Clinical Applications and Validation of an Innovative Wound Score

Objective. This study demonstrates the applicability of an innovative wound score that summates 5 assessments using 2-point (best) to 0-point (worst) grades based on specific findings to generate a 0- to 10-point wound score for categorizing diabetic foot ulcers as well as validates its effectiveness. Materials and Methods. Long Beach Wound Scores (LBWS) were determined prospectively over an 18-month period in 105 hospitalized patients, with or without diabetes, with lower extremity wounds. Wounds were categorized as healthy, problem, or end-stage from their initial LBWS. Outcomes were graded as good or poor using a 5-level scale. Outcome information was available and statistically analyzed for comparisons with initial evaluation LBWSs in 85 patients. Results. In the healthy category, 66.7% healed or improved and were designated as good outcomes. In the problem category, 83.3% had good outcomes. In the end-stage category, 50.0% had good outcomes. Outliers for poor outcomes in the healthy category were due to the patients’ comorbidities, and good outcomes in the end-stage category were explained by successful revascularizations and/or healing of minor amputations. The accuracy of the LBWS for predicting good versus poor outcomes was 75.3%. Conclusions. The 0- to 10-point LBWS utilizes objective criteria for grading wounds, has validation data to confirm its efficacy for predicting outcomes, categorizes wound management, and is a practical tool to use for Comparative Effectiveness Research of wound care products and quantifying Minimal Clinically Important Improvement.

A user-friendly wound scoring tool (Figure 1) incorporates the important features of 2 predominantly used diabetic foot ulcer (DFU) scoring methods and the widely accepted wound depth staging system. The Long Beach Wound Score (LBWS) is determined by summating the grades of five 0- to 2-point assessments to generate 0 to 10 (best possible) wound scores. Instead of an assessment based on either being present or absent, which is a predominant feature of DFU scoring systems, the investigators rated each assessment on a continuum of objective findings for each grade. A previous publication¹ demonstrated an interobserver reliability of 0.81 using the LBWS system in a prospective series of 83 patients. 

This validation paper serves 2 purposes: First, it quantitatively categorizes wounds as healthy, problem, or end-stage. This provides immediate guidance for management based on the initial examination of the wound. Second, it validates the LBWS for predicting outcomes in a prospective series of hospitalized patients with foot and leg wounds in patients with and without diabetes. Ancillary benefits of the LBWS are its usefulness for Comparative Effectiveness Research (CER) of wound dressing agents and for objectifying Minimal Clinically Important Improvement (MCII). 

It is essential to be as precise as possible for evaluation and management of wounds due to increasing scrutiny and authorization requirements. With more than 3000 wound dressing agents available, intelligent decisions need to be made about what dressing agent is most appropriate for the wound.² These challenges also apply to the use of hyperbaric oxygen (HBO), biologic dressings, and negative pressure wound therapy/subatmospheric wound dressings. Patient comfort for wound care, time needed for dressing changes, ease of application, and cost of the product also must be considered. 

Can a wound scoring system be a useful tool for making essential decisions for wound management? A previously published article by the authors¹ discussed concerns about the Wagner, National Pressure Ulcer Advisory Panel (NPUAP), University of Texas San Antonio Diabetic Wound Classification (UTSADWC), and Infectious Diseases Society of America-International Working Group on the Diabetic Foot (IDSA-IWGDF) wound scoring systems.^1,3-9 Each system only focuses on a particular wound assessment as a primary consideration, such as perfusion (Wagner), depth (NPUAP), or severity of infection (IDSA-IWGDF). Assessments are only recorded as present or absent. These scoring systems do not suggest using metrics to establish progress with subsequent observations. The UTSADWC classification, although frequently used in the diabetic foot literature, utilizes the components of the other 3 scoring systems, namely depth on a horizontal scale as an extension of the NPUAP system and perfusion coupled with infection as utilized by the Wagner and IDSA-IWGDF systems on a horizontal scale, to generate a 16-permutation matrix. The system only considers whether the criterion for each component or components of the matrix is present or absent without consideration for its severity or extent. This system is somewhat validated by the observation that as one moves down and to the right of the 16-permutation matrix, outcomes worsen.¹⁰ Finally, none have compelling reliability or validation data or both to support their usefulness. A not unexpected finding of the IDSA-IWGDF classification reported increased risks for amputations, higher-level amputations, and infection-related hospitalizations with increasing severity of infection.¹¹ 

