Abstract: Wounds are commonly encountered at the end of life. There is a need for palliative wound care, but little exists, and public policies do not focus on palliation. Understanding the magnitude and scope of the problem can provide a basis for developing palliative treatments and influencing public policy. Method: A cross-sectional study and retrospective records review from patients in a large urban/suburban hospice were conducted to describe (a) the prevalence of wounds and (b) the types of wounds and characteristics of hospice patients referred for wound care. Among 383 hospice patients, 35% had skin wounds, of which 50% were pressure ulcers. In a case series of 192 consecutive patients referred for wound consultation, the average age of patients was 82, 67% of patients were female, patients had numerous co-morbidities, and 40% of all wounds were pressure ulcers. Despite short treatment periods, palliative wound care measures resulted in significant healing for nearly half of wounds identified. Conclusions: These studies shed light on a problem of immense proportion. Wounds afflict more than one-third of the nearly 1 million hospice patients in the United States and many more patients at the end of life. Knowing characteristics of the patients and wounds as presented here can provide an evidence-based foundation to foster development of appropriate palliative treatment for these patients.

I

n 2001, a large local urban hospice delved into a bold arena—engaging the efforts of a consulting physician to help with the problem of wounds in their hospice patients. Hospice administrators perceived that wounds were common in the hospice population; they estimated approximately 10% or more of their patients had wounds and reasoned that an organized approach with palliative care specific for wounds would be beneficial for their patients and set them apart from other hospice organizations. Standard wound care focuses on wound healing and closure of wounds with an average cost of $1,600/patient/month.¹ This level of expenditure is unaffordable for a hospice that is caring for patients on a per diem of $118 to cover all expenses.² Additionally, wound healing and wound closure are unrealistic goals for hospice patients at the end of life.³ Skin is the largest organ of the body and is as susceptible to failure as any other organ system;⁴ it is not logical to expect healing of skin to occur while other organ systems are failing. The goal in hospice is always to provide comfort, relieve suffering, and improve the quality of living and dying.³
       A dedicated effort ensued to define palliative wound care of which little, if any, was known at the time and deliver that care within the per diem of hospice reimbursement. Using an interdisciplinary approach with input and feedback from nurses, physicians, aides, and family caregivers, a palliative wound care protocol for prevention and treatment was developed using simple, inexpensive products and techniques. The goals of wound care included relief of pain, elimination of odor, prevention of infection, maintenance of function, and, where possible, healing. Healing, however, was viewed as a side effect of palliative treatment and taken to be a surrogate indicator for quality of life. Patients with wounds were referred for physician consultation when problems or concerns with the wound or wound care occurred.
       Part of developing a palliative program involved defining the magnitude and scope of the problem. Wounds are a known problem in the elderly population with 10–25% of nursing home residents reported to have pressure ulcers,⁵ but there is little information on wounds at the end of life. What is the magnitude of the problem? How many wounds are occurring at the end of life, and what types of wounds are occurring? How severe is the problem at the end of life? Is it the 10% estimated by local hospice administrators? There is scant published data, and research is needed.⁶ Who has wounds, and what can be done? Is it realistic to expect healing of wounds as is the traditional approach to wound care and wound treatments? Or should palliative care with different goals and treatments than traditional care be instituted?
       This article is a report of 2 studies conducted in 2003 and 2004 that attempted to answer these questions or at least shed light on the issue. The first study (A) was a prevalence study of wounds in patients enrolled in the hospice, simply documenting at 1 point in time the number of patients with wounds and noting the types of wounds. The second study (B) was a retrospective chart review of the hospice patients with wounds referred to the consulting wound physician. This study examined the characteristics of patients referred and types of wounds. All patients in both studies were enrolled in hospice and received palliative care for their diseases and wounds. Disease management and palliation were per standard hospice protocols. Wound care protocols that were developed in house were followed for palliative care of wounds.

Methods

       Study A. A prevalence study of patients in a large urban-suburban hospice with over 400 patients enrolled at the time of the study was conducted. Primary care RNs on 8 clinical care teams were asked to report how many patients in their care had pressure ulcers (stage I–IV), tumors, ischemic ulcers, stasis ulcers, or other skin issues (ie, skin tears, bruises, surgical wounds, burns, itching). Data were collected during a 2-week period (mid-March through early April 2003). No data were collected on whether wounds were present on admission or acquired after admission to hospice.
       Study B. This study was a retrospective chart review of patients in a large urban-suburban hospice who were referred to a physician wound consultant during a 2 1/2 year period, September 2001 through March 2004, for wound evaluation and treatment. Data from 192 consecutive patients referred to the hospice physician wound consultant were analyzed. For each chart, data were collected on age, sex, race, type and location of wound, comorbidities, length of stay from time seen by the wound consultant, and whether the wounds were healing or healed after wound consultation. Most charts contained patient demographic information, medication lists, problem sheets, physician notes, and pictures of the patients’ wounds. These were all clinical encounters reviewed retrospectively.

