Longitudinal Study of Stage III and Stage IV Pressure Ulcer Area and Perimeter as Healing Parameters to Predict Wound Closure
Documentation of healing progress is central to the plan of care for patients with a pressure ulcer. Several studies have shown that a reduction in wound area is a predictor of chronic wound healing, but data about pressure ulcers are limited.
Furthermore, consensus is lacking as to which wound characteristics such as volume, area, and perimeter should be measured and what methods or tools to use when collecting measurements. This hampers comparisons among research studies and their eventual translation into clinical practice. The purpose of this longitudinal, repeated measurements study was to calculate healing parameters using wound area and perimeter measurements and evaluate their potential to predict closure. Twenty-seven (27) patients with 31 Stage III and Stage IV pressure ulcers participated in the 42-day study.
Wound length, width, and perimeter were measured at 15 time points or until healing, and the following healing parameters were calculated: absolute area, percent area reduction, mean percent area reduction, trajectory, and three variations of the linear healing parameter. Ulcer size at day 0 was a significant predictor of time to heal (P = 0.0231). Smaller wounds required less time, but initial size did not influence wound outcome (P = 0.3537). Among ulcers that closed 81% or more of their original area, the initial linear healing rate (4 weeks) was 0.16 ± 0.02 cm/week and mean percent area reduction was 35.37% + 4.83, compared to 0.021 + 0.02 cm/week and – 4.66% + 6.24, respectively, for ulcers that closed 40% or less of their original area. Percent area reduction and linear healing parameters all were predictive of wound outcomes. Percent area measurements are easiest to determine but sensitive to initial wound size. The linear healing parameter requires calculation of both wound area and perimeter, but it is independent of initial wound size and yields rates directly comparable among wounds. These findings confirm that change in wound size after 4 weeks of care is a predictor of healing Stage III and Stage IV pressure ulcers. Future research studies should include other wound characteristics and patient comorbidities to further refine acceptable rates of wound closure.
Potential Conflicts of Interest: This work was supported by the Telemedicine and Advanced Technology Research Center (TATRC) at the US Army Medical Research and Material Command (USAMRMC) through award W81XWH-05-1-0401.
The goal of pressure ulcer research is to evaluate optimal methods to provide care and facilitate healing. Evaluation of treatment effectiveness is accomplished through assessment and documentation of changes in the pressure ulcer until closure. Complete closure or accelerated closure are often endpoints in clinical studies, but chronic wounds usually require extended periods of time for complete closure, making trials lengthy and expensive.1 Surrogate markers (early predictors of healing or delayed healing) would enhance the assessment of treatment effectiveness and allow practitioners to modify and tailor treatments according to the individual patient’s response.2
Researchers have suggested a range of potential predictors of pressure, venous, and diabetic foot ulcer healing, including perfusion and dietary protein3; ulcer stage and wound area4; wound width, length, area, and perimeter5; and rate of healing. 6 Healing assessment tools such as the Pressure Ulcer Scale for Healing (PUSH© Tool) or the Bates-Jensen Wound Assessment Tool (BWAT) utilize exudate, tissue type, and periwound characteristics in addition to wound area to evaluate wound healing progress. 7-10 More recently, proteome analysis for potential biomarkers has been suggested as a potential method to assess wound healing. 11-13
Wound area, volume, perimeter, and linear margin advance all are used as parameters to assess healing. However, a consensus for which measurements are most effective to assess healing is lacking. 14,15 Most commonly, wound area measurements are used to track healing progress. Wound area changes can be tracked using raw data or transformed into percentages. Area expressed as a percentage may be calculated relative to the first or baseline measurement (trajectory) or as a percent relative to the previous measurement or clinic visit — ie, percent area reduction (PAR). Gowland et al16 used a linear healing parameter (LHP) to measure the daily advance of the wound margin toward the wound center in venous leg ulcers. The LHP uses a combination of area and perimeter measurements and compares values between each wound assessment. The LHP also has been calculated relative to baseline or the initial measurement as well as between consecutive weeks. Gowland et al,16 demonstrated that a LHP is compromised only when applied to very small or rapidly healing wounds where the linear advance exceeds the perimeter. Gilman17,18 employed the LHP to study closure in chronic wounds of varying size and shape. Refinements of the LHP include a mean adjusted method to reduce week-to-week variations.19 In a longitudinal study20 of 39 patients with 49 venous stasis ulcers observed for an average of 49.3 days and a retrospective analysis21 of healing rates for 10 Stage IV pressure ulcers in nine patients, wound area measurements were found to be sensitive to original wound size. A study22 of 160 patients with diabetic neuropathic foot ulcers observed for 20 weeks found that sensitivity to original wound size may be partially accommodated by using the trajectory parameter, which standardizes changes in wound area by expressing them as a proportion of the original wound area. The LHP uses perimeter measurements to remove the influence of wound size and shape. Methods to measure and assess wound healing have been the subject of recent reviews.23-25
Currently, studies suggest that healing progress measured by surface area changes may be the best predictor of wound closure, especially the initial healing rate (IHR) or changes in wound size during the first 2 to 4 weeks after injury or alteration in wound care. In a retrospective analysis of 165 patients observed for 12 weeks, Phillips et al26 found that the baseline venous ulcer area and duration of the wound were predictors of complete healing and time to heal. Venous ulcers that were at least 44.1% healed at week 3 could have outcomes predicted accurately (77%). In a prospective study4 of venous ulcers, a 30% or greater reduction in wound area within 2 weeks was predictive of healing (P = 0.002). Retrospective analyses2,27 using the Curative Health Services database that involved 56,488 venous leg ulcers and 39,918 diabetic neuropathic foot ulcers found that log healing rate, log wound area ratio, and percentage change in wound area (trajectory) at 4 weeks were predictive of healing at 12 or 24 weeks. A prospective, randomized controlled trial28 found that among 203 diabetic foot ulcers, a 53% reduction in area after 4 weeks was predictive of healing at 12 weeks. In a prospective study29 that included 48 patients with 56 Stage III or Stage IV pressure ulcers, trajectory, percent reduction in wound size, at 2 weeks was used successfully to predict time to healing.
