|
[Page 511 continued] The Clinical Presentation of DID Before
and After Treatment The Clinical Presentation of DID Before and After TreatmentAlthough proponents of the PTM have sometimes been hard-pressed to address the question of why the DID diagnosis was rarely made prior to 1970 (Piper, 1997), they have typically responded by contending that the signs and symptoms of DID are subtle, covert, and easily missed. Moreover, they have contended that individuals with DID often hide or minimize their symptoms (Ross, 1997). As a consequence, these authors have suggested, the diagnosis of DID was frequently overlooked by clinicians of previous generations, because these clinicians
Gleaves's arguments are similar. He asserted that Spanos's (1994) description of many DID patients, namely, "that of someone who openly calls herself or himself by different names and behaves like different people on different occasions" (Gleaves, 1996, p. 44), is at variance with what is reported in the DID literature. He further argued that DID often goes unrecognized for many years and that "a florid, obvious presentation of the disorder is atypical" (p. 45). It is unclear, however, how these findings are best interpreted. On the one hand, they may help to explain why DID was presumably under-diagnosed for many decades (Ross, 1997). On the other hand, if a florid and obvious presentation is atypical prior to therapy and becomes typical only during therapy, these findings raise the possibility that iatrogenic factors play an important role in DID. Kluft (1991) estimated that only 20% of DID patients exhibit clear-cut indications of this condition at the beginning of therapy and that the remaining 80% exhibit only specific "windows of diagnosability," namely, transient periods during which the classic features of DID are evident. Although there is disagreement concerning the exact percentages, virtually all authors in this literature have concurred that a large proportion - perhaps a majority - of DID patients in their samples exhibit few or no unambiguous signs of this condition prior to therapy (Kluft, 1984; Putnam, Guroff, Silberrnan, Barban, & Post, 1986; Ross, 1997). [Page 512] Moreover, although systematic data are lacking, numerous advocates of the PTM (e.g., Kluft, 1984; Ross, 1997; Schafer, 1986) have contended that DID patients themselves are frequently un- aware of their alters prior to therapy. This is a point that Gleaves ( 1996) did not clearly acknowledge and that is consistent with an iatrogenic explanation. Putnam (1989) estimated that 80% of DID patients possess no knowledge of their multiplicity before beginning treatment, and Dell and Eisenhower (1990) reported that all 11 of their adolescent patients with DID professed no awareness of their alters at the time of diagnosis. Lewis, Yeager, Swica, Pincus, and Lewis (1997) similarly reported that none of the 12 murderers with DID in their sample reported any awareness of the existence of their multiple personalities. Although Gleaves maintained that DID patients "appear to have experienced their symptoms most of their lives, well before they were ever in treatment for a dissociative disorder" (p. 49), the only published evidence he offered for this assertion was the reports of Coons et al. ( 1988) and Fahy, Abas, and Brown (1989), both of which are uncontrolled studies that did not provide either
Moreover, the pretreatment symptoms reported by the patient in Fahyet al., which included "blackouts," seizures of apparent psychogenic origin, depersonalization, memory gaps, auditory hallucinations, depression, and anxiety, were nonspecific and consistent with a number of diagnoses other than DID, including somatization disorder (which is sometimes characterized by both unexplained physical symptoms and amnestic periods; APA, 1994, p. 449) and BPD. Although proponents of the PTM (e.g., Ross, 1997) have often maintained that the essential features of DID are frequently "latent" and therefore difficult to detect prior to therapy (see Piper, 1997, for a discussion), this proposition raises important concerns regarding the falsifiability of the PTM. When confronted with evidence that DID patients often exhibit few clear indications of multiple identity enactments prior to therapy, advocates of the PTM could argue that these features were present but had not yet been elicited. Without independent evidence of the existence of these features, however, this assertion is difficult to refute. Several authors have also reported that the number of alters tends to increase over the course of treatment (see, e.g., Kluft, 1988; Ross, Norton, & Wozney, 1989). In addition, although the number of alters per DID case at the time of initial diagnosis has apparently remained constant over time (Ross, Norton, & Wozney, 1989), the number of alters per DID case in treatment has increased (North, Ryall, Ricci, & Wetzel, 1993 ). Although these findings are consistent with Gleaves's hypothesis that DID patients tend to hide their dissociative symptoms prior to treatment, they are also consistent with an iatrogenic hypothesis. Moreover, proponents of the PTM will again need to make explicit what data could potentially falsify the former hypothesis. We are hard-pressed to identify another DSM-IV disorder whose essential feature (viz., multiple identity enactment)