Results

       Study A. Reports were received on 383 patients (6 of 8 care teams reporting) and combined into an overall report. A total of 35% of patients in the hospice program had some type of skin issue. Wound prevalence was 17.5% pressure ulcers, 7% vascular wounds, and 10.5% other. These results are shown in Figure 1. Of these skin issues, 50% were pressure ulcers, 20% were ischemic ulcers, and 30% were other skin issues (ie, stasis ulcers, burns, skin tears, tumors). These results are shown in Figure 2. Two-thirds of patients were nursing home residents, and one-third were home care patients.

Table 1

       Study B. Patient demographic results are shown in Table 1. Average age of patients was 82, with a median age of 83. Two-thirds of patients were female, and 18% of patients were Black. The average length of stay (LOS) from first physician visit was 80 days, with a median of 31 days and a mode of 4 days. Average LOS in hospice was reported at 82 days, which was not significantly different. Malnutrition was present in 71% of patients, and two-thirds of patients had multiple wounds.

Figure 3

Figures 1,2

       Figure 3 shows patient comorbidities with frequency of conditions/diseases. The most common comorbidity was dementia, reported in over 45% of patients. Dementia was followed by stroke, peripheral vascular disease, and diabetes/cancer in order of prevalence. Figure 4 shows comorbidities by gender, and Figure 5 shows comorbidities by ethnicity. Several key differences were apparent, specifically increased rates of diabetes, renal failure, and anemia in Black patients and increased fractures in women.

Figure 5

Figure 4

       Figure 6 shows types of wounds seen, and Figure 7 shows location of wounds. The majority of wounds were pressure ulcers—70% of patients had pressure ulcers, with pressure ulcers representing 40% of all wounds. Most wounds were on the sacrum, followed by foot/heel and leg. Figure 8 summarizes wound types by race and gender. There were several differences in types of wounds by race and gender: Black patients were more likely to have gangrene or arterial wounds. White patients and women were more likely to have skin tears and traumatic wounds. Men were more likely to have surgical wounds, tumors, or neuropathic wounds.

Figure 9

Figure 8

Figure 7

Figure 6

       Figure 9 shows healing of wounds by wound type. Approximately half of wounds healed or were healing.