Utilizing a mean-adjusted healing rate in a study of 14 patients with venous ulcers observed for 24 weeks, Tallman et al19 suggested that healing can be predicted within 3 weeks of treatment initiation. Cardinal et al30 assessed wound margin advance, IHR, percent wound surface reduction, and wound healing trajectories at 4 weeks. Their retrospective analysis of 306 patients with venous leg ulcers observed for 24 weeks and 241 patients with diabetic foot ulcers observed for 32 weeks found all parameters were predictors (P <0.001) of complete healing by 12 weeks. Kantor et al’s31 retrospective cohort study found rate of healing was not predictive of closure in 104 venous ulcers observed for 24 weeks; instead, change in area from baseline to week 4 was predictive. In a longitudinal study32 evaluating the efficacy of hydrocolloid dressings, a 30% or greater reduction in area was predictive of time for healing, as well as treatment outcome, for 61 patients with 72 full-thickness leg ulcers observed for an average of 56 days. Four-week percent wound area reduction was predictive of healing in a 16 week randomized clinical trial33 evaluating standard moist wound therapy versus negative pressure wound therapy among diabetic foot ulcers following partial amputation. A prospective study34 of 704 diabetic foot ulcers found a 50% reduction in surface area within 4 weeks is associated with a higher long-term probability of healing and should be considered a valid surrogate end point for treatment efficacy. In a longitudinal study35 evaluating the efficacy of a hydrocolloid dressing among 119 patients with 153 full-thickness pressure ulcers, wounds that did not decrease 45% in area after 2 weeks or 77% after 4 weeks were significantly (P < .01) less likely to heal during the study. In a retrospective evaluation36 of the Pressure Sore Status Tool (PSST), among a subset of 80 pressure ulcers that closed, 76% had a 41% decrease in size at week 2, which supports the findings of other researchers.
These studies have led to attempts to develop predictive healing trajectories and mathematical or statistical models of healing in both acute and chronic wounds.37-41 Numerous studies have compared healing time to variables such as stage and initial size, but as yet no minimum time to healing has been specified for pressure ulcers regardless of their initial wound status. In a prospective multicenter study, Bolton et al42 found that partial-thickness pressure ulcers (Stage II) and venous ulcers healed in half the time of full-thickness venous ulcers or pressure ulcers (Stage III and Stage IV), providing general outcome information for healing time.
The purpose of this longitudinal, repeated measures, clinical study was to: 1) calculate parameters of pressure ulcer healing using wound area and perimeter, and 2) evaluate these parameters for their potential to predict wound closure for pressure ulcers.
Material and Methods
Population. All study participants were inpatients recruited from long-term skilled nursing facilities. Institutional Review Board approval for the study and consent from Daemen College and the US Army Medical Research and Materiel Command (USAMRMC) Human Research Protection Office were obtained. Potential volunteers were first identified and approached by the skin team of each facility; persons interested in learning about the study met with a clinical researcher. Residents were eligible to participate if they had a pressure ulcer for a minimum of 4 weeks and if the ulcer was not currently treated with negative pressure wound therapy (NPWT), enzymatic debriding agents, topical growth factors, or dressings containing proteins. After obtaining written informed consent, resident gender, age, ethnicity, height, weight, smoking and alcohol habits, allergies, and comorbidities including but not limited to hypertension, diabetes, and coronary artery disease were noted. Subjects were assigned study identification codes to protect patient confidentiality and pressure ulcer location, stage, duration, current dressing, and care protocol were recorded.
