Pressure Ulcer Incidence and Risk Factors Among Hospitalized Orthopedic Patients: Results of a Prospective Cohort Study
A variety of pressure ulcer (PU) risk factors has been identified, but not all information is consistent and some applies to specific patient populations only. The purpose of this prospective cohort study was to assess the incidence of and ascertain potential risk factors for PU formation among hospitalized orthopedic patients in the Philippines.
All patients consecutively admitted to an orthopedic ward of a tertiary hospital between March 1 and October 31, 2009 were eligible to participate if they were >19 years of age and expected to be confined to a bed or chair for at least 5 days while hospitalized. Eligible patients were seen within 48 hours following admission, and after providing informed consent, patient demographic variables, laboratory values, and other potential risk factors were recorded, including the presence of nonblanchable erythema or dry skin and Braden Risk Assessment Scale variables (sensory perception, moisture, activity, mobility, nutrition, friction shear).
The total Braden Risk Score was calculated and dichotomized to at risk (low, medium, high risk) and not at risk. Subsequent skin inspections for the development of a Stage II or greater ulcer occurred weekly. Of the 256 persons admitted, 43 met the inclusion criteria and 40 (31 men and nine women) agreed to participate in the study. Eight patients developed an ulcer a median of 7 days after admission for an overall incidence of 20%; seven ulcers were Stage II, and one ulcer progressed to Stage III. The only significant difference between patients who did and did not develop a PU was the presence of constant moisture (P <0.008) and impaired sensory perception (P <0.0001). The odds ratio (OR) for sensory perception was 45 (95% CI: 5.3–385, P = 0.0001) and for constant moisture was 10.8 (95% CI: 1.8–65, P = 0.008). On multivariate analysis, the adjusted OR was 43 (95% CI: 3.5–513) for sensory perception and 10 (95% CI: 0.85–117) for moisture. The likelihood ratio test showed that both impaired sensory perception and constant moisture were predictive of PU formation (P = 0.0465). Most (six out of eight) patients who developed a PU had a history of spinal cord injury. Study results confirm that limited sensory perception and the presence of constant moisture are risk factors for PU development among hospitalized orthopedic patients. This is the first PU incidence and risk factor study conducted in the Philippines, and the results confirm some of those reported from orthopedic patient studies in other countries.
Potential Conflicts of Interest: none disclosed
Pressure ulcers (PU) occur as a consequence of prolonged skin contact against a surface, whether the patient is in a recumbent or prone position1-4; they are among the most common complications in healthcare. Reported incidence varies,5 but usually ranges from 0.4% to 38% in acute care settings and from 2.2% to 23.9% in chronic care facilities.1 Numerous studies to investigate the factors predictive of PU formation have been conducted.6-14 Among these risk factors, impaired mobility or limited activity was found to be the most important.7-13
Populations at risk for PU include the elderly, persons with spinal cord injury or other neurological deficits or degenerative processes, trauma patients, and persons with any condition that limits the ability to move freely in response to the perception of discomfort.3 Specific subpopulations have been reported to have a high prevalence of PUs: hospitalized quadriplegics, elderly patients with femoral fractures, and patients in critical care units.5 The presence of PU in these populations was found to be a risk factor for poor overall prognosis and premature mortality in some patients,14 decreased quality of life,15 frequent and longer hospital admissions, and more intensive nursing and medical care, as well as a substantial burden to the healthcare system.15,16 These studies show that PUs in admitted patients result in a poorer prognosis, premature mortality, longer hospital stay, and increased costs.
