E. Andrea Nelson, BSc (Hons), RN, PhD.
Senior Research Fellow, Centre for Evidence Based Nursing, Department of Health Sciences, University of YorkPressure ulcers, also known as pressure sores or decubitus ulcers, are caused by unrelieved pressure, shear or friction acting on the skin (Dealey, 1991). Studies in UK hospitals have found that 6-10% of inpatients have a pressure ulcer (O'Dea, 1993). They are most commonly found at bony prominences such as the sacrum, heels and hips.
Pressure ulcers, also known as pressure sores or decubitus ulcers, are caused by unrelieved pressure, shear or friction acting on the skin (Dealey, 1991). Studies in UK hospitals have found that 6-10% of inpatients have a pressure ulcer (O'Dea, 1993). They are most commonly found at bony prominences such as the sacrum, heels and hips.
There is no reliable body of evidence from high-quality randomised controlled trials (RCTs) concerning the best strategies for preventing pressure ulcers and most of the important questions in pressure ulcer prevention and treatment remain unanswered (Cullum, 2001). Support surfaces such as special beds, mattresses and overlays are commonly used in pressure ulcer prevention and treatment. These surfaces can be classified in a number of different ways: mattress replacements, whole bed replacements, overlays that go on top of a standard mattress, or cushions. Within these categories of surface there are further distinctions based on mode of operation.
Specialist support surfaces for pressure ulcer prevention act by either moulding around the patient to distribute the patient's weight over a larger area (constant-low-pressure devices, also described as pressure-reducing devices), or by mechanically varying the pressure beneath the patient, so reducing the duration of the applied pressure (alternating-pressure devices, also described as pressure-redistributing devices).
A systematic review of the evidence for support surfaces in pressure ulcer prevention and treatment was commissioned by the NHS Health Technology Assessment (HTA) Programme and is available on the HTA website (www.hta.nhsweb. nhs.uk)(Cullum et al, 2001).
The systematic review indicates that using various high-specification foam alternatives to the standard foam hospital mattress can result in a relative reduction in the incidence of ulcers of 73% (95% confidence interval from 62 to 81%). Subsequent comparisons between different foam alternatives have been inconclusive; therefore, while foam alternatives to the standard hospital foam mattress have been associated with significantly fewer pressure ulcers in at-risk groups, there is no obvious 'best alternative'. The systematic review found eight small and poor-quality trials of other low-technology constant-low-pressure devices (for example water, air or gel-filled surfaces) and was therefore unable to draw conclusions about their effects.
Alternating-pressure surfaces are in widespread use. It is important to establish whether they offer any advantages over constant-low-pressure surfaces (for example those filled with foam, water and air) and whether there is a difference in effectiveness between the more expensive mattress replacements (typical purchase price £3300) and mattress overlays (typical purchase price £800). There is conflicting evidence as to the relative benefits of alternating and constant-low-pressure devices. The relative merits of the different alternating-pressure devices are unclear. We do not know, therefore, whether there are important differences in terms of clinical and cost effectiveness between foam surfaces and alternating-pressure surfaces in the prevention of pressure ulcers in medium-risk to high-risk patients. Furthermore, we do not know whether there are important differences in terms of clinical and cost effectiveness between alternating pressure overlays and alternating pressure mattresses in the prevention of pressure ulcers in medium-risk to high-risk patients.
In 1997, the NHS HTA Programme called for research to help answer some of this uncertainty. Bids were invited from researchers and clinicians across the UK. After extensive peer refereeing and short listing, a number of teams were asked to submit full research proposals in December 1998. A team of researchers from the University of York (Centre for Evidence Based Nursing (CEBN) in the Department of Health Sciences) and the University of Leeds (Northern and Yorkshire Clinical Trials and Research Unit - NYCTRU), as well as clinicians from hospitals across the north of England decided to collaborate on a trial of alternating-pressure mattress replacements versus alternating-pressure overlays in people at moderate to high risk of developing pressure ulcers (PRESSURE: Pressure RElieving SUrfaces, a Randomised Evaluation.
We were informed that our bid was successful in 1999 and began work on the trial in April 2000. Recruitment to the trial started in 2001 and will continue until late 2003. This paper describes the trial and challenges faced by the researchers.
