The physical environment of the psychiatric ward differs from that of the general hospital ward in a number of ways, but perhaps one of the more subtle is its relative lack of greetings cards (e.g. ‘get well soon’ cards). This observation has recently received empirical support from two, small case-control studies comparing psychiatric with medical in-patients (Reference Wiener, Wessely and LewisWiener et al, 1999; Reference Bromley and CunninghamBromley & Cunningham, 2004). In both studies, the former received fewer cards than the latter, findings that were interpreted in terms of stigma. Given the likely impact of stigma upon the patient’s well-being, we wished to test the central hypothesis that psychiatric in-patients receive fewer cards than others, through assaying larger samples, in cross-sectional surveys, repeated over time.
Method
On the weekends of 20-21 March, 1-2 May and 5-6 June 2004, we counted cards on the wards of the Northern General and Royal Hallamshire Hospitals. Both these Sheffield hospitals comprise many medical and surgical wards of different (‘adult’) specialties (the paediatric wards are elsewhere). We also counted cards on the two in-patient sites of the local psychiatric service (Sheffield Care Trust; SCT), where wards accommodate both working age and older adult in-patients. As there are far fewer psychiatric beds than general hospital beds in Sheffield we combined the figures for the SCT sites. All the units are teaching hospitals.
On each ward, we approached nursing staff to obtain their permission to count greetings cards. No patients were directly interviewed. Staff on most wards agreed to the count and the local research ethics committee approved the study under the category of ‘audit’ (Dr S. Brennan, personal communication).
The same protocol was followed on each of the three weekends and the following data acquired: the number of beds on each ward, the number of beds at which there were cards and the total number of cards (at each bed).
Results
On the days of the study, the precise numbers of beds to which we had access across wards varied owing to clinical activity and bed closures, but they were of similar magnitude on each occasion (Table 1). Relative to the numbers of beds on each unit, the mean numbers of cards recorded were lowest for the psychiatric wards at each time point.
Table 1. Bed, card and recipient numbers on wards in general and psychiatric hospital settings
| Site | Beds | Cards | Card/bed (mean) | Recipients1 | Recipients/bed (mean) | Cards/recipient (mean)2 |
|---|---|---|---|---|---|---|
| Time 1 | ||||||
| Northern General Hospital | 825 | 1730 | 2.09 | 333 | 0.40 | 5.19 |
| Royal Hallamshire Hospital | 570 | 1215 | 2.13 | 229 | 0.40 | 5.30 |
| Sheffield Care Trust | 183 | 61 | 0.33 | 28a | 0.15 | 2.17d |
| Time 2 | ||||||
| Northern General Hospital | 813 | 1399 | 1.72 | 280 | 0.34 | 4.99 |
| Royal Hallamshire Hospital | 493 | 1090 | 2.21 | 180 | 0.36 | 6.05 |
| Sheffield Care Trust | 179 | 75 | 0.41 | 26b | 0.14 | 2.88e |
| Time 3 | ||||||
| Northern General Hospital | 795 | 1103 | 1.38 | 237 | 0.29 | 4.65 |
| Royal Hallamshire Hospital | 510 | 853 | 1.67 | 163 | 0.31 | 5.23 |
| Sheffield Care Trust | 191 | 78 | 0.40 | 29c | 0.15 | 2.68f |
When compared formally, there were significantly fewer card recipients on the psychiatric wards on each occasion (χ2 test, each comparison significant at P<0.001, Table 1). Similarly, individual recipients on psychiatric wards exhibited significantly fewer cards at each time point than those recipients on general wards (Kruskal-Wallis test, each comparison significant at P<0.05, Table 1).
Discussion
Our data support the subjective impression that psychiatric wards exhibit fewer greetings cards. On all the measures we obtained, the numbers were significantly lower in the psychiatric setting. Hence, as well as there being relatively fewer cards in evidence, there were significantly fewer recipients of cards, and those who received cards at all evidenced fewer than recipients in general hospitals.
