Awareness of being ill (insight) is often absent in individuals with psychotic disorders. Lack of insight has been shown to be related to non-adherence to medication, involuntary hospitalisation and poor prognosis (Reference Small, Small and HaydenSmall et al, 1965; Reference Lin, Spiga and FortschLin et al, 1979; Reference Bartko, Herczeg and ZádorBartko et al, 1988; Reference McEvoy, Freter and EverettMcEvoy et al, 1989). Low insight is a predictor of both relapse and readmission in first-episode non-affective psychosis (Reference Drake, Dunn and TarrierDrake et al, 2007). There is no clear association between the resolution of acute psychotic symptoms and improvement in level of insight, suggesting that insight varies independently of psychotic symptoms (Reference McEvoy, Freter and EverettMcEvoy et al, 1989; Reference David, Buchanan and ReedDavid et al, 1992; Reference Amador, Strauss and YaleAmador et al, 1993). Insight has several components: recognition that one has a mental illness, recognition of the social consequence of the disorder, acknowledgement of the need of treatment, and recognition of the pathological nature of specific psychotic symptoms.
A previous randomised controlled trial using video self-observation in a group of 18 found a significant effect (P<0.05) of video self-observation (n=9) on the Insight and Treatment Attitudes Questionnaire (ITAQ) score in the experimental group (Reference Davidoff, Forester and GhaemiDavidoff et al, 1998). We investigated the role of video self-observation in raising the level of insight using a clinician-assessed measure of insight during admission and after discharge from hospital. The study was given ethical approval by the local research ethics committee.
Method
Participants were recruited from the wards of the West London Mental Health National Health Service (NHS) Trust - those recently admitted to hospital with a diagnosis of a psychotic disorder were assessed with the cooperation of the staff on the ward. Participants had to give written informed consent to be involved, which meant that the researchers were confident that once given all the relevant information the patient had the capacity to consent to participate in a low-risk research study. It seems likely that the most unwell in-patients were unable to take part in the study and they may well have had lower levels of insight than those able to participate. Consent was obtained using the recommended procedure and consent form of the Royal College of Psychiatrists (Royal College of Psychiatrists, 1998). Written explanation regarding the study was given to those willing to participate in the video interview. They had to be available for the initial interview, a self-observation session and a follow-up interview. Participants were paid an honorarium of £10 for their time.
Inclusion criteria were: aged between 18 and 65 years old, an in-patient in the acute psychiatric ward, and meeting the DSM-IV diagnostic criteria for schizophrenia, schizoaffective disorder or mania with psychotic features. Exclusion criteria were comorbid substance misuse, organic brain syndromes and learning disability.
There were 22 in-patients (male and female) recruited into the study, 17 of whom completed all stages of the study. Characteristics and diagnoses of the sample are presented in Table 1. Patients were interviewed on video by a psychiatrist within a few days of hospital admission when the patient was showing clear psychotic symptoms. The Schedule for Assessing Insight (SAI; Reference DavidDavid, 1990) and the Positive and Negative Syndrome Scale (PANSS) for schizophrenia (Reference Kay, Fiszbein and OplerKay et al, 1987) were the rating scales administered as part of the video session. The patients received standard antipsychotic treatment at the discretion of the treating team. They were seen again prior to discharge and the SAI and PANSS rating scales were administered prior to the intervention (video self-observation) and again following the intervention. Following this, each participant discussed their views on the recorded material with the researcher.
Demographics
| Characteristics | n |
|---|---|
| Gender | |
| Male | 13 |
| Female | 4 |
| Diagnosis | |
| Schizophrenia | 10 |
| Bipolar affective disorder | 5 |
| Schizoaffective disorder | 2 |
| Age | |
| Minimum | 18 |
| Maximum | 64 |
| Mean | 36 |
Participants acted as their own control group as their level of insight was being assessed at four points in time: at recruitment, before and after self-observation, and at follow-up.
Rating scales used in the study
The PANSS is a 30-item rating instrument evaluating the presence or absence and severity of positive, negative and general psychopathology of schizophrenia.
The SAI measures three dimensions of insight:
-
• recognition of an illness and that it is a mental illness (three questions, scored 0-6)
-
• treatment adherence (two questions, scored 0-4)
-
• re-labelling of psychotic phenomena as pathological (two questions, scored 0-4).
A maximum score of 14 suggests good insight. The scale takes approximately 10 min to complete and does not require specific training.
