Ir Med J. 2007 Mar;100(3):391-3


Abstract
The environment of a hospital can have a significant impact on the experiences of patients. In March 2003 a new purpose built acute psychiatric admission unit opened on the site of Kilkenny General hospital, while the admission wards of the 2 local stand-alone psychiatric hospitals closed. We sought to compare admissions before and after the move, hypothesising that there would be lower levels of aggression, sedative prescribing and intoxicant abuse in the new unit. Details of 98 acute admissions that occurred during the first 3 months of 2002 were compared to 97 acute admissions that occurred during the first 3 months of 2004. Average daily diazepam and chlorpromazine equivalents were calculated for each patient. The Modified Overt Aggression Scale (MOAS) was used to compare levels of aggression. Compared to 2002 fewer patients left the hospital against medical advice in 2004 (OR 0.35, p=0.027). Overall levels of aggression fell significantly (p=0.001). Levels of benzodiazapine prescribing also fell (Mean diazepam daily dose 5.75mg in 2002 versus 4.14mg in 2004; p=0.003). There were trends towards reductions in involuntary admissions, admissions of intoxicated people, patients abusing intoxicants in hospital and in antipsychotic prescribing. It is likely that the more pleasant, better designed and less stigmatising environment of the new unit together with the renewed energy and optimism of clinical staff contributed to the changes observed.



Introduction
In the past 30 years mental health policy in most of Europe has dictated that acute inpatient psychiatric services be increasingly provided from units located on the sites of general hospitals.¹ In Ireland the 1984 mental health policy document, “The Psychiatric Services - Planning for the Future”, set out a vision of a community orientated mental health service with in-patient beds provided in general hospitals.² Progress towards this goal has been slow with the proportion of psychiatric beds in general hospital units in Ireland growing from 0.9% to 14.2% between 1971 and 2001.³ However, by 2004 the proportion of new public psychiatric admissions that were to general hospital psychiatric units had grown to 47%.⁴

Typically, older stand-alone psychiatric hospitals were large institutions that were poorly designed for patient comfort and privacy.³ It is well recognised that the environment of hospitals can have a negative impact on patient experiences and outcomes.^5,6 Violence and drug misuse are widespread in some acute psychiatric inpatient facilities and reasons for this have been attributed to ward environments.^7,8 User surveys have also been highly critical of the environments of older stand-alone psychiatric hospitals⁹ and such units are perceived as being more stigmatising.^10,11 The Royal College of Psychiatrists set down recommendations for the design of new psychiatric inpatient units in their 1998 report, “Not just bricks and mortar”.¹² According to these, accommodation standards should bear comparison with a comfortable modern hotel; the unit should be landscaped sensitively with no long corridors and natural daylight views, and should have individual bedrooms with ensuite facilities for all patients.

Patients treated in psychiatric admission units within general hospitals have been found to have shorter hospital admissions¹³ and to be less likely to require long term psychiatric care.¹⁴ However, there has been very little research demonstrating any objective evidence of a benefit to patients in terms of their actual inpatient experience, associated with a move to a new general hospital unit.

In March 2003 a new 44-bedded acute inpatient psychiatric unit opened on the grounds of Kilkenny General Hospital, while at the same time the 63 acute admission beds in two local psychiatric institutions closed. The new unit was designed according to best architectural practice and included a 4-bedded intensive care unit. The move provided an opportunity to review protocols in relation to admissions and the management of aggressive behaviour. Staffing personnel, roles or numbers did not change in the move. We set out to compare features of admissions before and after the move. Our hypothesis was that an improved patient experience in this new, bright, purpose designed unit on the site of a general hospital, would be reflected in a reduced frequency of patients leaving hospital against medical advice, reduced use of mental health legislation to hold patients in hospital, reduced use of seclusion, reduced intoxication on the wards, reduced prescribing of sedating medication and fewer aggressive incidents.

Method
We used the hospital’s computerised admissions information system to generate a list of all acute admissions to St. Canice’s Psychiatric Hospital, Kilkenny for the first 3 months of 2002 (excluding any patient not from the catchment area of the hospital), and a list of all patients from the St. Canice’s Hospital catchment area that had been admitted to the new unit in Kilkenny General Hospital during the first 3 months of 2004. We then sought the records for all of these admissions. Two researchers (LF & AK) extracted information from medical records, nursing records, seclusion records and incident reports according to a proforma. Each researcher analysed 50% of the records for each year and the authors discussed anything unclear. Demographic and diagnostic information was extracted from the records. Any incidence of the use of mental health legislation, leaving the ward against medical advice, use of seclusion, and intoxication on the ward was recorded. Prescription records were analysed and average daily diazepam and chlorpromazine equivalents were calculated based on the literature.^15,16 Hypnotic use was also recorded.

