Implementing the PUSH Tool in Clinical Practice: Revisions and Result

   Evidence-based medicine promotes the use of systems and interventions that have been shown to be effective, allowing clinicians to rely on solid evidence (when available) for treatment decisions. Payor sources and hospital systems need efficacy data to guide their selection of appropriate services to offer.

The Pressure Ulcer Scale for Healing (PUSH) tool was introduced in 1997 by the National Pressure Ulcer Advisory Panel (NPUAP). Four years later, Stotts and Rodeheaver¹ published a report describing how the tool was validated and why slight changes to the tool were made. They concluded that the PUSH tool should be used to "document healing of pressure ulcers as an alternative to reverse staging. Further refinement of the PUSH tool will likely result from prospective clinical studies... Additional testing is needed to confirm these preliminary findings."

   Instructions for using the PUSH tool and its original forms are available on the NPUAP web site (see Figure 1).² Despite the well documented need to provide evidence-based care, the PUSH tool does not appear to be widely used or researched. Reasons for this may include time limitations and/or perceived limitations of the tool itself.

   The purpose of this study was to refine the PUSH tool for use in clinical practice and to test the use of the amended forms.

Methods

   A prospective cohort study using the amended PUSH tool was conducted by wound teams in two long-term, acute care (LTAC) hospitals from January 1, 2001, to December 31, 2001. All wound care was directed by one general surgeon (MP) who specializes in wound care. Measurements were obtained by people designated at their respective hospitals as totally dedicated to wound care - at one hospital, the wound team is managed by a registered nurse and at the other location two physical therapists lead the wound care team. The PUSH tool scores were recorded for each patient on admission and again on discharge to document overall progress of healing for each patient.

   Data forms. Review of the original PUSH forms (see Figure 1) shows that a separate tracking chart for each wound is required. Because many patients have multiple wounds, use of a single form for each patient is helpful and more efficient. Therefore, updated PUSH forms were created for this study.

   Individual Data Form. The Individual PUSH Data Form 1 was used for each individual wound patient. This form allows all wound data for a single patient to be collected on one page. The bottom of Form 1 consists of the scoring grid from the original PUSH forms, but unlike the original, it does not require a separate page for each wound. This is important to increase efficiency in collecting the data. The total PUSH score per patient represents the sum of the individual wound PUSH scores and is recorded at admission and again at discharge. Two blank boxes on the form enable customization based on hospital preference. For example, they can be used to record admission risk assessment (such as Braden or Norton Score). The number of days a patient was in the hospital also can be recorded on this form.

   For the purpose of this study, the PUSH Healing Score, defined as, "the change in the patient's admission total PUSH score and discharge total PUSH score," was developed. The score represents the improvement (positive score) or worsening (negative score) in a patient's overall wound burden. The PUSH Healing Rate, the healing score divided by the number of days hospitalized, also was developed and included in Form 1. The PUSH healing rate captures differences, if any, in wound change as a function of time. For example, Mr. S is admitted on May 2 with two wounds. Wound A on his sacrum has an admission PUSH score of 10, and Wound B on his left heel has a PUSH score of 8. His total admission PUSH total is, therefore, 18. Mr. S is discharged May 31 with a Wound A PUSH score of 4 and a Wound B PUSH score of 6. Therefore, his discharge total PUSH is 10, his PUSH Healing Score is 8 (18-10) and his PUSH Healing Rate is 0.276 (8÷29). = 0.276. The PUSH Healing Score per ulcer is 8÷2 = 4. Form 1 contains simple reminder formulas for each calculation.

   Monthly Summary Form. Form 2 is a monthly report designed to compile and summarize the data from individual patient forms. All data from wound patients discharged during that month is included. Form 2 is used to give an overall monthly evaluation of the wound care at any given hospital. Line C provides the percentage of wound patients by month of discharge. Hospitals may use this information to record the prevalence of wound patients in the hospital. The middle section of the Monthly Summary calculates the healing score from the total change in the Admission and Discharge PUSH wound scores. The Healing per Ulcer Score is the healing score divided by the number of ulcers to provide an average change in PUSH score for each ulcer during the month.

   The Healing Rate score is the healing score divided by the number of inpatient days of patients with wounds, which yields an average of the change in the PUSH scores per day for each patient during their stay. As for individual patients, the PUSH Healing Rate score measures outcome as a function of time. Also added to this worksheet is information about the nosocomial rate of wounds, because this information is important to any wound program. The total number of patients discharged for the month (with and without wounds) is necessary to calculate the nosocomial rate.

   Finally, Form 2 gives an average admission PUSH score per patient, which is a rough estimate of the average wound burden in that hospital for the month. Like Form 1, Form 2 has calculations on the form to aid in completion.

