsley (1987) we measured both degree of belief conviction, and preoccupation, using a modified form of Shapiro's (1961) Personal Questionnaire (Phillips, 1977). A Personal Questionnaire (PQ) was also used to measure the degree of anxiety experienced by the client while thinking about his or her beliefs. We offered each client five statements of intensity of belief conviction, preoccupation, and anxiety, respectively, and both accepted these statements as valid descriptions. Table 1 details the specific wording used for each client and shows that, while the conviction measure was concerned with how certain the client was feeling at that particular point in time, the preoccupation and the anxiety statements referred to the level of

T A B L E 1

THE WORDING USED FOR THE PERSONAL QUESTIONNAIRE MEASURES OF CONVICTION,

PREOCCUPATION, AND ANXIETY

S c o r e C o n v i c t i o n P r e o c c u p a t i o n A n x i e t y

0

M y b e l i e f is a l m o s t O v e r t h e l a s t w e e k I T h i n k i n g a b o u t m y b e l i e f

d e f i n i t e l y f a l s e t h o u g h t a b o u t m y I ge t v e r y s l igh t ly

b e l i e f o n c e a n x i o u s

1 • . . p r o b a b l y f a l s e . . . 3 t i m e s . . . s l i gh t ly a n x i o u s

2

• . . m a y o r m a y n o t be t r u e . . . o n c e a d a y . . . f a i r l y a n x i o u s

3

• . . p r o b a b l y t r u e . . . 4 t i m e s a d a y . . . v e r y a n x i o u s

4

• . . a l m o s t d e f i n i t e l y t r u e . . . o n c e a n h o u r . . . e x t r e m e l y a n x i o u s

5

N . B . P e r s o n a l Q u e s t i o n n a i r e s c o r e s fa l l b e t w e e n t w o v e r b a l d e s c r i p t i o n s ; o n e m i g h t r e s p o n d

more to p r o b a b l y t r u e b u t less t o a l m o s t d e f i n i t e l y t r ue .

464 LOWE A N D C H A D W I C K

symptom intensity experienced over the preceding week. The five statements for each measure were written on separate cards and ranked by the client. At the administration stage each card was presented randomly, and the client was required to say whether the intensity was more or less than that stated on the card. The score on each occasion was provided by the number of cards to which the client replied that the symptom intensity was greater than that stated on the card. The scale allowed for comparisons to be made for each subject across time. Following Hole, Rush, and Beck (1979), we measured degree of belief conviction by asking the client for a percentage rating of conviction, and this was taken after the PQ conviction score These four measures were administered at the close of every session throughout the entire study.

Again, in keeping with Brett-Jones et al. (1987), accommodation and reaction to hypothetical contradiction (RTHC) were assessed. The accommodation measure was concerned with the awareness demonstrated by the client of actual occurrences that were contradictory to his or her belief, and with how these had affected the belief. Accommodation was measured at the start of every session by asking the client whether anything had happened over the past week to alter his or her belief in any way. RTHC was measured following accommodation at Weeks 2 and 4 of baseline to evaluate the client's potential for accommodation of evidence at odds with his or her belief. A plausible but contradictory occurrence was posed and the client was asked how, if at all, this would change the belief. In each experiment RTHC was conducted in the case of only one of the three beliefs.

Two measures were included to cover at least some of the possible clinical ramifications of the loss or partial loss of a delusion. These were the Beck Depression Inventory (BDI; Beck, 1967), and a short symptom checklist comprising those items from Wing's Present State Examination (Wing, Cooper & Sartorius, 1974) describing the various forms of schizophrenic delusions and hallucinations. It should be stressed that the symptom checklist was employed not in any diagnostic capacity, but solely for descriptive purposes. The BDI and symptom checklist were administered in the session prior to challenging each of the three beliefs and in the final session of the intervention phase (i.e., in the last baseline session and on three occasions during the intervention phase) and at each follow-up.

Procedure Sessions lasting approximately 90 min each were conducted once a week

throughout the study. All interviews were conducted by the second author. Phase 1: Preliminary Interviewing. Two interviews were conducted with each

client and these satisfied the dual purpose of ascertaining the belief to be modified and establishing a rapport.

Phase 2: Baseline. Throughout the baseline we accumulated as much data as possible about the beliefs, and in particular the evidence, both past and present, that had helped to establish and maintain the beliefs. In the case of each belief, during the final baseline session the client was presented with every piece of evidence he or she had cited and was asked to rank them in order

VERBAL CONTROL OF DELUSIONS 465

of importance to the belief system. At no point was any aspect of a belief challenged until it was subjected to the verbal challenge.