The LBWS integrates important features of the above scoring systems plus adds 2 essential assessments, appearance of the wound base and size, to score a wound comprehensively. In addition, each of the 5 assessments is graded on a 0-point (worst possible) to 2-point (best possible) continuum of objective findings to reflect the seriousness of each assessment. The result is an easy to interpret 0- to 10-point LBWS. The goal of the LBWS is to resolve the concerns inherent in the other predominant DFU scoring systems with a single easy-to-use and interpret wound scoring tool that predicts outcomes and guides management. The LBWS facilitates: (1) quantifying and categorizing the seriousness of a wound from the initial examination, (2) guiding management, (3) objectifying CER, and (4) ascertaining MCII.

To improve precision for evaluating and managing wounds, the first author formulated a scoring system in 1989 that paralleled the simplicity and objectivity of the Apgar newborn vitality score.^12,13 The initial scoring system has been continually modified and revised with additions of supplemental Wellness and Goal Scores to formulate a comprehensive Institutional Review Board-approved wound scoring project. 

This study analyzed the ability of the initial LBWS to predict outcomes. Wounds were categorized as healthy, problem, or end-stage based on the initial LBWS. Healthy wounds have LBWSs of 7.5 to 10 points, problem wounds have scores of 3.5 to 7 points, and end-stage wounds score between 0 to 3 points. In the majority of the healthy category, the patients were admitted because of concurrent medical problems and their lower extremity wounds were incidental. The LBWS was then used to quantify wound changes at reevaluations. Outcomes were graded on a 0- to 2-point scale, with 2 points being a healed wound, 1.5 points an improved wound, 1 point an unchanged wound, 0.5 point a worsening wound, and 0 points if a lower limb amputation or death occurred. Criteria used to label a wound as improved included development of a healthy granulation tissue base, transition to easy-to-do wound care, ability to resume prewound level of ambulation, and amelioration of odor or discharge from the wound. Outcomes were recorded for each healthy, problem, or end-stage category. In addition, healed and improved wounds were labeled as good outcomes while unchanged, worsening, and amputation/death groups comprised poor outcomes. 

Performance measures for sensitivities, specificities, predictive values, and accuracy were determined and analyzed for differences between each wound category. In addition, outcomes were computed for the healthy combined with the problem wound categories for contrasts with the end-stage wound category. 

Of the 105 patients, 20 (19.0%) had unavailable outcome scores for the analysis. Of the remaining 85 patients who had both LBWS and outcome scores, 48 (56.5%) were male (mean age, 57.9 years; standard deviation [SD], 14.4 years), 34 (40.0%) were female (mean age, 57.0 years; SD, 15.4 years), and 3 (3.5%) had unspecified gender information. This analysis detected no effects of sex or ethnicity in the findings. The population was analyzed for healthy, problem, and end-stage wounds; outcome; and whether the patient had diabetes. Sixty-one (71.8%) of the 85 patients had diabetes. The mean LBWS of the patients with diabetes was 4.9 (SD, 1.7) and those without diabetes had 5.8 (SD, 1.5). This difference was statistically significant (Welch’s 2-sample t test; t = -2.22, df = 48.60, P = .0314).

Nine of the 85 patients (10.6%) had initial LBWSs in the healthy category, 60 (70.6%) in the problem category, and 16 (18.8%) in the end-stage category (Figure 2). In the healthy category, 6 out of 9 patients (66.7%) had good results with failures attributed to comorbidities and compliance issues. In the problem category, 50 out of 60 (83.3%) had good results. In the end-stage category, 8 out of 16 patients (50.0%) had good outcomes; all good outcomes in this category were attributed to successful distal foot amputations and/or revascularizations, which immediately improved their LBWS. Overall, 64 out of 85 patients (75.3%) had good outcomes. 

Performance measures showed the prediction accuracy of the LBWS for healing in the healthy and problem wound categories and failure in the end-stage wound category was 75.3% (Table). The positive predictive value (PPV) and accuracy for the problem wound category were each 83.3%. Of the 64 patients with good (healed or improved) outcomes, 8 (12.5%) were in patients whose initial LBWSs were in the end-stage category. Conversely, of the 21 patients from the 85 with poor outcomes (24.7%), 10 (47.6%) were in patients with initial LBWSs in the problem range. The outcome differences from patients in the combined healthy and problem wound categories were significantly different than from the end-stage category (Pearson Chi-square test: P = .023; Fisher’s Exact test: P = .020).