Wound data. The duration of the study was 6 weeks, and study participants were seen on days 0, 1, 2, 3, 4, 7, 8, 9, 10, 11, 14, 21, 28, 35, and 42. At each visit, wounds were digitally photographed. Images were analyzed and measured using VEV MD Wound Measurement Software (Vista Medical Ltd., Winnipeg, Manitoba, Canada). This software, which employs a stereophotogrammetry technique, has been shown to be more accurate and less biased than linear measurement or planimetry. 43 The accuracy of the software provides an objective technique for wound evaluation, measurement, and tracking.22,43-45
The final change in wound area (or area measured on the last day of participation for those wounds that closed or withdrew from the study) expressed as a percentage of area on day 0 (trajectory) was used to place patients in one of three categories for wound outcome: healed (wounds with a decrease in area of 81% to 100%), moderate healing (wounds that exhibited a 40% to 80% decrease in area), or delayed healing (wounds that increased in area or decreased <40%).
Data analyses. The relationship between original ulcer size and time to healing was tested using a General Linear Model (GLM) in SAS (Proc GLM, SAS Institute 2002, Cary, NC). GLMs are widely accepted as a good way to test the effects of both categorical and continuous independent variables on continuous dependent variables. These models are used because they are robust and support the modeling of interactions between and among independent variables. The time to heal was defined as the day on which the wound area was reduced by 81% or more of its original area; 10 healed wounds were available for analysis. Initial wound size was tested to determine if small wounds were more likely to heal than large wounds. The size of wounds at day 0 was tested against outcome in a logistic regression analysis using SAS (Proc Logistic, SAS Institute 2002). To determine if large wounds close faster than small wounds or vice versa, a survival analysis in SAS (SAS Institute 2002) was employed. A correlation between initial ulcer size and average daily healing rate expressed as cm2/day was tested using a GLM in SAS (Proc GLM, SAS Institute 2002).
Healing parameters. Wound surface area measurements were used to calculate parameters to assess progress of healing among pressure ulcers. Seven parameters of healing using wound area and perimeter were analyzed to compare their potential for predicting closure (see Table 1): absolute area, PAR, mean PAR (MPAR), trajectory, and three variations of the LHP. All can be described as a rate of healing by dividing the parameter by the time between wound assessments (1 week) and can be calculated relative to day 0 or the previous clinic visit. The authors defined PAR as the change in percent of wound area calculated relative to the previous wound assessment. Wound trajectory is PAR standardized to baseline or day 0 values. MPAR was calculated by averaging values to reduce week-to-week variation. The LHP uses both area and perimeter measurements to calculate the advance of the wound margin toward the wound center. The formula was calculated relative to baseline (BLHP) and between weekly clinic visits (WLHP), as well as a mean adjusted formula (MLHP) that averages weekly values. Table 1 summarizes the formulae used for each parameter. Mean and standard error were calculated weekly (SAS, SAS Institute 2002) for each parameter and outcome.
Outcome assignment predictability. The predictive ability of these parameters to correctly assign outcome using up to 4 weeks of data was tested with both two- and three-outcome models using Discriminant Function Analysis (DFA). DFA is a statistical method used to predict categorical outcomes based on continuous predictor variables. It is robust and can provide a model for classification of new data. The DFA was performed with SYSTAT (SYSTAT version 7.0 SAS Inc). Potential predictors included parameter values at weeks 1, 2, 3 and 4. Predictors were tested for colinearity before combination in DFA analysis. The three-outcomes model included healed, moderate healing, and delayed healing, while the two-outcomes model combined moderate healing and delayed healing into a single category: chronic. Models were evaluated based on the percent of correct predictions in a resubstitution matrix using that model.
Percent healing predictability. The ability of these parameters to correctly predict percent healing using up to 4 weeks of data was tested with both two- and three-outcome models using a GLM. For the GLM models, time points tested included 2 weeks and 4 weeks or IHR. The IHR for absolute area is represented as the mean of the first 4 weeks. Baseline adjusted parameters (trajectory and BLHP) use the week 4 value. For weekly parameters, mean adjusted MLHP and MPAR were tested. The GLM analysis of predictive power was tested for each parameter independently using the percent closure (trajectory) at week 6 as a dependent variable. The value of each parameter at week 2 and week 4 (the initial healing rate) were modeled. The cut-off value to reach 40% or 81% of the original wound size was calculated for significant models.
Thirty-four residents (10 men, 24 women, average age 72.3 years) with a total of 46 pressure ulcers were enrolled. Hypertension, diabetes, and coronary artery disease were present in 52.2%, 54.3%, and 30.4% of residents, respectively. Five residents (12 ulcers) withdrew from the study before 14 days, one resident and ulcer were excluded due to size (<0.6 cm2), and one resident and two ulcers were removed due to unclear etiology, leaving 27 volunteers with measurements from 31 pressure ulcers for wound surface area change analysis. This subset included nine men and 18 women (average age of 72.3 years). Pressure ulcers stages collected from medical records did not reflect the breadth of tissue damage upon enrollment in this study. When assessed at day 0 using the National Pressure Ulcer Advisory Panel staging system,46 the 31 ulcers included in healing parameter analyses were Stage III (74%) and Stage IV (26%). Hypertension, diabetes, and coronary artery disease were associated with 58.1%, 48.4%, and 22.6% of these 31 pressure ulcers, respectively.