Phase 3: Verbal challenge. Throughout the intervention the client was en- couraged to view a deluded belief as being only one possible interpretation of events. The discussion was conducted within an atmosphere of "collabora- tive empiricism" (Beck, 1967): the clients were not told that their beliefs were wrong, but were asked to suspend their conviction that the beliefs were un- deniably true and to consider alternatives. A non-confrontational approach of this kind has been supported by Milton et al. (1978). Initially, following Watts, Powell, and Austin (1973), we challenged the evidence for the belief in inverse order of importance. In each case the client's interpretation of events was countered by a non-deluded interpretation put forward by the researcher. An integral part of this discussion involved the researcher making clear to the client the way beliefs can exert a profound influence over the interpretations placed upon events - tha t is, the client was made aware of the regulatory function of language (see Vygotsky, 1962).

After discussing all the evidence for the belief, we challenged the belief it- self in three stages, which in practice overlapped. First, we drew attention to any inconsistency and irrationality within the client's belief system; this was tantamount to posing the question, "Would it make sense for things to be as you say they are?" We then presented an alternative explanation for what had been happening to the client. In accordance with the continuum view of delusions, and in deference to the growing dissatisfaction with the concept of schizophrenia (see Bentall, Jackson, & Pilgrim, 1988), a feature of the present research was that clients were not told that their beliefs were a sign of illness. Rather the beliefs were interpreted as having developed in reaction to, and as a way of making sense of, particular experiences that the clients had en- countered (see Maher & Ross, 1984; Maher, 1988). Finally, we argued that the available evidence offered strong support for the alternative interpretations.

Phase 4" Follow-up. To assess for maintenance of behavior change, 1-month, 3-month, and 6-month follow-up meetings were conducted. At these sessions all the measures were administered (with the exception of RTHC) in the order and manner described earlier.

Phase 5: Independent Assessment. An independent clinical psychologist who was kept blind with respect to the introduction of the invention interviewed the clients to assess their degree of conviction in each of their delusional beliefs and to obtain their observations on the study. The principal procedural difference between Experiments 1 and 2 was the number and timing of these external assessments and, accordingly, this aspect of the procedure is specified separately for each experiment below.

EXPERIMENT 1 Method

Subject CP, a 29-year-old white British male, had a psychiatric history extending

over 5 years. He attended school for 11 years and at the time of the study


was single and unemployed. CP presented with three distinct delusional beliefs, which he stated as having held for two years (Belief 2) and for four years (Beliefs 1 and 3); the case notes supported this claim. These were: Belief 1 - that he was to be married to a woman called Amanda, with whom he had not been in contact for many years, and who was reading his mind and controlling many of the things that happened to him; Belief 2 - that he had been Jesus Christ in a prior life; and Belief 3 - that he had been Leonardo da Vinci in a prior life.

Procedure This was in every respect the same as described in the General Method sec-

tion except that only one external assessment was conducted, and this took place after the final follow-up session.

Results (a) Belief Conviction

The main measure of recovery from delusional thinking was the degree of belief conviction. Figure 1 shows CP's percentage belief conviction in his three beliefs during each phase of the s t u d y - percentage belief conviction and PQ conviction score correlated very closely (a Pearson's r of .98) so the PQ score is not presented separately. Belief conviction in all three beliefs was very stable during the first five weeks of the study. At Week 6 we challenged CP's first belief ('Amanda'); there was an instant drop in conviction to only 50% certainty. Over the next three sessions conviction in Belief 1 fell further to only 10% certainty. Degree of conviction in Belief 2 ('Jesus') and Belief 3 ('Leonardo') remained at the baseline level of 100% until week 10 of the study. Discussion of Belief 1 was continued throughout the intervention phases.

At week 10 we challenged Belief 2, producing an immediate fall in conviction to only 70% certain. There was no change in conviction in Belief 3. How- ever, at week 11, conviction in Belief 3 fell to 80°70 certain and conviction in Belief 2 rose slightly, also to 80% certain. At the next session conviction in Belief 2 fell again, to only 50% certain; and conviction in Belief 3 rose slightly to 90°70. At week 13 Belief 2 was totally rejected and conviction in Belief 3 fell dramatically; of his belief that he had been Jesus, CP said, "I was be- lieving things that I'd wanted to believe and not things that were true." Be- cause of the apparent generalization effect that was taking place, we deferred challenging Belief 3 to see whether belief conviction would continue to fall. This caution was fully justified; at session 14 conviction in Belief 2 remained at 0% (i.e., the client saw the belief as definitely false) and conviction in Belief 3 fell to only 10%. CP stated at Week 14 "I feel a lot b e t t e r . . . I feel I'm making a lot more sense, a lot more rational". Time constraints required that we challenge Belief 3 at Week 15, when it was completely re jected- indeed this was true of all three beliefs. CP's degree of conviction in each belief remained at 0°70 for the remaining four sessions of the verbal challenge phase.

