Defining Patient Safety in Hospice: Principles To Guide Measurement and Public Reporting

Abstract

Despite progress towards safer care in most settings, there has been much less attention to improving safety in hospices, which care for more than 1,500,000 patients every year. In this article, we describe three serious conflicts that arise when safety measures from other settings are applied to hospice. First, safety measures that are imposed in order to reduce morbidity and mortality may be irrelevant for a hospice patient whose goals focus on comfort. Second, safety measures that are defined in patients with a life expectancy of years can be inappropriate for hospice patients whose typical survival is measured in days. Third, it can be very difficult to assign responsibility for the safety of hospice patients, whose care is provided mostly by family and friends. Therefore, generally accepted safety measures are often inappropriate for hospice care, and can lead to unintended consequences if they are applied without critical evaluation or modification. Instead, we suggest three principles that can guide the development of hospice-appropriate safety measures by considering a patient's goals and life expectancy, and the degree to which responsibility for a patient's care is shared.

Introduction

Patient safety, one of the anchors of high quality medical care,¹ can be defined as “freedom from accidental or preventable injury produced by medical care.”² Common examples of accidental or preventable injury include wrong-side and wrong-patient surgeries, transfusion errors, and bloodstream infections from intravenous catheters. Increasing awareness of these and other injuries has helped to make a compelling case for a culture of safety, mandatory disclosure and reporting, and system-wide interventions to prevent errors in the acute care setting,^3–7 as well as in nursing homes and home care.^8,9 These efforts in turn can catalyze efforts to improve care, and have produced some successes,^10–12 although overall evidence of progress has been mixed.¹³

However, three serious conflicts arise when safety measures from other health care settings are applied directly to hospice without consideration of the unique characteristics of hospice patients, or the philosophy of hospice care. First, some safety measures to reduce morbidity and mortality may be irrelevant for a hospice patient whose goals focus on comfort. Second, safety measures that are developed for patients with a life expectancy of years or decades can be inappropriate for hospice patients whose typical survival is measured in days or weeks. Third, virtually all existing safety measures hold the health care provider or institution responsible for safety failures, which is illogical for hospice patients whose care is provided mostly by family and friends.

The lack of regulatory safety standards in hospice is a significant concern because hospices care for large numbers of patients every year. For example, in 2009 approximately 5000 hospices provided care to 1,560,000 people.¹⁴ Therefore, although industry guidance exists,¹⁵ there is probably considerable variation in the degree to which hospices measure safety, and large numbers of vulnerable patients receive care without the protection of mandatory standards. On the other hand, there is a very real risk of unintended consequences for patients, families, and providers, if existing safety measures are imported to hospice without careful evaluation.

In this article, we use three cases of outcomes that would be classified as serious reportable adverse events in other health care settings¹⁶ to illustrate ways in which generally accepted safety measures are inappropriate for the hospice population. In addition, we suggest that their application can lead to unintended consequences and care that is inconsistent with hospice goals. Finally, we describe three principles to guide the creation of new safety standards that are consistent with hospice goals. Specifically, we suggest that appropriate safety measures should consider a patient's goals, a patient's life expectancy, and the degree to which responsibility for a patient's care is shared with informal caregivers.

The importance of patient preferences in hospice care

JK is an 89-year-old hospice patient with widely metastatic prostate cancer living in a nursing home. Despite education by members of the hospice team regarding the rationale for repositioning, he has refused to be repositioned because of severe pain with movement. Non-sedating pain medications have been ineffective and he has refused any increase in his opioid dose because of concerns with sedation and cognitive impairment that he has experienced in the past. Despite the use of a pressure-relieving mattress, JK develops a stage III pressure ulcer, a National Quality Forum (NQF)-endorsed serious reportable adverse event.^{16(NQF #4f)}

The Institute of Medicine defines quality medical care as the “degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with professional knowledge.”¹ Although most patients desire health outcomes that include improved function, independence, and survival,^17,18 hospice patients accept a plan of care that focuses on comfort.¹⁹ In other settings, JK's case would be defined as a safety failure (i.e., the selection of the wrong plan to achieve an aim),¹ because this reportable adverse event was directly attributable to a plan of care that did not include repositioning. However, JK's goals were to preserve his alertness and cognition, and thus the development of a stage III pressure ulcer cannot be said to reflect substandard or unsafe care.