We used the Modified Overt Aggression Scale (MOAS) to measure aggression.¹⁷ The MOAS is a modified version of the Overt Aggression Scale (OAS) devised by Yudofsky et al.¹⁸ This scale has been validated for use in retrospective chart review. The scale comprises four subscales: verbal aggression, aggression against property, autoaggression and aggression towards others. Based on review of case notes, each individual was rated on a scale between zero (no aggression) and four (maximum score) on each subscale. The subscale scores were weighted as described by Kay et al¹⁷ in order to calculate the total MOAS score: the verbal aggression score was multiplied by 1; the aggression against property score was multiplied by 2; the autoaggression score was multiplied by 3; and the aggression towards others score was multiplied by 4. The sum of these weighted scores is the MOAS total score. A high level of inter-rater agreement was achieved for the MOAS scores (kappa=0.94). The results were analysed using the Statistical Programme for Social Sciences (SPSS Version 12).¹⁹ The data were not normally distributed, so continuous data were analysed using Mann-Whitney U tests, while dichotomous data were analysed using Chi-square tests and odds ratios were generated.

Results
There were 107 admissions to the acute admission wards of St. Canice’s Psychiatric Hospital in the first 3 months of 2002, while there were 99 admissions to the new admission unit based in Kilkenny General Hospital during the first 3 months of 2004. In 2002 51.4% of admissions were male, whereas in 2004 63.6% were male. The mean age at admission was 44.9 years (SD=16.7) in 2002 and 45.1 (SD=15.4) in 2004. The mean length of stay was 34.6 days (SD=60.4) in 2002 versus 27.9 days (SD=39.4) in 2004. None of these differences were statistically significant (p>0.05 in all cases; t-tests). The proportion of 1st time admissions was similar in each year. We found the admission records for 98 (91.6%) of the 2002 admissions and 97 (98.0%) of the 2004 admissions. There were no statistically significant differences between admission diagnoses for 2002 and those for 2004 (table 1) (p>0.05 in all cases; chi-squares).

In 2004 significantly fewer patients (n=8) left the hospital against medical advice than in 2002 (n= 20) (OR 0.35 (95% CI 0.15-0.84), p=0.027). In 2004, compared to 2002, there were reductions in involuntary admissions, conversions to involuntary status, the use of seclusion, abuse of intoxicants on the wards and in the proportion of patients who were intoxicated on admission, but none of these reductions were statistically significant (p>0.05 in all cases; chi-squares). There was a non-significant increase in the proportion of patients who had been seen in the emergency department of Kilkenny General Hospital immediately prior to admission (p=0.145; chi-square). These results are displayed in table 2.

The mean daily diazepam equivalent dose prescribed was 5.75mg in 2002. This was significantly higher than the mean dose of 4.14mg prescribed in 2004 (Z= -2.925; p=0.003). The mean daily chlorpromazine equivalent dose prescribed was 166.5mg in 2002 versus 141.3mg in 2004 (Z= -0.992; p=0.32). The mean percentage of nights on which hypnotics were administered was 67.6 in 2002 as opposed to 67.3 in 2004 (OR 1.014 (95% CI 0.84-1.22); p=0.92).

Overall aggression levels fell significantly between 2002 and 2004. The mean weighted MOAS score was 1.89 (median 0; range 0-19) in 2002 versus 0.84 (median 0; range 0-27) in 2004 (Z=-3.32; p<0.001). In 2002 42 (42.9%) patients scored 1 or more according to the MOAS scale, while in 2004 24 (24.7%) patients did (OR 0.44 (95% CI 0.24-0.81); p=0.011). A breakdown of the subtypes of aggression according to the MOAS scale is given in table 3

Discussion
According to our findings patients were significantly less likely to behave aggressively in the new acute psychiatric admission unit in Kilkenny General Hospital in 2004 than they had been in the admissions unit in St. Canice’s Hospital, Kilkenny during the equivalent period in 2002. Most of this change was accounted for by reductions in verbal aggression, physical aggression directed at property and self-directed aggression. Physical aggression towards others was lower in this study at baseline than might have been expected from other published studies.²⁰ Patients were also significantly less likely to discharge themselves against medical advice and were prescribed significantly lower quantities of benzodiazapine medications in 2004 than 2002. These figures support our hypothesis that the new unit would provide an overall improved patient experience.

Our study has a number of shortcomings. It was retrospective. Many aggressive incidents may have gone unrecorded. The researchers were not blinded to the year of admission due to the difficulty of concealing all references to date in patient records, nor were they blinded to the study’s hypothesis. This is a potential source of bias. However, most of the outcome measures used were objective and thus less subject to bias. We were unable to find 8.4% of the files for 2002. There is nothing to suggest that these patients would have differed from those whose files we did find. There was a higher proportion of male patients admitted in 2004. If anything this should have caused aggression levels to be higher in that year. Our study was adequately powered to demonstrate differences in aggression and prescribing levels between 2002 and 2004, but was underpowered for most of the other outcome measures employed due to their infrequent occurrence. A power calculation should have been carried out in advance. We could not control for the many different factors that may have contributed to our findings. The opening of the new unit involved the closure of the admission wards of 2 local psychiatric hospitals. We were only able to compare admissions from one of these hospitals. Staff from both of these old psychiatric hospitals now work together in the new unit. However, besides this change, there were no real changes in staffing levels, roles or numbers between the 2 periods. Although there were more acute beds in the stand-alone hospitals, admission numbers were similar between the 2 periods, meaning that in 2002 there were more long stay patients on the ward.