Discussion

       These 2 studies offer significant insight into the problem of wounds at the end of life. Study A demonstrated that more than one-third of hospice patients suffer from skin wounds. With nearly 1 million hospice patients in the US,² over 300,000 people suffer from wounds in the hospice population alone. Since most persons are not in hospice at the time of death,⁷ extrapolating these results to the general population means that many more frail end-of-life persons suffer from wounds and could benefit from palliative care. A third study conducted by this author, but not yet completed, examines wound referrals of non-hospice patients. Demographics of this group are similar to the hospice group, with average LOS the same at 81 (versus 80), and median LOS slightly longer at 48, but interestingly, three-fourths of patients died within 6 months of being seen, which would have qualified them for hospice care. This suggests wounds are predominantly an end-of-life phenomenon. With 34 million people over the age of 65, and an aging population,³ the magnitude of the problem is staggering with a great need for palliative interventions.
       Clearly, wounds are a significant problem at the end of life. The patients are elderly, malnourished, have multiple comorbidities, and are severely ill. Unsurprisingly, the majority of wounds are pressure ulcers, but ischemic wounds and gangrene are significant. Stasis ulcers are very infrequent, which is curious given the reported prevalence of stasis ulcers,^3,8 perhaps suggesting that stasis ulcers are often misdiagnosed. It is possible, but seems less likely, that hospice patients do not have stasis ulcers.
       The most common wounds are pressure ulcers, afflicting 17.5% of the population and accounting for 50% of all skin wounds in Study A and representing 40% of all wounds in Study B, with 70% of patients having 1 or more pressure ulcers. Pressure ulcers, which are a major concern in nursing home and hospital populations, are almost inevitable in this frail population. Pressure ulcer risk is increased with poor nutrition, immobility, loss of cognitive function, and incontinence⁹—all of which are seen in the end-of-life population. Reifsnyder¹⁰ found a 15–27% prevalence of pressure ulcers in a home hospice study, very similar to the Study A result (17.5%). Median LOS (31 days) was the same; however, average age of the home hospice patients was 75 versus 82. Cancer was the number 1 diagnosis in the home study, but it was fourth in the author’s study. Comorbidities reported in Study B show high incidence of cognitive-related disorders with dementia being the primary risk factor, representing over 45% of patients. Reifsnyder also found dementia as the number 1 factor, with 40% of patients affected. The majority of patients in the author’s prevalence study were nursing home (two-thirds), and all of the retrospective chart reviews were nursing home patients. Similarities in these studies imply that patient location and age are not as important as disease condition or end-of-life frailty in engendering pressure ulcers.
       Interestingly, the most common wound location was the sacrum, with 40% of patients having wounds in this area, representing over 25% of all the wounds examined. The majority of these wounds were pressure, indicative of the immobility and frailty of this population. The majority of other wounds were heel, foot, and leg—again reflecting the immobility of this population and comorbidities of circulatory disorders and diabetes.
       It is unknown whether the percentage of Black patients is consistent with the population. If the 18% Black patients in Study B is truly higher than the reported 15% base, this begs the question, why? In looking at comorbidities, Black patients have higher prevalence of diabetes with concomitant renal failure and anemia, which could increase their likelihood of having wounds referred to the consulting physician. Or are there psychosocial issues related to ethnicity? Certainly, hospice is underutilized by the non-White population with Black patient referrals comprising only one-tenth the amount of hospice referrals compared to White patients.¹¹ In this author’s experience with non-hospice wound referrals, Black patients account for nearly one-fourth of referrals, compared to less than one-fifth of referrals in Study B. Further investigation is warranted to examine ethnic differences in wounds and referral patterns.
       It is not clear why nearly one-half of wound patients have dementia. Is dementia primarily a cardiovascular disease, associated with stroke and peripheral vascular disease? Lack of cognition is a known risk factor for development of pressure ulcers,⁹ but does dementia cause wound development, or do wounds develop because of the same reason a person is demented? Do demented persons receive the same level of care as non-demented persons? This warrants further investigation to elucidate the relationship of dementia to wounds and development of possible preventive strategies.
       Less than one-third of patients had diabetes, making it the fourth ranked disease, on par with cancer. Dementia, stroke, and peripheral vascular disease were involved much more often. Wounds are strikingly a cardiovascular phenomenon. Figure 3 indicates total frequency of patient diseases and conditions. When analysis was completed by gender and ethnicity, several compelling differences were evident. Women and White patients had much higher prevalence of hip and leg fractures (6–7% versus 1–2%). White patients also had more than double the amount of heart disease (10% versus 4%) and higher prevalence of cancer and dementia. Women had less strokes but more cancer and dementia. Black patients had higher prevalence of renal failure (nearly double), peripheral vascular disease, anemia (also twice as much), stroke, and diabetes (twice as much with 12% versus 6%). The increased renal failure and anemia were consistent with a higher prevalence of diabetes. Black patients were also more likely to have gangrene or arterial wounds, again consistent with higher prevalence of diabetes and peripheral vascular disease. Black patients also had more than double the number of infected wounds compared to White patients, perhaps also consistent with higher levels of diabetes and arterial insufficiency. Recognizing racial and gender differences could be key in developing preventive and palliative care programs for these frail patients.
       Healing results are very intriguing and challenge clinicians’ paradigm of what to expect in chronic wounds. It is possible that the percent healing/healed is even higher, as many patients did not have follow-up visits prior to their death, and wound status is unknown. It should be noted that all wounds seen were chronic in nature; pressure ulcers were almost exclusively stage III and stage IV; and necrosis and gangrene were highly prevalent. These wounds were fairly typical, however, of wounds in hospice patients, at least in this author’s experience. Over 70% of patients were malnourished, and in review of the records, only 37% received nutritional support. This further challenges understanding of the role of nutrition in wound healing.⁶ Nutritional support was provided where possible with protein, vitamin C, and zinc supplementation, but for the majority of patients, this was not achievable due to patient inability, refusal, or lack of availability. In examining the results, over half of pressure ulcers were healed or healing, almost one-third of arterial wounds were healed or healing, and more than two-fifths of neuropathic wounds were healing. Nearly half of all wounds were rated as healed or healing. No time frames were collected for this, but patients in the healed/healing cohort had median LOS 47 days, mode 31 days, and average 99 days, compared to 31, 4, and 80 days, respectively, for the overall group, ie, patients who lived longer were more likely to demonstrate healing. To achieve this amount of healing in this population with severe wounds is remarkable and speaks eloquently for good palliative care with the implication of improved quality of life for these patients.
