Outcomes Research—Wound Healing Outcomes: The Impact of Site of Care and Patient Stratification

Disclosure: Dr. Ennis is on the Medical Advisory Board for Accelecare Inc and Celleration Inc. No funds were received from any sources to perform this study or to prepare this manuscript.

Healing rates have become both clinical and marketing tools for many wound care centers. The recording and reporting of clinical results are imperative in today’s health care marketplace. Published articles, however, rarely stratify a patient’s risk in outpatient settings. Patients not seen in clinic for 30 consecutive days are considered lost to follow-up and that clinical episode is closed out from the data set. For example, a patient presents for a consult with a venous leg ulceration of 1-year duration that was unsuccessfully treated in a primary care physician’s office. The wound carries a heavy bioburden and fibrin load requiring an office-based debridement followed by 3 weeks of moist dressings and compression. The patient fails to improve and is admitted to the hospital for 1 week of treatment including intravenous (IV) antibiotics and surgical debridement. On discharge from the hospital, the patient is transferred for a 6-week stay in a sub-acute unit.A total of 8 weeks later, the same patient returns to the outpatient clinic from home due to a plateau in healing after 2 weeks of home health therapy and is re-enrolled in the clinic as a new patient. The wound is now 75% smaller in area than at the time of the original consult. Compression and moist dressings are again applied and the wound completely heals in 5 weeks. The various outcomes from this single case include, 1 year of treatment without improvement in a primary care office, a new consult lost to follow-up without healing in 3 weeks in the wound clinic, a hospital stay of 5 days with an increase in wound size status post surgical debridement, a 6 week sub-acute stay with a 50% volume reduction in 6 weeks, home health care for 2 weeks with no change in wound size, and a completely healed wound treated in 5 weeks in the outpatient clinic. The “episode of care” outcome, however, describes a venous leg ulceration that required 69 weeks of therapy including primary care visits, wound clinic treatment for 8 weeks, home health care for 2 weeks, a sub-acute care stay of 6 weeks, and 1-week hospitalization. This clinical example is a common scenario and represents the importance of defining the site of care when analyzing wound healing data. Each point along the continuum of care acts as a “silo” of care and not part of a larger system of care. Additionally, current reimbursement policies create the potential for each site of care to maximize economic outcomes that may not make clinical or economic sense if the entire “episode of care” was integrated across settings. Concepts such as pay for performance are a step in the right direction, but also focus on achieving benchmarks from individual sites of care and, therefore, fail to achieve true integration across care settings.
The authors have created an integrated care approach to wound healing using a combination of strategically aligned groups that do not function under the same corporate umbrella. Patients are seen in an outpatient, not for- profit, hospital-based wound clinic. In addition, inpatient wound care is provided for a second hospital that belongs to a completely different not-for-profit organization. The clinic admits directly to both of the hospitals in which inpatient care is provided. The authors’ sub-acute wound unit is a privately owned for-profit center with no formal business relationship to either hospital. The home health agencies are operated by each of the 2 previously described hospitals. The authors’ prosthetics and orthotics group provide services at all locations and are a small privately held firm. The specialists that consult and work with the authors’ team are mainly from private practice models. The entire team, including physicians, are salaried and have no volume or procedure-driven economic incentives at any of the sites of care. The single most difficult aspect of providing care in this model is case management. Not only does it represent the most time consuming component of overall patient care, it is the least economically productive. It seems paradoxical that the most critical piece of the final clinical outcome carries with it no form of reimbursement. The current model encourages procedure volume and fails to reward outcomes.
With that as a backdrop, if there is to be any change in the financial structure of wound care, wound care clinicians need to collect, analyze, and publish outcomes from all sites of care and for all strata of patient risk and wound complexity. Therefore, the authors set out to validate earlier published results from a sub-acute wound program run by the authors a few years ago. Validating those outcomes would confirm reproducibility of the clinical model across settings. The hypothesis was that a systematic approach to patient care could be reproduced in a similar, but new facility, given that the clinical wound team was held constant.
Objective. The primary objective of the study was to identify the healing outcomes both as a percentage of the total population and in relation to time of treatment. Another primary objective of the study was to validate the results from an original pilot program conducted at a similar private, for-profit facility that was part of a large, national chain of nursing homes. A secondary objective was to identify any patient characteristics on admission to the unit, using the US Centers for Medicare and Medicaid Minimum Data Set 2.0, which could be used to predict final outcomes. (The Minimum Data Set 2.0 is a US Centers for Medicare and Medicaid [CMS] initiative that was introduced as a part of the Nursing Home Reform Act of the Omnibus Budget Reconciliation Act of 1987).
Study design. The entire study was conducted at the authors 25-bed, sub-acute wound unit dedicated to the treatment of patients admitted with nonhealing wounds. The unit resides within a 300-bed, privately owned, for-profit nursing home/long-term care facility, one of several facilities owned by the company. The executive team was highly involved and encouraged the development of a clinically focused, outcomes driven unit. The nurse and physical therapists on the unit treat only patients with chronic wounds and have worked with the authors implementing prior programs. Patients originate from a hospital based wound clinic that receives more than 2500 patient visits per year (on average). A previously published systematic approach and treatment protocol is utilized for all patients with nonhealing wounds of any etiology.¹ The clinic has generated an average of 150 hospital admissions per year for the past 6 years. A percentage of patients are then transitioned from the inpatient unit to the sub-acute wound unit for advanced inpatient care. This group of patients is either too debilitated or their wound care requirements exceed those fulfilled in the home or outpatient clinic. Results obtained from this sub-group of patients constitute the basis of this report.
The current sub-acute program was created January 1, 2006. All patients admitted to the program between January 1, 2006 and May 1, 2007 were included in intent to- treat analysis, with a statistical significance criterion of P ≤ 0.05. The only patients excluded were those with either no wound (ie, a wound recently closed with a myocutaneous flap, admitted for offloading and antibiotics) or patients who had a single visit (patients seen once and then transferred to an acute care hospital). This group contained 8 patients. Each patient was admitted to an internal medicine or family medicine unit for general medical care by the attending physician.A wound physician saw the patient on admission and weekly on rounds. A comprehensive wound treatment plan was designed and care was delivered by the nurse and wound care physical therapists. All advanced wound care products were available, as well as IV antibiotics, total parenteral nutrition, rehabilitation services, pharmacy, and nutritional consultative support. Wound modalities including negative pressure wound therapy, ultrasound—both Megahertz and MIST^™ Ultrasound Therapy (Celleration, Eden Prairie, Minn)—electrical stimulation, and ultraviolet light, were employed when indicated. The physical therapist or physician performed wound debridement as needed. Each week the wound physician selected the modality and dressing treatment plan to meet the needs of the wound and the patient as identified by standardized assessments. Orthotic and prosthetic consultation was available on site for offloading and compression. A physician specializing in infectious disease evaluated all patients weekly to monitor culture results and antibiotic levels. The wound nurse and wound clinic staff communicated daily to organize patient transition to and from the hospital for staged procedures, and ensured that on discharge patients were transitioned smoothly back to the outpatient program.
 
