Should hospitals provide all patients with single rooms?

Infection prevention and control

Opponents to the universal provision of single rooms in hospital have claimed that there is no evidence that they reduce hospital acquired infections. Such evidence is hard to get because it cannot be established by experiment and because confounding effects make controlled trials difficult to do and interpret.2 But the evidence that physical barriers are good at preventing the spread of microbes is strong. A recent study in Canada on the effect of changing intensive care unit arrangements from multibed to single rooms showed falls in rates of acquisition of Clostridium difficile of 43% (a rate ratio of 0.57) and meticillin resistant Staphylococcus aureus of 47% (rate ratio 0.53) after the change—which was also followed by a 10% lower adjusted rate ratio of length of stay.3

Norovirus is by far the commonest infection imported into hospitals. It is also the most transmissible. It produces its most severe clinical effects in elderly people with underlying health conditions—the sick get sicker.4 The prevention of aerosol spread associated with projectile vomiting is a key control measure, and so is environmental decontamination.5 The universal provision of single rooms will considerably reduce the risk from the first and greatly facilitate the second.

The transmission of tuberculosis in hospitals re-emerged in the early 1990s.6 Single rooms are an absolutely basic requirement for control of this infection.7 A patient with undiagnosed disease in a single room is less likely to infect others. The same applies to the prevention of nosocomial influenza.8

But the provision of single rooms in new hospitals and rebuilds must also anticipate the future. It is certain that new pathogens will appear, probable that this will happen without warning, and quite possible that it will take some time to identify the novel cause and work out its transmission potential. Such events must be prepared for. During the outbreak of severe acute respiratory syndrome in 2002-3, a patient in hospital in Toronto infected the patient in the next bed, 1.5 m away, and another three beds away, 5 m distant.9

Patient choice

A Scottish government Social Research Public Attitude Survey in November 2008 sought the views on single rooms in hospitals of 990 representative adults across Scotland10; 41% preferred them, 22% would rather be in small (not more than six beds) multibed units, 3% in larger multibed units, and 27% didn’t mind. Those who favoured single rooms gave privacy as the main reason (93%) followed by less noise (42%), and 78% of those favouring multibed unit preferred them because of more companionship and less isolation. The popularity of single room was greater and enthusiasm for shared accommodation much less than in a survey done by Ipsos MORI in March 2007 for the Department of Health in England,11 which found 35% preferred single rooms, 40% small multibed units, and 9% larger units, with 14% having no preference. Although it is not possible to interpret these results as showing a trend in public expectations towards a wish for more privacy, they are compatible with it. This shift, which started in hospitals with the abolition of Nightingale wards, is still in progress, and should be accommodated.

Arguments about “companionship” run that “many patients in hospital are comforted by ‘looking out’ for one another, knowing that someone will buzz for help if anything goes wrong and do their best to make sure that needs for water, good care, and toileting are met.”12 The evidence from inquiries into the high mortality rates at the Mid-Staffordshire NHS Trust shows that this “comfort” is misplaced.13 Most patients there were in multibed units; but for many their needs for water, care, and toileting went unmet.

Patient safety is paramount. Privacy is desired by many. Delivering these things needs single rooms.

Where is the evidence?

Jason Leitch, the Scottish government’s national clinical lead for patient safety, has said that patients in single rooms will be better monitored (Radio Scotland, 4 February 2013). I assume he means alarms so that a nurse at a console can tell if the patient has stopped moving or fallen out of bed rather than electrocardiography, blood pressure, and heart rate monitors. Continuous physiological monitoring is necessary for only a small proportion of admitted patients, most of whom will be in a single room in a high dependency ward. The majority will be waiting for the results of investigations, for physiotherapy or occupational therapy assessment, or for care packages. Some of these patients will be mobile and able to move to the ward’s socialisation space; others will not. No one should underestimate the time it can take to get an older person out of bed and into a chair.

Alex Neil, the Scottish health secretary, recently stated that 26 of the country’s 218 hospitals already had 100% single rooms, saying there was “strong evidence” of their effectiveness.15 I obtained a copy of the names of these hospitals and only the Golden Jubilee in Glasgow was a hospital in the sense that most of us recognise. Two of the 26 hospitals are in Dumfries and Galloway. Between them they have 10 single rooms for teenagers with learning difficulties and challenging behaviour. I struggle to understand the relevance this might have for shaping policy when building a new 300 bed acute hospital.

The health secretary has further said, “There is evidence of the effectiveness of single rooms in minimising risks of healthcare associated infections.”15 Health Protection Scotland drew my attention to a survey from Canada that reported an 11% increase in risk of Clostridium difficile infection with each exposure to a new hospital room-mate.16 This has been seized upon by the “frankly, it’s common sense” brigade as justification for having all single rooms,17 though a more detailed analysis of the results suggests that this enthusiasm may be misplaced. In this study, the risk of C difficile infection among 35 697 patients admitted over five years was 0.5%. Single room enthusiasts should note that 11% of nothing is nothing, and 11% of 0.5% is not much more. They are, of course, assuming that the act of isolating a patient in a single room is what makes the difference and not some other infection control measure such as hand washing.

One room does not fit all

I do not believe that this debate should be about all single rooms or all shared accommodation, and I strongly support the need for more single rooms when the new Dumfries Infirmary is built. We have only 21% single rooms at present and need many more to isolate sick patients and patients with hospital acquired infection while meeting the needs of those who express a preference to be on their own. Different specialties within a hospital are also likely to have different preferences. Patients admitted for elective surgery, who are in and out of hospital within 2-3 days might well prefer the privacy of a single room. Those admitted as medical emergencies, who are usually older, have multiple comorbidities, often live alone, and need to stay in hospital for a week or more for assessment or arrangement of care packages, could well prefer a bit of company. We could probably achieve this with an 80:20 or 70:30 single to shared split.

This brings me to what I hope might be an acceptable compromise. Instead of insisting on 100% single rooms for new hospitals, the Scottish government should allow health boards to decide what proportion best meets the needs and preferences of the populations they serve, provided that this is at least 50% and, importantly, that all four bedded rooms have separate showers and toilets for each patient. This would surely satisfy all parties. Dignity would be improved and all the advantages of having company would be preserved if that is what a patient prefers. England want at least 50% single rooms,18 and Wales encourages single rooms but does not specify a proportion.19 France’s Ministry of Health favours single rooms but doesn’t specify limits, its spokesman Jean-Philippe Roux told me. What I don’t understand, given the lack of strong evidence in favour of 100% single rooms, is why the preferences of so many Scots are being ignored.