Our findings appear reliable over time and benefit from internal replication. However, certain caveats remain. First, the data relate to Sheffield teaching hospitals and may not reflect conditions elsewhere (although they are consistent with the findings of Reference Wiener, Wessely and LewisWiener et al, 1999; Reference Bromley and CunninghamBromley & Cunningham, 2004). Second, we have selected only three time points within 1 year. It is possible that we have missed aspects of seasonal variation. Third, card counts extending over weekends, i.e. 2 days, may have incurred errors in so far as they were not obtained on one occasion; hence, some patients may have been discharged or admitted while the count was ongoing. However, we think the latter would have militated against our hypothesis rather than promoting a type 1 error. Assuming that the rate of bed turnover is higher in the general hospital setting (true for most of the wards we examined), we should expect more cards to be removed by those leaving hospital on the general sites, and more patients to be admitted (‘without’ cards) over the weekends. Such movements would serve to minimise the effect we describe.
A fourth potential caveat arises from our counting only those cards that were visible on the wards; we did not take any account of those that might have been concealed, discarded or destroyed. Although the latter possibility might be posited to have greater face validity in the psychiatric setting, the number of missing cards required to nullify our findings would be considerable (approximately 300 on each occasion). If this were the case, i.e. if in-patients were hiding/discarding this many cards, then we might expect it to become apparent. Nevertheless, we cannot exclude this possibility on the basis of our current data.
Notwithstanding the above, we think that our findings really do reflect a phenomenon that is readily apparent on psychiatric wards: a notable absence of cards wishing patients well. In the following we offer some, admittedly speculative, hypotheses to account for this finding.
Isolation of psychiatric patients
Many psychiatric disorders impact upon social function, rendering individuals withdrawn or isolated. Hence, it might be that they have fewer potential correspondents than other in-patients; indeed, re-admissions and prolonged admissions of psychiatric patients have been found to be associated with such an absence of friends (Reference De Falcode Falco, 1975; Reference Salokangas, Loikkanen and SantalaSalokangas et al, 1990). However, when case-control studies have controlled for size of social network they have still found that psychiatric in-patients receive fewer cards and gifts (Reference Wiener, Wessely and LewisWiener et al,1999; Reference Bromley and CunninghamBromley & Cunningham, 2004), suggesting that the problem is not primarily one of network size.
‘Self-stigmatisation’
It is possible that psychiatric in-patients do not tell others that they are in hospital. Such ‘self-stigmatisation’ is reported by service users, who may be ashamed by their illness and may not wish to be seen when ill. They may also fear loss of employment: ‘Because people don’t understand, if they know that I’ve been off work because of mental ill health they may choose to use somebody else rather than me’ (a service user interviewed by Reference Dinos, Stevens and SerfatyDinos et al, 2004).
Therapeutic nihilism
It may be that where in-patients inform friends and family the latter regard it as unlikely that their friend or relative will recover. Hence, the absence of ‘get well’ messages may reflect a tacit therapeutic nihilism on the part of those known to patients.
Provocative message
In the case of psychiatric disorder, messages such as ‘get well soon’ or the like are perhaps more contentious than in most other medical settings. These words imply a recognition that the recipient is ‘ ill’, something which many patients may not accept (particularly those who are formally detained). It might be provocative to send a ‘get well soon’ message (though neither this, nor the therapeutic nihilism hypothesis, would preclude the sending of other, well-meaning messages).
There may be other possible explanations for the findings we describe. Nevertheless, the latter clearly imply that even such a simple social transaction as the giving and receiving of a greetings card exposes the discontinuity between the world of the general hospital patient and that of the psychiatric in-patient. Every card denotes a sequence of actions: card selection; the message written with a purpose (generally altruistic), directed towards the person who is unwell; the sending or delivering of that card. The notion that such actions occur less often within the social milieu of psychiatric in-patients might appear trivial but it provides an index of the way our patients are treated by others. It suggests that despite a contemporary emphasis upon ‘care in the community’, integration, and the reduction of stigma (Reference Crisp, Cowan and HartCrisp et al, 2004), following their admission to hospital, psychiatric in-patients are indeed treated differently to others; and this treatment implicates those who are ‘closest’ to them.
Acknowledgements
We thank the ward staff who allowed us to conduct these surveys and Dr V. Ganesan for his assistance in data preparation.
Open access
eLetters
No eLetters have been published for this article.