As part of the admitting process, the nursing home conducted a formal intake history and completed the Minimum Data Set (MDS 2.0) forms. The nursing home MDS coordinator completed this form on admission, every 90 days, and each time a patient was either admitted to the hospital setting or when a significant clinical event occurred. The MDS data were captured electronically and translated into a format compatible with SPSS^™ software. Each weekly clinical visit was entered into this electronic database. Parameters including wound length, width, depth, dressings utilized, procedures performed, admissions to the acute care setting, as well as wound area and volume, were recorded. When the patient’s wound was either healed, or the patient was ready to be transitioned to the next site of care, a final disposition was recorded electronically—wound outcomes were reported as “healed,”“more than 50% volume reduction,” or “50% or less volume reduction. ”A large database was constructed combining the admitting MDS information with the complete clinical record for each patient’s entire sub-acute care stay. After all information was entered for an individual patient, all identifiers were eliminated to protect patient privacy.
The study was a prospective, longitudinal, outcomes analysis from a sub-acute wound care unit. Patients were not randomized.All patients with wounds and more than 1 visit were included in the intent-to-treat analysis. The comprehensive wound assessment and treatment system was utilized as standard of care.¹ The results were compared to previously published outcomes as a historical control.²

Results

A total of 109 patients were enrolled in the database. There were 101 patients with 209 evaluable wounds, excluding 8 patients who had 1 visit or a closed wound on admission. Demographics are described in Table 1. Although there were more women enrolled in the study, the number did not achieve statistical significance.There were statistically more patients in the 61- to 70-year-old group than in any other group (P = 0.039). Neither age nor sex was correlated with healing outcomes (Table 2). Outcomes were divided into 2 distinct groups. Group 1 included those patients achieving complete healing and those with marked improvement (defined as > 50% reduction in wound volume). Group 2 consisted of patients who improved but did not achieve 50% volume reduction, patients whose wound size remained unchanged, and those with a deteriorating wound. Kaplan-Meier derived median time to healing was 7.9 weeks for patients in Group1 (Figure 1). A statistically significant difference was noted when wound outcomes were separated by wound etiology (Table 3). Traumatic wounds were noted to heal in a short time interval compared with other wound etiologies.Wound location and initial wound volumes are described in Table 4.

