Participants

The participants were 39 patients admitted consecutively to the rehabilitation ward of the London Health Sciences Centre, University Campus, between 1 September 1987 and 31 October 1998. These patients were from an original sample of 130 patients referred to the Physical Medicine and Rehabilitation Unit by tertiary care specialists (generally neurologists) and examined by the second author (R.W.T.) as out-patients or during their in-patient admission to the Neurology Unit of this tertiary care facility. Although all 130 patients were referred with a (provisional) diagnosis of conversion disorder, 75 patients were subsequently diagnosed as having a chronic pain disorder with secondary, un-explained motor symptoms and were not deemed suitable for our programme (Reference Shapiro and TeasellShapiro & Teasell, 1997b ; Reference Teasell and ShapiroTeasell & Shapiro, 1997). Fifty-five patients were told that in-patient admission to our rehabilitation unit could ‘get them functioning and walking normally’. Fourteen patients declined treatment.

Forty-one patients were admitted for an initial week of evaluation to rule out an organic basis for their symptoms. If not already performed, this evaluation included central nervous system imaging with magnetic resonance imaging and/or computed tomography, extensive blood testing and, in some cases, electromyography/visual and somatosensory evoked potential testing. The vast majority of patients had been assessed by more than one neurologist and many had been evaluated by multiple specialists. Two patients were subsequently diagnosed with organic disorders (transverse myelitis, renal phosphate-wasting osteomalacia). The diagnosis of non-organic motor disorder thus was based upon the presence of paralysis or paresis, astasia basia and/or ataxic-like symptoms with no apparent neurological or other organic disorder. Astasia basia is characterised by an unsteady gait (ataxia) with a bizarre lack of coordination, even though all leg movements can be performed normally while sitting or lying down. To confirm the diagnosis, during the evaluation period the remaining 39 patients were closely monitored by rehabilitation staff who observed and documented dramatic inconsistencies in symptom presentation between that demonstrated during formal examination v. informal (unobtrusive) observation when patients were engaged in distracting activities. We avoid using the diagnoses of conversion or factitious disorder. The criterion for distinguishing between them is whether patients are aware of intentionally producing their symptoms (i.e. whether symptoms are under conscious control). An observer must infer that which is conscious v. unconscious, an inference that is impossible to make definitively. This diagnostic issue is addressed more fully in the discussion section below.

Patient demographics and presenting symptoms are summarised in Table 1. Two groups were readily discernible on the basis of the length of time the symptoms were present. Nine patients had acute motor disorder (onset within 2 months of admission) and 30 patients had chronic motor disorder (a symptom duration of more than 6 months). The primary symptoms were paresis and/or paralysis of one or more limbs and astasia basia. These symptoms were considered ‘primary’ because they were readily apparent to an observer and therefore critical for conferring the status of being disabled. The elimination of these observable symptoms was the focus of treatment. Patients frequently presented with secondary symptoms without a discernible organic basis. These included leg shaking, tremors, reports of pain, speech abnormalities and difficulty with bladder or bowel function. None of the patients reporting pain considered pain to be the primary reason for their disability. Inspection of Table 1 reveals that patients with chronic motor disorder more often presented with multiple symptoms.

Table 1 Patient characteristics

	Acute motor disorder (n=9)	Chronic motor disorder (n=30)
Median age of symptom onset (years)	39	42
Male	11% (1)	20% (6)
Paresis	100% (9)	63% (19)
Paralysis	11% (1)	7% (2)
Astasia basia	0% (0)	40% (12)
Pain	33% (3)	70% (21)
Bladder/bowel dysfunction	0% (0)	27% (8)
Other secondary symptoms	11% (1)	60% (18)
More than one non-organic symptom	11% (1)	77% (23)

