Prevalence of Pressure Ulcers in Canadian Healthcare Settings
Clinicians working in wound care appreciate how life for individuals with wounds is disrupted by care, cost issues, and the pain associated with treatment. Although managing pressure ulcers is often a passion for wound care specialists, the majority of the population is unaware of the challenges involved and many healthcare professionals place little emphasis on identifying and treating skin ulcers.
The number of individuals seeking wound care services continues to grow, which suggests that pressure ulcers are a relatively common healthcare concern and an escalating problem. Statistics are available regarding the number of pressure ulcers in the US1 as well as for other countries of the world.2 However, little information is available about the number of individuals in Canada who have pressure ulcers. National estimates for the number of pressure ulcers in various healthcare settings in regions across Canada are nonexistent. Without this information, estimating costs to the Canadian healthcare system associated with managing chronic wounds is not possible.
Pressure ulcers are not nationally recognized as an important healthcare problem in Canada. Currently, little national or provincial funding is available to provide coordinated healthcare delivery programs for their prevention and management or to promote the development of educational programs for healthcare professionals. Furthermore, nationally funded grants to support research programs for identifying the underlying cause of chronic wounds and establishing new interventions and innovative healthcare delivery models are rare. To focus national attention and resources on this serious and growing healthcare problem in Canada, national awareness about pressure ulcers must be raised. Gathering facts and statistical data that describe the extent of the problem in Canada is essential to the success of lobbying healthcare administrators and government officials and for informing the general Canadian population about the extent of the problem.
Recently, the National Pressure Ulcer Advisory Panel (NPUAP) completed a large study describing the prevalence of pressure ulcers in the US.3 In Canada, many national organizations that support other common disease conditions such as diabetes, cardiovascular disease, and cancer have invested significant human and financial resources to develop and maintain large national registries. Clearly, gathering this information is no small task. It takes years to organize, collect, and collate the data.
Given the enormity of the task, the authors believed an important first step was to systematically search and identify existing data on the prevalence of pressure ulcers in Canada. Specifically, the goal of this project was to determine, from current available information, the prevalence of pressure ulcers in different care settings in regions across Canada. Prevalence refers to the proportion of a group (patients not ulcers) that has a pressure ulcer at a given single point in time or time period during which the cases are counted. A cross-sectional study is the appropriate design for determining the number of patients with pressure ulcers from the number of patients assessed.
Data collection. Between January 2003 and November 2003, all available data from prevalence studies conducted between 1990 and 2003 were collected. Several sources for locating studies on the prevalence of pressure ulcers in all healthcare settings and in the general population were investigated, including peer-reviewed published reports, unpublished studies, and wound care company databases. For studies in which prevalence could be separated by facility or facility type, each facility or facility type was treated as a distinct study.
Systematic computer and manual searches of library databases PubMed (Medline®) and CINAHL® were conducted using the keywords ulcers, Canada, and prevalence to locate studies involving all healthcare settings and populations published in peer-reviewed journals. Few published articles describing the prevalence of pressure ulcers in Canada were found. After locating the articles, all references were reviewed and researchers in this field were contacted to locate peer-reviewed articles that might have been missed.
Many members of the Canadian Association for Wound Care (CAWC) responded to a general website request for information and provided unpublished Canadian study data. Approximately 50 people were contacted; often, the search for studies required contact with several people before the correct contact was made.
Several wound care companies have large databases of prevalence and/or incidence studies performed as a service for their customers. One company, Kinetic Concepts, Inc. (KCI Medical Canada, Inc., Mississauga, Ontario) expended great effort to contact individual consumers to address proprietary and confidentiality issues in order to share this valuable information. The results of individual studies conducted in Ontario and Quebec were made available. In addition, aggregate data (without facility names and without patient information) were provided by KCI and Hill-Rom Canada (Mississauga, Ontario).
Definitions and critical appraisal. Because of regional differences and recent changes in terminology, the term non-acute care has been used in this report to include the following types of care setting: subacute care, chronic care, complex continuing care, long-term care (LTC), and nursing home. The term mixed health settings refers to prevalence studies in settings that consist of a mixture of acute, non-acute and/or community care healthcare delivery models; the prevalence estimate is expressed overall rather than broken down by specific setting type.