The LBWS establishes the seriousness of wounds as healthy, problem, or end-stage with a 0- to 10-point score using objective findings to establish the score. With serial observations, it quantifies improvement, no change, or deterioration. This provides criteria for continuing present management or changing interventions. It also provides a numerical tool for CER. The result is the ability to objectively compare the effectiveness of wound dressing agents with similar LBWSs. A subsidiary benefit is that MCII becomes objective as LBWS improves. 

The study group consisted of patients with lower extremity wounds serious enough to require hospitalization. Three findings were associated with 91.9% of the wounds in this study and included underlying deformity; deep infection involving bone, bursa, and/or cicatrix; and ischemia-hypoxia.¹⁴ This explains why the majority (70.6%) were in the problem wound category where interventions such as debridement, offloading, and revascularization when feasible (with or without adjunctive HBO) favored good outcomes.

The failure to achieve good outcomes in the healthy wound category cohort is attributed to small vasculitic ulcers, uncontrolled deformities that interfered with mobility, and/or patient compliance. In almost all instances, wounds that were in the healthy category required admission for reasons other than their lower extremity wounds, and their wounds failed to improve during the observation period. The cases where end-stage wounds at the time of the initial evaluation had good outcomes were explained by the wounds being distal in the forefoot with successful distal foot or toe amputations after revascularization and/or with HBO treatments. All the end-stage wounds that had poor outcomes either could not be revascularized or, if done, did not improve with HBO treatments. Obviously, successful revascularizations improved the perfusion assessment of the LBWS and moved the wounds into the problem or healthy wound category. However, for the investigators’ analysis, the end-stage category designation was based on the initial LBWS.

The PPV (81.2%) and accuracy of the LBWS (75.3%) validate the effectiveness of this tool to predict outcomes in lower extremity wounds. Accuracy in predicting good outcomes from the initial LBWS in the combined healthy plus problem wound categories also was 81.2%. The prediction between good and poor outcomes using LBWS categories was statistically significant (Fisher’s exact test: P = .020). This information enables clinicians to predict outcomes from the initial LBWS for patients hospitalized with lower extremity wounds. Outliers that deviated from expectations based on the LBWS had identifiable explanations.

While randomized controlled trials are considered the gold standard, increasing importance is being given to prospective studies that utilize CER and MCII. Most wounds where biologics are recommended correspond to the present healthy wound category. Can equally good results be obtained with less costly wound dressing agents? The LBWS is particularly applicable to the problem wound category, which had more than 80% accuracy in this study for predicting good outcomes. It is the category of wounds in which appropriate interventions make the greatest differences between good and poor outcomes. Comparisons of outcomes of wound interventions for similar LBWS provide objectivity for CER.

Of the criteria for MCII, reduction in wound surface area may be the least important. Patients live and function well with small, chronic, stable, vascular-based wounds. More importantly, MCII of lower extremity wounds includes control of sepsis, alleviation of pain, cessation of exudate production, development of a granulating wound base, simplification of wound care, and restoration of function. With respect to MCII of lower extremity wounds, it reflects more than whether the wound has healed. Three biologics (Appligraf [Organogenesis, Canton, MA], Dermagraft [Organogenesis], and REGRANEX [Smith & Nephew, Fort Worth, TX]) were given pre-market approval (PMA) by the US Food and Drug Administration for DFUs. The approvals were based on reduction in wound surface areas over a 12-week period.^15-17 The reductions ranged from 30% to 50% versus the controls using normal saline dressings being 18% to 35%. Small wounds with healthy bases corresponding to the LBWS healthy category were used, for the most part, in these PMA studies where perfusion was adequate and infection was controlled. 

How can clinicians reconcile that a LBWS of ≤ 3 is an end-stage wound, while a score of ≥ 3.5 points is a problem wound with more than an 80% likelihood of healing? To do such, providers must label wounds with scores of 2.5 to 4 points as transition zone wounds. Before making a decision to salvage versus recommending lower limb amputation in the transition zone wound, clinicians who use this LBWS system utilize information from the Wellness and Goal Scores (Figure 3).