In three instances, data substitutions were allowed to prevent exclusion of the wound data from statistical analyses due to the lack of a single data point. A day 35 value actually was collected on day 36, and one day 8 value and one day 10 value were used for missing day 7 data.
The average area of all pressure ulcers was 10.97 ± 2.09 cm2 (range 0.62 to 35.5 cm2) upon enrollment. Ulcers included in the healed category had an initial size of 8.88 ± 3.3 cm2 (range: 0.62 to 30.8 cm2) and closed 94.6% ± 2.0% of their initial surface area after an average of 38 days. Two subjects achieved closure before 42 days. Five ulcers included in the moderate healing outcome group had an initial size of 9.76 ± 5.16 cm2 (range 0.8 to 24.54 cm2) and closed 51.2% ± 4.2% of their original surface area after an average of 36 days. Two subjects in the moderate healing outcome group withdrew from the study (week 3 and week 5) before completion. Sixteen ulcers exhibited delayed healing. Their average initial size was 12.66 ± 3.21 cm2 (range 0.88 to 35.5 cm2), and they were followed for an average of 35 days. Six subjects in the delayed healing outcome group withdrew prematurely. Nine ulcers in this outcome group increased in size over 6 weeks, and wound closure ranged from a 36.8% decrease to an almost 400% increase in wound area. Outcomes included 10 healed, five moderate healing, and 16 delayed healing ulcers as defined previously using wound surface area changes. The mean and standard error for each parameter and time point are shown in Figure 1. The trajectory, BLHP, MPAR, and MLHP parameters show separation of outcomes without the need for statistical modeling. Weekly calculated PAR, WLHP, and absolute area changes varied widely and overlap between outcomes using untransformed data. Among wounds with a healed outcome, two had nearly 0 but negative LHPs (-0.05 and -0.08) for week 1, but all remaining weeks had positive LHPs. The IHR (4-week) for all parameters by wound outcome is presented in Table 2.
Original ulcer size and time to healing as defined by a decrease in area of at least 81% of the initial size were correlated. Ulcer size at day 0 was a significant predictor of time to healing (P = 0.0231); smaller wounds required less time. Initial size did not influence wound outcome (P = 0.3537). The relationship between wound size and rate of closure using trajectory data was investigated using a survival analysis; however, preliminary tests were inconclusive, most likely a factor of the low N value for these tests. Average daily healing rate calculated using the change in absolute area and expressed as a daily rate was positively correlated with initial wound size (P <0.000). Larger wounds had higher average daily healing rates than smaller wounds.
Parameter values at weeks 1, 2, 3, and 4 were used in various combinations to predict wound outcome using a DFA. Parameters are presented in Table 3. The two-outcomes approach (wounds classified as healed or chronic) had higher predictive ability than the three-outcomes approach (wounds classified as healed, moderate healing, or unhealed) for all models. In general, increasing the number of predictors or using variables from later weeks increased the accuracy of the models. Significant GLM models were found for all parameters except absolute area. Models using week 2 and week 4 (the initial healing rate) data to predict wound outcome are shown with the resulting minimum values required at each week for inclusion in the moderate healing and healed outcomes (see Table 4).
The authors believe pressure ulcer treatment would benefit from an easily measured parameter that is positively correlated with eventual wound healing or closure. In addition, the goal of complete wound closure often is not achieved during research studies or clinical trials because of protracted healing times for chronic ulcers and the time and financial constraints imposed by funding agencies, public and private. The complex etiology of chronic wounds makes the selection of appropriate parameters to measure and serve as surrogate markers difficult, although progress has been made. 15,27 In order to compare the efficacy of different interventions, appropriate endpoints or outcomes are necessary to design trials.
Outcome classification. The categorization of pressure ulcer outcomes into three groups (healed, moderate healing, and delayed healing) was designed to capture potential proteomic changes11 and has been utilized in other wound healing studies. 47 Not all wounds closed during the 42-day study, but they also did not significantly increase in size or exhibit delayed healing. Although it complicates predictive modeling efforts, categorizing wound healing into three outcomes rather than two may help distinguish helpful therapeutics and/or wound care procedures. The parameter plots of both the healed and delayed healing groups are somewhat intuitive. The high variability associated with each time point in the delayed or chronic group supports the absence of a single factor in delayed healing. In fact, the delayed healing process is different for each patient and the data reflect this wide scope. Chronic wounds with comorbidities are difficult to describe mathematically, and although some wounds respond rapidly to treatment and continue until closure, others have a significant initial decrease in size but are delayed in final closure.48
Wound depth consideration. Pressure ulcers are divided according to depth of tissue damage. Staging is not used to assess healing; rather, it is a descriptor of the wound appearance on initial examination.36 Wound stages are not reassigned during wound healing as the depth decreases. Wound depth changes are reflective of new granulation tissue as it fills a wound from the base upward.49 Wound depth alone is not commonly used as a parameter to measure healing, but is combined with area as an indirect estimate of ulcer volume. Utilizing volume as a parameter for predicting chronic wound outcome is uncommon due to the inability to easily obtain measurements. The typical volume measurement is based on the depth at a single location in the wound, and the depth at that site may vary dramatically from the rest of the wound bed, resulting in substantial measurement inaccuracies.6 For wounds with significant depth, wound depth may be better recorded independent of area as an indication of vertical healing. This would avoid the error in volume associated with depth changes across the wound bed.50,51 Because wounds heal from the bottom up, changes in depth appear as a delay in healing and are eventually accounted for by area.18 In this study, two-dimensional area and perimeter measurements have been used successfully to track healing progress and predict outcome without the need to quantitatively measure depth or volume of a wound.
