Guidelines on the treatment of obsessive–compulsive disorder (OCD) from the National Institute for Health and Clinical Excellence propose a six-stage model of stepped care, with patients with the most severe and resistant problems being treated as in-patients using psychological and drug treatments or other intensive treatment programmes (National Institute for Health and Clinical Excellence, 2006). Treatment-resistant OCD is associated with high levels of dependency on mental health services and high levels of comorbidity, with suicide reported in up to 15% of patients (Reference Angst, Gamma and EndrassAngst et al, 2005). Recently, mental health services have been increasingly provided locally, with less emphasis on highly specialist units. There is only one fully staffed National Health Service in-patient unit in England that treats patients with severe OCD using predominantly psychological methods and another in Scotland that specialises in neurosurgical techniques. This dearth of services is not surprising since only about 1% of the population have clinically relevant OCD (Reference Karno, Golding and SorensonKarno et al, 1988).
Modern psychological and pharmacological treatments have revolutionised the outcome for these patients (Reference Mohammad, Fineberg and DrummondMohammad et al, 2000). Approximately three-quarters will improve with graduated exposure treatment (Reference Drummond, Wilkinson and SteinDrummond, 1998) and about two-thirds will respond to selective serotonin reuptake inhibitors (SSRIs; Reference McDougle and GoodmanMcDougle & Goodman, 1991).
Most patients referred to the Behavioural Cognitive Psychotherapy Unit at Springfield Hospital receive either intensive home-based or out-patient treatment with telephone monitoring. However, the 10-bedded in-patient unit treats those patients with severe, chronic, resistant OCD who cannot be managed as out-patients owing to the nature or extent of their condition and the need for care by nurses trained in the management of severe mental illness (Reference DrummondDrummond, 1993). Following referral each patient is comprehensively assessed to confirm the diagnosis, specific problems, history (including the possible reasons for the previous failure of treatment) and risks to ensure appropriate strategies are implemented. Patients are evaluated for severity of illness using observer and self-assessment validated rating instruments on admission and at discharge.
The mainstay treatment is behavioural, but this may be combined with medication change, cognitive reattribution or psychoeducational methods (such as danger ideation reduction therapy; Reference Jones and MenziesJones & Menzies, 1998). All patients have individualised treatment from a trained therapist who continues the treatment programme on a daily basis. The content of therapy is dependent on functional analysis of the problem. Here we report a naturalistic cohort study of recent treatment in our unit.
Method
A psychiatrist interviewed all patients who were admitted to the in-patient unit between 30 June 2004 and 1 July 2005. ICD–10 diagnoses were made following a detailed psychiatric interview. All patients fulfilling ICD–10 criteria for OCD (World Health Organization, 1992) were included in this study.
Demographic data were collected at the time of admission. The severity of OCD symptoms was measured using the Padua Inventory (Reference SanavioSanavio, 1988) and the Yale–Brown Obsessive Compulsive Scale (YBOCS; Reference Goodman, Price and RasmussenGoodman et al, 1989) and depression using the Beck Depression Inventory (BDI; Reference BeckBeck, 1978) at the time of admission and at discharge. Data were analysed using the Statistical Package for the Social Sciences, version 14.0 for Windows.
Results
Twenty-seven patients were admitted over 12 months but one was excluded owing to a primary diagnosis of somatoform disorder. Twenty-six patients had severe, chronic, resistant OCD (11 men, 15 women). Five patients discharged themselves without participating in therapy. Discharge data were not available for a further 8 patients. On admission there were no significant differences in either severity of depression (measured by BDI) or OCD (measured by Padua Inventory) between these 13 patients who failed to complete the measures at discharge and those who did.
The mean age of all patients was 37 years (s.d.=13.8, range 18–61 years) and the mean duration of OCD was 18.4 years (s.d.=10.9, range 4–40 years). At the time of admission the mean YBOCS score was 29.6 (s.d.=5), which indicates very severe OCD. Patients had moderate to severe depression (mean BDI score=25.4).
Previous treatment
Only 1 patient had not previously received behavioural and cognitive psychotherapy from an accredited therapist and most had received several trials of therapy. This 1 patient had refused treatment owing to his obsessive fears. He discharged himself from the unit before treatment started.
Twenty-one patients had received a trial of treatment with SSRIs and 9 had received clomipramine. A trial of therapy was defined as receiving at least 12 weeks of therapy at the maximum recommended doses (British Medical Association & Royal Pharmaceutical Society, 2005). Four patients had been prescribed SSRIs but had not completed 12 weeks at recommended doses (2 refused to increase the dose and 2 stopped owing to reported side-effects). Fourteen patients had received antipsychotic drugs to augment the SSRIs. Other drugs prescribed at the time of admission included non-SSRI antidepressants (8 patients), benzodiazepines (9 patients), sleeping pills (6 patients), buspirone (1 patient), mood stabilisers (3 patients) and propranolol (2 patients). None of the current cohort had had previous psychosurgery.
Comorbidity
Comorbidity included severe depressive episode (5 patients), agoraphobia (1), social phobia (1), hypochondriacal disorder (2), hypochondriacal disorder (body dysmorphic disorder) (2), post-traumatic stress disorder (1), alcohol dependence syndrome (2), bulimia nervosa (1), movement disorder (1) and schizophrenia (1). About half of the patients admitted to the in-patient unit have delayed sleep phase syndrome, whereby they stay awake for much of the night and sleep for most of the day, thus rendering attendance at regular out-patient appointments impossible (Reference Drummond, Fineberg and MukhopadhyayDrummond et al, 2005).
