The Braden Scale is today's most widely used PU risk assessment tool. The theoretical framework is based on a physiological model depicting factors contributing to the development of PUs. It includes factors affecting the intensity and duration of pressure (decreased mobility, decreased activity, and decreased sensory perception), which combine with intrinsic factors (age, nutrition, vascular perfusion) and extrinsic factors (increased moisture, increased friction, and increased shear forces) that affect tissue tolerance. The Braden Scale is publicized as the most extensively tested and studied of the assessment tools. The Braden Scale has a potential score ranging from 6 to 23 derived from the total scores of its six subscales (sensory perception, mobility, activity level, moisture/incontinence, nutrition, and friction/shear). Lower scores on the Braden Scale indicate a greater risk for PU development. For example: Very high risk = 9 or below; High risk = 10-12; Moderate risk= 13-14; and Mild risk = 15-18. There is literature by Braden to suggest that if a person has other major risk factors present (age, fever, poor nutrition, hemodynamic instability), their score should be advanced to the next highest level of risk (i.e., next lower score), yet observational studies suggest nurses do not routinely do this.
Clinical Pearl:Nurses who are being tasked with conducting daily pressure ulcer risk assessments should take them very seriously and make sure to score them accurately, taking into account any change in the patient’s condition and risk not assessed by the specific tool being used. It is always better to overestimate risk than to underestimate risk! And sometimes this means assigning a lower score, if lower scores indicate higher risk on the risk screening tool you are using (such as the Braden Tool). If the person’s sub-score falls between numbers, always select the lower number to try and capture any possible risk. For instance, if the patient’s nutritional sub-score falls between a “2” and “3” on the Braden Tool, always mark it a “2”. |
Reports of sensitivity and specificity of the Braden and Norton Scales may be misleading. Jalali and Rezaie report sensitivity to be "the percent of individuals who developed a pressure ulcer who were assessed at being at risk" for a PU by the tool (Norton, Braden, Waterlow, or Gosnell scales), and specificity to be "the percent of individuals who do not develop a pressure ulcer who were assessed to not be at risk" (p.94). They report sensitivity and specificity (respectively) for the Norton scale to be 49% and 100%, the Braden scale was 53% and 100%, the Gosnell scale was 85% and 83%, and Waterlow was 63% and 82.5%. As Norton posits, in using the tool at all, you may be providing an intervention; therefore, the "predictive validity" scores may accurately depict a lower percent sensitivity for the more effective tools because they effectively lower the incident number of PUs in those identified as higher risk. The specificity of 100% (Braden and Norton scales) depicts the tool's ability to correctly identify those not at risk for developing a PU and suggests the Gosnell and Waterlow scales were not as successful. Therefore, a word of caution should accompany research studies seeking to 'validate' risk assessment scales in this manner rather than investigating the impact of their use on the incidence of Pus (Jalali & Rezaie, 2005).
Defloor criticized Braden & Bergstrom's inclusion of "tissue tolerance" in their conceptual model's causal pathway toward PU development, stating, "Tissue tolerance cannot cause pressure sores. The existence of pressure or shearing force is needed. Tissue tolerance is considered an intermediate variable, not a causal factor" (p.207). Defloor said that the Braden & Bergstrom conceptual model did not include factors identified in other studies as strongly associated with PU development, such as "specific diseases, dehydration, protein deficiency, body build, position, etc." He described his conceptual scheme of pressure sore formation, utilizing known risk factors and pathophysiology and expanding on the factors listed in the Braden & Bergstrom model. Defloor's 1999 model depicts Compressive Force and Shearing Force as independent variables that interact and Tissue Tolerance for Pressure as well as Tissue Tolerance for Oxygen as intermediate (moderating) variables toward the dependent variable of Pressure Sores. Defloor also noted that more research is needed, especially regarding factors such as smoking and low protein, as well as the influence of preventive measures. Defloor concluded (p.214) that many authors on the subject of PU development limit themselves to identifying risk factors, stating, "it is important to gain insight into how these risk factors interact, not only for a better understanding of the pathophysiology of pressure sores and preventive measures but also for the development of valid risk scales."
Jalali & Rezaie tested the "predictive power" of the 4 most common PU risk assessment scales (RAS) side by side in a prospective study evaluating incident PUs in 3 educational hospitals in Iran between 2000 and 2002. They examined 230 patients (100 men + 130 women) over 21 years old admitted to the hospital without a PU. They used 4 common PU risk scales (Norton, Braden, Waterlow, and Gosnell) and a uniform skin assessment tool requiring the researcher to document the skin condition of all bony prominences for every patient within 48 hours of admission and every 24 hours afterward for 14 days. All incident PUs were staged and recorded according to the AHCPR (now AHRQ) PU treatment guidelines that contained Stage I – Stage IV (no unstageable or deep tissue injury stages). Four separate researchers each evaluated the patient using one of the 4 RAS, but there were some limitations associated with the study. Deep tissue injury discoloration could be misclassified as Stage I. Researchers report that the patients have been assessed for a minimum of 14 days under their methods section and a maximum of 14 days under the procedure section.
The average number of days the sample was followed was unclear; however, it seems apparent that patients were followed for only 2 weeks. This is a limitation of the study, especially regarding the negative predictive value (NPV) and positive predictive value (PPV) they report regarding the scales. Norton and Braden demonstrated 100% PPV in this study, but Gosnell and Waterlow demonstrated only 59 and 61% PPV, while Gosnell and Waterlow demonstrated 95% and 84% NPV and Norton and Braden only had 52% and 58% NPV. Assuming these values to be true for the argument, perhaps patients developed PUs on day 15 or 20. If more patients developed ulcers later, it would alter some of these results. If the study followed the patients longer, it would be more supportive of the author's premise regarding the scales.
Furthermore, the Youden's Index (developed in 1950 to evaluate the accuracy of diagnostic tools/tests) was used to report predictive validity or "predictive power" for the four scales, but the index data was not reported uniformly. Jalali & Rezaie reported that Youden's Index "assumes that sensitivity and specificity have equal importance" (p.94); however, it is unclear if sensitivity and specificity of PU risk scales do or should have equal importance. Is it more important that PU risk assessment scales predict who will get a PU or who will not?