Obtaining measurements. The potential for a simple area measurement to be used to predict outcome or impact of a current treatment protocol could prove consequential. Several methods are used to measure wounds, including rulers, acetate tracings, planimetry, stereophotogrammetry (SPG) and structured light technique. Considerations to maximize accuracy and consistency of wound size measurements were recently summarized.9,49 The current study used images captured by digital cameras and the VeV analysis software to calculate wound area and perimeter. Analysis of wound area using calibrated digital images by trained individuals is reliable and reproducible.52,53 Limitations include error associated with defining the wound margin, body position, and lighting inconsistencies.5,54 Many studies have compared the results of wound measurements taken by rulers, planimetry, and SPG.43,52,55,56 Length and width ruler area of wounds was correlated with planimetry area measurements, but the correlation weakens as wounds exceeded 40 cm2 or become irregular in shape.5,56 Unfortunately, the common practice of using length and width of a wound bed to measure the area of a wound makes accurate measurements difficult to obtain. Simply put, few wounds are perfect rectangles.
Predictability of data. Using untransformed absolute area data has limited utility to predict wound outcome, and current study results confirm this trend.27 Absolute area used as a parameter to assess healing is correlated to initial wound size, making accurate comparison of different or unbalanced size class groups impossible.21 A percent transformation to baseline values (wound trajectory) is a common parameter used to track wound closure.20 Healing trajectories for 100% closure and nonclosure of diabetic foot ulcers over 20 weeks treated at different centers appear similar.22 Both trajectory and PAR are sensitive to initial wound size, especially small wounds where changes can translate to large percentage increases.30 Comparisons of a wide range of wound sizes using trajectory is contraindicated.18 The findings in the current study support the use of the trajectory parameter to predict wound outcome. A limitation of trajectory is its single-study data dependency and specificity.18 Although the basis for this dependency is legitimate, a number of studies using the trajectory approach all have reached similar conclusions, supporting the findings of this healing parameter.
Standardizing area changes as a percentage of the previous measurement (PAR) is not common but has advantages over using a trajectory approach. The strength of the PAR model in predicting outcome is that by utilizing percent area changes between visits and not previous week’s data in the analysis, data from the more recent time points are not confounded by the baseline or initial wound size. Results of a retrospective study30 have shown that the average PAR for the first 4 weeks was predictive of healing for venous leg ulcers and diabetic foot ulcers. In another retrospective cohort study,31 the average PAR for the initial 4 weeks was not tested; however, individual PAR of week 1–2 and week 3–4 distinguished the healed from the nonhealed venous leg ulcers but was not predictive of wound outcome. In the current DFA analysis, PAR models for three outcomes performed worse than trajectory but two-outcome models were similar in predictive ability for both PAR and trajectory parameters. MPAR GLM models required an approximate 20% reduction in wound surface area per week to be included in the moderate healing outcome, but for wounds with healed outcomes, the values were doubled. Similarly, trajectory GLM model minimum values for healed wounds were double those of moderately healing ulcers.
The LHP was proposed by Gilman et al17 to eliminate the effect of original wound size and shape and make healing rate for wounds of all etiologies directly comparable. It is especially recommended when wounds of various sizes are included in a single analysis. Many analyses have confirmed the correlation between healing rates using raw wound surface area measurements and initial wound size and the lack of correlation when using the LHP.20,57,58 The advantages of the LHP were recently assessed.18,23,24,48 In brief, a comparison of the healing rates of two wounds, one small and one large, using absolute area, percent area (trajectory), and LHP results in an inflated rate of closure for the larger wound, an inflated rate of closure for the smaller wound, or equal rates of closure for both, respectively.
Margolis et al58 used Gilman et al’s method to calculate the initial LHP with the first week and the fourth week used as endpoints (baseline adjusted, BLHP) in a study of 23 patients with 27 chronic venous leg ulcers observed for an average of 8.3 weeks. Among wounds that healed, the initial BLHP was positive. Importantly, a positive initial BLHP did not guarantee a good outcome, because some wounds that failed to heal also had positive initial BLHP. Margolis noted variable rates for the BLHP between weeks and suggested 4-week intervals for calculations.
