ADVERTISEMENT
Incidence of Hospital-Acquired Pressure Ulcers in Acute Care Using Two Different Risk Assessment Scales: Results of a Retrospective Study
Abstract
More than one million people develop pressure ulcers (PU) every year in the US — a major problem that leads to increased morbidity and mortality. A hospital-acquired PU (HAPU) is defined as any ulcer noted 24 or more hours after hospital admission. The purpose of this retrospective study was to compare the incidence of HAPUs between two 1-year periods of time (March 1, 2007 through March 17, 2008 and March 18, 2008 through March 31, 2009) at an acute care facility. Records from all admitted adult medical-surgical and critical care patients along with routinely submitted quality control reports on patients who acquired PUs during hospitalization were reviewed and abstracted. A locally developed PU risk assessment instrument (SST) was used during the first time period and the Braden Scale (BS) was used during the second time period. The overall PU incidence was 368 in 41,840 patient admissions (8.80 per 1,000); 9.49 in the SST and 8.08 in the BS patient sample (P = 0.125). No significant differences in patient age, race, and PU severity or location were observed. However, the incidence of more severe ulcers (Stage III and Stage IV) was lower in the BS group (0.5667 in the SST group compared to 0.2419 per 1,000 admissions in the BS group) even though median patient length of stay was longer (28 days and 34 days in the SST and BS groups, respectively), suggesting that use of a valid and reliable PU risk assessment instrument may reduce the incidence of severe PUs. Study design and sample size limit the ability to generalize the results of this study. Controlled clinical studies comparing the effect of different PU risk assessment instruments on PU incidence are needed.
Potential Conflicts of Interest: none disclosed
More than one million people develop pressure ulcers (PUs) every year in the US.1 Healthy People 20102 identified the need for healthcare providers to reduce PU incidence in nursing home residents, and one of the goals of Healthy People 20203 is to reduce PU-related hospitalizations among older adults. In October 2008, the Centers for Medicare and Medicaid Services4 stopped reimbursement for hospital-acquired PUs (HAPUs), defined as any ulcer noted after 24 or more hours of hospital admission; this action has impacted hospital budgets. The National Quality Forum5 has included Stage II and Stage IV PUs acquired after admission to a healthcare facility in its list of Serious Reportable Events.
A group of international PU experts from Europe and the US6 developed definitions for incidence in the context of PUs. Incidence measures the rate of new PUs that develop over a specified time sample. The incidence of PUs in hospitalized patients varies greatly, ranging from 0.4% to 38%.7 Using a risk assessment scale (RAS) to identify patients at risk for developing PUs and implementing prevention interventions can reduce PU incidence, patient morbidity and mortality, and costs to the healthcare system.8
For a PU prevention program in an acute care setting to be effective, patients who are at risk for developing PUs must be identified in order to initiate the necessary interventions to prevent PUs. The use of a valid and reliable RAS is recommended over a locally developed tool, and risk assessment based on clinician judgment alone has shown inconsistent implementation of PU prevention interventions.9 In a seminal study10 on the Braden Scale (BS), using a standardized risk assessment tool with intervention protocols based on PU risk level has been found to decrease PU incidence up to 60% and decrease PU severity and costs of care.
The purpose of this retrospective study was to compare pre- and post-BS implementation PU incidence, PU severity, and PU-related patient hospital length of stay to data.
Background
There are multiple PU RASs with varying levels of validity and reliability. The BS is the most commonly used PU RAS in the US. The BS was developed by Barbara Braden and Nancy Bergstrom in 1987 and consists of a six-point scale to assess sensory perception, moisture, activity, mobility, nutrition, friction, and shear subscales. In Spain and the Netherlands, the BS was found to show good inter-rater reliability (ranging from 83% to 99%) in multiple clinical settings, a required characteristic of a risk assessment instrument intended for use by many clinicians to obtain the same assessment findings. The reliability and validity of the BS have been more widely examined than the Norton and Waterlow RAS instruments.11,12 Sensitivity and specificity of the BS are high (0.83 to 1.00 and 0.64 to 0.90, respectively) when the cut-off point of 16 is used for the possible BS score range of 6 to 23.9,13,14 For BS scores ranging from 12 to 20, the sensitivity of the scale ranges from 29% to 93% and the specificity ranges from 67% to 97%. The predictive value of positive test ranges from 52% to 80%, and the predictive value of negative test ranges from 78% to 96%.15 Overall, the BS is user-friendly, and each subscale has a detailed explanation to assist users.16 Evidence to support use of the BS to assess for PU risk and implementation of appropriate prevention interventions is extensive.
