Case illustration: Frank

Frank presented with a 19-year history of panic disorder with agoraphobia, GAD, social anxiety and depression. He was aged 44 and married with a grown-up son. He had developed panic attacks in his mid-twenties with no previous mental health or developmental problems. He started avoiding situations out of fear he would have further panic attacks. His avoidance increased and he eventually lost his job. Frank retreated from normal life, became increasingly worried and pre-occupied about panicking and being seen by others. As his negative affect increased (e.g. anxiety, embarrassment, shame) he became more avoidant and his positive affect decreased. His motivation to find other work and maintain a social life gradually became suppressed. Anxious avoidance interacted with depressive withdrawal and Frank became reclusive: within a year of losing his job he rarely left his house and had only occasionally done so in the previous 18 years. Prior to his referral Frank had received six courses of counselling, clinical psychology or psychotherapy. These had produced minimal benefit.

Frank was a challenging case because of multiple co-morbidities, long-term chronicity and resistance to prior treatments. Figure 1 depicts the biopsychosocial factors associated with Frank's depression. The question is: were these factors complications? The social and work factors were not really complications: his lack of social life and unemployment were a consequence of persistent depression and they did not directly affect his access to treatment, readiness for it, or capacity to form a good working alliance. However, the four co-morbid anxiety disorders were a potential source of complexity: panic disorder, agoraphobia, social anxiety and GAD. In spite of the apparent complexity, Frank's therapist proposed a simple strategy of focusing consistently on overcoming his agoraphobia, nested within desired life-goals to help sustain Frank's motivation. Frank and his therapist were aware of the potential influence of panic, social anxiety and GAD but made overcoming agoraphobic avoidance the priority. It was hoped that these behavioural changes would also have an activating effect towards desired goals, creating positive reinforcement as indicated within behavioural activation (Martell et al, Reference Martell, Dimidjian and Herman-Dunn2010). This iterative process was approached through three recurring steps:

1. (1) Establish concrete approach goals: i.e. Frank's desires of how he would like to live;

2. (2) Conduct exposure to feared situations while increasing behavioural activation towards desired goals;

3. (3) Reflect on the emotional and interpersonal consequences of engaging in new behaviours.

Frank generated four approach goals and these formed a natural hierarchy: (a) go out of his house without fear; (b) go out for a coffee or meal; (c) train to be a painter/decorator; (d) get a job as a painter/decorator. Figure 2 summarizes the behavioural changes across the first phase of Frank's therapy, showing the graded tasks that he approached at each stage. A slow pace was established with lots of repetition of the target behaviours. The aim was to conduct full and thorough exposure to feared situations and to only move onto the next step when the last one had become comfortable, not just tolerable. The first stage focused on Frank stepping a few yards out of his front door into his garden, then holding his head high when he did this, then walking in the local street, and so on.

Figure 2. Frank's progress in the first phase of therapy

As might be expected, at various points progress was slowed by barriers to change that at the time Frank felt were insurmountable, for example, worry about being stopped and asked questions, self-critical rumination, intense shame and severe anxiety symptoms. The strategy was to attend to these factors if they created a block on progress, doing sufficient work on them to enable the change process to continue. Using a disorder-specific strategy, it is important to remain aware of potential interactions with other complexity factors just as Frank and his therapist did. Their focus only switched if another factor had a limiting effect on the main target and created a block on progress. For example, at various points worry, rumination and shame all needed attention because they were limiting further behavioural change. Therapy was still tethered to behavioural change, but these other factors were given time-limited attention to overcome their limiting effects. Having received the attention they needed, this had a consolidating effect on the primary target and treatment then returned to focus directly on it.

The principles underlying the disorder-specific approach can be summarized as follows:

1. (i) Tether to a primary target (A) and keep it the main priority;

2. (ii) Monitor the influence of, and effects on, other factors;

3. (iii) If another factor (B) limits progress on A, give B time-limited attention;

4. (iv) Attend only to those aspects of B that are limiting progress on A;

5. (v) Return the focus to A once those aspects of B have been resolved;

6. (vi) Review progress on A after a sustained dose of treatment has been delivered (usually around 5–6 sessions);

7. (vii) Only change the primary target:

   1. (a) When A is resolved and other problems require further attention;

   2. (b) If limited progress has been made on A in spite of targeting it in a consistent way.

The evidence base in CBT for major depression suggests that this strategy will be effective most of the time because disorder-specific treatments have trans-diagnostic effects: co-morbid disorders are often interdependent and mutually reinforcing. This is a potential source of complexity but it can mean that not all disorders need direct treatment. Assuming there are sympathetic links between disorders, a number of problems will improve through trans-diagnostic and indirect effects (DeRubeis et al., Reference DeRubeis, Hollon, Amsterdam, Shelton, Young and Salomon2005). In this context sympathetic means that A improves when B improves, and A worsens when B worsens. For example, worry might not need separate treatment if changes in rumination indirectly improve the worry process. Consequently, most cases with multiple problems do not need complicated treatment. Co-morbidity is not necessarily complicated, depending on the interactions between the disorders. When problems are sympathetically linked it may not matter what the primary target is, as long as it is manageable and agreed collaboratively. There can be multiple paths to recovery: if A is targeted B improves, and if B is targeted A improves. What matters is tethering to a primary target and monitoring the influence of, and effects on, other problems.