These two observations probably help explain why iatrogenesis has long been a serious concern in the DID literature (e.g., Aldridge-Morris, 1989). Although Gleaves (1996) acknowledged that "additional alters can be iatrogenically created" (p. 54) once the disorder has begun, he denied that iatrogenic influences playa role in DID's onset. Although it is difficult to refute this hypothesis given the absence of relevant data, Gleaves's argument hinges on the critical assumption that iatrogenic factors can lead patients with one or more alters to develop additional alters but cannot lead patients without alters to develop one or more alters. Although the theoretical basis underlying this assumption was not articulated by Gleaves, a clear explication of the grounds for this assumption appears necessary for evaluating the assertions of the PTM's proponents. Hypnosis and the Creation of MultiplicityGleaves (1996) took issue with the claim that hypnosis plays a causal role in a number of cases of DID. He cited studies (Coons et al., 1988; Ross, Norton, & Wozney , 1989) indicating that most DID patients have never been hypnotized, as well as studies that reported no differences in the diagnostic features of DID patients (e.g., number of alters, number of diagnostic criteria) who had and had not been hypnotized (see, e.g., Putnam et al., 1986; Ross & Norton, 1989). According to Gleaves, these results refute predictions derived from the SCM. Nevertheless, the SCM does not maintain that hypnosis is necessary for the creation of DID, because hypnotic procedures do not possess any inherent or unique features that are necessary to facilitate responsivity to suggestion (Barber, Spanos, & Chaves, 1974; Spanos & Chaves, 1989). Other methods, such as leading interviews and suggestive questions, may be equally likely to induce clients' adoption of multiple roles (Barber, 1979; Spanos, 1996). Moreover, many of the features of DID may derive from widely available societal scripts concerning the characteristics of this condition. Thus, the SCM would not necessarily predict differences between hypnotized and non-hypnotized individuals in their rates of DID or DID symptoms, particularly if both groups were subjected to suggestive therapeutic procedures. It might nonetheless be argued that
If so, the findings of Putnam et al. (1986) and Ross and Norton (1989) may warrant closer examination. Nevertheless, for two reasons, these two studies do not, as argued by Gleaves, provide evidence against iatrogenesis. First, because all patients in these studies had DID, the finding that hypnotized and non-hypnotized patients did not differ in the number of diagnostic criteria for DID is difficult to interpret in light of ceiling effects (see also Powell & Gee, in press). For example, all of Ross and Norton's ( 1989) patients met the criteria for DID given in the revised third edition of the Diagnostic and Statistical Manual of Mental Disorders (APA, 1987), and the percentages of these patients who met the three additional DID criteria from the third edition (DSM-III; APA. 1980) and the National Institute of Mental Health criterion sets ranged from 94.4% to 95.7% (Putnam et al., 1986, did not report descriptive statistics for the number of DID criteria met in their sample). A more relevant question, which- has yet to be examined, is whether patients who initially present without symptoms of DID and are then hypnotized subsequently exhibit more symptoms of DID than do comparable patients who are not hypnotized. [Page 513] Second, contrary to Gleaves's claims, the results of Ross and Norton (1989) did reveal differences between hypnotized and non-hypnotized patients with DID. In a reanalysis of Ross and Norton's data, Powell and Gee (in press) found that hypnotized patients exhibited greater variance in the number of alters at the time of diagnosis and in subsequent treatment. Although the meaning of this finding is not entirely clear, the authors conjectured that this finding might reflect bimodal attitudes regarding iatrogenesis among practitioners who use hypnosis, with some practitioners (i.e., those who believe that hypnosis is potentially iatrogenic) using hypnosis with caution and others (i.e., those who believe that hypnosis