Defining the research question
Our first challenge was to define a research question that had the potential to excite all the collaborators (so that enthusiasm could be maintained), was not already answered (as undertaking a trial to answer a question to which the answer was already known is unethical), and was feasible.
It was not possible to answer the question 'What is the benefit of alternating pressure devices compared with constant-low-pressure devices?' as clinicians reported that, in their centres, people at moderate to high risk of developing a pressure ulcer would be provided with an alternating system. This meant the clinicians had a clear preference for one treatment over another. One of the fundamental principles for designing ethical trials is that clinicians should have real doubt about whether one treatment is preferable to another, also called being in 'equipoise' (Lilford and Djulbegovic, 2001). The clinicians in some of the trial sites were not in equipoise: they had a clear preference for alternating-pressure support systems for patients at high risk of pressure ulceration, and therefore they would not be able to participate in a trial where patients might be allocated to a constant-low-pressure surface such as a high-specification foam mattress, even though the research evidence is equivocal.
The trial objectives are set out in Box 1. The study is a multi-centre, randomised controlled trial. People newly admitted to participating hospitals and identified as suitable for inclusion in the trial, by ward staff or research nurses, are approached by a research nurse and invited to participate. The research nurse must discuss the trial within 24 hours of hospital admission, so that any pressure ulcers can be ascribed to the trial surfaces, and not to the bed or mattress used at admission.
If the patient agrees to participate, then he or she is randomised to either an alternating-pressure overlay (Alpha Xcell or Debut) or an alternating-pressure mattress replacement (Nimbus II or Debut MR). The randomisation is done by telephoning an automated randomisation service at the NYCTRU.
Randomisation ensures that the only systematic difference between the two treatment groups is the pressure-relieving surface itself. The randomisation process should ensure that people allocated to the two support surfaces are evenly matched for factors, such as mobility, which influence pressure ulcer development. Without randomisation it might be tempting to place the more ill patients on the alternating mattress replacement, and the less ill patients on the overlay. If this happened then one could not make a valid comparison between the overlay and mattress group because people in the mattress group are less likely to do well because of their concurrent illness. This is called an allocation bias and minimising this bias is the key reason for undertaking a randomised trial (Smith, 1989; ter-Riet et al, 1998).
People in the trial
Alternating-pressure support surfaces are usually used for people at moderate to high risk of developing a pressure ulcer. We defined people at moderate to high risk of pressure ulcers in the following manner:
- Acute or elective patients aged over 55 years admitted to orthopaedic, elderly and vascular specialties with an anticipated length of stay of at least seven days or
- People who are chairfast or bedfast and are described as having very limited mobility or completely immobile on the Braden scale (Braden and Bergstrom, 1987). This equates to a score of 1 or 2 on the activity or mobility scale or
- Acute and elective patients aged over 55 admitted to orthopaedic, elderly and vascular specialties who have a grade 2 pressure ulcer on admission.
Calculating the sample size Before the trial began, data from the clinical trial centres indicated that the pressure ulcer incidence rate (of at least a break in the skin) in these moderate or high-risk patients was up to 10%. We decided that a clinically meaningful change in this incidence rate was to reduce it to 5% (halving the incidence rate). Any smaller changes in the rate of development of pressure ulcers, for example from 10% to 7%, were deemed to be too small a change to warrant changing from an alternating-pressure overlay to an alternating-pressure mattress. Owing to the small difference in pressure ulcer rates we planned to detect, 2100 patients must be recruited in order to meet the trial's primary objective.
Clinical effectiveness and cost effectiveness
We will report the number of people developing new ulcers, the grade and position of these ulcers, and the time they take to heal.
The costs of the overlays/mattresses, dressing materials used to treat pressure ulcers and the hospital stay of people in the trial will be recorded so that we can report the costs of preventing a pressure ulcer on both overlays and mattresses. This allows us to determine the cost-effectiveness of the two systems being evaluated.