This is a good summary of Frank's treatment. Frank's depression was challenging and difficult to change because of its chronicity and stuckness. Progress was slow. It took Frank two years to achieve his goal of going to college. After three years he was looking for work, functioning well and managing anxiety and depression symptoms that had reduced to a moderate level. However, neither Frank's depression nor its treatment was particularly complicated; progress was difficult but not complex. Although there were multiple co-morbidities the disorders were sympathetically linked. Overcoming agoraphobic avoidance had a simultaneously activating effect on his depression and these changes had a trans-diagnostic effect on his other problems.

Case illustration: Jeff

At the time of his referral, Jeff was aged 48 and married with two daughters. He was a public service employee with a 20-year history of obsessive compulsive disorder (OCD) with recurrent major depressive episodes (MDE). Jeff had previously received three courses of CBT:

Treatment 1: CBT for OCD (National Collaborating Centre for Mental Health, 2005). Jeff received 14 sessions over a 4-month period. Due to unpredictable changes in mood and affect, it was difficult keeping the focus on OCD. During this treatment a psychiatric assessment resulted in a diagnosis of bipolar II disorder. Jeff disclosed his diagnosis to his employer, believing this was the right thing to do, and this initiated a lengthy process of occupational health assessment. The bipolar diagnosis coincided with staffing changes and Jeff was offered a new therapist who specialized in the area of unstable mood.

Treatment 2: CBT for bipolar disorder (Meyer and Hautzinger, Reference Meyer and Hautzinger2012). Jeff received 53 sessions over a 3-year period. The therapeutic dose reflected the severity of Jeff's depression, mood swings and occasional hypomanic episodes. Treatment was strongly affected by the occupational health process. Jeff was assessed by a number of medical practitioners, not all with psychiatric training. The stress and uncertainty provoked by these assessments interacted with Jeff's unstable moods and made it challenging to progress. Over an extended period there was a gradual improvement and stabilization of Jeff's mood. OCD was not the priority during this period and it was not targeted explicitly within treatment.

Treatment 3: Consolidation of CBT for bipolar disorder. Jeff received a further 23 sessions over a 2-year period to consolidate progress and help him cope with the occupational health process. In the intervening period, his employer had reacted in a risk-averse way, moving him to a less senior role within the organization. His manager informed him: ‘your mental health problems mean you cannot talk to the public’. This had a devastating effect on Jeff's self-esteem and shattered his assumptions about his employer's competence, loyalty and duty of care. It provoked extreme anger, reactivated Jeff's OCD and led to further mood de-stabilization. The context of acute occupational stress was a complicating factor because it increased uncertainty and made it difficult for Jeff to maintain a consistent treatment focus.

In spite of the benefits gained from these treatments, Jeff relapsed into moderate depression with occasional expansive moods and his OCD had also been re-activated. He was pre-occupied by anger with his employer and this was interfering with his functioning at work. Disorder-specific protocols had not had a lasting effect so an alternative approach was adopted. Jeff was offered a new therapist and subsequently received 44 sessions of CBT over a 2-year period. The strategy was firstly to map the non-depression factors that could be complicating his presentation, illustrated in Fig. 3.

Figure 3. Biopsychosocial map of complexity factors associated with Jeff's depression

This revealed three social factors and three psychological factors. The therapist did not assume that these were complicating Jeff's mood disorder; rather, there was a process of guided discovery to find out whether or not they were having a complicating effect. There was a potential healthcare factor concerning Jeff's repeated relapses that may have reduced his confidence in CBT or the clinical service. In fact, this was not the case. Jeff reported experiencing the benefits of the previous courses of CBT but something stopped them having a sufficiently potent and lasting effect. There was a potential interpersonal factor following significant tensions between Jeff and his wife; however, they had managed to overcome these difficulties and they were no longer problematic. There was a clear and pronounced occupational factor that had a complicating effect on Jeff's depression both during and after the occupational health process.

The occupational stress interacted with three psychological factors: hypomanic moods, OCD and post-traumatic anger. Occupational stress over the previous five years had a traumatic impact on Jeff. He was not suffering from PTSD but he continued to feel ‘devastated’ by his employer's actions. They had a profoundly disruptive impact on his self-identity and Jeff felt extremely hurt and angry several years after the most stressful events had occurred. Jeff's work had been a significant and self-defining aspect of his self-identity throughout his adult life.