is not iatrogenic) using hypnosis relatively indiscriminately and producing a large number of alters. In addition, Powell and Gee reported that practitioners who used hypnosis reported a a significantly higher number of DID patients in their caseloads than did practitioners who did not use hypnosis. Although this finding is correlational and open to multiple interpretations (e.g., specialists in DID may be more likely to use hypnosis), it is potentially consistent with iatrogenesis. Thus, Ross and Norton's (1989) data do not argue against an iatrogenic hypothesis and may in fact provide suggestive evidence for this hypothesis. Current Treatment Methods for DIDGleaves criticized Spanos's (1994) characterizations of the DID treatment literature as "at best, lacking in support" (Gleaves, 1996, p. 47). He contended that Spanos's assertions that
are not borne out by an examination of the DID treatment , literature. Instead, Gleaves claimed, this literature discourages therapists from treating DID patients as though they possessed genuine personalities and encourages them to treat patients' alters as self-generated. He contended that
Nevertheless, an examination of the widely available treatment literature on DID reveals that much, and arguably most, of this literature is oriented around such techniques as mapping the system of alter personalities and establishing direct contact with alters (e.g., see Ross, 1997, pp. 305-315). These "reifying" techniques appear to be especially common in the early phases of therapy, although the later phases pf therapy often focus on unreifying alters and achieving integration among them (Ross, 1997). Moreover, many prominent authors do in fact appear to treat DID patients as harboring multiple discrete personality-like entities, if not fully developed personalities (Piper, 1997). A sampling of quotations from five of the most influential and widely cited proponents of mainstream treatment methods for DID illustrates this point.
An inspection of the mainstream DID treatment literature reveals that these quotations are representative of those of many other authors (see Piper, 1997, pp. 61-68, for similar examples). These quotations appear to contradict Gleaves' s ( 1996) assertions that "alters are explained and conceptualized as part of a whole peron, not as separate people or entities" and that the "general recommendation is that one speaks with alters to understand all parts of the person in therapy but not as if they were different people" (p. 48). As is evident from the preceding quotations, many or most influential authors in the DID treatment literature treat alters as independent entities or even personalities, at least in the early phases of treatment, although systematic data are needed to ascertain the prevalence of these practices among therapists in the community. Moreover, although Gleaves (1996, described the therapeutic practices of most DID clinicians as a relatively passive process of "acknowledging [that] a patient with DID [has the] genuine experience of alters or real people or entities" (p. 48; emphasis in original), many of these practices (e.g., summoning alters that have not yet appeared, naming alters) appear to be quite active or potentially suggestive, particularly if, as noted earlier, many DID patients have no conscious awareness of multiple identity enactments prior to therapy. From a behavioral or social learning perspective, this reification of alters may adventitiously reinforce DID patients' displays of multiplicity. [*5]
[Page 514 ] Gleaves (1996) also committed a logical error by confusing the absence of appropriate treatment ("benign neglect") with the behavioral technique of extinction. The potential utility of extinctio! techniques in the treatment of DID was illustrated by Kohlenberg (1973) using a single-subject design. Kohlenberg found that systematically ignoring and attending to a DID patient's behavioral expressions of multiplicity reduced and increased, respectively, the frequency with which this patient presented with an alter personality. To argue against the efficacy of extinction, Gleaves cited reports (e.g., Ross, Norton, & Wozney, 1989) indicating that many patients with DID whose condition went unrecognized (and whose DID was presumably not addressed in treatment) for many years exhibited little improvement. He then used this evidence to contend that the approach advocated by proponents of the SCM - not attending to or otherwise reinforcing the patient's displays of multiplicity - is counter-therapeutic. Gleaves (1996) asserted that "of these hundreds of patients with DID, not addressing and treating the dissociative condition did not lead to clinical improvement" (p. 49). For three reasons, however, these data do not provide evidence against the SCM.
