ADVERTISEMENT
Computer-based Testing of the Braden Scale for Predicting Pressure Sore Risk©
Pressure ulcers — also known as bedsores, pressure sores, and decubitus ulcers — are localized areas of tissue destruction caused by compression of soft tissue between a bony prominence and an external surface, such as a bed or chair, for a prolonged period of time. They have plagued mankind for centuries and vary in severity from superficial tissue damage to deep craters exposing muscle and bone. Most pressure ulcers are preventable if patient risk is recognized in time for preventive actions to be initiated.1
Achieving the best quality care requires that healthcare be safe, evidence-based, patient-centered, and timely. One aspect of the National Quality Forum (NQF) Healthcare Safety Practices is evaluation of patients for pressure ulcer risk. Pressure ulcer prevalence is cited as an example of nursing-sensitive patient outcomes.2
Determining patient risk for pressure ulceration is greatly assisted by the availability of research-based instruments such as the Braden Scale. The Braden Scale for Predicting Pressure Sore Risk© is a formal screening tool used to identify persons who are likely to develop pressure ulcers (see Figure 1). This clinically validated tool allows nurse providers to reliably score a patient’s level of risk for developing pressure ulcers.3,4 The Braden Scale is composed of six subscales that reflect the degree of sensory perception, skin moisture, physical activity, nutritional intake, and friction and shear, along with the ability to change and control body position independently. Each of these subscales is rated from 1 to 4 except friction and shear, which is rated from 1 to 3. Each rating is accompanied by a brief description of the criteria for assigning the rating. Nurses assess the patients and match their clinical observations to the rating criteria on the Braden subscales. When the six subscales are added together, the total Braden Scale risk score ranges from 6 to 23. A Braden score of 18 or below in an adult patient is considered predictive of pressure ulceration unless preventive measures are taken.5,6 Standard pressure ulcer prevention measures are associated with each Braden subscale.6-9 Pressure ulcer prevention is an ongoing task.
Accurate use of the Braden Scale facilitates prevention by distinguishing individuals who are, or are not, at risk for developing pressure ulcers. Under-scoring patient risk on an assessment scale precludes the opportunity to initiate prevention measures on patients likely to develop pressure ulcers. Conversely, over-scoring patients’ risk on a pressure ulcer risk assessment scale may entice nurses to unnecessarily use prevention equipment or resources for patients not likely to develop pressure ulcers. Using the Braden Scale accurately in clinical practice would allow nurses to implement pressure ulcer prevention strategies on patients at risk while judiciously using precious resources.6
Periodic patient evaluation with the Braden Scale is prudent because a patient’s level of risk may change as this/her condition changes. Generally, declining Braden risk score indicate declining functional status. The frequency of risk assessment depends on the healthcare setting and patient’s healthcare status. Many acute care facilities require nurses to use the Braden scale on a daily basis. However, a search of the literature did not reveal any training programs on how to use the Braden Scale.
Purpose
The Detroit Medical Center (DMC) is an urban tertiary healthcare system that provides teaching hospitals for students from Wayne State University. The DMC nursing documentation system requires all staff nurses to complete the Braden Scale for Predicting Pressure Sore Risk© on an Acute Care Flow Record or a Critical Care Flow Sheet on every patient, every day.
Historically, the DMC hospitals have conducted quarterly or semi-annual house-wide pressure ulcer audits. On pressure ulcer audit day, staff nurses use a data collection tool to record pertinent data on their assigned patients.10 Nurses calculate a current Braden Scale score and assess the skin over the bony prominences of their assigned patients. They record these variables and any pressure ulcer prevention measures in place at the time. Advanced Practice Nurses (APNs) and ET Nurses validate any pressure ulcers found by staff members. Concern arose as to whether staff nurses accurately used the Braden Scale to calculate pressure sore risk. During validation of patient’s skin integrity, APNs and ET Nurses noted that staff nurses were rating patients higher on Braden Scale scores (lower levels of patient risk) than was warranted by the patient condition. Additionally, it was unclear whether the pressure ulcer prevention measures were related to Braden subscale scores and accounted for the increase in pressure ulcer risk.
Detroit Medical Center nurses were familiar with the Braden scale because an abbreviated version appears on the Acute Care Flow Record and Critical Care Flow Sheet for daily use. A videotape of instructions for using the Braden Scale by Bergstrom and Braden (available at https://www.Bradenscale.com)7 was purchased and shown to staff nurses on all shifts. Despite this inservice education, the APNs and ET Nurses remained concerned that the nursing staff continued to rate patients at too low a risk level.