Despite these methodological weaknesses it seems likely that there were real differences between 2002 and 2004. The finding that there was a higher proportion of patients who were reviewed in the emergency department of the general hospital before psychiatric admission may indicate that severely intoxicated people were less likely to come directly to the psychiatric unit, resulting in fewer incidences of aggression. However there were no differences observed between the numbers of intoxicated persons admitted in 2002 and 2004. The differences observed could also represent a general trend towards reduced aggression levels on psychiatric inpatient facilities, unrelated to the move to the new unit. It would be necessary to examine samples from different time points to explore this possibility. However, data published based on other psychiatric inpatient units suggest a trend towards increased levels of aggression and substance misuse.^{21, 22} The most likely explanation for the welcome finding that there was reduced levels of aggression, sedative prescribing and discharges against medical advice associated with the move to the new inpatient unit, is the more pleasant, better designed and less stigmatising environment of the new unit, the fresh approach to practice afforded by the move, and the renewed energy and optimism of clinical staff. It is impossible to measure the specific contributions of these diverse factors and it will be important to examine to see if these positive changes are sustained. This study indicates that properly designed inpatient psychiatric facilities can contribute to a therapeutic environment, thus improving experiences for patients and staff alike. All patients who require psychiatric admission should be admitted to a well designed, well maintained, well staffed facility, with plenty of space for therapy, recreation and other activities.

References

1. Hall P. & Brockington IF (ed.s). The Closure of Mental Hospitals. London: The Royal College of Psychiatrists, 1991.2 Department of Health. The Psychiatric Services - Planning For The Future. The Stationery Office, Dublin, 1984.

3. Walsh D, Daly A. Mental illness in Ireland 1750-2002: Reflections on the Rise and Fall of Institutional Care. Dublin: Health Research Board, 2004.
4. Daly A, Walsh D, Comish J et al. Activities of Irish Psychiatric Units and Hospitals 2004. Dublin: Health Research Board, 2005.
5. Goffman E. Asylums. Essays on the Social Situation of Mental Patients and Other Inmates. Harmondsworth: Penguin Books, 1968.
6. Barton R. Institutional Neurosis. Bristol: J. Wright, 1976.
7. Kho K, Sensky T, Mortimer A, Corcos C. Prospective study into factors associated with aggressive incidents in psychiatric acute admission wards. British Journal of Psychiatry 1998, 172, 38-43.
8. Phillips P, Johnson S. Drug and alcohol misuse among in-patients with psychotic illnesses in three inner-London psychiatric units. Psychiatric Bulletin 2003, 27, 217-220.
9. Sainsbury Centre for Mental Health. Acute Problems: A Survey of the Quality of Care in Acute Psychiatric Wards. London: Sainsbury Centre for Mental Health, 1998.
10. Furlong RC. Asylum in the Community. London: Tomlinson & Carrier, 1996.
11. Lyons D, El Sayed OA, Matthew VM. New unit on old ground or general hospital: where do patients want inpatient treatment? Irish Journal of Psychological Medicine 2001, 18, 129-131.
12. Royal College of Psychiatrists. Not just bricks and mortar: Report of the Royal College of Psychiatrists Working Party on the size, staffing, structure, siting, and security of new acute adult psychiatric in-patient units (Council Report CR62). London: Royal College of Psychiatrists, 1998.
13. Mahadevan S, Forster DP. Psychiatric units in District General Hospitals and traditional mental hospitals: some recent evidence. British Journal of Psychiatry 1982, 140, 160-165.
14. Baker AA. Psychiatric unit in a district general hospital. Lancet 1969, i, 1090- 1092.
15. Valenstein M, Taylor KK, Austin K, Kales HC, McCarthy JF, Blow FC. Benzodiazapine use among depressed patients treated in mental health settings. American Journal of Psychiatry 2004, 161, 654-661.
16. Woods SW. Chlorpromazine equivalent doses for the newer atypical antipsychotics. Journal of Clinical Psychiatry 2003, 64, 663-667.
17. Kay SR, Wolkenfeld F, Murrill LM. Profiles of aggression among psychiatric patients: I. Nature and prevalence. Journal of Nervous and Mental Disorders 1988, 176, 539-546.
18. Yudofsky SC, Silver JM, Jackson W et al. The Overt Aggression Scale for the objective rating of verbal and physical aggression. American Journal of Psychiatry 1986, 143, 35-39.
19. Statistical Programme for Social Sciences (SPSS). SPSS for Windows, Release 12. 2003, Chicago, Il: SPSS.
20. El-Badri SM, Mellsop G. Aggressive behaviour in an acute general adult psychiatric unit. Psychiatric Bulletin 2006, 30, 166-16.
21. Lelliott P. Acute inpatient psychiatry in England: an old problem and a new priority. Epidemiologia e Psichiatria Sociale 2006, 15, 91-94.
22. McMahon C, Butwell M, Taylor PJ. Changes in patterns of excessive alcohol consumption in 25 years of high security hospital admissions from England and Wales. Criminal Behaviour and Mental Health 2003, 13, 17-30.