Gleaves in effect equated a systematic psychological treatment (viz., extinction) with the absence of psychological treatment and then erroneously extrapolated from the literature on the latter to evaluate the effectiveness of the former. As an analogy, Patterson (1982) would not equate the absence of adequate treatment for a child with conduct disorder (CD) with extinction. Instead, he would almost certainly contend that the antisocial behaviors of an untreated child with CD were being intermittently reinforced by parental attention and submission to the child's actions and that extinction of such behaviors by eliminating this pattern of reinforcement was necessary for behavior change. As Gleaves noted (1996, p. 54), there exist no controlled studies on the treatment of DID. Ellason and Ross (1997) reported that a sample of hospitalized patients with DID showed improvement after a 2-year period following discharge, but this study did not include either a randomized or a matched group of DID patients who received either no treatment or an alternative treatment. Nor was the nature of the treatment received by DID patients made explicit. Further complicating the interpretation of Ellason and Ross's findings is the fact that their original sample comprised 135 patients, of whom only 54 (40% ) were located and reassessed at follow-up (see Powell & Howell, 1998a, 1998b, for additional methodological criticisms of Ellason and Ross's design). Controlled treatment studies of DID will be necessary to better evaluate the relative merits of competing therapeutic approaches (e.g., extinction, traditional treatment methods emphasizing integration among alters). The Distribution of DID Diagnoses Across CliniciansTo address the SCM's assertion that iatrogenesis is an important factor in the etiology of DID, Gleaves (1996) disputed Spanos's (1994) claim that a disproportionate number of DID diagnoses are made by a small number of therapists. Gleaves cited data indicating that the DID cases described by three investigative teams were referred by a large number of different clinicians. Careful inspection of these studies, however, reveals serious selection biases. Putnam et al. (1986) distributed 400 questionnaires to "clinicians ...who had previously indicated an interest in multiple personality disorder" (p. 291) and received responses from 92 individuals. Schultz, Braun, and Kluft (1989) mailed questionnaires "to 676 clinicians who had indicated an interest in MPD" {p. 47) and received 355 responses. The mean number of DID patients seen by each of the responding clinicians ranged from 1 to over 100. Ross, Norton, and Wozney (1989) mailed questionnaires to 515 members of the International Society for the Study of Multiple Personality and Dissociation (ISSMD) and to 1729 members of the Canadian Psychiatric Association (CPA). The 236 cases of DID examined by Ross, Norton, and Wozney were referred by 154 members of ISSMD and 49 members of CPA. Thus, members of ISSMD were between 10 and 11 times more likely to report having seen a case of DID than were members of CPA. Thus, the results of these studies do not refute Spanos's (1994) contention that a disproportionate number of DID diagnoses are made by a small number of therapists, because
Along similar lines, Mai (1995) found evidence for considerable variability in the number of DID diagnoses across psychiatrists and concluded that diagnoses of DID derive mostly from a relatively small number of psychiatrists. These findings dovetail with those of Qin, Goodman, Bottoms, and Shaver (1998), who found that reports of satanic ritual abuse (which are closely associated with DID; Mulhern, 1991) derive primarily from a small number of therapists. Contrary to Gleaves's ( 1996) claims, the results of these studies are thus consistent with the possibility that iatrogenesis is a key factor in the genesis of DID. Moreover, they provide one important test of the SCM, because if DID diagnoses were not made disproportionately by a subset of therapists - namely, those who are ardent proponents of the DID diagnosis - the iatrogenic hypothesis would be called into question. Nevertheless, these findings are causally indeterminate and do not prove iatrogenesis, because they are also consistent with the hypothesis that specialists in DID receive referrals for a disproportionate number of DID cases. Longitudinal studies examining whether patients tend to experience the symptoms of DID, particularly multiple identity