Although the degree of conviction in Beliefs 2 and 3 remained at 0% at each of the follow-up meetings, at the 1-month follow-up conviction in Belief

"O

.<

100

80

B

60 40

20

0

VERBAL CONTROL OF DELUSIONS

VC FU

467

Z

Z

=

100

80

60 40

20

10~

-z 8c

6( 4(

2( (

2 4 6 8 10 12 14 16 18 20 1 3 O

W E E K M O N T H

FIG 1. Percentage conviction for Subject CP during each phase of the study: baseline (B), verbal

challenge (VC) and follow-up (FU).

1 returned to the baseline level of 100070. In fact the "Amanda" belief came back as part of a larger belief system developed over the previous month. This was: First, that he and Amanda were no longer meant for one another because his brother had discovered their relationship and had 'screwed things up', and second, that he was now meant for a famous sportswoman and it was she, not Amanda, who was presently reading his mind and influencing his life. In 1983 CP voiced similar persecutory ideas, holding his brother and father responsible for the failure of an earlier "relationship".

At the 3-month follow-up CP expressed his disappointment that the sportswoman had not come to him. However, he said he now realized she had "done the right thing" by staying away a little longer to train for the Olympics. He

468

5 ¸

< 1 " 0

LOWE AND CHADWICK

B VC 1

FU

5 4

O r~ 1

o

5

2

- 0 0 2 4 6 8 10 12 14 16 18 20 1 3 6

WEEK MONTH

Fro 2. Preoccupation (PREO) and anxiety (ANX) scores for Subject CP during each phase of the study: baseline (B) verbal challenge (VC) and follow-up (FU). Because both measures are retrospective, applying to behavior in the week before the measure was taken (see text), the shading to mark the introduction of the intervention appears 1 week later than in Figure 1.

was still certain that in the past Amanda had read his mind and influenced his life and had been meant for him. By the 6-month follow-up CP reported having completely rejected his belief about the sportswoman and being only 50°70 certain of his belief about Amanda. He reported having challenged these beliefs in a similar manner to which his beliefs had been challenged during the experimental intervention. Unfortunately, he seemed quite depressed, saying he could not do anything for himself, and that he was always blaming other people.

(b) Secondary Measures Preoccupation and Anxiety. The PQ scores for preoccupation and anxiety

are shown in Figure 2. There was no consistent relationship between these two


TABLE 2 BECK DEPRESSION INVENTORY SCORES DURING THE BASELINE PHASE (B) AND ON EACH OF

THREE OCCASIONS DUmNG THE VERBAL CHALLENGE PHASE (WE) AND AT 1 MONTH (1M FU), 3 MONTHS (3M FU) AND 6 MONTHS (6M FU) FOLLOW-UP.

Subject B VCI VC2 VC3 1M FU 2M FU 3M FU

CP 13 4 4 5 2 6 24 BG 17 12 12 6 9 11 8

PQ measures across the three beliefs. For example, although for the "Amanda" belief baseline preoccupation scores were stable and anxiety scores varied, in the case of the two reincarnation beliefs it was anxiety which was stable and preoccupation which varied. Taking Figures 1 and 2 together, conviction and preoccupation appeared to be related for Belief 1 but not for Belief 2 or 3.

Accommodation. At no point did CP report an experience of disconfirmation for a belief which had not already been subjected to the verbal challenge. Thus, for example, it was not until Week 9 of the study that CP reported an experience which he interpreted as contradicting his belief about Amanda. He had been expecting to see Amanda on television on a specific day and her non-appearance was interpreted as further evidence that the belief was false. Similarly, at Week 17 of the study he reported an experience of disconfirmation concerning his belief about having been Jesus. Also, at Week 8 CP reported having "reality-tested" his Amanda belief by asking a friend whether he did in fact understand and accept his link to Amanda, as CP had believed he did.

Reaction to Hypothetical Contradiction. RTHC was only performed for the first of CP's beliefs. He was asked how, if at all, his belief about Amanda would be affected if he met her and she said that none of that belief was true. On both occasions CP stated that such an occurrence would lead him to reject the idea that he and Amanda were meant for one another and to doubt that she had ever read his mind or influenced his life.

Beck Depression Inventory. The BDI was administered seven times in total. Table 2 shows that there was a steady downward trend in BDI scores over the course of the intervention phase, from 13 during the baseline to 5 at the close of the intervention. Although at the 1- and 3-month follow-up meetings the BDI scores remained low, there was a sharp rise to 24 at the 6-month follow- up meeting, by which time CP had rejected his belief about the sportswoman entirely and was becoming far less sure of the Amanda belief again.

Symptom Checklist. In the present study CP did report a new belief following the loss of the three beliefs. The fact that the new belief shared many common features with the "Amanda" belief suggests that it was the loss of this latter belief, in particular, that contributed to the formation of the new belief about the sportswoman.