Method
Phase one. The DMC Pressure Ulcer Committee members developed a computer-based learning module on the Braden scale as a process improvement project. A computer expert from the medical center’s Information Technology department worked with the content expert group to construct the computer-based program using Net Learning© (Thomas Delmar Learning, Knoxville, Tenn.), a commercially available software package. The purposes of the computer-based module were: 1) to teach the nursing staff to accurately evaluate pressure ulcer risk by using the Braden Scale and its subscales and 2) to identify preventive interventions based on the patient’s Braden subscale scores. Case study examples and examination questions were developed to test clinical use of the Braden Scale. The American Nurses Credentialing Center (ANCC) Item Writers’ Guide11 was used as a reference for constructing the test questions.
All DMC clinical nurses were automatically enrolled in the computer-based Braden Scale learning module. Nurses were able to access the computer-based learning program from either the patient care units or from home. The learning module contains both Braden Scale content and associated pressure ulcer prevention measures. Five acute care case study exemplars of patients at different levels of pressure ulcer risk follow the learning module (see Mr. WG, Figure 2). The case studies contain only enough information to make a clinical judgment about the patient’s level of pressure ulcer risk and to determine appropriate preventive interventions for that case. Nurses had to assign the correct level of pressure sore risk and then correctly answer several questions about appropriate preventive interventions. Each case study has a separate test that contains between 10 and 14 questions. When an answer to a multiple-choice or true-false test question is selected, the learner is immediately informed if the answer is correct. The Net Learning computer-based program automatically collects data on how participants answered the questions and how long it took them to complete the tests. Each test stands alone and can be completed at a time determined by the learner. Nurses must test successfully on all five case studies to demonstrate Braden Scale competency. An evaluation form is electronically attached to each test to allow nursing staff to comment on both the content and the testing process.
Inter-rater reliability was conducted by having each of the APN and WOCN item writers take the actual tests for each of the four case study exemplars not written by them. All answers for each of the five case study exemplars were compared for consistency among all five item writers. The initial content of the Braden Scale module was then pilot-tested on members of the pressure ulcer committee and several staff nurses.
The Pressure Ulcer Committee members conducted an item analysis of the Braden Scale and preventive intervention test questions, which revealed a wide variation of responses to certain questions. Committee members tried to identify a reason for the broad range of answers on some questions. Many of the incorrect answers were based on patient scenarios written in abstract terms. The more abstract the clinical evidence, the greater the variability in nurses’ assignment of subscale scores. During discussions about why nurses answered the way they did, experienced nurses acknowledged that they read more into a case study than was written. The nurses made assumptions that if a case study patient “sounded bad” to them, the patient “surely must be incontinent or malnourished” even if this was not stated in the case study.
The question writers determined that the case studies and/or the multiple-choice answers contained too many distracters. Where a wide variation of answers was offered, data analysts/researchers assumed the questions were poorly written and/or the nurses lacked knowledge needed to answer the question. A printed copy of the DMC Pressure Ulcer Assessment and Prevention Guideline brochure was used to emphasize areas that nurses consistently answered incorrectly. Questions or answers that were ambiguous on the computer-based test were rewritten. For example, in Case Study 1, additional descriptive verbiage was added to the patient information to make it match the criteria for “very limited” mobility. In other cases, extraneous patient information and admitting orders that did not contribute to making a decision or answering the test questions were eliminated.
Phase two. Computer-based testing of the Braden Scale went “live” for the DMC nursing staff in July 2002. Computer testing was not new to DMC nurses, although previous tests had been on regulatory compliance only. The Braden Scale Module was the first computer-based learning program that required clinical judgment. The module was much longer compared to previous computer tests and required critical thinking skills. The logistics of using the computer for testing posed a problem for some nurses. They needed to be able to use the scroll bar to review patient case studies while answering succeeding questions on the Braden subscales. Because nurses expressed difficulty mastering use of the scroll bar to go back and forth, printed copies of the five case studies were placed near the patient care unit computers. Having hard copies of the case studies visible during computer testing seemed to ease nurses’ frustration with having to use the scroll bar.
Results
More than 2,500 DMC nurses completed the computer-based Braden Scale learning module during the third quarter of 2002. Each of the five case studies took nurses between 6 and 10 minutes to complete. On average, nurses correctly rated the Braden Scale level of risk 75.6% of the time. “Not at risk” and “very high risk” levels were rated correctly more often than “mild risk,” “moderate risk,” and “high risk” levels (see Table 1). Rank order and percentage of correct responses by risk level within the Braden subscales are shown in Table 2. The highest percentages of correct nurse ratings within the Braden subscales were “not at risk” (Braden score = 19 to 23) and “very high risk” (Braden score = ≤9). The lowest percentage of correct nurse ratings within the Braden subscales was in the “mild risk” level (Braden score = 15 to 18) (see Figure 3). This was true across five of six subscales. The next lowest percentage of correct nurse ratings within the Braden subscales was the “moderate risk” level (Braden score = 13 to 14). This was true across four subscales.