Procedure

Prior to admission, patients were told that, regardless of the origin of their disorder, full recovery was possible with intensive in-patient rehabilitation. Among the 39 consecutive patients who agreed to treatment, 37 underwent standard behavioural treatment and two patients underwent the strategic-behavioural programme from the outset. If after 4 weeks of treatment there was no progress, the strategic protocol was implemented. Progress was defined as clearly observable improvements in gait and posture during physiotherapy. The first 20 patients were provided with individual counselling to ‘help them through a difficult rehabilitation process’. As patients began to improve, they were encouraged to discuss concerns related to discharge. The vast majority consistently maintained that they did not have emotional concerns (i.e. they failed to engage in a meaningful therapeutic process). Accordingly, individual counselling was rarely provided for subsequent patients. In anticipation of discharge, all patients were encouraged to consider follow-up counselling to help them adjust to the transition to a non-disabled status. Almost all declined. The few patients who initially agreed to follow-up counselling never attended the scheduled appointments.

Outcome measures

Outcomes were defined with respect to outward signs of disability. The usual index of improvement in rehabilitation - increasing functional independence - was not considered an appropriate outcome because patients readily improved in their ability to perform tasks independently while still exhibiting little or no change in the abnormal way they used their limbs. It was this apparent effort of patients to maintain control over the nature and extent of their improvement that prompted the development of the strategic protocol (Reference Teasell and ShapiroTeasell & Shapiro, 1994). Outcomes were established from chart review by both authors independently and defined as follows.

1. (a) Complete/near complete improvement. Patients displayed no overt signs of abnormal movement or posture suggestive of disability, nor did they complain of any symptoms that would render them disabled from the perspective of an outside observer. This included complete resolution of secondary symptoms. The only exceptions were very subtle residual signs such as the use of a one-point cane, but with normal posture and gait.

2. (b) Significant improvement. Overt symptoms of disability were significantly reduced relative to admission status. However, an outside observer would still view the patient as disabled. Only one patient met this criterion - the first patient for whom we developed the strategic-behavioural protocol. She was admitted with quadriplegia and requested discharge once she progressed to fully independent paraplegia (Reference Teasell and ShapiroTeasell & Shapiro, 1994). The strategic protocol was altered with subsequent patients so that failure to achieve complete resolution of symptoms was deemed proof of a psychiatric aetiology.

3. (c) Minimal/no improvement. Outward signs of disability were not significantly reduced relative to admission.

Methodological limitations

This paper is best viewed as a series of repeated case studies, with the many associated methodological limitations. Arguably, the most significant flaw is the absence of systematic follow-up, thus precluding any conclusions about long-term outcome. The majority of patients came from a significant distance and were lost to follow-up. Based on an initial attempt to collect follow-up data on the first 20 patients, a conservative estimate of relapse among patients with chronic motor disorder would be 30-40%. The simulation of disability represents an extreme solution to life problems and most likely reflects the existence of certain variables - whether conceptualised as personality dimensions, skill deficits, unconscious conflict or family systems factors - that render patients at risk for retreating back into the sick role. In some patients, these predisposing variables are likely to exert relatively greater influence. Patients with the most severe form of factitious disorder, Munchhausen's syndrome, most likely fall into this category. Elsewhere (Reference Shapiro and TeasellShapiro & Teasell, 1997a ) we suggest a more systematic and potentially more effective strategic intervention to require patients to continue to prove the legitimacy of their original disorder by remaining symptom-free post-discharge.

The absence of long-term follow-up is particularly problematic because patients diagnosed with conversion disorder may demonstrate a fluctuating course and thus it is conceivable that the observed improvement reflected this natural fluctuation. This is especially true of the patients with acute motor disorder, who might have remitted without any intervention. However, among our patients with chronic motor disorder almost all reported a history (typically greater than 2 years) of non-remitting symptoms that, if anything, became progressively worse over time. Only one patient reported a history of temporary symptom remissions and this patient did not improve following the strategic intervention. It is also possible that among patients with chronic motor disorder the differential effectiveness of the strategic v. the behavioural intervention was a function of increased time in hospital. We believe that this is unlikely. For virtually every patient with chronic motor disorder who was eventually discharged symptom-free, introduction of the strategic script was immediately followed by improvements in physiotherapy that were qualitatively different from any improvements observed previously. Thus, patients would exhibit, for the first time, a more normal gait and/or posture whereas previous improvements were in the form of increased ability to accomplish daily tasks without any concomitant observable decrease in the highly abnormal ways they used their bodies.