(c) External Validation When CP was interviewed shortly after the final follow-up meeting by an

independent clinical psychologist, he confirmed that the experimental inter-


vention had led him to reject his beliefs, adding, "I was able to sort things o u t . . . I was clearer in my mind". At the time of the independent assessment he was still free of any thoughts of being either Jesus or Leonardo. He reported that his "Amanda" belief had started to bother him again soon after the close of the intervention. However, he was very positive about the long-term use- fulness of the intervention; although the ideas about Amanda were still present he reported coping better with them again by drawing on the arguments and techniques employed during the intervention. He reported that this process was ongoing and that he was able to control his thinking when the beliefs troubled him most. When asked to rate the extent of his present conviction in the "Amanda" belief, CP reported a 30°70 certainty.

Discussion We challenged three delusional beliefs which had been held for between

2 and 5 years over 6 to 14 sessions. Each belief was rejected totally over the closing four sessions of the intervention phase. For 2 of the 3 beliefs maintenance was good; however, at the first follow-up meeting conviction returned to the baseline rate for one of the beliefs. In fact the "Amanda" belief came back in an encapsulated f o r m - that is, CP reported that although he was absolutely certain that Amanda had read his mind and influenced his life in the past, it was now not she but another woman who was carrying out these func- tions. To account for the failure of his prediction that Amanda and he were to be married, he also reported persecutory ideation not raised before during the course of the study, although similar paranoid beliefs had been expressed in 1983.

Although the challenge countering the "Amanda" belief had no effect upon either of the reincarnation delusions, a generalization effect appeared to occur between them. Although both Belief 2 and Belief 3 concerned a person CP claimed to have been in a prior lifetime, there were good reasons to challenge the two beliefs separately. Not only were they formed at different points in time, but we had elicited no common evidence. When challenging CP's "Jesus" belief we made explicit that it was his claim to have been Jesus that was in question and not the possibility of reincarnation. The complex nature of the observed generalization effect goes some way towards endorsing the decision to challenge the two beliefs separately.

In addition to the reduction in conviction ratings there were other beneficial changes. During the verbal challenge phase, as CP became less sure of his belief about Amanda, he also became progressively less preoccupied with the belief. Again, the follow-up meetings and independent assessment suggested that he was better able to regulate his delusional thinking following the interventions. It might be argued that he did not really come to doubt his beliefs, but simply told the experimenter what the latter wanted to hear. Certainly the danger of demand characteristics is particularly acute when dealing with so-called "private" events. While the independent assessment militated to some extent against this possibility, in Experiment 2 the independent validation was conducted on separate occasions throughout the study, as further precaution. Otherwise a direct replication of the first experiment was undertaken.


EXPERIMENT 2 Method

Subject BG was a 51-year-old married woman with no children. Her psychiatric his-

tory extended back over 20 years and included frequent auditory hallucinations and an attempted suicide. Three beliefs, which had been held for at least 10 years, were challenged. These were: Belief 1 - that she was not yet 20 years old; Belief 2 - that she was the daughter of Princess Anne; Belief 3 - that her life (during which, she believed, she had been raped on a number of occasions and had given birth to several children, each of whom had immediately been taken away from her) was being planned and controlled by the government.

Procedure The external validation was ongoing: BG was interviewed between sessions

4 and 5, 9 and 10, 13 and 14, and 20 and 21. Also, as in Experiment 1, a final external assessment was conducted following the 6 month follow-up. Other- wise the measures and procedure were the same as those used in Experiment 1.

Results (a) Belief Conviction

Figure 3 shows BG's percentage conviction scores in each of her three beliefs during each phase of the study. In the case of Belief 1, conviction fluctu- ated between 80 and 100°70 certainty during the baseline phase. When we challenged Belief 1 at Week 6, there was a substantial drop in belief conviction to 25070 certainty: BG stated, "I look 50 and I tire more quickly than I used to; I must be 50". Conviction remained at this level at Week 7. At Week 8, in spite of observing an instance of disconfirmation during the preceding week, BG's conviction rose very slightly to 30070 certainty. At Weeks 9 and 10 conviction fell to only 10°70. At Week 11, when the belief was totally rejected for the first time, BG said of the possibility of her being in her late teens, "It would be very nice, I know, but it's not true". Over the subsequent six sessions, conviction was rated at between 0 and 10070 certain; at the closing three sessions of the verbal challenge conviction was rated 007o. This was maintained at each of the three follow-up meetings. At the 3-month follow-up BG said "I look in the mirror and I can't possibly be t w e n t y - t h e mirror doesn't lie".