Impact of pressure ulcers
There are currently no validated, published, health-related quality-of-life tools specific for people with pressure ulcers. In order to inform the trial about the non-monetary costs associated with the development of a pressure ulcer, the trial will be supplemented with a qualitative study involving a purposive sample of 20-30 patients who develop pressure ulcers, in order to assess the impact of the pressure ulcers on their well-being. Indepth, semi-structured, audio-taped interviews of patients with pressure ulcers both in hospital and after discharge at home will be used to identify the impact of pressure ulcers and their treatment on patients' quality of life.
As this trial includes 10 sites, the trial protocol had to be submitted to a multi-centre research ethics committee (MREC) for approval. After MREC approval was obtained, the trial was submitted to the local research ethics committees (LRECs) at each site.
In line with the Medical Research Council's good clinical practice guidelines (1998), an independent trial steering committee (TSC) was set up to oversee the conduct of the trial. The TSC has a chairperson, a pressure ulcer specialist, a person with experience of pressure ulceration, and a senior statistician. The TSC meets six monthly to provide overall supervision, ensuring that the trial is conducted to the highest standards. The TSC monitors recruitment, adherence to the protocol and any new relevant research findings. For example, if new evidence became available that showed that the trial was unnecessary, then it would be unethical to continue randomising patients.
The trial timetable
Although the need for a trial to answer these questions was identified as far back as 1995 (Cullum et al, 1995) there is a considerable lag between this and the trial being commissioned, undertaken and reported.
Recruitment to the trial started in 2001 and will continue until 2100 patients are recruited or the end of 2003. The results of the trial will be analysed, peer refereed and then published both in journals and on the NHS HTA website, from which the full report will be available (www.hta.nhsweb.nhs.uk).
The trial sites
Ten trial sites have been involved, including teaching hospitals and district general hospitals. In this way, the trial will recruit a representative sample of people at risk of pressure damage. The sites are listed in Box 2.
Clinical research nurses (CRNs) at these sites are responsible for identifying people suitable for inclusion in the trial, inviting them to participate in the trial, obtaining consent, randomisation, ensuring that the person receives the overlay/mattress to which they were allocated, and twice-weekly skin assessments to confirm the presence of pressure damage. A number of these nurses have been seconded from the clinical setting and plan to return to their clinical posts after the trial has finished. A senior trial CRN works between the sites supporting the research nurses and helping to identify and resolve problems that might limit recruitment, or affect the conduct of the study, such as delays in notifying the research nurses that patients have been admitted, or lack of trial overlays/mattresses.
The CRNs are constantly monitoring patients recruited to the trial as well as recruiting more patients and they have developed their time-management and problem-solving skills. Being a part of the trial has also immersed them in clinical trials and the importance of accurate data collection and the difficulties faced in day-to-day research management. A number of the CRNs have moved on to other jobs and been promoted after spending nine to 12 months as a CRN and they report that working in the trial has helped them develop professionally.
Challenges for the trial
Recruitment of 2100 patients over 18 months means that each site has a target of around 15 new patients a month. The CRNs work extremely hard to identify people potentially suitable for the trial within hours of their admission. Once people have been in hospital for 24 hours they cannot be recruited to the trial as it is impossible to determine whether the ulcer had started while they were on the non-trial support surface. Current recruitment is running at around 50% of target. This is due to various factors such as ward closures, reduction in elective surgery and bed-blocking.
As of April 2003, 1358 people have been recruited to the trial, making it, we believe, the largest pressure ulcer trial ever conducted.
Identifying a pressure ulcer
We have decided that only pressure ulcers of at least grade 2, that is a blister or break in the skin, should be classified as a pressure ulcer (Nixon et al, 1998). This is because the initial signs of skin damage such as erythema do not indicate irreversible skin damage (Nixon et al, 1998). There are great problems in inter-rater reliability for grading skin damage of grade 1a and 1b, blanching erythema and non-blanching erythema (Nixon et al, 1998). We are continually monitoring the accuracy of skin assessments by the CRNs to ensure that there is a high level of agreement between the sites over what constitutes a grade 2 or grade 3 pressure ulcer, for example. This involves the lead CRN double checking the CRNs' skin assessments on a sample of patients every few months to ensure that their assessments are all in agreement.