At Step 4 the strategy is not to deliver a disorder-specific protocol for any one of these problems. Instead, an individualized formulation is developed of the interactions between these factors, i.e. the functional links where the output of A is the input to B, and so on. At the outset of therapy a simple map is used with the main presenting problems on it and this is used to explore the interactions between them. In Jeff's case, as the map grew and became better specified there were three main benefits to having an over-arching formulation. Firstly, because Jeff's problems were inter-dependent and highly changeable the formulation helped to identify his mental state at any point and enabled reflection on how it was affecting his other difficulties. Secondly, the emerging map made it more straightforward to agree the optimal treatment focus at that point. Thirdly, by paying close attention to the interaction between Jeff's difficulties, functional links became apparent and these could then be targeted for change.

Using this approach it is important to ensure that therapy is evidence-based in two key respects: firstly, the formulation should be tested empirically through guided discovery, collaboration, behavioural experiments, etc., so that hypotheses about functional links are disconfirmed and not just assumed. Secondly, only treatment components from evidence-based protocols should be used, even if they are delivered in a novel sequence or innovative way. In Jeff's case occupational stress, depression, hypomania, OCD and post-traumatic anger were interacting in a way that complicated the usual maintenance of depression and made it harder for disorder-specific protocols to have a lasting effect.

The formulation evolved gradually over several sessions and the final version is presented in Fig. 4. Jeff's difficulties were re-triggered by continuing to work in the same organization. Memories of how he had been treated were re-activated on a daily basis and this provoked intense upset, anger and occasional rage. The anger triggered vengeful thoughts and images in which he would retaliate against his managers with extreme physical violence. A complicating factor was the unpredictable effect of those thoughts on Jeff's obsessional and hypomanic tendencies. Sometimes he would experience the thoughts as unwelcome intrusions and an OCD cycle would be triggered with inflated responsibility to prevent harm, anxiety about acting on the thoughts and then various forms of neutralization, including behavioural avoidance and thought suppression. On other occasions Jeff would experience the thoughts as liberating: they enabled him to emotionally process traumatic memories but this could also have a disinhibiting effect, leading to grandiose ideas about being powerful and dominant, resulting in elated mood and raising concerns about risks of harm to others.

Figure 4. Jeff's individualized case formulation

Jeff had learned the harmful consequences of hypomania during earlier treatments and he was wary of acting inappropriately with other people. Memories of past hypomanic behaviour were sometimes activated, leading to guilt and shame-based rumination which had the potential to depress Jeff's mood and spiral into helplessness and social disengagement. Guilt and shame also activated Jeff's felt-sense of responsibility to act dutifully as a husband, father and public servant and this had a reinforcing effect on the OCD cycle. As Fig. 4 illustrates, by mapping out the links between disorders Jeff and his therapist were able to make sense of his different mental events and mood states. The formulation helped to de-mystify and explain what Jeff otherwise experienced as ‘mental turmoil’ or ‘going mad’.

The main complication was anger simultaneously triggering an inhibiting process (OCD) and a disinhibiting process (hypomania) that produced unstable moods, internal conflict and cognitive disorganization. In this respect the maintenance of Jeff's depression was subtly different from many clinical presentations. His depressed moods had classic features such as social withdrawal, rumination and self-critical thinking. However, targeting depressed moods did not automatically improve interactions between depression, hypomania, OCD and post-traumatic anger.

Three trans-diagnostic factors also became apparent; firstly, across different mood states Jeff would often engage in unhelpful repetitive thinking. The content tended to be self-critical in the depressed mode, self-actualizing in the hypomanic mode, retaliatory in the traumatized mode and obsessional in the OCD mode. The common process was repeated dwelling on emotionally charged topics that was generally unproductive and created further distress. Secondly, Jeff's affect was unstable irrespective of which disorder was dominating his current mental state. Thirdly, across different mental states Jeff's self-identity was often uncertain and confused. There was a degree of self-identity confusion prior to treatment that had been exacerbated by the occupational health process. These trans-diagnostic factors were placed in the centre of the formulation as a reminder of higher-order processes that could be targeted in any mood state, for example, encouraging Jeff to be reflective rather than ruminative, encouraging acceptance of affect in favour of emotional avoidance, and affirming experiences in which he felt grounded in his ‘true self’.

In addition, Jeff's treatment included a series of sessions delivering exposure and response prevention (ERP) for his obsessional thoughts (McKay et al., Reference McKay, Sookman, Neziroglu, Wilhelm, Stein, Kyrios, Matthews and Veale2015). Their content was primarily based on the perceived threat of acting on his anger and harming his employer. Jeff was able to expose himself to these thoughts, tolerate the distress they provoked and gradually recognize that they were only thoughts and not intentions. The only deviation from standard protocol was monitoring for hypomania and behavioural disinhibition; this occurred in some of the early exposures. This was used to elaborate the formulation. Overall, the formulation had an integrating effect on Jeff's learning; he was able to identify which processes were active at any point and developed greater mastery in counteracting them with different cognitive and behavioural strategies. Jeff has since made a successful transition into early retirement and his treatment has ended.