In the case of BG's belief that she was the daughter of Princess Anne, belief conviction was stable over the first six sessions of the study. At Weeks 7 and 8 conviction fell to 80070 and 7007o respectively. BG attributed this drop in belief conviction to the effect of the intervention on her belief that she was only in her late teens; she now felt that it was probably the case that she was 51- years old, and Princess Anne appeared to be younger than this. At Week 9 conviction returned to the baseline rate of 10007o certainty. At Week 10 we challenged BG's belief that she was of royal blood. By the close of this session the belief was completely rejected; indeed the belief was rejected at all subsequent sessions with the exception of Weeks 14 and 16, when BG expressed

472 LOWE AND C H A D W I C K

her opinion that it was just possible that the belief was true and rated conviction at 10%0 certainty. At the 1- and 3-month follow-up meetings BG stated that she was absolutely certain that she was not the daughter of Princess Anne. At the final follow-up meeting BG gave a percentage conviction rating of 20%0.

In the case of Belief 3, that her life was being controlled by the government, conviction remained at 100%0 throughout the 13 week baseline phase: on one occasion BG stated, "I've been coming here (the hospital) for years now, and I still believe these things are true". At Week 14 we challenged Belief 3 and conviction fell to 80%. BG stated of her belief, "It's like a s h i e l d . . . I've got to try to face up to what I'm hiding from". At the subsequent session conviction fell again, to 70070: BG said of her belief, "It's just like a justification • . . you're hurt and you seek solace". At Week 16 there was a substantial drop in belief conviction to only 10%; BG stated that she accepted the alternative interpretation of her beliefs, adding, "I think you're doing me good talking to me". At the following session BG entirely rejected Belief 3, and also Beliefs 1 and 2, and stated that "realization is half the cure, isn't it?" The three beliefs were also rejected in the remaining two sessions of the intervention phase. At Week 19 of the study BG gave a clear example of self-regulation: she reported that over the preceding week the 'voices' told her that she was the mother of a different member of the royal family; she refused to believe the news, however, telling herself that this could not possibly be true and that she should not be "so stupid". In the final session of the intervention phase BG offered the following explanation for the emergence of her beliefs, "I think I made a mistake, I misinterpreted the whole thing". At the 1-month follow-up BG was still completely certain that Belief 3 was false. At the 3-month follow-up meeting BG stated that she thought it just possible that the government was controlling her life and gave a 5% conviction rating. At the final follow-up BG reported that she was absolutely certain that the government was not controlling her life. She provided further evidence of self-regulation, adding that on those occasions when she thought about her belief, "logic always came to the fore".

(b) Secondary Measures Preoccupation and Anxiety. The PQ scores for preoccupation and anxiety

are shown in Figure 4. During the baseline phase scores on both measures were similar across the 3 beliefs, with fluctuations with respect to any one belief being mirrored in scores for the other two beliefs. The effect of the intervention was to disrupt this pattern somewhat, although there was still a fair degree of overlap. Taken together, Figures 1 and 2 suggest that conviction and preoccupation scores were related in the case of each belief, with a fall in one set of scores generally corresponding to a fall in the other.

Accommodation. BG reported instances of disconfirmation of Beliefs 1 and 2 during the baseline phase which led to her doubting both beliefs. In the case of Belief 1, at Week 2 of the study BG reported having felt tired and old; certainly not how a teenager should feel. In the case of the "Princess Anne" belief, BG reported an experience of disconfirmation at Week 7 of the baseline phase which concerned seeing Anne on television and feeling too old to


X

Z

FIG 3.

100

80

60

40

20

0

100 80

60 40

20

0

100

8O

"- 60 40

o 20

" 0

B VC FU

W E E K M O N T H

Percentage conviction for Subject BG during each phase of the study.

be her daughter. This instance was probably due in part to BG's new-found view that she was more than likely not in her teens. At no point during the baseline phase did BG report an instance of disconfirmation of her belief about being controlled by the government. Following the introduction of the interventions BG reported several instances of disconfirmation in Beliefs 1 and 2 but not of Belief 3.

Reaction to Hypothetical Contradiction. At Weeks 2 and 4 BG was asked whether Belief 2 would be affected if Princess Anne herself told BG that the belief was completely false. On both occasions BG replied that this would affect neither her degree of belief conviction nor the belief content.

Beck Depression Inventory. As in Experiment 1 the BDI was administered

474

[...

B

LOWE AND CHADWICK

VC FU

L)

5 =. 4

3 2 1

0

h o 4.

5 r~ 4

" 3 J

~" 2 = 1

0

WEEK MONTH

Preoccupation and anxiety scores for Subject BG during each phase of the study.

7 times in total. Table 2 shows that BDI scores fell steadily from a baseline level of 17 to 6 at the end of the intervention phase, and remained low over the three follow-up meetings.

The Symptom Checklist. This measure proved sensitive to the improvements in BG's thinking brought about by the intervention. No new symptomatology was reported during the course of the study.