In the early 1980s, a group of wound ostomy continence nurses (WOCNs) at Sentara Healthcare developed a RAS, the Skin Scoring Tool (SST) (see Figure 1). The SST includes the following domains: sensorium, mobility, continence, skin integrity, activity, and nutrition. Each one of the six SST criteria was scored on a four-point scale (where 1 is favorable and 4 is unfavorable) with a score range of 6 to 24. Unlike the BS, higher scores on the SST indicate increased risk for PU development. If a SST score was 10 or higher, a skin care protocol was initiated, which included interventions for PU prevention. The SST was not tested for sensitivity, validity, or reliability and in 2007 it was decided that the SST should be replaced with the BS. In February 2008, 1 month before implementing the BS at Sentara Norfolk General Hospital (SNGH), an extensive staff education program developed by the hospital WOCNs was launched. The purpose of the program was to introduce the clinical nursing staff to the BS. The computer-based program was delivered via the Sentara Performance Learning Management System (PLMS). Completion of the education program was mandatory for all inpatient nurses and monitored via the PLMS. A multiple-choice test also was administered via the PLMS to assess nurses’ knowledge acquisition; a passing score of 80% was required to demonstrate proficiency. Of the 6,355 nurses who have utilized the program, 4,657 (73%) achieved a passing score on their first attempt; the remaining 27% achieved a score of 80% on either their second, third, or fourth attempt. It was hypothesized that, after implementation of the BS, PU incidence and severity, as well PU-related hospital length of stay, would decrease significantly.
Methods
SNGH is a 543-bed tertiary care facility, designated as a Level I Trauma Center. Specialty services include a heart hospital, cancer institute, transplant program, highly advanced electronic monitoring system for intensive care patients, hospitalist program, specialized level II nursery, minimally invasive surgery techniques, neurosciences, and reconstructive surgery. For this study, the author reviewed quality control reports routinely submitted by nurses on patients who acquired PUs during hospitalization. Data from adult patient units (two surgical, three medical, four step-down, seven critical care, and one oncology unit) were included. Patients who developed more than one HAPU were counted once and only the most severe PU was included in the database. Data from all patients admitted between March 1, 2007 and March 17, 2008 were included in the SST sample and reports from all patients admitted between March 18, 2008 and March 31, 2009 were included in the BS sample. The following variables were abstracted: age, gender, race, nursing unit, length of stay, admission date, discharge date, PU stage, location, and date the ulcer was recorded.
Because the National Pressure Ulcer Advisory Panel (NPUAP) staging definitions changed between the start and end of the review period,17 ulcers classified as suspected deep tissue injury (sDTI) following the NPUAP change in definitions were counted as Stage I ulcers. PU incidence (number of patients with HAPU divided by total number of patients), severity (incidence by ulcer stage), and patient length of stay (calculated by midnight census) of the two consecutive samples were calculated and compared.
Descriptive statistics were used to analyze patient demographic variables. Pre- and post-BS implementation outcomes were analyzed as follows: percent change in PU incidence rate (Pearson’s chi-square), percent change in PU severity (Fisher’s Exact Test), and percent change in length of stay (independent t-test). A continuity correction was used to improve the accuracy of the results. The alpha level for each statistical test was set at 0.05. This project was approved by the University of Virginia Institutional Review Board (HSR# 14836) and the Eastern Virginia Medical School Institutional Review Board (IRB# 10-04-WC-0098).
Results
During the two study periods, 368 HAPUs were reported — 201 pre-intervention (SST) and 167 post-intervention (BS). The average age of the total study sample was 62 years (range 17–93 years, median 65). During SST use, the average age was 62 years (range 17–93, median 65); during the BS period, average age was 60 years (range 17–93, median 65). Most patients were male (224, 60%) — 121 (54%) were in the SST and 103 (46%) were in the BS samples. Among all patients, 180 (49% — 101 in the SST and 79 in the BS group) were Caucasian and 173 (47% — 94 in the SST and 79 in the BS group) were African-American. No patients were identified as Native American or nonspecified race in the SST sample; in the BS group, four patients were Hispanic, one was American Indian, seven were Asian/Pacific Islander, and three were listed as other/not specified. Seventeen adult nursing units were represented in the sample. Critical care units accounted for 192 patients, medical-surgical units 100, and step-down units 71 (see Table 1).
Ulcers were noted on the sacrum, heel, hip, ischial area, buttock, occipital, and “other” (see Table 1). All stages of PUs were represented in both samples (see Table 1). Approximately 70% of the ulcers in each sample were Stage II PUs. The incidence of each stage of PU decreased in the BS sample except DTIs and unstageable PUs.