Results

   A total of 374 patients with 989 wounds were included in the study. These patients represented 13,737 inpatient days of care with an average of 2.65 wounds each and an average PUSH score of 26 on admission and 16 on discharge. Thus, the average PUSH Healing Score was (26-16) = 10. The average number of inpatient days per patient was 36.7, resulting in an average change in PUSH score per day (PUSH Healing Rate) of (10÷36.7) = 0.27. The average change in PUSH score per ulcer or PUSH Healing Score per ulcer was (10÷2.65) = 3.77 (see Table 1).

Discussion

   How was the PUSH Score meant to be used? From the original PUSH forms, it appears the PUSH Score was meant to measure healing of individual wounds on an ongoing basis during patient hospitalization. One of the reasons for creating the PUSH tool was to provide an alternative to reverse staging; indeed, one of the original graphs plots PUSH totals for each wound at regular intervals. The author's study was designed to provide a broader picture. Given the extreme wound burden seen at these centers, calculating PUSH scores for each wound and each patient on a weekly or biweekly basis is very time intensive. This is presumably true elsewhere, considering the paucity of PUSH usage nationwide. Instead of using the PUSH to check progression of healing during patients' stay, in this study, PUSH scores for each patient were obtained only on admission and at discharge. The answer to the logical question, "Is the PUSH tool necessary to decide if an individual wound is not doing well?" is, of course, no. Failure of a wound to improve on any of the individual PUSH variables (size, drainage, quality of tissue) would alert a concerned wound team that something may be amiss. Most wound programs will do well if they concentrate on regularly measuring and documenting basic wound information (size and tissue type). Waiting until data are tabulated and graphed (as required when using the PUSH) to realize that a wound care plan may need to be revised or to document reasons for failure to heal would be inappropriate.

   How should the PUSH Score be used? The PUSH scores' greatest utility may be found more in a system model because it has been found to work well in quantifying patient healing over an extended time period for the purpose of monitoring overall healing outcomes of a wound program.

   Current use of data. Both LTAC facilities involved in this study now use PUSH data to make a monthly check of their wound programs. By tabulating monthly summaries of the various categories listed above, graphs are created and reviewed at monthly wound meetings to evaluate performance and help make decisions about resource allocation.

   Use of benchmark data. Merriam Webster's dictionary defines "benchmark" as a point of reference for measurement; also: standard. One of the potential benefits of using the PUSH tool is that a wound program can compare monthly (or yearly) patient outcomes with other wound programs, national averages, and/or against its own historical data. Before this study, however, no detailed published data were available to provide a benchmark for comparison. Benchmark usage is becoming more prevalent in all areas of medicine to assess quality of care, according to Joint Commission mandates.

   Using benchmark data to compare PUSH healing rates and PUSH healing scores per ulcer is the most obvious parameter in defining the efficacy of one wound program versus another. The utility of this approach is that the data are expressed in terms of healing rate per ulcer or per day, regardless of number of patients treated.

   For example, using the benchmarks from this study, data from a hypothetical facility ("New Hope Hospital") was entered onto the graph for illustrative purposes to discuss interpretation of these type graphs and their potential implications for wound programs. This study helped demonstrate how to use data from the two wound programs for benchmarking purposes.

   Monthly admission PUSH score total. The Monthly PUSH score total offers a glimpse of the trend in wound burden admitted to the hospital. The score may be increased by more patients with wounds, more severe wounds, or more per patient. Figure 2 indicates a wound program with fairly high volume (similar to the volume of the centers in this study) that became even greater as time progressed. This could be seen as a positive reflection on the program - as marketing and/or outcomes makes the program more widely known, volume increases. At the same time, higher volume might alert administrators that the program is growing and may need more clinical resources allocated.

   Average healing score per ulcer. The average healing score per ulcer is a measure of the efficacy of a program to heal wounds. This parameter is independent of the volume of the program. In Figure 3, the data closely compare to the benchmark data.

   Healing rate. Probably the most useful and intriguing of the available comparisons, healing rate not only examines outcome, but also the rate of improvement. Wound care is already one of the most expensive components of inpatient care, and the cost of wound care is closely tied to the individual patient's wound burden.³ In addition to healing rate, adding cost data to this updated PUSH methodology makes it possible to determine not only healing rates, but also healing rates per unit of cost. To use a physics analogy, it would be fair to say the power of a wound program would be defined by its healing rate, but the efficiency of a program would be based on the healing rate per unit of cost. This type data could, for instance, justify slightly more expensive dressings that could increase efficiency if they promoted more healing with less overall (labor) costs. Such information on efficiency is greatly needed by healthcare in general and wound care in particular.

