A Comparison of Private and Public Psychiatry in Three Hospitals

METHOD

A retrospective casenote review was performed on one hundred consecutive patient admissions to each site. Where patients had been admitted more than once within the same group of one hundred patients, only the first admission was counted. This was to allow for equivalent sample sizes, and it was appreciated that the data may thus be skewed.

The survey was conducted in three separate psychiatric facilities and permission to proceed with the survey was granted by the Ethics Committees of each of those three hospitals. The units were chosen to give a representative picture of the different types of psychiatric facilities available in South Australia, and are probably typical of facilities in Australia. All of the units are within a four kilometre radius of the central business district of Adelaide.

The private hospital was The Adelaide Clinic (TAC), a private psychiatric unit. At the time of the casenote review, TAC comprised 91 beds and had specialised services such as a 22-bed elderly assessment unit, but the bulk of its work was related to general adult psychiatry. TAC was not gazetted to accept patients detained under the Mental Health Act during the survey period, although that has since occurred.

The public hospitals were Glenside Hospital (GH), which at the time of the study was a 335-bed state institution which incorporated acute admission wards for general psychiatric patients, as well as rehabilitation and closed intensive care beds, and The Royal Adelaide Hospital (RAH), which is a 740-bed major general teaching hospital with a 24-bed psychiatric ward.

Demographic and clinical data, including the reasons for admission as far as these could be determined from the casenotes, the number of admissions in the past year, the total number of previous admissions, the socioeconomic status (estimated from the patient's postcode) and the patient's likely discharge placement were collated. In addition, the patient's length of stay and the treatment modalities, including psychotropic medications and electroconvulsive therapy were recorded. The medical and psychiatric diagnoses of each patient were taken from the notes. It was necessary to combine some of the codes for psychotropic medications and diagnoses for statistical purposes.

Data were analysed using Chi-Squared and analysis of variance statistics.

RESULTS

Significant difference between the hospitals emerged with respect to age, gender, employment, marital and socioeconomic status. TAC had an older population who were more likely to be employed. GH had a higher population of males who were less likely to be married. RAH had the greatest number of patients in the lowest socioeconomic status group (see Table 1).

No significant difference was found between the hospitals for the mean length of stay, although the median length of stay at GH was appreciably shorter (6 days) compared to that of TAC (15 days) and RAH (17 days). There were no differences between the three hospitals in terms of the number of admissions for each patient or the time since their previous admission (see Table 2).

Significantly fewer patients from GH were discharged to their own home, with them being more likely to stay with their parents, compared to patients from TAC and RAH (χ²=20.3, df=4, p ≥ 0.01).

Table 3 demonstrates the use of psychotropic medications on discharge. The majority of patients from the three hospitals were prescribed psychotropic medication at discharge (89%, 81% and 85% of patients from TAC, GH and RAH respectively), and there was no significant difference between the hospitals (χ=2.4, df=2, ns).

As can be seen from Table 4, TAC had more cases of depressive disorder than did the other two hospitals. In contrast, GH had considerably more cases of schizophrenia and schizoaffective disorder.

DISCUSSION

There are a number of methodological shortcomings in this study. Because it was a casenote review, it was reliant upon the observations recorded by other people. The nature of the information collected by the three hospitals varied considerably, from sparse (TAC) to extensive (RAH). In addition, there was a great deal of individual variation, with some staff members recording very little useful information, and others supplying highly detailed data. This variation was especially prevalent in the socioeconomic information recorded, and made it difficult to compare the three hospitals. Therefore the patients' postcode was used as a guide to their socioeconomic status (SES), using statistics compiled by the SA Health Commission.

It is also acknowledged that the small sample size diminished this survey's ability to detect significant differences between the three hospitals, and thereby increased the risk of Type 1 errors. Furthermore, the differing size of the hospitals could have resulted in a bias related to seasonality of illnesses and admissions, as at GH it took only one month to collect data on one hundred patients, compared to 3 and 8 months at TAC and RAH respectively.

Notwithstanding these limitations, we have no reason to doubt the validity of the results and a number of differences did emerge.

More females were admitted to TAC and to the RAH (65% and 69% respectively) than to GH, the state psychiatric hospital (43%). This finding parallels that of other Australian studies [2,4] and probably reflects the higher incidence of affective and neurotic disorders, which are more commonly found in females, at the private and general hospitals. Similarly, the higher incidence of males admitted to GH was probably related to the increased incidence of severe psychotic disorders in young males.

Not unexpectedly, the private hospital, TAC, had the highest proportion of patients who were employed, although that was not reflected in the assessment of socio-economic states (SES), where TAC had 12 patients at the lowest SES compared to the RAH which had 19, and GH which had only 9. Fewer patients at the state psychiatric hospital were married, which was consistent with the higher incidence of psychotic illness which tends to affect people in their late teens and early twenties, when long-term relationships are most likely to develop.

There was a striking similarity between the 3 hospitals for the mean length of stay (LOS), the number of admissions both lifetime and within the previous 12 months, and the mean number of days between admissions. This is perhaps surprising given the diversity of demographic and diagnostic groups catered for by the 3 hospitals. It has been argued [5] that the median LOS is a more useful measure because it is unlikely to be skewed by outlying data points, as a mean score can be. When the data were examined in this manner there was a reduced median LOS at GH, with only 6 days compared to 15 days at TAC and 17 days at RAH, a finding which was more in keeping with clinical intuition about the respective hospitals.

The majority of patients from each hospital were prescribed psychotropic medication at discharge (89% at TAC, 82% at GH and 85% at RAH). Antipsychotic agents were the most frequently prescribed drugs in all settings, with more being used at the state psychiatric hospital. It is of interest that the actual diagnosis of psychosis was made in a smaller number of patients than were prescribed for, and this presumably reflects the diverse range of uses of antipsychotic medication as antipsychotic, anxiolytic, antiaggression, sedative and hypnotic agents.

Antidepressants were used significantly less frequently at GH, reflecting the lower diagnosis of depression at that hospital. Of note was the fact that all hospitals were prescribing the newer selective serotonin reuptake inhibiting agents more frequently than tricyclic antidepressants.

A considerable proportion of patients was prescribed hypnotic agents at discharge. TAC was over-represented, with 53% of patients being prescribed a hypnotic at discharge. In addition, a significantly greater number of patients at TAC were prescribed a minor tranquilliser (also generally a benzodiazepine) at discharge. These findings are potentially of concern because of the risk of dependence with benzodiazepines, although such concern has been tempered in a recent commentary by Williams and McBride [6].

With regard to diagnosis, depression was more likely to be the discharge diagnosis at TAC compared to the other 2 hospitals. In contrast, more patients at GH were diagnosed as suffering from a psychotic illness. When the diagnoses of schizophrenia, schizoaffective disorder and bipolar affective disorder were combined, there were 49 cases at GH, 16 at TAC and 27 at RAH. However, when the numbers of patients suffering from depressive disorders were added, the numbers were surprisingly similar, with 67 patients at TAC, 62 at GH and 65 at RAH. Although this result cannot be considered a reliable index of the clinical acuity of illness, it does indicate a similarity between hospitals in terms of those diagnoses generally considered to comprise serious mental illness.

CONCLUSION

There were a number of demographic and clinical differences between patients in the three hospital settings, the majority of which were readily understandable in terms of issues associated with illness and economic factors. However, it was unexpected that the overall diagnoses of those patients generally considered to have serious mental illness was similar between the three hospitals, and it cannot be assumed that patients in one sector are necessarily more or less severely ill than those in another sector. Further comparative studies are indicated, particularly those in which measures of severity of illness are utilised.