At present, no data are available regarding the prevalence of PU and the costs of treatment in the Philippines. In 2001, the average cost of treating PU in the US ranged from $5,000 to $70,000 depending on PU stage. The total national cost in the US for treatment was estimated to be $1.3 billion annually and rising.15 A recent study17 in nursing and residential facilities showed that establishing a PU quality improvement collaborative (QIC) resulted in a decreased PU incidence and improved patient quality of life. Similar studies showed that a multidisciplinary18 or a collaborative19 approach may decrease PU incidence. However, the long-term cost effectiveness of implementing the QIC for PU prevention is still not known.17
The Braden Risk Assessment Scale generally has been found to be a valid and reliable predictor of PU development.6,7,20,21 The scale comprises six subscales that reflect sensory perception, skin moisture, activity, mobility, friction and shear, and nutritional status. Results of a recent study22 among 233 elderly patients indicate the scale has a sensitivity of 74.1% and a specificity of 75.4%.
Because the treatment of established PUs is difficult and costly, the general recommendation from the literature was that use of a risk assessment scale always should be supplemented by a targeted preventive approach.13 The purpose of this prospective cohort study was to determine if the following risk factors were associated with the occurrence of PUs among hospitalized orthopedic patients: the presence of nonblanchable erythema, Braden Scale score parameters (sensory perception, mobility, activity, nutrition, moisture, and friction shear), and dry skin.
Patients. This study comprised a prospective cohort of all patients admitted to the orthopedic ward of a tertiary acute care hospital between March 1, 2009 and October 31, 2009. The protocol was approved by the Institutional Review Board of the hospital. The ward nurses were informed of the study by the investigators, and a research assistant (a registered nurse with at least 1 year of experience with patient care in a hospital setting) identified potential study participants by visiting the wards three times a week or after being informed by nursing staff that an eligible patient had been admitted during the past 48 hours. Inclusion criteria for the study were: patients 19 years old who expected to be confined to chair or bed for at least 5 days and expected to be hospitalized for at least 5 days. Persons admitted with a Stage II or greater PU23 or active skin disease that would interfere with PU assessment were not eligible to participate.
Upon admission to the orthopedic ward, all patients were placed on a standard mattress and a standard turning protocol (every 2 to 3 hours) was implemented for all admitted patients with limited mobility. Once a participant was determined to be eligible for the study, the research assistant provided him/her a copy of the informed consent document.
Study variables. At study baseline (following admission), demographic variables, including patient age, gender, height, weight, and body mass index, and laboratory findings (lymphocyte count and albumin levels), were recorded. The research assistant assessed the following: presence of nonblanching erythema (skin with localized nonblanchable redness over a bony prominence upon depression) and dry skin (a paucity of moisture or the presence of scaling over a bony prominence). Bony prominences included occipital, scapular, spinal, sacral, trochanteric, ankle, and heel areas. Braden Risk Assessment Scale parameters (sensory perception, activity, moisture, mobility, nutrition, and friction shear) were assessed; participants were categorized as at risk (score of £18) or not at risk (score of >19).22 A score of 18 was the cut-off value for the risk assessment. The research assistant assessed the skin over all bony prominences of enrolled participants two times a week until discharge for a maximum of 8 weeks. Standard care protocols, including wound dressing and/or debridement, were initiated when a Stage II or greater PU was identified.
Outcomes evaluation. PUs were defined as epithelial loss or skin breakdown over a bony prominence2,23 and were graded according to the European Pressure Ulcer Advisory Panel.23 The primary outcome or endpoint in this prospective study was the occurrence of a Stage II or greater PU.
Data analysis. Data were encoded using Microsoft Excel 2007 and EpiInfo™ version 3.5.1 (Atlanta, GA) for Windows and analyzed using EpiInfo™ for Windows and STAT™ version 10.0 (College Station, TX). Continuous data were expressed as means and standard deviations. Categorical data were expressed as frequencies for populations with and without pressure ulcers. The t-test and the Fisher’s exact test were used to determine whether a difference was found between the demographic factors and eight hypothesized risk factors of persons who did and did not develop PUs.
To determine which variables were predictors for PU formation, each hypothesized risk factor was analyzed using a crude logistic regression analysis with 95% confidence intervals (CI) in which a P value of <0.10 was considered significant. Significant factors on crude analysis were included in the multiple regression analysis using the hierarchical method of selection with a P value of 0.05 set for statistical significance.