Subscales with the lowest percentage of correct answers were moisture and sensory perception. Part of the Braden moisture subscale is measured in terms of “very moist” (linen changes at least once per shift) or “occasionally moist” (requires an extra linen change approximately once a day). Many nurses disagreed with a stated number of linen changes. They had difficulty differentiating “linen changes at least once a shift” from an “extra linen change once a day.” The nurses did not know how often the linen was changed on another shift. They stated they would have just changed the linen as needed if the patient had a moisture problem.
Part of the Braden “sensory perception” subscale is based on patient communication of discomfort and the ability to feel pain over “half of the body” or in “one to two extremities.” Unless these descriptions were spelled out exactly in the case study, nurses tended to score the patient higher (with less risk) than the test writers did.
In addition to the Braden Scale pressure ulcer risk factors, evaluation of patients’ skin is an important concept for evaluating risk for pressure ulcers. Even if patients are rated “not at risk” according to the Braden Scale score, Braden recommends placing them in an “at-risk” category if they have actual pressure ulcers, healed pressure ulcers, or persistently reddened areas of skin over bony prominences.7 Having actual or healed pressure ulcers or persistent erythema of intact skin indicates a low tissue tolerance for pressure. In these cases, pressure ulcer preventive measures need to be initiated automatically.
Nurses did well identifying appropriate pressure ulcer prevention measures for the case study examples. They did not score well on clinical identification of a Stage I pressure ulcer as it was described in writing. Nurses gave correct answers to questions about blanchable and nonblanchable erythema only 53% of the time.
Discussion
Nurses apparently had the easiest time determining patients generally not at risk and patients at highest levels of risk for pressure ulcers. They had the most difficulty differentiating mid-level pressure ulcer risk categories. Possibly, nurses who identify patients at a Braden Scale “mild risk” level of 15 to 18 would not see the need for prevention measures. This would be consistent with DMC pressure ulcer audits that show most nososcomial pressure ulcers at the facility are Stage II ulcers that occur in patients who are rated in the “mild risk” level (Braden score = 15 to 18) of pressure ulcer risk.10
Nurses also did not score well identifying Stage I ulcers; this concept may be better demonstrated in the learning module by including actual photographs rather than descriptions of erythema. The next version of the Braden Scale learning module will have diagrams and photographs explaining how to test for blanchable and nonblanchable erythema of the skin. This will help nurses differentiate Stage I pressure ulcers from reactive hyperemia and/or deep tissue injury — important concepts for pressure ulcer risk reduction.
Although the APNs and ET nurses thought staff nurses were under-estimating patient pressure ulcer risk, it appears that was not necessarily so. In some patient case studies, nurses overrated pressure ulcer risk and in others they underrated risk. It would be interesting to compare the written patient case studies with actual inpatients. It is not known if the nurses would improve their Braden Scale rating accuracy if they had a real patient to assess. In all probability, nurses would find it easier to assess the Braden Scale risk factors in a “live” person. In order to simulate a “live” person, visual scenarios will be included with the case studies in the next version of the computer-based learning program. Additionally, plans to give feedback to learners on why each test answer is correct or incorrect are underway. This would improve retention of instructional material.
Nurses in specialty areas of practice such as the emergency department (ED) and the operating room (OR) indicated they could not relate to the inpatient case study scenarios. Accordingly, pressure ulcer case studies suitable for the OR, ED, home care, and long-term care have been added for more widespread application.
Computer-based testing of the Braden Scale raised awareness of pressure ulcer risk and pressure ulcer prevention techniques. Since implementing the module, nurses can be heard conversing about the Braden Scale case studies throughout the hospital. It will be interesting to note whether this translates into fewer nosocomial pressure ulcers.