Another potential methodological concern is that outcome was based upon retrospective chart review by the study authors. However, to be categorised as a treatment success the patients had to demonstrate virtually complete normalisation of what previously was a dramatically abnormal gait and/or posture. They also had to report resolution of any (non-pain-related) pseudo-neurological symptoms that were not readily observable this included normalisation of bowel and bladder function. These outcomes were clearly documented in patients' charts, as was the referral to psychiatry when patients continued to exhibit or report symptoms and were deemed treatment failures. This ‘all-or-none’ outcome criterion (notwithstanding the very subtle residual symptoms temporarily allowed) thus left little room for observer or investigator bias. Ultimately, definitive conclusions would be possible only after replication of these findings using a randomised controlled design and incorporating rigorous outcome measures (v. chart review) and systematic follow-up. Given the inherent difficulties in identifying and recruiting this patient population and the deceptive aspect of the strategic protocol, few clinical researchers are likely to consider such an undertaking.

What factors maintain conversion/factitious disorders?

The strategic protocol was developed and refined based upon observations of the first three patients (Reference Teasell and ShapiroTeasell & Shapiro, 1994) who appeared to try to control the nature and extent of their disabled status by maintaining the abnormal way they used their limbs. In designing this protocol, we were guided by strategic therapy's conceptualisations of behavioural disorders and its approach to patient resistance (Reference Watzlawick, Weakland and FischWatzlawick et al, 1974; Reference Fisch, Weakland and SegalFisch et al, 1982). Strategic therapy is, in turn, an outgrowth of the 1960s-70s family therapy movement and the pioneering work of Milton Erikson, Don Jackson and Jay Hayley (Hayley, Reference Hayley1973, Reference Hayley1976). For strategic therapists, life problems become disorders when, as a problem is not resolved, more of the same ineffective solution is applied. A vicious circle then ensues, with the problem escalating in size and nature to a point where it may have little apparent similarity to the original difficulty. We can speculate how this process might unfold and account for the differential impact of standard behavioural and strategic treatments.

In the case of most acute ‘conversion’ disorders, one would expect symptoms to resolve quickly as the precipitating stressor is no longer a factor and if patients are told that their symptoms will dissipate and do not require further attention. However, if symptoms persist the patients move into a sub-acute stage where further medical investigations rule out organic conditions and the possibility of a psychiatric aetiology often is raised. Once the spectre of the disorder being ‘feigned’ is raised, relinquishing the symptoms may be seen as confirming their non-organic nature. The more resources that have been provided to the patient in the form of time, finances and emotional support, the more frightening the anticipated reaction of family who have made these significant sacrifices to accommodate the patient's needs. Thus, although the original stressor may have become inconsequential, a more immediate problem has emerged with potential for longterm, negative interpersonal consequences. The availability of a face-saving medical intervention at this sub-acute stage (e.g. in-patient behavioural intervention) may allow the patient to get better. However, if no such intervention is offered, and/or the precipitating stressor is still present, a safer response would be to maintain the symptoms. From the patient's perspective, lingering doubts about the veracity of symptoms may be preferred to confirming these suspicions by getting better. An even more effective response may be to develop more symptoms, which would be sure to generate a new round of medical investigations. The more seriously the patient is affected by these new symptoms, the more difficult it would be for others to question their veracity. The finding that patients with acute motor disorders typically presented with only one symptom whereas patients with chronic motor disorders evidenced multiple problems is consistent with this conceptualisation.