   With more study, PUSH data combined with cost data also could be used to predict prospectively the overall wound-related costs for the care of any given patient with a PUSH score.

   Unfavorable results. Figure 4, which shows PUSH healing rate, is an example of a wound program that does not compare favorably to the benchmark data. The reason for the decreased rate compared to the benchmark could be less overall healing and/or too long a time frame. In clinical terms, less overall healing could represent factors such as poor nutrition, ineffective dressings, or poor selection of specialty surfaces. The increased time frame could be due, for example, to failure to surgically debride in a timely fashion or ineffective discharge planning. A look at the average number of days per patient would be the first step in discerning the factors responsible. Such data are readily available using this methodology.

   These graphs are fairly easy to create using the results of Form 2 entered into an Excel spreadsheet on a monthly basis. A software program also has been created that can take the basic PUSH wound data and calculate all of the data for the monthly summary as well as graph the charts used for monthly wound meetings.

Limitations

   PUSH methodology is arguably the best available at this time for comparing healing outcomes, but it will always require careful interpretation by thoughtful wound providers.

  The biggest limitation to interpreting the PUSH score for a wound program is that it does not identify factors that may be responsible for the outcomes. However, if the patient mix stays relatively unchanged, the trend of the PUSH score may offer some indication regarding the positive or negative affect of changes in protocol or practice. For instance, if a wound program changes the composition of the wound team, the ratio of nurses to patients, the type of surfaces provided, or the wound products available, the monthly PUSH data may be the most impartial and complete arbiter of the effects of these changes on outcomes.

  In comparing healing outcomes between different programs, the PUSH methodology, as demonstrated in this report, is not affected by the number of patients with wounds, but certainly may be less valid if the patient mixes are too dissimilar. For instance, if one program primarily included patients with respiratory failure on ventilators and another program primarily included young spinal injury patients, differences in healing might not be solely due to the care rendered. In this type of situation, the within-hospital trend is again a more valid comparator, and a more appropriate benchmark. Figure 5 provides an example of a hospital comparing monthly data against national and its own historical benchmarks. National benchmark data may someday be available if programs adopt standardized PUSH methodology and agree to share their results with a central database in return for receiving the national benchmark data. For now, the data from this study are presented as benchmark data to stimulate interest until more work is completed.

   Possible future use of data. Large hospital systems could conceivably use this data to compare wound-healing outcomes in different locations. The PUSH score is best utilized in any setting that routinely provides care that encompasses several weeks or more, such as LTACs, home care, or skilled nursing facilities. Realistically, the PUSH tool will most likely be used in LTACs because their personnel most commonly have the necessary expertise. The PUSH score is probably not useful for measuring outcomes over short time frames, making its use in acute care hospitals unlikely. Because wound care is poorly standardized, some measurement of efficacy is needed to better identify "centers of excellence" and to try to duplicate their systems and techniques. These data are the first nationally recognized benchmark on which to gauge the wound healing efficacy of an entire system.

   Can this update save the PUSH tool? The PUSH score was a good idea, especially with the move toward evidence-based medicine, but it has gone nowhere, at least in terms of widespread usage, additional research, and published reports of its usefulness. The reason for this is most likely related to the fact that it puts an additional strain on an already burdened healthcare system. Most programs are doing well even to measure wounds on a consistent basis, let alone allocate time for additional data collection and manipulation. While many facilities seem to realize the need for a wound specialist on the floor, few seem to understand that much can be learned by looking at the big picture. The updated forms used in this study simplify the data collection process and provide a framework for using the information (see Form 2).

Conclusion

   Wound healing is an amalgamation of many factors, including nutrition, offloading, wound care, and the effects of comorbidities and quality of nursing care. Wound science lags behind many other areas of medicine in terms of standardization of treatments and measurement of outcomes. This report describes the use of the PUSH tool on a large number of wound patients to quantify healing outcomes. The terms healing score and healing rate are defined and illustrated, and new forms are introduced to make using the PUSH tool more practical. The PUSH score is probably best suited for measuring overall patient and facility outcomes. Hopefully, this report will spur a new interest in its use.

Acknowledgments

The author expresses special thanks to the dedicated wound teams of IHS and LifeCare Dallas who did (and do) all the hard work.

1. Stotts N, Rodeheaver GT, Thomas DK. An instrument to measure healing in pressure ulcers: development and validation of the Pressure Ulcer Scale for Healing (PUSH). J Gerontol A. 2001;56(12):795-799.

2. National Pressure Ulcer Advisory Panel. PUSH Tool 3.0. Available at: http://www.npuap.org. Accessed July 15, 2003.

3. Pompeo M. The role of 'wound burden' in determining the costs associated with wound care. Ostomy/Wound Management. 2001;47(3):65-71.