Similar potential conflicts arise when applying other safety measures to hospice care. For instance, safety measures that reflect mortality rates (e.g., NQF-endorsed safety measures such as mortality within 30 days of hospital admission for acute myocardial infarction^16(#23) or pneumonia^16(#468)) should not apply to hospice patients, whose goals focus on comfort rather than on survival. Similarly, hospice patients receiving warfarin therapy might be willing to tolerate some risk of bleeding or clotting in order to avoid the monitoring that safety guidelines suggest.^{16(#555 and #556)}

Of course, not all hospice patients have goals that focus exclusively on comfort, and patients' goals necessarily reflect the heterogeneity of illnesses and hospice programs. For instance, a patient with advanced cancer who enrolls in an “open access”²⁰ hospice program that allows chemotherapy might have goals that are similar to those of a home care patient. Similarly, some patients may choose to revoke their hospice benefit if their goals change. Nevertheless, reducing mortality is not likely to be the most important outcome for most hospice patients.

The reality of limited survival in the hospice population

HW is a 71-year-old man with frequent seizures due to glioblastoma multiforme who is imminently dying and is admitted to a free-standing hospice inpatient unit for terminal care. When he becomes increasingly confused and is unable to swallow pills, he is given liquid lorazepam 3 milligrams orally every 6 hours for seizure prophylaxis. The following day, he climbs out of bed and falls, striking his head on a nearby table. A neurological examination is non-focal, but his level of consciousness declines and he dies 24 hours later. Because his fall may have resulted in his death, HW's case is flagged as a severe reportable adverse event.^16(5d)

Although falls resulting in severe disability or death are a straightforward measure of safety in most health care settings, this safety measure is difficult to apply to a patient like HW who was very near the end of life. Approximately half of hospice patients die within 3 weeks, one third die within a week, and 10% die within 1 day.¹⁴ This very limited survival creates two challenges for safety measurement.

First, monitoring and reporting adverse outcomes in a hospice population is difficult. In HW's case in particular, it is virtually impossible to determine whether his fall resulted in his death. Similarly, because disease progression and organ dysfunction are expected outcomes in hospice, many indicators of safety problems (e.g., fever due to infection or somnolence due to medication overdose) may be indistinguishable from disease progression.

Second, the very short survival of hospice patients means that harms produced by some adverse events are much less common than they are in other settings. For instance, a serious drug side effect that has an annual incidence of 5% may be of substantial concern in a general population with a median survival of several years, but would have a negligible impact in a hospice population in whom exposure to the same risk is measured in days. For the hospice population with the same frequency of use of that drug, the number of adverse events would be extremely small. Similarly, although a series of guidelines and recommendations have identified drugs like lorazepam that should be used rarely, if ever, in older adults,^21,22 these may be the only way to keep a patient like HW seizure-free during the last days of life.

This is not to say, however, that an adverse event in hospice is less concerning because patients are already near the end of life. Although it is true that the net number of years of life lost is less for a fatal adverse event in hospice than in virtually any other setting, it does not follow that adverse events are therefore less serious in hospice. Moreover, focusing less carefully on events that result in a smaller loss of life would create a dangerous ethical precedent. Finally, it would be difficult to build a culture of safety if responses to adverse events are motivated by outcomes rather than by the degree to which they may be preventable.

Of course, the risk of premature death must be taken seriously, regardless of its magnitude. In addition, families could suffer psychological harm if they perceive that a patient's death was hastened. Therefore, there is a need for some internal processes to identify potential adverse events so that providers can learn from them and potentially avoid them in the future. Nevertheless, strict adherence to safety guidelines developed for chronically ill patients could lead to substandard care for hospice patients.