As symptoms become more chronic, simply providing ‘an excuse to get better’ may no longer suffice for the majority of patients. The longer the problem has continued and the more severely affected the patient, the greater the familial resources that have been devoted to the patient. The fear that symptom resolution would cast doubt on the veracity of the disorder would be even more pronounced at this stage. With chronicity comes a history of increasingly sophisticated investigations and an impressive array of specialists, thus requiring an even more impressive rationale and intervention before a patient can be confident that recovery would not cast doubt upon the nature of the disorder. The ‘muscle patterning’ explanation that accompanied our behavioural intervention likely was not sufficiently impressive for most patients with chronic motor disorders. The longer the symptoms persist, the greater the likelihood that additional issues emerge, making it even harder for patients to relinquish their symptoms. The longer the sick role has enabled patients to bypass many of life's difficulties, the greater the fear of being unable to function adequately outside this sick role. Over time, many relationships dissipate to give way to new relationships formed on the basis of the patients' disability. Recovery carries with it the possibility of significant isolation, especially if it casts doubt on the legitimacy of the disability. Another factor is the admiration that most, if not all, of our patients received for apparently coping so well with their disability. Thus, patients usually appeared happy and well adjusted and were viewed as pillars of emotional strength. Undoubtedly, this view would be put to the test even if recovery did not cast doubt on the legitimacy of the disorder. With chronicity, most of our patients received long-term disability benefits, the potential loss of which may have been less anxiety-provoking than the prospect of having to function successfully in the workforce. Although the strategic intervention did not eliminate these barriers, it left patients with little choice but to accept the face-saving option of relinquishing their symptoms and confronting the feared consequences of being well.

Symptom duration was the critical factor predicting the outcome of the standard behavioural intervention. Patient characteristics associated with treatment failure in the strategic programme were less clear. However, clinical observation suggested that failure to recover most often occurred in patients whose families either would not believe or were relatively unaffected by the psychiatric diagnosis. For instance, one male patient who appeared unconcerned about a psychiatric diagnosis came from a strongly patriarchal culture in which the entire family readily accepted their duty to care for him. He likely also wielded sufficient influence over his family's understanding of his difficulties that he could successfully dismiss the strategic script presented in the family conference. In contrast, the strategic programme appeared to work especially quickly when one or more family members appeared both angry and sceptical about the nature of the symptoms. Thus, the effectiveness of the strategic protocol appeared to be a function of the degree to which a psychiatric diagnosis carried with it the potential for significant negative responses from family. Although all our patients had family with whom they were living, it may well be the case that patients living alone and without family support would not have benefited from this treatment.

Notwithstanding its methodological limitations, these results, along with the discussion of possible factors maintaining conversion/factitious disorders, suggest a relatively cost-effective approach to early intervention. Patients who do not initially respond to reassurance and a relatively simple medically oriented intervention (e.g. out-patient physiotherapy) could be presented with an out-patient strategic protocol, thus increasing the likelihood that they will avail themselves of an early, face-saving opportunity to relinquish their symptoms.

Conversion v. factitious disorder

The foregoing discussion of the conditions under which patients will ‘relinquish’ their disability may leave the impression that patients maintain conscious control over their symptoms. Indeed, our original development of the strategic protocol (Reference Teasell and ShapiroTeasell & Shapiro, 1994) was in response to what we viewed as patients' attempts to actively control the rehabilitation process by limiting the nature and extent of their improvement. The distinction between conversion and factitious disorder is that the former is not consciously produced whereas in the latter the patients are intentionally simulating their symptoms. Accordingly, it could be argued that our patients are more appropriately diagnosed as having a factitious disorder. The DSM-IV (American Psychiatric Association, 1994) indicates that the judgement that a symptom is intentionally produced should be based on direct evidence, as happens when a patient with haematuria is found to possess anticoagulants and blood studies are consistent with anticoagulant ingestion. However, for pseudoneurological symptoms such as paralysis or paresis, the DSM-IV provides no criteria to distinguish conscious from unconscious intent and thereby automatically relegates these patients to the conversion category. Accordingly, it was of no surprise that all our patients previously had been diagnosed as having conversion disorders. Ultimately, the observer must infer that which is conscious v. unconscious, an inference that is arguably impossible to make definitively, especially in the case of pseudoneurological symptoms. Miller (Reference Miller1988) argued that the criterion of whether patients are consciously aware of producing their symptoms should be dropped from the diagnosis of conversion disorder. This would eliminate the distinction between conversion and factitious disorders. Our experience with the in-patient rehabilitation of non-organic motor disorders also leads us to question the usefulness of this distinction.