Analysis
Initially, 22 in-patients consented to the video interview; 3 did not wish to view their video on discharge from hospital and so were withdrawn from the study. Of these, 2 felt it would be distressing to watch their video as they did not wish to be reminded of their illness now that they were feeling well; 1 did not respond to two invitations to view the video. A further 2 in-patients were withdrawn from the study by the researchers: 1 became increasingly agitated and paranoid during the initial interview and so the interview had to be terminated; the other in-patient revealed that he had not been open about his symptoms with his clinical team and since this raised doubts about his openness generally, he was withdrawn from the study. Thus, 17 patients continued in the study (Table 1).
Of the 17 participants, 5 were unavailable for the 6-month follow-up: 2 had left the country, 1 had been readmitted to hospital, 1 did not respond to invitations to attend an interview and the carer of one refused contact, not wishing to distress them.
The data were analysed using the SPSS version 10 for Windows (Table 2). As we were comparing observations at different time points in a single sample where results were normally distributed, the paired t-test was used to analyse the data.
PANSS and SAI scores during the study period
| Patient number | Initial interview (PANSS 1) | Initial interview (SAI 1) | Discharge interview (PANSS 2) | Pre-video self-observation (SAI 2) | Post-video self-observation (SAI 3) | Follow-up (SAI 4) |
|---|---|---|---|---|---|---|
| 1 | 80 | 3 | 49 | 4 | 9 | 10 |
| 2 | 80 | 12 | 52 | 13 | 13 | 13 |
| 3 | 60 | 4 | 37 | 4 | 7 | in-patient |
| 4 | 63 | 4 | 42 | 4 | 10 | 14 |
| 5 | 88 | 9 | 55 | 6 | 6 | 11 |
| 6 | 74 | 3 | 33 | 14 | 14 | abroad |
| 7 | 73 | 3 | 31 | 14 | 14 | 14 |
| 8 | 105 | 2 | 32 | 12 | 13 | abroad |
| 9 | 77 | 7 | 38 | 14 | 14 | 14 |
| 10 | 85 | 4 | 60 | 8 | 12 | refused |
| 11 | 84 | 2 | 70 | 11 | 12 | 13 |
| 12 | 82 | 3 | 61 | 7 | 9 | no contact |
| 13 | 87 | 1 | 54 | 8 | 10 | 2 |
| 14 | 87 | 3 | 74 | 5 | 5 | 13 |
| 15 | 110 | 1 | 77 | 3 | 3 | 1 |
| 16 | 101 | 4 | 57 | 12 | 13 | 14 |
| 17 | 76 | 4 | 53 | 9 | 11 | 14 |
Results
The PANSS scores at recruitment and discharge are shown in Table 3. These show a mean improvement of 31.6 points (95% CI 24.3-38.9), representing a statistically significant improvement in mental state (P=0.001). The comparison of SAI scores at recruitment (SAI 1) and at discharge, prior to watching the video (SAI 2) shows a statistically significant (P=0.005) mean improvement of 4.6 (95% CI 2.4-6.8).
Comparison of PANSS and SAI scores at different times during the study
| Mean | s.d. | Standard error (mean) | 95% CI | P | |
|---|---|---|---|---|---|
| PANSS | |||||
| At initial interview (PANSS 1)–at discharge (PANSS 2) | 31.59 | 14.16 | 3.44 | 24.3 to 38.9 | <0.005 |
| SAI | |||||
| At pre-video self-observation (SAI 2)–at initial interview (SAI 1) | 4.65 | 4.29 | 1.04 | 2.44 to 6.86 | <0.005 |
| At post-video self-observation (SAI 3)–at pre-video self-observation (SAI 2) | 1.53 | 1.97 | 0.49 | 0.52 to 2.54 | 0.006 |
| At follow-up (SAI 4)–post-video self-observation (SAI 3) | 1.17 | 3.92 | 1.13 | –1.33 to 3.66 | 0.33 |
After video self-observation (SAI 3) there was a further statistically significant (P<0.006) improvement in the insight score, a mean increase of 1.5 points (95% CI 0.5-2.5).
Twelve of the participants were followed-up at 3 to 6 months and their insight was assessed (SAI 4). There was a mean improvement of a further 1.2 points on the SAI scale, although this change was not significant (P=0.33). Instrument scores did not differ as a function of age, gender or diagnosis.
Discussion
Psychometric data
This study further supports the potential benefits of video self-observation in improving insight in individuals with psychotic illness. Both symptoms and insight improved during admission, and a further improvement in insight occurred after the video self-observation. There was no significant change (improvement or decline) in insight scores on assessment approximately 3 months after discharge.