Shared responsibility for hospice care

GL is an 85-year-old woman with advanced heart failure who lives with her daughter. She has a large, non-healing sacral decubitus ulcer that causes severe pain for which she takes 300 milligrams of long-acting morphine every 12 hours, as well as frequent 60-milligram breakthrough doses. The hospice staff educate GL and her daughter about GL's pain medication and appropriate dosing. However, when her daughter leaves the house to run errands, GL is left in the care of a neighbor, who mistakenly gives GL four long-acting morphine tablets for breakthrough pain. GL becomes increasingly somnolent and has a respiratory arrest 6 hours later. When the error is discovered, the hospice reports GL's death as a serious reportable adverse event.^16(4a)

A nursing home or hospital should be held accountable for GL's death if her medication overdose were the result of staff error. However, responsibility for GL's care was shared among a large group of formal and informal caregivers, a common hospice scenario since most hospice care is delivered in the home.¹⁴ At home, families take on the majority of caregiving responsibilities, including bathing, feeding, and dressing,^23,24 and may also administer medications and help manage intravenous and Foley catheters.

This shared responsibility limits hospices' ability to prevent adverse outcomes. Although hospice staff could provide additional education to GL's family, and could observe future medication doses, hospice staff cannot always be present. Moreover, unlike a hospital that can select staff, hospices cannot select family members to provide care. Finally, hospices' ability to prevent adverse outcomes is limited because the population of informal caregivers is constantly changing, making systemic interventions impossible.

Responsibility for care is also shared for non-hospice patients, including those receiving home care. However, hospice patients are more seriously ill and closer to the end of life. They may have more needs for care, and have greater family involvement. Thus, family caregivers are often partners in the plan of care, and families' involvement is therefore more intensive.

Principles to guide the rational measurement of safety in hospice

Although measures of safety in hospice are essential, standards from other domains of health care should be applied to this population with caution. Safety measurement in hospice must take into account the unique characteristics of the hospice population and must consider the potential for unintended consequences. We propose three principles to guide the conceptualization and measurement of patient safety in the hospice setting.

Create measures of safety that incorporate patients' goals and preferences

Meaningful safety measures need to incorporate the Institute of Medicine's definition of quality as the achievement of a desired outcome,¹ and ensure that these desired outcomes are patient-centered. In JK's case, this would mean defining “safe” care as care that achieved his goals of avoiding repositioning and maintaining an optimal level of alertness. Using this definition, a pressure ulcer was an anticipated consequence of his preference to avoid repositioning, and the hospice should not be held accountable.

However, the hospice should still be held accountable for ensuring that JK and his family understood that minimal repositioning increased his risk of developing a pressure ulcer. This in turn would require that the hospice have policies to define this kind of education, and to ensure that such teaching is documented. Finally, the hospice should measure the degree to which JK and/or his family felt that they were adequately informed about this risk. Therefore, hospices could still be held accountable for most safe practices related to pressure ulcers such as a risk assessment (NQF #540), documentation of a skin care plan (NQF #538), and implementation of a risk reduction plan (NQF #539), if those practices are consistent with the patient's preferences.

This principle should prompt the development of new safety measures that are optimally tailored to hospice care. For instance, hospices should have policies to ensure that patients' preferences to avoid cardiopulmonary resuscitation or hospitalization will be honored and that symptoms are controlled. Similarly, hospices should be held accountable for the frequency with which patients' preferences are discussed and recorded, and the frequency with which patients receive aggressive treatment against their wishes.

Exclude imminently dying patients from the denominator of some safety measures

Because hospice patients have a very limited prognosis, many safety measures should not be applied in a hospice population. For instance, measures of mortality are inappropriate. Measures that focus on the prevention of long-term harms (e.g., monitoring of the adequacy of glycemic control for diabetics) are also inappropriate.

Other safety measures will require the case-by-case exclusion of patients with an extremely limited prognosis. For instance, lorazepam should generally be avoided in older adults and in hospice patients who are likely to survive for several months. However, its use is appropriate for a patient like HW who is imminently dying. These patients should be excluded from the denominator in calculating rates of inappropriate medication use.

Similarly, other measures such as weight loss, the development of new pressure ulcers, and urinary tract infections, should not be applied to patients who are imminently dying, nor should hospices be held accountable for these adverse events. Instead, for these patients hospice staff should clearly document the patient's illness trajectory and likely prognosis to justify their exclusion. Other adverse events such as medication errors should be applied to all patients, even those who are imminently dying.