The overall incidence of PUs was 368 in 41,840 patient admissions (8.80 per 1,000 admissions) (see Table 2). Incidence was 201 per 21,174 (9.49 per 1,000 admissions) and 167 per 20,666 admissions (8.08 per 1,000 admissions) in the SST and BS samples, respectively, a Pearson chi-square value of 2.35 with a continuity correction value of 2.192. The difference in the SST and BS sample rate was not statistically significant (P = 0.125).
In the total sample, four patients were included twice. Each of these patients had two separate admissions during the study in which they acquired PUs. A random coin toss was used to remove one admission of each of the four patient sets. After the four admissions were randomly removed, the Pearson chi-square value was 2.717 with a continuity correction value of 2.546 (P = 0.099). The overall incidence of Stage III and Stage IV PUs in the total study population was 17 — 0.5667 per 1,000 in the SST and 0.2419 per 1,000 admissions in the BS group (Pearson chi-square 2.714, continuity correction value 1.974, P = -.099) (see Table 1).
Differences in level of PU severity according to stage between the BS and SST samples also were compared (Pearson chi-square 4.295) (see Table 2). Because some values were <5, Fisher’s Exact Test was conducted. The differences were not significant (P = 0.574) and remained not significant after the four admissions of duplicate patients were removed (P = 0.644). An independent t-test conducted to evaluate the hypothesis that patients assessed for PU risk utilizing the BS would have a decreased length of stay as compared to patients assessed with the SST found no significant differences in patient length of stay between the two groups neither before (P = 0.773) nor after the four admissions of duplicate patients were removed (P = 0.721) (see Table 2). When outliers were excluded (5% of admissions with the longest lengths of stay), average length of stay was 32.93 days (SD 45.484) in the SST and 38.57 (SD 41.366) in the BS group (P = 0.064). 
Discussion
The results of this study suggest that use of the BS may decrease PU incidence, especially of more severe Stage III and Stage IV PUs as compared to the SST, even though the two contain similar variables, but overall differences were not statistically significant. Gender and PU location distribution was similar for each sample and the largest percent of ulcers in both the SST and BS sample were located in the sacral area.
The incidence of each PU stage decreased and unstageable ulcers and DTIs increased in the BS sample. These findings were not statistically significant. However, the observed overall lower incidence and 58% difference in the rate of Stage III and Stage IV PUs after implementing the BS may be clinically significant, especially because the mean patient length of stay was 4.51 days longer in the BS sample than in the SST sample. One possible explanation for this difference in length of stay may be an increase in patient acuity between the data collection periods. Data on patient acuity variables or PU risk scores was not collected.
Limitations
Multiple factors limit this study’s findings. During November 2008, the hospital transitioned from a manual quality reporting process to electronic reporting accessed via the hospital intranet. This change could have increased or decreased reporting of PU reporting during the BS sample. Another major system change occurred in March 2009 — the hospital transitioned from a partially electronic nursing documentation system to a fully integrated electronic medical record. Even though this change only affected the last month of the study, it had the potential to change nursing documentation. The transition from the SST to the BS also may have been a limiting factor. Mandatory education for nurses was provided; as such, reporting could have increased or decreased related to the staff education and a general increase in awareness of skin assessment, PU staging, the BS, and PU prevention interventions. However, the Hawthorne effect was minimized because these data were collected for quality management purposes before this retrospective study.
Due to the length of the study, other confounding variables were not controllable. The patients in each time sample were not matched and patient characteristics may have changed over time. This may possibly explain the difference in length of stay between the two samples. If the sample size were larger and PU risk scores had been collected, a stronger relationship between the SST and BS samples may have been established.18
Staff characteristics also could have changed during the study time period, leading to observation bias. Although all new staff are required to attend a nursing orientation that includes a didactic presentation on PU prevention, this may not equalize skill sets among nurses. Another consideration is that staff may be reluctant to complete quality reports. Even though the facility offers a fair and just culture, there may be fear of retribution.
PUs discovered after the first 24 hours of admission are considered hospital-acquired at the study facility. Consistent and timely admission skin assessments were not verified. Staging accuracy is another concern. All Stage III and Stage IV PUs were validated by a certified WOCN; however, Stage II PUs were not validated.
Finally, the effect of prevention measures is a common weakness in RAS studies because these measures impact the ability of the risk assessment tool to measure PU risk.19
Conclusion
A study was conducted to compare PU incidence in two consecutive patient samples using two different assessment instruments — a hospital-developed risk assessment tool (SST) and the BS. Using the BS, 201 patients developed a HAPU (9.49 per 1,000 admissions) and using the BS, the HAPU rate was 8.08 per 1,000 admissions. Differences in PU rates were not statistically significant but fewer severe ulcers (Stage III and Stage IV) were documented and the average patient length of stay was longer during the time the BS was used. The importance of using a risk assessment tool as an integral component of clinical practice to identify patients at risk so prevention measures may be implemented is well established.20 The availability of more patient-specific information might have provided more insight into the results obtained. Studies comparing the effects of different PU risk assessment instruments are needed.