Demographic characteristics. Of the 256 persons admitted in the 8-month study period, 43 were eligible for inclusion. The majority of patients had an average hospital stay of >14 days. Most of the patients (four out of eight) who developed PUs were 24 to 33 years of age. Three (3) of the 43 eligible patients did not want to provide informed consent. Of the 40 remaining participants (31 men and nine women), eight developed a Stage II or greater PU for a total incidence of 20%; of the 31 men, six developed PU (P = 0.59) and of the nine women, two developed PU (P = 0.59), all in the sacral area. The median time for PU development was 7 days. Seven people developed Stage II and one developed Stage III PUs. More patients with spinal cord injuries (five out of six) developed a PU compared to those with lower extremity fractures (three out of 31). None of the baseline patient variables was significantly different between patients who did and did not develop a PU (see Table 1).
Risk factors. Of all patient and skin variables assessed, only the variables sensory perception and moisture were significantly different between persons who did and did not develop a PU (P <0.05) (see Table 2). Similarly, crude logistics analysis showed that reduced sensory perception and constant moisture on admission were significant predictors of PU occurrence (see Table 3). Patients with impaired sensory perception were 45 times more likely to develop a PU (95% CI: 5.3–385, P = 0.0001) and patients with constant moisture were more than 10 times more likely to develop an ulcer (OR = 10.7, 95% CI: 1.8–65, P = 0.008) as persons not experiencing these issues. On multivariate analysis using the hierarchical method of selection, sensory perception and moisture remained predictive of PU development. Adjusted OR showed that, while controlling for moisture, persons with impaired sensory perception at baseline were 43 times (95% CI: 3.5–5.13) more likely to have a PU. Likewise, adjusting for sensory perception, the adjusted OR for constant moisture was 10 (95% CI: 0.85 –117). Among the eight patients who developed PUs, six had impaired sensory perception, six had constant moisture, and five had both impaired sensory perception and constant moisture. The likelihood ratio test showed both factors were predictive of developing a PU (P = 0.0465).
Incidence variability among studies. In this prospective study involving patients admitted to an orthopedic ward, the total cumulative PU incidence was 20% (eight out of 40) with a median time of 7 days from time of admission to study outcome. This incidence was relatively high compared to previous studies by Perneger et al24 (14.2%) and Allman et al1 (12.5%) with median occurrence of 9 days, and Chan et al21 (9.1%) and Reed et al25 (14.7%). However, the incidence in this study was lower compared to Verschueren et al26 (36.5%). The differences in the incidence rate may be due to the variability in the age of the study population; the average age of current study participants was lower than that of other studies.1,26
Determining risk factors. Although few studies have investigated risk factors for orthopedic patients, reduced sensory perception, small body build for height, and the presence of moisture were found to be predictors for PU development in a prospective cohort study involving 197 orthopedic patients.21 The current study results confirm the role of sensory perception and the presence of moisture in PU occurrence.