(c) Independent Assessment The independent assessor was kept blind with respect to the introduction

of the intervention. In the interviews no reference was made to the present


study until the fourth assessment, conducted at the close of the intervention phase, when BG raised the matter for the first time. The initial interview, conducted between sessions 4 and 5 of the study, revealed the following. First, BG reported holding the following beliefs: (1) that she was only in her late teens, (2) that she was the daughter of Princess Anne, (3) that the government was monitoring her constantly, and (4) that she had been raped and had had children. Second, BG said that she had held these beliefs for over 10 years. Third, BG stated that at no point in time over the past 12 months had she doubted the truth of her beliefs, adding "the government wouldn't have gone to all this trouble if it wasn't true". Fourth, BG reported being absolutely certain of each of these beliefs.

The second interview was conducted prior to the 10th session of the study. When discussing her belief that she was not yet 20, BG said that although the voices told her she was young she was unable to reconcile this with the fact that she looked like a 50-year-old woman: BG stated being virtually certain that she was 51. Although BG stated having doubts that she was the daughter of Princess Anne, she remained absolutely certain that her life was being controlled by the government. The third assessment was conducted prior to session 14. When discussing her belief about being of royal blood, BG said that although the voices told her this was the case her own logic was now telling her that she could not be Anne's daughter. BG reported having no doubt what- soever that the government was controlling her life.

The fourth assessment was conducted prior to the 20th and final session of the intervention phase. BG reported being absolutely certain that each of her beliefs was false, a response corresponding to 0°70 certainty. To ensure that this was the case for each of BG's beliefs, the interviewer asked BG to list those beliefs which she now rejected totally. BG recalled each of the three beliefs challenged during the intervention. At this point the interviewer asked BG why she now rejected these beliefs. BG attributed the change to her participation in the present study, saying, "We've thrashed it out, we've discussed it, it's all come to a head and I feel tons b e t t e r . . . I realize now it was all in my mind, it wasn't anything from outside doing it to me. The whole thing was in my mind". BG stated that the discussions had occasionally been slightly upsetting.

A final assessment was conducted prior to the final follow-up meeting. At this meeting BG reported still being absolutely certain that Beliefs 1 and 3 were false. When asked whether she believed she was the daughter of Princess Anne, BG replied, "I am not certain this is true, but I think it could be". When asked to quantify this she stated being 20070 certain that she was Anne's daughter. BG attributed this increase in her conviction that she was of royal descent to "things going against her at home". She stated that she was aware the belief might be an escape route. BG reported having been able to control her delusional thinking by applying "logic". She was specifically able to re- mind herself of the alternative explanation put forward to account for her beliefs during the verbal challenge phase. BG also reported reminding herself of the evidence running counter to her beliefs, which she had discussed with the experimenter in the present study, and she said that this helped her to regulate her delusional thinking.

476 LOWE AND C H A D W I C K

Discussion Three longstanding delusions were challenged and each was rejected fol-

lowing between 6 and 15 intervention sessions. The maintenance of change in each belief was good. BG reported being able to regulate her thinking effectively as a consequence of the intervention, and this included being able to resist the promptings of her auditory hallucinations. For example, during the follow-up stage she did not accept that she was the mother of a younger member of the royal family, as the voices would have her believe. Also, as in the first experiment, there seemed to be a generalization effect across 2 of the 3 beliefs. In some respects this was to be expected: BG's belief that she was only in her teens had a direct bearing on whether she could be the daughter of Princess Anne. However, a generalized reduction in conviction was not the only possible outcome. She could have maintained that Princess Anne was far older than she looked; that people could successfully disguise their true age which was an integral part of the belief that she was not yet 20 years old.

GENERAL DISCUSSION In each experiment three delusional beliefs, held for between 2 to 10 years,

were rejected as a consequence of the verbal challenge intervention. For 5 of these 6 beliefs maintenance was good. Although one of CP's beliefs did re- turn at the 1-month follow-up, conviction fell again at the final follow-up meeting and subsequently at the independent assessment. CP attributed these two reductions in belief conviction to his ability to draw on the discussions that had taken place during the verbal challenge.

In both studies there appeared to be a generalization effect. This might reflect the clients' ability to think rationally about their beliefs; certainly, in many people diagnosed as schizophrenic the ability to reason is not lost (Bleuler & Bleuler, 1986). Alternatively, the generalization effects raise a question mark over the extent to which the beliefs of both clients were distinct. For example, two of the first client's beliefs concerned reincarnation, and might therefore have been challenged together as one complex belief. However, because they shared no common evidence, there seemed sufficient reason to challenge them separately. There is also the problem of whether a given belief could be sub- divided into 2 or more discrete delusions. For example, in the case of the second client, the third belief (about the government) was more complex than the other two, and might have merited further sub-division. This was not done, however, since the different components of the government belief shared common evidence, suggesting that they were not independent; it is, of course, a requirement of multiple-baseline methodology that behaviors or beliefs challenged separately should be independent (see Kazdin, 1982). Nevertheless, further research is certainly needed to clarify the criteria for establishing the in- dependence of delusions.