The methodology used to collect prevalence information in all studies, published and unpublished, was critically appraised using a modified version of recommended criteria for evaluating prevalence studies.4 The original critical appraisal work by Loney et al4 relates to patients with dementia and consists of a series of questions that are more appropriate for health problems that can be evaluated using population surveys than for chronic wounds, which are generally evaluated in healthcare facility settings. Hence, it was necessary to modify the questions to reflect the most common study situations in which patients in healthcare facilities were evaluated, often by physical examination over a relatively short time. The critical appraisal questions used in the present report are listed in Table 1.
The authors independently determined a methodological score for each study by assigning each of the nine numbered questions a score of one (1) point if appropriate methods were used. The final methodological score for each study was obtained by consensus, with higher score studies representing more appropriate and rigorous research methods and less potential bias in the results. Therefore, prevalence estimates from studies with higher methodological scores can be accepted with more confidence.
Questions 2, 4, and 5 of the critical appraisal (see Table 1) were the most vital. Studies were considered to lack validity if: 1) responses indicated that skin ulcers were counted by methods other than direct physical skin assessment, 2) outcome measures used to identify pressure ulcers were not those accepted as the standard, and 3) the presence of ulcers was determined by healthcare professionals responsible for patient care rather than by unbiased assessors. Studies for which the responses to all three questions were negative and studies with scores <2 were excluded from the summary statistics generated for this report. The relationship between prevalence and methodological score was investigated to ascertain the extent to which control of bias might affect prevalence estimates.
Statistical analyses. The published and unpublished studies included were summarized within the relevant healthcare setting. The point estimate of prevalence is expressed as a percentage of the total population at risk. Construction of the 95% Confidence Interval (CI) around the estimate allows that the authors are 95% confident the true prevalence is between the confidence limits. The formula for the 95% CI has been published previously.5
Data received. The number of individual published and unpublished studies obtained for four healthcare settings (acute care, non-acute care, community care, and mixed healthcare) for the years 1990 to 2003 is shown in Table 2. As few as four studies for community care and as many as 23 for non-acute care were received. Across 65 healthcare facilities/institutions, 14,102 patients were evaluated in prevalence studies, demonstrating a wide range of sample sizes (between 29 and 2,384). No studies of pressure ulcers prevalence in the general population were found.
In addition to individual studies, aggregated information about numerous individual studies was received from two wound care companies. The estimates were based on aggregate data and have not been combined with individual studies because information about facility type, location, and patient population was insufficient. Nevertheless, because the methods used by each company are consistent within the company and are of high quality, the prevalence estimates are reported.
Published studies. Literature searches yielded seven prevalence studies conducted in Canadian healthcare institutions. An eighth article, published in 1994 but conducted before 1990, was excluded.6 Information about each of these studies, including the assigned methodological score, is shown in Table 3.
The earliest study of pressure ulcers in this time period, reported by Foster,7 was conducted in three tertiary care hospitals, one community hospital, two LTC facilities, and two community agencies in Ontario. Overall pressure ulcer prevalence for all health facilities was 25.7%7; separate prevalence estimates for acute care, extended care, and community agencies were gathered from a figure in the original article.
Two prevalence studies conducted in acute care facilities achieved high methodological scores using an appropriate cross-sectional study design and careful methods (eg, standard definitions of pressure ulcer presence and staging) to ensure that assessments were done reliably and without bias.8,9 These two studies produced prevalence estimates of 29.7% and 23.9%, respectively.
The only Canadian study estimating the prevalence of pressures ulcers in a chronic care facility in Ontario was conducted before 1995, with pre and post wound care protocol evaluations.10 The study’s methodological score of 5 was affected by the relatively small sample size and by the fact that patients with ulcers were identified based on patient survey responses obtained by the nurses before direct physical examination.
The study of pressure ulcers in LTC with the highest methodological score (7) produced high estimates of prevalence — 36.8% and 56.2%.11 A lower estimate was obtained in another LTC study in which ulcers were assessed using non-standard measures.12 A third study in LTC, which was not conducted primarily to determine prevalence, created an unfavorable situation. Specifically, the Quebec LTC database was analyzed retrospectively to determine the relationship between presence of pressure sores and nursing workload.13 The use of a retrospective analysis rather than a cross-sectional study design provided an imperfect situation for determining prevalence because specific data on the presence of pressure ulcers were not recorded. The presence of pressure ulcers was inferred from an evaluation of required nursing actions for the treatment of pressure ulcers. The retrospective study design and nonstandard measurement of outcome adversely affected the methodological score obtained in the critical appraisal. These two latter studies had poor methodology scores and were excluded from the summary data.