Overall, 41.6% of patients achieved complete healing while 31.6% achieved greater than 50% volume reduction. Success within a sub-acute wound program (defined as the combination of complete healing and > 50% volume reduction) was achieved in 73.2% of all treated wounds. An additional 14.8% of wounds improved, but achieved less than the 50% volume reduction threshold.A total of 7.7% of wounds worsened during the treatment course (Figure 2). Comorbidities were common, but there were no statistically significant correlations between disease state and final clinical outcomes (Table 5).

An aggressive surgical approach was taken to achieve healing in these complex wound cases. In total, 87 wounds were completely healed and of those, 39% were surgically closed. Three wounds were closed using myocutaneous flap procedures and 31 wounds were closed using a skin graft. There were 2 below-knee amputations (1.9% of the evaluable population).

Discussion
Previously published data from a single clinical team working from two distinctly different hospital based wound programs, demonstrated that consistent outcomes are achievable using a uniform clinical approach to wound care.³ Patients in that study were analyzed from a 200-bed community hospital based wound program, and a 700-bed level 1 tertiary center. Those outcomes, while consistent, were taken from a single care setting. Even prior data² from the present authors single sub-acute program can be further analyzed. For example, the overall healing and 50% volume reduction group totaled only 53% in a study of 346 patients with 680 wounds published by the authors using the same clinical team that conducted the present study (Program 1 in Table 6).³ An overall comparison of 3 data sets from sub-acute wound programs clinically managed by the authors and their clinical team is shown in Table 6. The data² from Program 2 and the current study represent patients who were accepted only from the authors’ two hospital programs. The wound care team case managed these patients starting from the outpatient center through the hospital admission, and subsequently in the sub-acute wound program. Long-term IV access, surgical debridement, and medical stabilization of the patient occurred before discharge to the sub-acute unit. These 2 programs demonstrated statistically significant increases in healing and 50% wound volume reduction compared to Program 1. Hospital length of stay can be minimized using these specialized units, which benefits the economics for the hospital while minimizing unnecessary risks of prolonged hospitalization for the patient. When patients are admitted to a sub-acute wound program from outside hospitals, as was the case in Program 1, it is difficult to achieve equivalent outcomes as the critical steps along the continuum of care may not have been optimized. The healing rates from hospital outpatient clinic programs range from 72%–74% compared to 41.6%–45.9% in the sub-acute programs.³ Evaluating these publications separately, one would conclude that the clinical success rates were excellent in the outpatient setting and sub-optimal in the sub-acute program, but might not notice that the publications were written by the same clinical team, applying the same standardized wound and patient assessments and protocols of care. Outcomes data in wound care, therefore, need to address the clinical team involved and its point of involvement along the continuum of care, the setting of care, and an analysis of the patient population (case mix/severity indexed, etc.)

