The efficacy of CBT for severe mental illness and the challenge of dissemination in routine care

Cognitive behavior therapy (CBT) for severe mental illness has acquired a solid scientific status. Although the effect sizes are small to moderate, they are also quite robust and re-established in many meta-analyses over the last decade. Small to moderate effect sizes are common in adjunctive therapies, since the adjunctive intervention has to add further effects to a treatment which is already efficacious.

The conclusion at this moment is that monotherapy with pharmacological agents such as antipsychotic medication and antidepressants does not constitute optimal treatment for severe mental illness. CBT can add to that with improved symptom reduction and fewer relapses in cases of depression (1). In children and adolescents, early intervention services for psychosis, delivering CBT and other interventions, are cost-effective (2). CBT has been found to be effective in the prevention of a first episode of psychosis in meta-analyses (3), and also cost-effective (4,5). So, CBT adds health gains for lower costs. Why not implement CBT in routine services?

Here we meet the biggest challenge for CBT in severe mental illness. Although CBT for schizophrenia has been recommended in several guidelines (e.g., 6) for over a decade, the accessibility in routine services even in England, where CBT for psychosis is most promoted and researched, is dramatically low (7).

The pharmaceutical industry has built an implementation infrastructure that is missing in psychosocial and psychological interventions. Whenever a new drug is released on the market, a substantial marketing budget is used on merchandising. Hundreds of salesmen will then visit doctors with brochures, free specimens and small gifts. In this way, the medication is successfully disseminated in the routine practices of the prescribers in a relatively short time. The challenge for CBT is that the professionals who will deliver the intervention first have to be trained and employed by the services, while a marketing budget is completely lacking. Training takes several years, and vacancies are not automatically open after training. This slows down the process enormously.

The caseload for a CBT therapist is much lower than for a medication-prescribing psychiatrist. Many more CBT therapists are needed to guarantee an access to CBT comparable to the access to drug treatment. Several therapists have been training nurses in CBT, but these nurses in general have too little knowledge of general psychopathology to develop adequate individually tailored case formulations, and permanent supervision is often needed.

Training of nurses and low-intensity therapists in CBT interventions directed at accomplishing a patient's personal goal has been successful. These interventions are more broadly accessible to service users, but the need for more specialized CBT for complex cases is not diminished (8). At some places in the Netherlands, CBT therapists work in teams with nurses, where the CBT therapist develops the individually tailored case formulation and the nurse is conducting parts of the therapy such as exposure exercises, self-esteem protocols, relaxation exercises. In this way, the shortage of CBT therapists working with patients with severe mental illness can partly be overcome.

On the other hand, CBT is a transdiagnostic intervention, successfully applied to almost all psychiatric conditions. The underlying mechanisms of (for instance) selective attention and avoidance behavior are common to several anxiety disorders, depression, eating disorders and also paranoid delusions. This is certainly an advantage. A therapist can use the principles of CBT in several kinds of disorders, and comorbid disorders as well. Training can be limited to the universal principles, and only small adjustments are needed for each disorder, such as different assessment instruments, a different speed of progress through the protocol, and different attitudes in case formulation dependent on the patient's illness insight. Because of the universal nature of generic CBT, therapists working in other psychiatric domains could easily enter the field of severe mental illness with little extra training.

This advantage of universality is, however, a disadvantage as well. Generic therapy might not be targeted enough for the specific problems of some patients. In recent years, more specific variants of CBT have been developed and successfully tested for psychotic symptoms, command hallucinations, and negative symptoms. Dissemination is not the dissemination of a final intervention. The intervention is at the same time improved and differentiated to address more specific problems with improved efficacy.

In conclusion, we can state that CBT has reached such a level of adjunctive effectiveness in severe mental illness that pharmacological monotherapy is not the optimal treatment anymore. CBT must be made available and accessible in routine psychiatric practice. This can only be accomplished if budgets are reallocated from clinical services with high 24-hour staffing needs to less labour-intensive outpatient services with more specialized interventions directed at symptomatic and social remission. The era of implementation has a slow start, but there is progress.

At the same time, there is room for improvement of CBT protocols by more specific and symptom-tailored interventions. The era of research has just begun for psychological treatments such as CBT in severe mental illness.