For several years, the DMC has conducted semi-annual hospital-wide pressure ulcer audits through direct observation of patients. Pressure ulcer point prevalence and percentage of hospital-acquired ulcers were trended over a several year period. Pressure ulcer point prevalence indicates the percentage of patients with a pressure ulcer in the hospital on the day of the audit. Detroit Medical Center pressure ulcer point prevalence has averaged 12% semi-annually for more than 15 years.10 Pressure ulcer incidence indicates the percentage of patients with nosocomial or hospital-associated pressure ulcers. Detroit Medical Center pressure ulcer incidence was determined by calculating the percentage of patients with newly acquired pressure ulcers during hospitalization. These data were reported at both the hospital and unit level. Over the last 5 years, the DMC pressure ulcer incidence ranged from 3% to 6%, data consistent with published statistics for acute care.12 Since the introduction of Braden Scale competency testing for nurses, the percentage of nosocomial pressure ulcers has declined but a direct relationship to the competency testing cannot be established. Continued tracking of pressure ulcer point prevalence and percentage of hospital-associated ulcers is planned. An electronic incidence reporting system to determine a more accurate hospital-acquired pressure ulcer rate has been initiated.
Nursing Implications
Demand is expected to increase for evidence-based practice that will reshape the work that nurses do and lead to improved health outcomes. This project on computer-based testing of the Braden Scale has several evidence-based practice implications for nursing. First, nurses need to use research-based tools, where they exist, in clinical practice. Simply placing a shortened or modified version of any scale on a hospital form will not ensure correct use; an abbreviated risk assessment tool may lead to inaccurate results when it is trialed with patients. When using the Braden Scale, nurses need to have available the entire risk assessment scale with definitions for each subscale along with the criteria for each level of risk in order to apply patient findings to the criterion related definitions on the subscales.
Nurses need education and practice in using research-based tools. Nurse managers can ensure more accurate patient risk ratings if they make certain nursing staff are well trained and competent in Braden Scale use. Even with training, inaccuracies in scoring can be anticipated; therefore, ongoing assessment of nurse competency on use of the Braden scale is important.
Managers also need to stress that patients at “mild risk” for pressure ulcers need pressure ulcer prevention measures instituted. In many facilities, these patients are likely to fall through the cracks and, at least at the DMC, patient audits show the greatest number of pressure ulcers occur in this “mild risk” category.
Conclusion
Accurate use of the Braden Scale facilitates prevention by distinguishing individuals who are at risk for pressure ulcers. In research studies where formal pressure ulcer risk assessment was implemented and levels of risk were directly related to preventive protocols, the incidence of pressure ulcers dropped by 60%.13
Evidence-based, cost-effective healthcare should be incorporated into clinical nursing practice. Before this can be accomplished, education programs that show clinical nurses how to use research findings need to be developed. In addition to the program discussed here, DMC Pressure Ulcer Committee members plan to include the Braden Scale for Predicting Pressure Sore Risk in their annual education and competency testing. A standardized approach to pressure ulcer risk assessment and prevention can have a great impact on the clinical outcome of patients and ultimately reduce hospital-associated pressure ulcers.
1. Maklebust J, Sieggreen MY. Pressure Ulcers: Guidelines for Prevention and Management, 3rd ed. Springhouse, Pa.: Springhouse Corporation;2001.
2. National Center for Nursing Quality (NCNQ). Available at: www.nursingquality.org. Accessed October 20, 2004.
3. Bergstrom N, Braden BJ. Predictive validity of the Braden Scale among black and white subjects. Nursing Research. 2002;51(6):398–403.
4. Bergstrom N, Braden BJ, Laguzza A, Holman V. The Braden Scale for Predicting Pressure Sore Risk. Nursing Research. 1987;36:205–210.
5. Bergstrom N, Braden B, Kemp M, Champagne M, Ruby E. Predicting pressure ulcer risk: a multi-site study of the predictive validity of the Braden Scale. Nursing Research. 1998;47(5):261.
6. Ayello EA, Braden, BJ. How and why to do pressure ulcer risk assessment. Advances in Skin Wound Care. 2002;15(3):125–131.
7. Pressure Ulcer Prevention Plus. Available at: www.BradenScale.com. Accessed October 20, 2004
8. Braden BJ, Bryant R. Innovations to prevent and treat pressure ulcers. Geriatric Nursing. 1990;11(4):182–186.
9. Bergstrom N. Strategies for preventing pressure ulcers. Clinics in Geriatric Medicine. 1997;13(3):437.
10. Maklebust J, Magnan M. Risk factors associated with having a pressure ulcer: a secondary data analysis. Advances in Wound Care. 1994;6:25–42.
11. American Nurses Credentialing Center (ANCC) Item Writers’ Guide, February 6, 2003.
12. Cuddigan J, Ayello AE, Sussman C, eds. Pressure Ulcers in America: Prevalence, Incidence and Implications for the Future. Washington, DC: National Pressure Ulcer Advisory Panel;2001.
13. Braden BJ, Bergstrom N. Clinical utility of the Braden Scale for Predicting Pressure Sore Risk. Decubitus.1989;2(3):44–51.