Unpublished studies. Thirty-three unpublished studies were received from 19 people. Of these, 30 were accepted for use in this report. Included with the unpublished studies were 11 individual studies provided by a company — three representing acute care and eight representing non-acute care. All unpublished studies were critically evaluated and the results combined with the published studies.
After excluding the studies with scores of 2 or less and those with negative answers to the three key methodological questions, the mean methodological scores for the published and unpublished studies across the four healthcare settings ranged from 5.6 to 6.6 (maximum = 9) (see Table 2).
Prevalence estimates. Estimates of prevalence from studies with poorer methodological scores tended to be lower than estimates obtained from studies in which bias was controlled (see Figure 1). Therefore, studies with scores less than or equal to 2 were excluded.
The prevalence estimates of the individual published, unpublished, and KCI studies were summarized for acute care, non-acute care, community, and mixed health settings (see Tables 4 to 7, respectively). The overall mean prevalence for the healthcare settings ranged from 15.1% (95% CI, 13.4% to 16.8%) in community care, to 29.9% (95% CI, 29.3% to 31.4%) in non-acute care, with mixed health setting at 22.1% (95% CI, 20.9% to 23.4%) and acute care at 25.1% (95% CI, 23.8% to 26.3%) (see Figure 2). The 95% confidence limits for each estimate were narrow (less than two percentage points from the estimate), reflecting the large sample sizes that resulted from the combination of studies. Conversely, a large range of values was noted between the minimum and maximum estimates reported for the pressure ulcer prevalence of each individual study. Because the confidence limits for different healthcare settings do not overlap, the estimates in these settings are significantly different. When all data are combined to report overall prevalence (regardless of setting and avoiding representing study data more than once), the mean prevalence is higher at 26% (95% CI, 25.2-26.8%), based on 10,911 subjects.
Aggregate data prevalence estimates. KCI provided yearly prevalence estimates for 61 acute care facilities from studies conducted from 1997 to 2003. Because similar research methods were used to collect these prevalence estimates, they can be compared over time (see Figure 3). Estimates of pressure ulcer prevalence were found to decrease gradually from 42% in 1997 to 29% in 2002.
Results of prevalence studies conducted by Hill-Rom in 35 acute and non-acute Canadian facilities — ie, mixed healthcare settings — from 2001 to 2003 produced a prevalence estimate of 14.9% based on 6,828 patients.
This project provided reliable estimates of pressure ulcer prevalence for four healthcare settings in Canada. The estimates, which have narrow confidence intervals, are based on individual studies that were critically appraised and found to be methodologically sound, yielding large combined samples from across the country for most healthcare facility types.
The prevalence estimates obtained and a desired narrow 95% CI width of 10% can be used to ascertain the sample size requirements for conducting a future prevalence study in each of the healthcare settings, using the formula suggested by Baumgarten.5 (The acceptance of a 95% CI wider than 10% results is a trade-off in precision for a smaller sample size requirement.) The required sample size estimates are: 289 acute care subjects, 322 non-acute care subjects, 197 community care subjects, and 265 mixed healthcare subjects. In other words, in settings in which the prevalence is anticipated to be above 25%, approximately 300 subjects are needed, while fewer subjects are needed if the prevalence is anticipated to be as low as 15%. These sample size calculations lend support to the methodological criterion suggested by Loney et al4 that appropriate sample size for prevalence studies is greater than 300 subjects to allow for the possibility that the prevalence is as high as 25%.
Comparing estimates with those from other countries suggests that the estimate for the number of patients with pressure ulcers in acute care (25.1%) is higher than two previously reported pressure ulcer prevalence estimates from across the US.1,3 The NPUAP pressure ulcer prevalence estimate ranged between 10% and 17.1%.3 Whittington et al1 reported a 15.1% prevalence of pressure ulcers from a series of studies conducted by KCI in acute care facilities across the US. The differences between Canadian and US estimates of prevalence of pressure ulcers in acute care facilities might be due to different methodologies employed and the time period over which the data were collected. However, US and Canadian estimates generated by KCI in 1999 using identical methods found a pressure ulcer prevalence estimate in Canada of 27% (see Figure 3) that was considerably higher than the 15.1% value reported in US.1 Therefore, these estimates may represent true differences between the Canadian and US healthcare systems. Other possible explanations include differences in the sample sizes and patient profiles. The US KCI estimate reported by Whittington was limited to adult patients in medical-surgical and intensive care units, while samples surveyed in Canadian healthcare institutions in this report included more units.