The results of this study represent strong patient selection bias. For example, over the past 8 years the authors have generated a hospital admission for every 20 clinic patient visits. An average patient accumulates 10 visits during therapy.With 2500 patient visits per year, this yields approximately 250 new patients to the program per year. On average, 100 patients are admitted to the hospital, with 150 admissions overall during a 1-year period (a 50% re-admission rate). Therefore, 40% of all patients seen in the clinic require at least a 1-day stay in the acute care setting. Approximately 25% of admitted patients are transferred to the sub-acute program. In total, therefore, only 10% of all patients seen in the authors’ outpatient wound clinic are admitted to the sub-acute program. It is fair to assume that these patients represent a medically complex subgroup from the original cohort.The healing rates in the acute care and sub-acute care settings, therefore, would not be expected to mirror each other.
Predictors of healing. Specialized wound centers have been shown to achieve improved healing rates compared to more fractionated care.⁴ However, comparisons between centers are difficult and possibly misleading. Would an outpatient center that primarily focuses on venous leg ulcers demonstrate an equivalent healing rate compared to a vascular surgery based program that treats critical limb ischemia? Keyser⁵ reported an 88% healing rate for diabetic foot ulcers, while others report healing rates of 38%.^6,7 One report in the literature quotes 100% healing in an outpatient clinic.⁸ Further research is needed to define the parameters that underlie such varying outcomes. Outcomes of success and mortality are now in the public domain for coronary artery bypass grafting at most hospitals. Clearly, there are differences in outcomes between high volume centers of excellence and those from less experienced centers. The wide variation in healing rates—reported in the literature from 38% to 100%—probably reflect case mix and reporting differences rather than wide clinical variation.^6–8
In addition to the outcomes reporting problem, there is a relative absence of wound care education for healthcare professionals.A recent survey of medical school curricula reveals, on average, a medical student receives only 9 hours of education on wound healing over a 4-year undergraduate medical degree program.⁹ As no single provider group has complete expertise in wound healing, it is not surprising that multidisciplinary teams can achieve improved clinical and economic outcomes, especially in the nursing home environment.¹⁰
In an effort to assist clinicians in determining which patients will respond to treatment, the authors hypothesized that there might be a clinical/functional “profile” that could be used as a predictive tool. The MDS 2.0 is a US Centers for Medicare and Medicaid (CMS) initiative that was introduced as a part of the Nursing Home Reform Act of the Omnibus Budget Reconciliation Act of 1987. MDS is part of a comprehensive resident assessment instrument (RAI) which contains information on clinical,behavioral, and social status of nursing homes.^11,12 The RAI consists of the MDS, utilization guidelines, and Resident Assessment Protocols (RAP).¹¹ When a patient has a particular MDS status that matches a trigger for a RAP,1 or more of 18 problem-based RAPs are performed. Several MDS criteria have been validated as predictors of pressure ulcer development. Vap et al¹³ determined that the MDS was less sensitive but more specific in predicting the development of a pressure ulcer compared with the more popular Braden Score. Bates-Jensen et al¹⁴ found problems using a patient’s bed bound status, one of the MDS quality indicators, because of significant underreporting in the facility. In another study, Bates-Jensen et al¹⁵ were unable to correlate nursing homes with low pressure ulcer prevalence with improved clinical care processes.
Despite conflicting reports it appears that overall quality of care has improved in nursing homes since the release of the MDS.¹⁶ Although Jones et al have recently tried to identify clinical and functional aspects of the patient history to predict pressure ulcer healing there is minimal published literature on the topic.¹⁷ Vap et al,¹³ found that specific MDS components correlate with pressure ulcer risk. We attempted to utilize some of these parameters to determine if they could be used as predictors of wound healing. Bedfast status, bowel incontinence, use of bed rails for transfer, and the use of a transfer aide, were not found to have statistical correlation with the ultimate outcome of healing or marked improvement (Table 7).

Conclusion
In this prospective, longitudinal, intent-to-treat study of patients within a comprehensive sub-acute wound program, a 41.6% healing rate and a > 50% wound volume reduction rate of 31.6% were achieved in a median time to healing of 7.9 weeks. These results were similar to published outcomes from a previous program managed by the authors.² The similar outcomes of this study support the primary conclusion that a wound program can be reproduced if the same clinical approach is taken, regardless of physical plant, staffing, and ownership of a nursing home. However, a comparison of these results to published healing rates from an outpatient, hospital based wound clinic run by the same clinical team using the same protocols of assessment and care, highlight the importance of identifying the population under study, early intervention, and the clinical site of care. Additionally, readily available MDS criteria may prove useful for the prediction of wound development, but are unlikely to assist a clinician in predicting who will respond to therapy.
It is evident that some form of wound indexing or severity indices are needed to help wound care clinicians maximize clinical outcomes and select therapeutic options from the myriad currently available to the wound care clinician.The wound care field is unique in that therapeutic options have outpaced diagnostic and predictive innovation. Part of the problem is that woundcare societies have been unable to translate cognitive and case management work efforts into meaningful, appropriate evaluation and management codes, and ultimately reimbursement.Wound care clinicians can look forward to wound profiling, bioassay development, wound severity scoring, and gene expression changes in wound tissue as described in the eloquent work by Brem et al,¹⁸ as a potential means to predict outcomes, prevent occurrence, determine debridement margins, and to select the most economical and clinical effective therapies available for patients. A useful wound scoring system will need to include patient comorbid conditions and quality of life parameters. The APACHE score (acute physiology and chronic health evaluation) score used in the critical care industry provides a useful analogy.¹⁹ Outcomes data are very important and there is a need for all clinicians in various sites of care to report on their work.The larger task at hand will be to string these outcomes together in order to provide accurate clinical and economical episode of care data.