       A very important question is whether the wound treatment used had an impact on healing rates and quality of life. Treatment in all cases was noninterventional and tailored for palliation. Further study of this is needed. The US Food and Drug Administration claims for new treatments require demonstration of healing and closure of wounds,³ and those who treat persons at the end of life are in need of ways to palliate wounds without necessarily healing. Proponents of palliative care are in need of data on prevalence and types of wounds and effectiveness of palliative treatments in order to properly advocate for change in programs and protocols.^3,7 Results presented in this article can provide a foundation of evidence to support further efforts to research and advocate for palliative wound care.
       Part of palliative wound care includes control of pain and reduction or elimination of odor. While these objectives were achieved, this article will not report on these factors, as the focus here is the scope and magnitude of the overall problem.
       Whether the results of the retrospective case review are reflective of the general hospice population is not known precisely. It would seem fairly representative, since percentages of pressure ulcers and vascular ulcers are similar (50/20 in Study A versus 40/22 in Study B). Two reasons can explain why slightly more vascular ulcers and fewer pressure ulcers were seen by the physician: there were differences in the 2 groups, or perhaps the physician diagnosis differs from the nurse assessment. The author favors the latter, since it is quite common for a wound to be assessed by a nurse as pressure when it is actually ischemic or traumatic.
       There were several sources of possible bias or error in both studies. Reporting for Study A may not have been consistent between nurses. There was no effort to train or coordinate with persons collecting data to ensure consistency. Therefore, it is likely that some wounds were over-reported, under-reported, or misclassified. However, results are similar to other studies,¹⁰ so one might suspect a negligible effect. Data was collected by RNs who have received basic training in skin assessment. The data reporting was similar to most standard skin assessments completed in any facility and subject to the same inconsistencies intrinsic to the method.
       All teams in Study A were consistent in the types of information reported, with fairly consistent reporting of 20–36% of skin issues. One home care team reported 72% of skin issues, but analysis showed they reported all cases of itching and bruising, which the other teams had not. When corrected for this, the team results were similar.
       The prevalence study included patients in multiple settings, including home and nursing home; however, hospital patients were not represented, as the hospital teams did not respond to the survey. Since for this hospice population, the hospital patients were comprised of nursing home and home patients that varied over time, it is probable that results would be similar. No attempt was made to differentiate home versus nursing home patients as to type or percent of wounds, though nursing home patients represented two-thirds of the patients. Additionally, Study A did not collect data on gender, age, or ethnicity of patients.
       This prevalence study was not conducted at a single specific point in time but over a 2-week period as nurses were able to see their patients. The hospice census was stable over this time period. Thus, it is unlikely that significant error was introduced using this method.
       Study B was a retrospective chart review and thus has inherent deficiencies related to this type of study. Not all parameters being assessed were noted for each patient or at each visit. Many patients had no follow-up visits, so outcome in terms of healing is unknown for these patients. Wound measurements, in particular, were often not available, and variation could be significant, depending on position of the patient, amount of help available to move or position patient, lighting, patient condition, etc. Often the clinician impression is the most reliable data available.
       A strength is that the same clinician performed all visits in the retrospective study, thus consistency can be expected in evaluation and diagnosis. Treatments for each wound type also tend to be more consistent with the same ordering physician, aiding in comparison, analysis, and drawing of conclusions. Having 1 clinician perform all visits could also introduce some individual bias, but this would be expected to be consistent.
       Study B included almost exclusively nursing home patients and thus provided little information on home or hospital patients.
       While high rates of healing were reported, evaluation of healing lacked strong objective measurements. In some cases where there was regular physician follow up, measurements were available; in many cases, there was no follow up or only subjective evaluation. Yet, with only palliative care treatments, over one half of wounds were judged to be healed or healing by physician report. Since LOS was very short in these patients, most would not live long enough to achieve healing. Healing that was not measured by wound size was assessed by wound progression in a normal healing mode, such as presence of granulation tissue and resolution of necrotic slough. For purposes of this article, healing is being used as a surrogate measure for quality of life.

Conclusion

       Results presented here are both provocative and enlightening. Wounds at the end of life are a problem of tragic proportion for the nearly 1 million hospice patients and millions of other frail, elderly persons living with chronic disease. Not only are wounds prevalent at the end of life, affecting more than one-third of patients, the development of wounds is an indication of the person’s frail and terminal condition. As the US population ages, and more people are living longer with fatal chronic conditions, the provision of quality palliative care should be a national priority. Cost of care at the end of life is rising precipitously, rapidly outpacing available resources.⁷ Providing palliative care becomes not only a caring, humane choice but also a choice that can allow responsible stewardship of limited resources. Wound care is a large and integral part of end-of-life care, and there is need for palliative protocols and products that will allow compassionate care and dignity at the end of life. Knowing the characteristics of the patients and wounds as presented in this article can help better identify patients at risk and provides an evidence-based foundation to foster development of appropriate palliative treatment to benefit these patients. Even at the end of life, with severe comorbidities and fatal disease conditions, palliative wound care can significantly enhance the quality of life for these patients.