Tallman et al19 proposed a modification of Gilman’s formula to eliminate observed healing rate instability, the MLHP, an average of weekly LHP (WLHP) values. Importantly, the WLHP compares each visit to the previous clinic visit rather than the initial or baseline visit. Tallman compares a BLHP to his MLHP. Tallman’s MLHP for 15 venous ulcers was found to be predictive of outcome as early as 3 weeks after treatment initiation, while the baseline method was not predictive. Hill et al59 tested the difference between the MLHP and the BLHP in a retrospective analysis of 17 patients with venous leg ulcers observed for 12 weeks. The calculated rates for both methods were almost identical. The BLHP and MLHP for the initial 4 weeks did not predict healing well, but for all patients that failed to heal, the initial BLHP and MLHP were negative. Both the MLHP and BLHP for the initial 4 weeks were predictive of venous leg ulcer 12-week outcomes.30
For the current analysis, the BLHP and MLHP were similar, but outcome trends are easiest to recognize in the raw data for the MLHP (see Figure 1). Similar to results from Margolis et al58 and Tallman et al,19 healing rate instability was observed, but these week-to-week differences among the outcome groups were diminished by the MLHP. Among chronic wounds, week-to-week changes in linear healing rate may reflect the comorbidities and complications common to many patients.17 All healed outcome wounds had IHRs that were positive for the BLHP and MLHP; however, a positive IHR did not guarantee a healed outcome (see Table 2). Among ulcers that closed 81% or more of their original area within 42 days, the IHR (4 weeks) was 0.16 ±0.02 cm/week, within the range of reported rates for other healing chronic diabetic, venous, and pressure ulcers.19,20, 30,51,58-61
At present, support is overwhelming for a 2- to 4-week allowable response period to assess wound healing progress for the purpose of treatment modification. That support is available for venous leg, diabetic foot, and pressure ulcers where healing is monitored using the LHP19,30,58-61 and trajectory parameters.2,27,28,30-35,62 Very few reports have compared multiple healing parameters on the same data set27,30,31,40,59 to establish which are best to track healing. This research confirms previous findings regarding percent change at 2 to 4 weeks as a predictor of Stage III and Stage IV pressure ulcer outcome.35 For an individual wound, the exact parameter used to measure the wound is equally as important as its consistent application over the weeks or months required to reach closure.9 However, in order to compare wound healing progress across wound types and scientific studies, the parameter used to measure wounds is as important as the methods used to identify the wound margin and to collect measurements.54 Only the LHP as described by Gilman et al17 yields a rate that is independent of initial wound size and shape and comparable among differing wound etiologies.
The surface area measurements used to test healing parameters were collected as part of a project to identify protein biomarkers of chronic pressure ulcers. The collection protocol, which included daily and weekly sampling for up to 6 weeks, was designed to investigate dynamic changes in protein concentration and therefore, not optimal for a temporal study of wound closure parameters. Not all ulcers were included in the temporal evaluation of wound healing parameters because of small initial wound size and study participant discontinuation. The data set used to evaluate parameters of healing had relatively low power (n = 31). Four individuals had two ulcers and both ulcers were included in the study and treated as independent wounds for statistical modeling purposes. No attempt was made to correlate comorbidities with wound outcome in the statistical models. In addition, all wounds were Stage III and Stage IV pressure ulcers, so it may not be possible to extrapolate the results to include ulcers that are less severe. The study duration was 6 weeks, an insufficient amount of time for most Stage III and Stage IV pressure ulcers to progress to complete closure. As a result, the inclusion criteria for an ulcer to be defined as healed was modified from 100% closure to 81% or more of its original surface area after 6 weeks. Despite these limitations, the results support a 4-week response period with healing progress measured using the LHP, trajectory, or percent area reduction.
vStudy data support the use of linear healing rates and percent area measurements to document healing progress in Stage III and Stage IV pressure ulcers. In a clinical setting, accurately recording validated wound healing parameters over a 4-week response time may help clinicians evaluate the effectiveness of the plan of care. Percent area measurements may be affected by initial wound size, especially for small wounds, but are simple to calculate and track in patient charts. Use of perimeter in LHP corrects for initial wound size and shape differences, but it requires slightly more sophisticated methodologies to calculate. Reports of LHP rates for chronic pressure ulcers are uncommon but necessary for validation of the parameter. Future research should include a larger data set for calculation of LHPs and area measurements. Inclusion of characteristics such as subject age, pressure ulcer stage, and wound number and duration, as well as comorbidities, may be useful to further refine acceptable rates of wound closure. A recognized range for rate of healing would provide an unequivocal standard for healthcare professionals and researchers to measure healing progress.
The authors thank Catholic Health System Partners in Rehab, ElderWood Health Care, and Kaleida Health for allowing them to enroll subjects from their facilities.