Ethical considerations

The strategic protocol may be viewed as deceptive and manipulative. Patients signed a standard consent form explaining that the multidisciplinary approach necessitates that patient information is shared both among team members and with patients' families. However, patients could not be informed about the exact nature of the programme. This raises ethical concerns related to informed consent and patient autonomy on the one hand, and undue medical influence and control on the other. Similar concerns were raised in the 1970s with the advent of behaviour therapy, particularly the use of contingency programmes, in institutional settings. Wachtel (Reference Wachtel1977) considered many of these same ethical issues in his classic text on the integration of psychoanalysis and behaviour therapy and his perspective is equally applicable to strategic therapy interventions. He started with the not uncommon view, at that time, that the use of reinforcement was a form of manipulation and coercion and antithetical to the psychotherapeutic process, which involved self-transcendence, a process of choice that originates within the person (Reference WheelisWheelis, 1973). Wachtel (Reference Wachtel1977) argued that the term ‘manipulation’ prejudges the issue and suggested that an alternative perspective is to view ‘the therapist as obligated to make a maximum effort to use his knowledge, skills, and understanding to help the patient free himself of a destructive cycle of events...’ (p. 274).

Wachtel also argued that part of the difficulty is that reinforcement is somehow viewed as having an inexorable controlling effect that renders the patient incapable of choice. He points out that, even when being reinforced, patients are always making choices. The strategic protocol also provided patients with choice, the opportunity to save face and confront the fear of getting well or accept the psychiatric nature of the disorder and seek another form of treatment. There may be a concern that by pitting a medical diagnosis against a psychiatric diagnosis we accentuated the stigma associated with the latter, thus making it more difficult for patients to choose psychiatric treatment. However, when discussing the psychiatric diagnosis we were careful to define it as simply another type of illness, based on an ‘unconscious’ need and requiring treatment that is ‘non-medical’. We emphasised that either diagnosis allowed for recovery, given the appropriate treatment. Wachtel (Reference Wachtel1977) reminds us that influence is inherent in all human relationships, including psychotherapy. Regardless of orientation, therapists' actions ultimately are designed to enable the patient to act and feel differently. He argues that the therapists' aim ‘is not to enable a pristine separation of outer from inner influences, but, rather, to enable the person to be more fully alive to the possibilities life offers’ (p. 247).

In considering use of the strategic protocol, practitioners may be less concerned about exerting influence and control but uncomfortable with having to lie to patients and their families in order to do so. In strategic therapy, the manner in which the therapist ‘frames’ the problem is critical and typically involves lying (Reference Fisch, Weakland and SegalFisch et al, 1982). This is particularly true of our intervention, which not only required the attending physician to lie but involved varying levels of deception on the part of the entire treatment team. Although some discomfort may be inevitable when presenting the strategic script, it is necessary to lie well (i.e. it is necessary to be unequivocal) - if a patient recovers completely and permanently, it is and always was a physical disorder. One must resist the urge not to lie by leaving the question of causation ambiguous. If patients believe that recovery can still be interpreted as reflecting a psychological aetiology, they may view it as a ‘no-win’ situation and elect the safe alternative of remaining disabled and avoiding the feared consequences of becoming well.

CLINICAL IMPLICATIONS

1. ▪ In-patient behavioural treatment is ineffective for chronic conversion/factitious disorders.

2. ▪ Factors maintaining conversion/factitious disorders change over time.

3. ▪ A strategic-behavioural approach may be effective for chronic conversion/factitious disorders and if introduced at the sub-acute stage it may prevent chronicity.

LIMITATIONS

1. ▪ Methodological limitations inherent in repeated case series preclude definitive conclusions.

2. ▪ Failure to distinguish between conversion and factitious disorders is problematic.

3. ▪ The strategic-behavioural treatment protocol raises significant ethical concerns.