When a person at moderate to high risk of developing pressure damage is admitted to a ward he or she might be placed on an alternating-pressure mattress replacement, for example if information about the patient is available from the accident and emergency department, and a mattress is in the ward. If the patient is subsequently recruited to the trial then they have a 50% chance of being randomised to the overlay instead. Some ward nurses have found it difficult to replace the mattress with an overlay to comply with the randomisation as they see this as 'stepping down' the pressure relief. As we actually have no evidence that a replacement mattress is better than an overlay (the reason behind the trial) then there is no reason to describe this as 'stepping down'. The CRNs liaise with the ward staff, reiterating the lack of any evidence to demonstrate that a mattress replacement is better than an overlay, to prevent this occurring.
Initially the trial protocol stated that to enter the trial a patient had to give either written informed consent or, in the case of a patient who was unconscious or confused, then their spouse, son or daughter would be approached and asked to provide written relative assent. We decided to ask for this because people admitted to hospital who are unconscious or confused are likely to be at high risk of developing a pressure ulcer. Had we excluded these people from the trial then the results might not be applicable to this group of patients. Furthermore, the trial would take much longer to achieve the sample size of 2100 patients.
During the trial it became apparent that a small number of people were not being recruited into the trial because a spouse, son or daughter was not available within 24 hours of hospital admission, often because they were not living nearby.
MREC approval was sought to allow the CRNs to approach the 'named next of kin' for assent for the trial. This meant that more people were eligible for inclusion in the trial as people without a spouse or child could be recruited provided that assent was given by the person identified as 'next of kin' in the hospital notes.
Decisions about the choice of pressure-relieving surface for people in hospital at moderate to high risk of developing a pressure ulcer is currently made on poor-quality information about the relative effectiveness and cost effectiveness of different systems. The PRESSURE trial will provide clinicians with information on the relative effectiveness and cost-effectiveness of alternating- pressure surfaces.
We are demonstrating that questions of the effectiveness of pressure ulcer prevention can be answered in large randomised controlled trials measuring clinically relevant outcomes. Collaboration between clinical collaborators and experienced researchers ensures that high-quality and clinically relevant research is undertaken.
The PRESSURE trial is funded by the NHS R&D Programme. Views and opinions expressed do not necessarily reflect those of the NHS Executive.
Braden, B.J., Bergstrom, N. (1987)A conceptual schema for the study of the aetiology of pressure sores. Rehabilitation Nursing 12: 1, 8-16.
Cullum, N., Deeks, J.J., Fletcher, A.W., Sheldon, T.A. et al. (1995)Preventing and treating pressure sores. Quality in Health Care 4: 289-297.
Cullum, N., Nelson, E.A., Flemming, K., Sheldon, T. (2001)Systematic reviews of wound care management: (5) beds; (6) compression; (7) laser therapy, therapeutic ultrasound, electrotherapy and electromagnetic therapy. Health Technology Assessment 5: 9, 1-77.
Cullum, N. (2001)What is the evidence for beds and dressings in treating pressure ulcers? Journal of Tissue Viability 11: 3, 121-122.
Dealey, C. (1991)The size of the pressure-sore problem in a teaching hospital. Journal of Advanced Nursing 16: 663-670.
Lilford, R.J., Djulbegovic, B. (2001)Declaration of Helsinki should be strengthened. Equipoise is essential principle of human experimentation. British Medical Journal 322: 7281, 299-300.
Medical Research Council. (1998)MRC Guidelines for Good Clinical Practice in Clinical Trials. London: Medical Research Council. Available at www.mrc.ac.uk/pdf-ctg.pdf
Nixon, J., McElvenny, D., Mason, S. et al. (1998)A sequential randomised controlled trial comparing a dry visco-elastic polymer pad and standard operating table mattress in the prevention of postoperative pressure sores. International Journal of Nursing Studies 35: 4, 193-203.
O'Dea, K. (1993)Prevalence of pressure damage in hospital patients in the UK. Journal of Wound Care 2: 4, 221-225.
Smith, E.B. (1989)Effect of investigator bias on clinical trials. Archives of Dermatology 125: 216-218.
ter-Riet, G., Kessels, A.G, Knipschild, P. (1998)Problems in the conduct of a randomised clinical trial. Journal of Wound Care 7: 5, 259-262.