Dr. Edsberg is the Director of the Center for Wound Healing Research, Director of the Natural & Health Sciences Research Center, and Associate Professor, Natural Sciences Department; and Dr. Wyffels is Research Scientist in the Natural and Health Sciences Research Center, Center for Wound Healing Research, Daemen College, Amherst, NY. Dr. Ha is an adjunct assistant professor, Biology Department, Millersville University, Millersville PA. Please address correspondence to: Laura E. Edsberg, PhD, Director, Natural and Health Sciences Research Center, Center for Wound Healing Research, Daemen College, 4380 Main Street, Amherst, NY 14226-3592; email: firstname.lastname@example.org.
1. Food and Drug Administration (FDA) [website]. Center for Biologics Evaluation and Research (CBER). Guidance for industry: Chronic cutaneous ulcer and burn wounds — developing products for treatment. Updated June 1, 2006. Available at www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guid.... Accessed September 20,2011.
2. Margolis DJ, Gelfand JM, Hoffstad O, Berlin JA. Surrogate end points for the treatment of diabetic neuropathic foot ulcers. Diabetes Care. 2003;26(6):1696–700.
3. Allman RM. Outcomes in prospective studies and clinical trials. Adv Wound Care. 1995;8(4 suppl):61–64.
4. Arnold TE, Stanley JC, Fellows EP, Moncada GA, Allen R, Hutchinson JJ, Swartz WM, Bolton LL, Vickers CF, Kerstein MD. Prospective, multicenter study of managing lower extremity venous ulcers. Ann Vasc Surg. 1994;8(4):356–362.
5. Kantor J, Margolis DJ. Efficacy and prognostic value of simple wound measurements. Arch Dermatol. 1998;134:1571–1574.
6. Gentzkow GD. Methods for measuring size in pressure ulcers. Adv Wound Care. 1995;8(4 suppl):43–45.
7. Bates-Jensen BM, Vredevoe DL, Brecht M-L. Validity and reliability of the Pressure Sore Status Tool. Decubitus. 1992;5(6):20–28.
8. Barber S. A clinically relevant wound assessment method to monitor healing progression. Ostomy Wound Manage. 2008;54(3):42–49.
9. Flanagan M. Improving accuracy of wound measurement in clinical practice. Ostomy Wound Manage. 2003;49(10):28–40.
10. Thomas DR, Rodeheaver GT, Bartolucci AA, Franz RA, Sussman C, Ferrell BA, Cuddigen J, Stotts NA, Maklebust J. Pressure ulcer scale for healing: derivation and validation of the PUSH tool. Adv Wound Care. 1997;10(5):96–101.
11. Wyffels JT, Fries KM, Randall JS, Ha DS, Lodwig CA, Brogan MS, Shero M, Edsberg LE. Analysis of pressure ulcer wound fluid using two-dimensional electrophoresis. Int Wound J. 2010;7(4):236–248.
12. Broadbent J, Walsh T, Upton Z. Proteomics in chronic wound research: potentials in healing and health. Proteomics — Clinical Applications. 2010;4(2):204–214.
13. Edsberg LE. Proteomic approaches for studying the phases of wound healing. In: Gefen A, ed. Bioengineering Research of Chronic Wounds. Berlin Heidelberg: Springer-Verlag;2009.
14. Cooper DM. Clinical assessment/measurement of healing; evolution and status. Clin Materials. 1991;8:263–271.
15. Mani R. Science of measurements in wound healing. Wound Repair Regen. 1999;7(5):330–334.
16. Gowland Hopkins NF, Jamieson CW. Antibiotic concentration in the exudate of venous ulcers: the prediction of ulcer healing rate. Br J Surg. 1983;70(9):532–534.
17. Gilman TH. Parameter for measurement of wound closure. WOUNDS. 1990;2(3):95–101.
18. Gilman T. Wound outcomes: the utility of surface measures. Lower Extremity Wounds. 2004;3(3):125–132.
19. Tallman P, Muscare E, Carson P, Eaglstein WH, Falanga V. Initial rate of healing predicts complete healing of venous ulcers. Arch Dermatol. 1997;133(10):1231–1234.
20. Gorin DR, Cordts PR, LaMorte WW, Manzoian JO. The influence of wound geometry on the measurement of wound healing rates in clinical trials. J Vasc Surg. 1996;23(3):524–528.
21. Brown GS. Reporting outcomes for Stage IV pressure ulcer healing: a proposal. Adv Skin Wound Care. 2000;13(6):277–283.
22. Robson MC, Hill DP, Woodske ME, Steed DL. Wound healing trajectories as predictors of effectiveness of therapeutic agents. Arch Surg. 2000;135(7):773–777.
23. Donohue K, Falanga V. Healing rate as a prognostic indicator of complete healing: a reappraisal. WOUNDS. 2003;15(3):71–76.
24. Jessup RL. What is the best method for assessing the rate of wound healing? A comparison of 3 mathematical formulas. Adv Skin Wound Care. 2006;19(3):138–147.
25. Ahn C, Salcido RS. Advances in wound photography and assessment methods. Adv Skin Wound Care. 2008;21(2):85–93.
26. Phillips TJ, Machado F, Trout R, Porter J, Olin J, Falanga V. Prognostic indicators in venous ulcers. J Am Acad Dermatol. 2000;43(4):627–630.