Dr. Jackson is Director of Patient Care Services, Sentara Norfolk General Hospital, Norfolk, VA. Please address correspondence to: Stephanie Smith Jackson, DNP, RN, ACNS-BC, Director of Patient Care Services, Sentara Norfolk General Hospital, Education Department, 600 Gresham Drive, Norfolk, VA 23507; email: sxjacks1@sentara.com.
1. Clay KS. Evidence-based Pressure Ulcer Prevention: A Study Guide for Nurses. Marblehead, MA: HCPro, Inc;2005.
2. US Department of Health and Human Services. Healthy People 2010. Available at: www.health.gov/healthypeople/. Accessed April 1, 2009.
3. US Department of Health and Human Services. Healthy People 2020. Available at: http://healthypeople.gov/2020/topicsobjectives2020/objectiveslist.aspx?topicid=31. Accessed April 3, 2011.
4. Centers for Medicare and Medicaid Services. CMS proposes to expand quality program for hospital inpatient services in FY 2009. Available at: http://www.cms.gov/.
5. National Quality Forum. National Quality Forum Serious Reportable Events in Healthcare, 2002 Consensus Report, Updated October, 2006. Available at: www.qualityforum.org/projects/sre2006.aspx. Accessed October 5, 2009.
6. Pressure Ulcer Prevention: Prevalence and Incidence in Context. A Consensus Document. London, UK: MEP Ltd;2009.
7. Lyder CH. Pressure ulcer prevention and management. JAMA. 2003;289(2):223–226.
8. Ayello E, Braden B. How and why to do pressure ulcer risk assessment. Adv Skin Wound Care. 2002;15(3):125–133.
9. VandenBosch T, Montoye C, Satwicz M, Durkee-Leonard K, Boylan-Lewis B. Predictive validity of the Braden Scale and nurse perception in identifying pressure ulcer risk. Appl Nurs Res. 1996;9(2):80–86.
10. Braden B, Bergstrom N. Clinical utility of the Braden Scale for Predicting Pressure Sore Risk. Decubitus. 1989;2(3):44.
11. Anthony D, Parboteeah S, Saleh M, Papanikolaou P. Norton, Waterlow and Braden scores: a review of the literature and a comparison between the scores and clinical judgement. J Clin Nurs. 2008;17(5):646–653.
12. Kring D. Reliability and validity of the Braden Scale for predicting pressure ulcer risk. J WOCN. 2007;34(4):399–406.
13. Balzer K, Pohl C, Dassen T, Halfens R. The Norton, Waterlow, Braden, and Care Dependency Scales. J WOCN. 2007;34(4):389–398.
14. Watkinson C. Inter-rater reliability of risk-assessment scales: comparing their validity when identifying patients’ pressure score risk. Prof Nurs. 1996;11(11):751.
15. Capobianco ML, McDonald DD. Factors affecting the predictive validity of the Braden Scale. Adv Wound Care. 1996;9(6):32–36.
16. Chan W, Pang S, Kwong E. Assessing predictive validity of the modified Braden scale for prediction of pressure ulcer risk of orthopaedic patients in an acute care setting. J Clin Nurs. 2009;18(11):1565–1573.
17. National Pressure Ulcer Advisory Panel. Updated Staging System: Pressure Ulcer Stages Revised by NPUAP. Available at: www.npuap.org/pr2.htm. Accessed September 14, 2009.
18. Bolton L. Which pressure ulcer risk assessment scales are valid for use in the clinical setting? J WOCN. 2007;34(4):368–381.
19. Pancorbo-Hidalgo PL, Garcia-Fernandez FP, Lopez-Medina IM, Alvarez-Nieto C. Risk assessment scales for pressure ulcer prevention: a systematic review. J Adv Nurs. 2006;54(1):94–110.
20. Polit DF, Beck CT. Nursing Research: Generating and Assessing Evidence for Nursing Practice, 8th ed. Philadelphia, PA: Lippincott Williams & Wilkins;2008.
21. Feuchtinger J, Halfens R, Dassen T. Pressure ulcer risk assessment im- mediately after cardiac surgery — does it make a difference? A comparison of three pressure ulcer risk assessment instruments within a cardiac surgery population. Nurs Crit Care. 2007;12(1):42–49.
22. Moore Z, Cowman S. Risk assessment tools for the prevention of pressure ulcer (Review). The Cochrane Collaboration. 2008;3:1–13.