Previous studies have shown that antipsychotic drug therapy is effective in improving insight when compared with psychological therapies (Reference MayMay, 1968). More specifically, treatment with atypical antipsychotic medication is associated with improvement in insight compared with typical medications (Reference Ghaemi and PopeGhaemi & Pope, 1994; Reference Pallanti, Quercioli and PazzagliPallanti et al 1999). We did not look at the particular treatment the participants were receiving during our study.
Our study increases the number in the group receiving the intervention of video self-observation compared with that of Davidoff et al (Reference Davidoff, Forester and Ghaemi1998). In our work, the study group is its own control group, thus avoiding complications of group matching within a relatively small sample. This resulted in a highly significant improvement in SAI score (P<0.006), which appears to be maintained over time.
Qualitative observations
Prior to watching the video, participants accepted treatment while in hospital but otherwise seemed to have little awareness of their illness or need for treatment. They had little recollection of their psychotic symptoms at the time of admission. They often did not think they needed to continue medication outside the hospital as they felt that they were well at the time of discharge.
Actually watching themselves experiencing hallucinations, expressing delusional beliefs or behaving in a bizarre manner resulted in participants feeling surprised and at times embarrassed. It helped to increase their awareness that they suffered from a severe mental illness and that treatment had benefited them.
Although participants acknowledged that they had a mental illness after watching the video, some were still reluctant to accept their diagnosis. They attributed their symptoms and illness to factors such as stress, depression and illicit drugs. However, after watching the video they seemed more convinced that medication would help them to remain well.
Apart from the three individuals withdrawn from the study, the participants found the intervention useful as they felt that being reminded of their symptoms was a strong incentive to take medication and engage with services. A few of them requested copies of their video as they felt it would help their families to understand their illness. The following are some of the comments made by participants after watching their videos:
‘I was on a different level - cannot believe it was me.’
‘Feel scared of becoming ill again - will take my medication.’
‘Was off my trolley - medication really works.’
‘Something was definitely wrong - probably the doctors are right.’
‘Mind was playing tricks and not working as it should.’
‘Cannot believe how argumentative I was - can my boy-friend view the video?’
One person said: ‘You should not remind people of bad things.’
The pattern of participation and comments made by participants seems similar to that observed by Davidoff et al (Reference Davidoff, Forester and Ghaemi1998) who also reported that patients seemed surprised and sometimes exhibited some displeasure while watching their video.
The instrument used to measure insight (SAI) was unable to measure qualitative influences in the improvement of insight. After watching the videos some participants felt they had a mental illness but did not understand the nature, chronicity or aetiology of their illness. One person said, ‘I was not well but that was because I was locked up in hospital.’ Another said, ‘I had a breakdown due to family stress and cannabis.’ Some acknowledged that they needed medication but attributed this to the need for sleep or to help lift their mood.
Future research
The study seemed to suggest that younger patients, at an early stage of their illness, responded more positively to the video self-observation, but a larger study is required to investigate this further. The study also indicated that the improvement in insight was sustained for up to 6 months, which also needs further investigation. A larger study may help identify the patient population most likely to benefit from such an intervention. Furthermore, there appears to be a need for a more detailed insight questionnaire able to measure the qualitative aspects of insight. Qualitative research on individual reactions to viewing videos of themselves while ill is likely to be of benefit in both better understanding the effects of video self-observation and also, perhaps, in the development of more sensitive measures of insight. Future research in video self-observation could use the Drug Attitude Inventory (Reference Hogan, Awad and EastwoodHogan et al, 1983) to assess the effect of the procedure on adherence to medication.
Clinical implications
The results of this pilot study are promising and warrant a larger randomised controlled trial to confirm the results. Few interventions are currently available to improve insight in psychosis and video self-observation is a relatively simple, inexpensive intervention that can be of benefit here. It could be targeted at the patient group most likely to respond - this may be individuals recently diagnosed with psychosis, as part of an early intervention programme.
Limitations
Individuals experiencing psychotic symptoms may become anxious at the prospect of being videotaped. They may find video self-observation distressing at a time when they are mentally stable, which may negate any potential benefits of the intervention.
Since the selection of participants for this study was based on written consent being given, those without capacity were excluded and their response to such an intervention is unknown.
Declaration of interest
None.
Open access
eLetters
No eLetters have been published for this article.