Risk assessment in PU prevention has been studied to help create a targeted approach to PU management and prevention.19 Several factors known to be associated with the occurrence of PU have been investigated. This study suggests that sensory perception and moisture were predictive of the development of Stage II or greater PUs; factors reported in other studies2,24-26 to be predictive of PU occurrence were not shown to be significant predictors of PU in this study. Nonblanching erythema,1,28 dry skin,1,8 friction-shear,22 nutrition,22,29,30 mobility,31 and the Braden Risk Assessment Scale scores6,7,20,21 were not shown to be predictive of PU in this subset of orthopedic patients after multivariate regression analysis. More patients with spinal cord injuries (five out of six) developed a PU compared to those with lower extremity fractures (three out of 31). In persons with spinal cord injuries, multivariate analysis showed both impaired sensory perception and moisture were predictive of PU formation.28,30
Challenges to accurate assessment. The large confidence intervals for the OR for factors studied may show some imprecision. Assessing the presence of nonblanchable erythema may be challenging in persons with fair or dark-colored skin, such as the Filipinos. In this study, only four patients presented with nonblanching erythema on admission and two of them eventually developed Stage II ulcers. Also, age and associated age-related changes and concomitant comorbidities make the older population prone to PUs.8,22 In addition, the evidence regarding the role of albumin in predicting PU formation remains inconclusive. In recent studies, lymphopenia as well as low serum albumin2,32 were shown to be predictors of PU formation in an elderly population. To the contrary, Gallagher et al31 did not find the albumin status of adult patients to be a predictor of PU formation after multivariate analysis. Yet in the current study, 70% of patients had low albumin levels on admission. This could explain why the variable was not significantly different. Some authors32 suggest that the association between hypoalbuminemia and PUs in past studies may be explained by protein losses in persons with PUs; this is especially true in cross-sectional studies and case-control or retrospective studies.2
The degree of mobility as part of the Braden scale has been consistently reported to be a predictor of PU.21,31,33 On univariable analysis in this study, only impaired sensory perception and moisture were significant, with P = 0.010 and OR = 10.7 (95% CI: 1.75, 65.2). Also, using a score of 18 as the cut-off for at risk per the Braden Risk Assessment Scale was not a predictor of PU in this study, contrary to other research,6,7,20-22 again potentially related to the sample. On multivariate logistic regression analysis, both impaired sensory perception (OR 43) and the presence of moisture (OR 10) were shown to be predictors of PU formation. Among the 40 patients in this study, six out of eight patients who developed PUs had impaired sensory perception. Likewise, among those who also developed PUs, six out of eight also had constant moisture and five of these patients also had impaired sensory perception.
Based on the results of this and other studies, protocols for patient care must address skin moisture, especially in persons such as spinal cord injury patients with impaired sensory perception. Although constant moisture was a factor in a study of patients in an intensive care unit,34 studies on moisture as an individual risk factor in spinal cord injury patients are not common. Moisture may have been incorporated in the Braden Risk Assessment Scale20,22,27 or may be related to incontinence because both urinary and fecal incontinence may cause constant soiling and moisture of the skin, especially in the sacral area.1,33,34
One of the main limitations of this study was that patient comorbidities were not included as study variables. Although other researchers7,11,29 consider comorbidities such as age, infection, diabetes, chronic diseases, and malignancy predictors of PUs, the current study was not able to include them as possible risk factors. Matching groups in the data collection stage or performing a subgroup analysis could have facilitated inclusion of comorbidities.
Another limitation was the small sample size. The large CI for the estimated OR underscored the lack of adequate sample size, even though most of the patients who developed an ulcer had a history of spinal cord injury. This should be addressed in future studies.
The purpose of this prospective study was to determine if nonblanchable erythema, dry skin, Braden Risk Assessment Scale parameters, and risk assessment per the Braden scale were risk factors for PU formation among patients admitted to an orthopedic ward. The incidence of PUs was 20% and impaired sensory perception and constant moisture were predictive of PU formation. Identifying these factors should help facilities institute preventive measures to avoid PU occurrence and subsequent expensive care. This may translate into improvement of the quality of care and possible reduction in costs.
Dr. Molon is a senior orthopedic resident and Dr. Estrella is a Clinical Associate Professor, Microsurgery Unit, Department of Orthopedics, UP-College of Medicine, Philipinne General Hospital; and Advanced Study and Research in Orthopedics (ASTRO) Study Group, National Institutes of Health (NIH), University of the Philippines-Manila, Manila, Philippines. Please address correspondence to: Emmanuel Estrella, MD, Microsurgery Unit, Department of Orthopedics, Philippine General Hospital, UP Manila, Manila, Philippines 1000; email: firstname.lastname@example.org.