The Wellness and Goal Scores are both 0- to 10-point scores based on 5 assessments analogous to the format of the LBWS. These scores quantify the ability to perform activities of daily living, mobility, comorbidities, smoking/steroid use, neurological deficits, comprehension, compliance, motivation to avoid major amputation, support from others, and insight. Scores over 4.5 points for each score justify salvage for the transition zone wound. The authors have observed that about 50% of patients in the transition zone category who meet the criteria for limb salvage heal primarily from their surgical procedures.¹⁸ Of the 50% whose wounds dehisce or remain infected after surgery, 90% eventually avoid lower limb amputations with optimal outpatient wound management. Wounds in the transition zone were not separately analyzed; that is, LBWS that were < 3 points were analyzed as an end-stage wound while those that scored > 3.5 points were analyzed as problem wounds without consideration for the ancillary Wellness and Goal Scores.

Although the data are analyzed from 85 patients prospectively, the numbers from any single LBWS were small, especially for the healthy and end-stage categories. With inclusion of outpatients with wounds, the investigators anticipate a larger proportion will be in the healthy category, which may improve the predictive values and accuracy of the LBWS. The small number of end-stage wounds is attributed to vascular surgeons primarily managing patients with ischemic wounds. If revascularizations failed or were not feasible, they did lower limb amputations without being included in this study. 

Another limitation to this study is that wound care was not specified. While most patients were managed by the first author, it was not the purpose of the study to specify specific treatment protocols but rather to grade and categorize wounds based on the initial evaluation and observe outcomes regardless of the management. 

In addition, time intervals between the initial evaluation and the final outcome grading were not considered. Times ranged from 2 weeks after minor (toe, forefoot, midfoot) surgeries to 1 year or more in patients whose wounds were not completely healed but were chronic, stable, and easily managed. 

The question of whether it is appropriate to include improvement but not complete healing as a good result in this analysis also may be considered a limitation. Based on MCII measures, the investigators feel it is appropriate to do such. Conversely, they included wounds that did not show changes in LBWS as poor results, although they could have been placed in a separate outcome result and perhaps made the statistics more robust. 

Lastly, it sounds almost contradictory that the purpose of this study was to predict outcomes based on LBWS by categorizing wounds but not recommending management. Yet, the categories healthy, problem, and end-stage offer immediate guidance as to what interventions are needed. For example, healthy wounds require only the simplest of interventions, while problem wounds require a strategic management approach to address the components of the Troublesome Triad.¹⁴ Strategic management includes 5 tactics: (1) management of the wound base with debridements, ostectomies, and/or amputations; (2) wound protection and stabilization; (3) optimal medical management of the patient’s comorbidities; (4) selection of appropriate wound dressing agents; and (5) mitigating wound ischemia-hypoxia.¹⁹End-stage wounds almost always require revascularizations if lower limb amputations are to be avoided.

In summary, the LBWS is a versatile tool that makes it possible to grade wounds easily and accurately whether in the foot and whether the patient has diabetes. The LBWS integrates perfusion, depth, and infection information from 3 predominantly used DFU and pressure injury scoring systems while adding the assessments of appearance of the wound base and size. The 0- to 10-point LBWS scoring system is intuitively obvious and uses a continuum of objective findings, with 2 points being best possible and 0 points being worst grade for each assessment. From the LBWS, objectivity is achieved for categorizing wounds as healthy, problem, or end-stage. From these categories, interventions for managing each category become obvious. The initial LBWS predicted good outcomes for healthy and problem wounds (LBWS of ≥ 3.5) and poor outcomes for end-stage wounds (LBWS of ≤ 3) with 75.3% accuracy in this study. While other wound scoring systems have face validity through their use, the LBWS is the first to have both comprehensive reliability and validation data.¹ The LBWS facilitates CER for wound care interventions by comparing outcomes with wounds having similar LBWSs. Finally, the LBWS is a tool for quantifying MCII with serial scorings of the wound. 

Affiliations: Long Beach Memorial Medical Center, Long Beach, CA; California State University Long Beach, Long Beach, CA; and VA Long Beach Healthcare System, Long Beach, CA

Correspondence: Long Beach Memorial Medical Center, Orthopaedics, 2801 Atlantic Avenue, Long Beach, CA 90801; straussdrs@gmail.com 

Disclosure: The authors disclose no financial or other conflicts of interest.