27. Gelfand JM, Hoffstad O, Margolis DJ. Surrogate endpoints for the treatment of venous leg ulcers. J Invest Dermatol. 2002;119(6):1420–1425.
28. Sheehan P, Jones P, Caselli A, Giurini JM, Veves A. Percent change in wound area of diabetic foot ulcers over a 4-week period is a robust predictor of complete healing in a 12-week prospective trial. Diabetes Care. 2003;26(6):1879–1882.
29. van Rijswijk L, Polansky M. Predictors of time to healing deep pressure ulcers. Ostomy Wound Manage. 1994;40(8):40–48.
30. Cardinal M, Eisenbud DE, Phillips T, Harding K. Early healing rates and wound area measurements are reliable predictors of later complete wound closure. Wound Repair Regen. 2008;16(1):19–22.
31. Kantor J, Margolis DJ. A multicenter study of percentage change in venous leg ulcer area as a prognostic index of healing at 24 weeks. Br J Dermatol. 2000;142(5):960–964.
32. van Rijswijk L. Full-thickness leg ulcers: patient demographics and predictors of healing. Multi-Center Leg Ulcer Study Group. J Fam Pract. 1993;36(6):625–632.
33. Lavery LA, Barnes SA, Keith MS, Seaman JW, Jr., Armstrong DG. Prediction of healing for postoperative diabetic foot wounds based on early wound area progression. Diabetes Care. 2008;31(1):26–29.
34. Coerper S, Beckert S, Kuper MA, Jekov M, Konigsrainer A. Fifty percent area reduction after 4 weeks of treatment is a reliable indicator for healing--analysis of a single-center cohort of 704 diabetic patients. J Diabetes Complications. 2009;23(1):49–53.
35. van Rijswijk L. Full-thickness pressure ulcers: patient and wound healing characteristics. Decubitus. 1993;6(1):16–21.
36. Bates-Jensen BM. The Pressure Sore Status Tool a few thousand assessments later. Adv Wound Care. 1997;10(5):65–73.
37. Franz MG, Kuhn MA, Wright TE, Wachtel TL, Robson MC. Use of the wound healing trajectory as an outcome determinant for acute wound healing. Wound Repair Regen. 2000;8(6):511–516.
38. Wallenstein S, Brem H. Statistical analysis of wound-healing rates for pressure ulcers. Am J Surg. 2004;188(1A suppl):73–78.
39. Cukjati D, Rebersek S, Karba R, Miklavcic D. Modelling of chronic wound healing dynamics. Med Biol Eng Comput. 2000;38(3):339–347.
40. Cukjati D, Rebersek S, Miklavcic D. A reliable method of determining wound healing rate. Med Biol Eng Comput. 2001;39(2):263–271.
41. Payne WG, Bhalla R, Hill DP, Pierpont YN, Robson MC. Wound healing trajectories to determine pressure ulcer treatment efficacy. Eplasty. 2011;11:1–8.
42. Bolton L, McNees P, van Rijswijk L, de Leon J, Lyder C, Kobza L, Edman K, Scheurich A, Shannon R, Toth M. Wound-healing outcomes using standardized assessment and care in clinical practice. J WOCN. 2004;31(2):65–71.
43. Langemo D, Anderson J, Hanson D, Hunter S, Thompson P. Measuring wound length, width, and area: which technique? Adv Skin Wound Care. 2008;21(1):42–45.
44. Salcido R. The future of wound measurement. Adv Skin Wound Care. 2000;13(2):54, 56.
45. Eriksson G, Eklund AE, Torlegard K, Dauphin E. Evaluation of leg ulcer treatment with stereophotogrammetry. A pilot study. Br J Dermatol. 1979;101(2):123–131.
46. National Pressure Ulcer Advisory Panel, European Pressure Ulcer Advisory Panel. Prevention and treatment of pressure ulcers: clinical practice guideline. Washington, DC: National Pressure Ulcer Advisory Panel;2009.
47. Polansky M, van Rijswijk L. Utilizing survival analysis techniques in chronic wound healing studies. WOUNDS. 1994;6(5):105–158.
48. Margolis DJ. Wound healing assessment: the clinical utility of wound healing rates. Ostomy Wound Manage. 1994;40(8):20–27.
49. Sussman C. Wound measurements and prediction of healing. In: Sussman C, Bates-Jensen BM (eds). Wound Care: A Collaborative Practice Manual, 3rd ed. Philadelphia, PA: Wolters Kluver Health/Lippincott Williams & Wilkins:2007.
50. Johnson JD. Using ulcer surface area and volume to document wound size. J Am Podiatr Med Assoc. 1995;85(2):91–95.
51. Schubert V, Zander M. Analysis of the measurement of four wound variables in elderly patients with pressure ulcers. Adv Wound Care. 1996;9(4):29–36.
52. Thawer HA, Houghton PE, Woodbury MG, Keast D, Campbell K. A comparison of computer-assisted and manual wound size measurement.