Treatment
Patients were treated on the in-patient ward for a mean of 104 days (s.d.=58.2, range 2 days to 6 months). Of the 21 patients treated, 18 had graduated exposure and self-imposed response prevention and 3 received danger ideation reduction therapy; 13 had cognitive reattribution and 2 danger ideation reduction therapy in addition to self-imposed response prevention.
Six patients had their medication optimised according to current guidelines (Reference Baldwin, Anderson and NuttBaldwin et al, 2005). This consisted of prescribing a low dose of sulpiride or amisulpiride as an adjunct to an SSRI for 4 patients, increasing the dose of SSRI towards maximum for 1 patient and changing to fluoxetine from another non-SSRI in another patient. In addition, withdrawal of benzodiazepine medication was commenced for 3 patients and 1 patient had propranolol stopped while on the ward.
Outcome
Patients showed an overall improvement on clinical measures of both depression and OCD (Table 1). Scores for obsessive symptoms on the Padua Inventory were significiantly reduced from a mean of 97.5 to 42.7. At the time of discharge, scores on the Padua Inventory were slightly above those found in a healthy population (Reference SanavioSanavio, 1988). The reduction in OCD symptoms as shown by the YBOCS represents a move from severe to moderate symptomatology. Depression, as measured by the BDI, improved significantly despite depression not being the focus of treatment. Patients moved from BDI scores indicating moderate depression to scores in the normal range.
Clinical measures of obsessive–compulsive disorder and depression at the time of admission and discharge
| Admission | Discharge | ||
|---|---|---|---|
| Measures | mean (s.d) | P | |
| Padua Inventory (n=13) | 97.5 (39.3) | 42.7 (20.1) | <0.0005 |
| YBOCS (n=9) | 29.7 (3.6) | 19.2 (8.4) | <0.005 |
| BDI (n=12) | 21.8 (13.7) | 8.3 (6.3) | <0.005 |
Discussion
Our study shows that in general patients admitted to our unit for severe OCD demonstrate clinical improvement using validated rating scales. Most studies of in-patient treatment for OCD were performed in the 1970s at the time of the development of modern exposure methods. These studies showed that between 75% and 80% of patients improved (Reference Marks, Hodgson and RachmanMarks et al, 1975; Reference Foa and GoldsteinFoa & Goldstein, 1978; Reference Wetzel, Bents and FlorinWetzel et al, 1999), but included all patients, not just the most severe, and thus did not yield significant differences from similar methods of out-patient treatment (Reference Marks, Lelliot and BasogluMarks et al, 1988). Foa (Reference Foa1991) reported a preference for treating patients in their natural environment as opposed to hospital. However, intensive home-based treatment was used with many hours of individual therapy every week (Reference FoaFoa, 1991).
In the Behavioural Cognitive Psychotherapy Unit at Springfield Hospital, we do try, where possible, to use home-based methods, but some patients are too severely ill to cope with therapy in the community. For these patients in-patient treatment seems more cost-effective as some therapies can be offered to a group. There are also possible beneficial effects of a peer group with similar problems undergoing therapy at the same time. Fewer home visits are required, thus reducing staff travelling time.
A previous study demonstrated that patients had spent a mean of 9.9 months in acute admission wards prior to referral to the unit (Reference DrummondDrummond, 1993). In the light of this, in-patient treatment might be considered very cost-effective. This previous study demonstrated that two-thirds of patients were discharged with clinical improvement (Reference DrummondDrummond, 1993). Overall the patients showed a 32% reduction in OCD symptoms on the Compulsion Activity Checklist (Reference MarksMarks, 1986) and a 32% reduction in depressive symptoms on the BDI (Reference Beck, Rial and RickelsBeck et al, 1974). Different measures were used with the current patient group, but they seem to have at least as severe symptoms and possibly better treatment outcome. Since the 1993 study, there has been refinement of cognitive reattribution techniques, the introduction of psychoeducational methods of treatment and an increased awareness of available drug treatment.
Despite these refinements, some patients still felt unable to undergo therapy following admission. The main reasons were that it would be too stressful to remain on the ward or that they did not feel able to commit to remain free of alcohol or non-prescribed drugs (which is a requirement for admission to the unit). These patients were given information about treatment and referred back to the referring teams.
The patients treated had all failed previous out-patient, home-based or in-patient treatments with their local services. This study demonstrates that there is a role for in-patient treatment using behavioural psychotherapy in combination with other approaches for these patients with the most severe form of OCD.
This study was naturalistic and the collation of data was retrospective. Thus not all data were available for analysis. Some patients refused to embark on therapy and others left treatment early and refused to repeat rating instruments. Moreover, follow-up data were not collected.
Previous studies have shown that gains made in hospital tend to persist following discharge (Reference DrummondDrummond, 1993). This may be related to the way the treatment is applied. Early on in treatment patients are expected to go home at weekends and to practise their therapy exercises in the home environment. Nearer to discharge, patients are asked to spend increasing periods at home, often with visits from staff at the unit or other healthcare workers.
In summary, specialised in-patient care can be a cost-effective means of treating a small group of patients with severe OCD.
Declaration of interest
None.
Open access
eLetters
No eLetters have been published for this article.