The Beck Depression Inventory was included to see whether the loss or partial loss of a strongly held long-term delusion might have a detrimental effect on the client's affective state, for both CP and BG the BDI score recorded at the close of the intervention phase, by which time each belief had been re-


jected, was lower than the BDI score recorded during the baseline phase, confirming the findings of Chadwick and Lowe (1990). It does not, therefore, appear to be the case that the loss of a long-term delusional belief leads to increased depressive feelings. The symptom checklist proved sensitive to the changes in delusional thinking over the course of the two experiments. Al- though in the case of CP the loss of his delusions was followed by the formation of a new belief, in the case of BG no new beliefs were reported following the loss of her three delusions. In the case of CP the large degree of similarity between the "Amanda" belief and the subsequent "sportswoman" belief suggests that it was the loss of the "Amanda" belief in particular which led to the emergence of the new delusion. These two beliefs perhaps served a common function. However, the argument that beliefs that are functional will always be replaced is not borne out by the evidence. The beliefs held by BG might also be seen as functional but they were not substituted; also, Chadwick and Lowe's (1990) study provides a number of cases where the weakening and re- jection of delusions are not followed by the formation of new beliefs.

The results of the reaction to hypothetical contradiction (RTHC) were ambiguous. CP was responsive to hypothetical contradiction and sensitive to the interventions, whereas BG was unresponsive to hypothetical contradiction and yet sensitive to the interventions. The accommodation measure was more in- formative. As in the Brett-Jones et al. (1987), and Chadwick and Lowe (1990) studies, very few instances of disconfirmation were reported during the baseline phase. Both clients reported numerous instances of confirmation, and interpreted ambiguous information as confirmatory-- further examples of continuity with the normal population (see Maher and Ross, 1984). Thus, for instance, CP reported an occasion when a dishevelled, messy girl, with no laces in her shoes, had asked him for some money at a bus stop. He did not recognize the girl but, in line with his core delusion, he subsequently deduced that she had in fact been Amanda "in disguise" and that she had planned the meeting as a way of reminding him of her presence. This tendency for individuals to use evidence in ways which confirm their existing rules is well known in the human learning literature (see Lowe, 1979).

Having come to doubt their beliefs during the verbal challenge phase, both clients reported instances of disconfirmation and appeared to view their beliefs more as hypotheses than as articles of faith. Clearly a reduction in belief conviction is more likely to be maintained if the client begins to observe disconfirmation. However, maintenance may also depend on the extent to which the client reacts to occurrences which prior to the intervention would have been taken as evidence that the delusion was true (see Johnson, Ross, & Mas- tria, 1977). In this vein BG reported no longer interpreting her auditory hallucinations as confirmatory. Maintenance may also be enhanced when clients initially interpret events in a delusional fashion, but subsequently reinterpret the experience. For example, following the intervention CP reported that he still occasionally thought that Amanda was reading his mind but that subsequently he was able to recognize that this was untrue and merely a spur-of- the-moment reaction. This is the reverse of CP's behavior prior to the intervention when, as with the case concerning the girl at the bus stop, he initially


interpreted an event in non-delusional terms, but subsequently reinterpreted the experience delusionally.

The reliance on self-report measures poses a problem with respect to demand characteristics, which is partly solved by using an independent interviewer, and it should be noted that in both our experiments the clients gave no indication that they perceived a link between the experimenter and the assessor. Behavioral measures may further reduce problems associated with demand characteristics although, as in the two cases reported here, some delusions do not reliably give rise to any particular overt non-verbal behavior. Clearly, such behavior should be measured whenever possible; in those cases where this cannot be done additional procedures to increase confidence in self- reports are desirable.

Overall, the present study is consistent with other modification studies (e.g., Alford, 1986; Chadwick & Lowe, 1990), and extends the approach to include multiple delusions. Collectively, these studies indicate that successful belief modification is possible for clients diagnosed as schizophrenic and receiving medication. Clearly, the treatment parameters need to be established; these interventions have yet to be tried with an inpatient group, or with an acutely psychotic group, or with individuals not on medication. However, the ability of CP and BG to regulate their thinking effectively is consistent with the growing literature on the role of language in guiding behavior within both normal and clinical populations. As such, it is yet further evidence of a continuity between the functioning of those people with delusions and those without.