The study prevalence estimate for patients in non-acute care facilities in Canada (29%) is close to the upper value in the range reported by the NPUAP, 2.3% to 28%,3 and to another recently published prevalence estimate (28%).14 Horn et al14 used a retrospective cohort sample of 2,420 patients who were at risk of developing pressure ulcers as indicated by Braden scores of 17 or less. One would have expected the current estimates for pressure ulcer prevalence in all patients within Canadian non-acute care facilities to be much higher than estimates from the US that examined only patients who had already been identified as at risk of developing pressure ulcers. Differences between current estimates in Canada and US pressure ulcer prevalence estimates may be due to the use of different methodologies — the US study used retrospective analysis rather than direct skin examination from which the Canadian estimates in this report were derived. The authors found that studies that used retrospective data analysis received lower methodological scores and tended to produce lower prevalence estimates; therefore, these studies may be more likely to underestimate true prevalence of pressure ulcers.
Differences between Canadian and US estimates of pressure ulcer prevalence also may be explained by the fact that the term non-acute care encompasses a potentially diverse population. The descriptors for healthcare facilities with non-acute patients have changed over the past several years and are not uniform in different regions across Canada. They include LTC, nursing homes, complex continuing care, skilled nursing facilities, rehabilitation, and geriatrics. Because of the varying terminology, the results were combined into a generic category: non-acute care. It is possible that the true prevalence in any particular subgroup within this classification may be masked by this healthcare setting’s diversity.
The national estimate of pressure ulcer prevalence in Canada (26%) is slightly higher than the estimate reported from a national study done in the Netherlands and considerably higher than the international aggregate estimate for 2003 provided by Hill-Rom. The epidemiological study performed in the Netherlands examined 16,344 patients and produced an overall estimate of 23.1% for all the health settings.2 The international estimate from Hill-Rom was 15.5%, based on 61,427 surveyed patients in 461 facilities of all types. The majority of these Hill-Rom international studies were conducted in the US.
The diversity of the non-acute care population that was combined for the present project may have resulted in an estimate for this generic sample that is not accurate for any of the subgroups, (eg, LTC facilities, nursing homes, complex continuing care, skilled nursing facilities, rehabilitation, and geriatrics).
Most studies, published and unpublished, reported insufficient information to answer all the methodological questions about the studies. Some gaps in information were filled by additional contacts with project authors but when the information was not available or authors could not be contacted, it was assumed that rigorous methods were not performed or that certain results were not obtained.
The data obtained do not represent all data collected from Canadian healthcare settings. No estimates of prevalence in acute and non-acute care came from the Prairie Provinces; therefore, one must assume that the prevalence of pressure ulcers in these two settings would be similar in these provinces. The information received is only a sample from relatively few institutions across Canada. It is possible that facilities electing to do prevalence studies do so when they suspect a potential problem and this might contribute to the higher prevalence reported in relation to other countries.
All pressure ulcer prevalence estimates obtained for this project are within the healthcare sector; no national estimate was secured for the general population. Several national databases, such as the National Population Health Survey (NPHS), were reviewed in vain to find a statistic for the population.
This project provides prevalence estimates for pressure ulcers in various Canadian healthcare settings of 15% to 30%, and an overall estimate of 26%. These estimates seem to be higher than estimates from the US and the Netherlands, perhaps because of the trend in the Canadian healthcare system to limit hospital admission and reduce length of stay; thereby, resulting in sicker patients within the system.
This information will be useful to clinicians, researchers, and policy makers in Canada and other countries to advocate for the needs of patients with chronic pressure ulcers. Additional information is needed about the prevalence of other types of wounds. In a recent systematic review of prevalence of lower limb ulcers, the prevalence of venous leg ulcers was determined to be 0.12% to 1.1% based on the studies that employed clinical validation of ulcers.15 More studies are needed to estimate the number people in the general population with chronic wounds of any cause.