enactments, before or after referrals to specialists would help to determine whether these data speak primarily to iatrogenesis or to differential referral patterns. [Page 515] The Epidemiology of DID in Adulthood and Childhood: Implications for IatrogenesisOne set of findings that is sometimes invoked as evidence against the SCM is the literature on the prevalence of DID in community and clinical samples (see, e.g., Ross, 1997). If it could be shown that a large number of individuals in the general population, for example, met criteria for DID prior to treatment and to extensive exposure to information concerning the signs and symptoms of DID, this finding would provide evidence against iatrogenesis and the SCM more generally. The study by Ross (1991) represents the only published study on the epidemiology of DID in the general population (see Gleaves, 1996, p. 50). Ross (1991 ) used a structured interview, the Dissociative Disorders Interview Schedule (DDIS; Ross, Heber, et aI., 1989), to establish diagnoses of DID and conducted interviews with 454 community residents in Winnipeg, Canada. These residents formed a subset of an initial sample of 1,055 respondents identified by a stratified cluster sampling method. Ross (1991) reported that 14 individuals (3.1 %) met criteria for DID according to the DDIS, 6 of whom reported histories of child abuse. Nevertheless, these findings are difficult to interpret for several reasons. First, the DDIS has not been validated for use in non- clinical (e.g., community) samples (Ross, 1991), and its false-positive rate in such samples is unknown. This issue is of particular concern because diagnostic measures developed for use in clinical samples often yield high false-positive rates when applied to samples with low base rates of the diagnosis (Finn & Kamphuis, 1995). This concern is heightened by the finding (Ross, 1991) that 13 out of the 14 respondents who met DDIS criteria for DID scored in the average range (10 to 20) on the DES. Because Ross (1991) did not follow up positive DDIS diagnoses of DID with blind diagnostic interviews by an independent assessor, the issue of false positives is difficult to evaluate. Second, because there is no information on whether the 14 individuals who met criteria for DID had received psychotherapy, the possibility of iatrogenesis cannot be excluded. Perhaps more important, Ross did not report data on the extent of respondents , exposure to explicit information concerning the features of DID (e.g., media coverage of DID). Such data would be helpful in evaluating the extent to which the SCM could account for these cases of DID. Similar problems apply to studies of the prevalence of DID in clinical samples (e.g., Ross, Anderson, Fleisher, & Norton, 1991), which do not provide data on the exposure of DID patients either to potentially suggestive treatment practices (e.g., repeated probing regarding the existence of potential alters) or to explicit information regarding the expected features of DID. A second source of data potentially relevant to evaluating the SCM is findings on the prevalence of DID in children. As noted earlier, data indicating that unambiguous cases of DID emerge in childhood prior to either treatment or extensive exposure to information regarding the features of DID would call the SCM into question. Although cases of DID have been reported in children (Putnam, 1997), there are no large-scale systematic studies of the prevalence of childhood DID in the general population (Ross, 1996). In addition, studies of the prevalence of childhood DID in psychiatric samples (e.g., Waterbury, 1991) have not provided data on the exposure of participants to either
The former issue is of particular importance given research demonstrating the heightened suggestibility of children compared with adults (Ceci & Bruck, 1993), although this literature focuses primarily on children's episodic memory rather than on their willingness to endorse the presence of latent personality structures (e.g., alters). Moreover, it is not known whether cases of DID in children tend to persist into adulthood. Such information would be helpful in evaluating whether such cases represent stable syndromes that are etiologically related to adult DID or instead represent transient conditions that differ qualitatively from adult DID. More detailed inforn1ation concerning both the antecedents and the course of childhood DID should prove useful in testing the predictions of both the SCM and the PTM.
|