REFERENCES Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans:

Critique and reformulation. Journal of Abnormal Psychology, 78, 40-74. Alford, B. A. (1986). Behavioral treatment of schizophrenic delusions: A single-case experimental

analysis. Behavior Therapy, 17, 637-644. Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical aspects. New York: Harper

& Row. Bentall, R. P., Jackson, H. E, & Pilgrim, D. (1988). Abandoning the concept of 'schizophrenia':

Some implications of validity arguments for psychological research into psychotic phenomena. British Journal of Clinical Psychology, 27, 303-324.

Bleuler, M., & Bleuler, R. (1986) Dementia praecox oder die Gruppe der Schizophrenien: Eugen Bleuler. British Journal of Psychiatry, 149, 661-664.

Brett-Jones, J., Garety, P. A., & Hemsley, D. R. (1987). Measuring delusional experiences: A method and its application. British Journal of Clinical Psychology, 26, 257-265.

Chadwick, P. D. J. & Lowe, C. E (1990). The measurement and modification of delusional beliefs. Journal of Consulting and Clinical Psychology, 58, 225-232.

Garety, P. A. (1985) Delusions: Problems in definition and measurement. British Journal of Medical Psychology, 58, 25-34.

Hayes S. C. (1987). A contextual approach to therapeutic change. In N. Jacobson (Ed.) Psy chotherapists in clinical practice: Cognitive and behavioral perspectives. New York: Guilford

Hole, R. W., Rush, A. J., & Beck, A. T. (1979). A cognitive investigation of schizophrenic delusions. Psychiatry, 42, 312-319.

Johnson, W. G., Ross, J. M., & Mastria, M. A. (1977). Delusional behavior: An attributional analysis of formation and modification. Journal of Abnormal Psychology, 86, 421-426.

VERBAL CONTROL OF DELU~ONS 479

Kazdin, A. E. (1982). Single case research designs. New York: Oxford University Press. Kelly, G. A. (1955). The psychology of personal constructs (2 vols.) New York: Norton. Lowe, C. E (1979). Determinants of human operant behavior. In M. D. Zeiler & P. Harzem (Eds.),

Advances in analysis of behaviour: Vol. 1. Reinforcement and the organization of behaviour. Chichester, UK: Wiley.

Lowe, C. E (1983). Radical behaviourism and human psychology. In G. C. L. Davey (Ed.), An- imal models of human behaviour. Chichester, UK: Wiley.

Lowe, C. E, Home, P. J., & Higson, P. J. (1987). Operant conditioning: The hiatus between theory and practice in clinical psychology. In H. J. Eysenck & I. Martin (Eds.), Theoretical founda- tions of behaviour therapy, New York: Plenum.

Maher, B. A. (1988). Anomalous experience and delusional thinking: The logic of explanations, In T. E Oltmanns & B. A. Maher (Eds.), Delusional beliefs. New York: Wiley.

Maher, B. A., & Ross, J. S. (1984). Delusions. In H. E. Adams & P. Sutker (Eds.), Comprehensive handbook of psychiatry. New York: Plenum.

Milton, E, Patwa, K. & Hafner, R. J. (1978). Confrontation vs. belief modification in persis- tently deluded patients. British Journal o f Medical Psychology, 51, 127-130.

Phillips, J. P. N. 0977). Generalized personal questionnaire techniques. In P. Slater (Ed.), Dimen- sions of intrapersonal space (vol. 2). New York: Wiley.

Shapiro, M. B. (1961). A method of measuring psychological changes specific to the individual psychiatric patient. British Journal of Medical Psychology, 34, 151-155.

Skinner, B. E (1969). Contingencies of reinforcement: A theoretical analysis. New York: Appleton- Century-Crofts.

Strauss, J. S. (1969). Hallucinations and delusions as points on continua function. Archives of General Psychiatry, 21, 581-586.

Vygotsky, L. S. (1962). Thought and language. New York: Wiley. Vygotsky, L. S. (1978). Mind in society. Cambridge, Mass.: Harvard University Press. Watts, E N., Powell, E. G., & Austin, S. V. (1973). The modification of abnormal beliefs. British

Journal of Medical Psychology, 46, 359-363. Wing, J. K., Cooper, J. E., & Sartorius, N. (1974). The measurement and classification of psy-

chiatric syndromes. Cambridge: Cambridge University Press. Zettle, R. D., & Hayes, S. C. 0982). Rule-governed behaviour: A potential theoretical framework

for cognitive-behavioral research and therapy. In P. C. Kendell (Ed.), Advances in cognitive behavioral research and therapy Vol. 1. New York: Academic Press.

RECEIVED: March 20, 1990 PINAL ACCEPTANCE: June 27, 1990

Reprint requests should be sent to C. F. Lowe, Ph.D., Department of Psychology, University Col- lege of North Wales, Bangor, Gwynedd LL57 2DG, Wales.

verbal control of delusions

Documents