Tailor hospices' accountability to reflect the degree to which they can control processes of care

When hospices provide care to patients at home, their ability to ensure a patient's safety is severely limited. Therefore, when patients are in their homes and are receiving care from family members, neighbors, or friends, different standards of accountability should apply. Nevertheless, hospices should still be required to provide family members and other informal caregivers with sufficient education and training to mitigate foreseeable risks and to provide patients with the safest possible environment.

Accountability must include policies for education, and printed educational materials. In addition, hospices should be responsible for teaching plans designed to provide caregivers with the knowledge and skills for necessary for safe caregiving. Furthermore, hospices should be held accountable for ensuring that families feel adequately prepared as caregivers. For instance, caregivers should feel comfortable that they were taught in a way that they could understand, and should feel confident in performing key caregiving tasks (e.g., monitoring symptoms and basic skin care). Finally, family members should feel that they were prepared to carry out these tasks safely.

Conclusion

Every year, hospices provide care for increasing numbers of dying patients. This is a vulnerable population that is at high risk of adverse outcomes due to unsafe care. However, it is essential that we select measures of safety that respect the unique circumstances of hospice. Specifically, safety measures in hospice must be sensitive to patients' goals for care, the short survival of this population, and the way in which responsibility for care is shared.

The risks of implementing the wrong safety measures in hospice are substantial. If measures are unrealistic or inappropriate, they will not be taken seriously by the growing hospice industry, and implementation will be difficult. Moreover, safety measures that promote care that is inconsistent with the goals of hospice, or of individual patients, may have the unintended consequence of promoting safety in a narrow area while decreasing the quality of care overall.

Finally, it is important to ensure that safety measures reflect the interdisciplinary nature of hospice care. Virtually all of the measures described in this article focus on medical aspects of care, such as the development of pressure ulcers, catheter infections, or aspiration pneumonia. Although these adverse outcomes are a reasonable starting point because they are known and understood, the boundaries of hospice care are much broader. A true culture of safety in hospice should reflect the philosophy of hospice care, and should include elements of care that span counseling, goals discussions, spiritual support, and sensitivity to cultural and religious beliefs.

Footnotes

Author Disclosure Statement

No conflicting financial interests exist.

References

Kohn

, Corrigan

, Donaldson

. To Err Is Human. Building a Safer Health System. Washington, DC: National Academy Press, 1999.

AHRQ glossary of safety terms. http://www.psnet.ahrq.gov/popup_glossary.aspx?name=patientsafety.

Nagpal

, Vats

, Lamb

et al. Information transfer and communication in surgery: a systematic review. Ann Surg, 2010; 252:225–239.

Pushkin

, Frassetto

, Tsourounis

, Segal

, Kim

. Improving the reporting of adverse drug reactions in the hospital setting. Postgrad Med, 2010; 122:154–164.

Waneka

, Spetz

. Hospital information technology systems' impact on nurses and nursing care. J Nurs Admin, 2010; 40:509–514.

Zwaan

, de Bruijne

, Wagner

et al. Patient record review of the incidence, consequences, and causes of diagnostic adverse events. Arch Intern Med, 2010; 170:1015–1021.

Stockwell

, Kane-Gill

. Developing a patient safety surveillance system to identify adverse events in the intensive care unit. Crit Care Med, 2010; 38:S117–S125.

Castle

, Wagner

, Ferguson

, Handler

. Nursing home deficiency citations for safety. J Aging Social Policy, 2011; 23:34–57.

Gruneir

, Mor

. Nursing home safety: current issues and barriers to improvement. Ann Rev Public Health, 2008; 29:369–382.

10.

Meredith

, Feldman

, Frey

et al. Improving medication use in newly admitted home health care patients: a randomized controlled trial. J Am Geriatr Soci, 2002; 50:1484–1491.

11.

Fulmer

, Feldman

, Kim

et al. An intervention study to enhance medication compliance in community-dwelling elderly individuals. J Gerontol Nurs, 1999; 25:6–14.

12.

Tinetti

, Baker

, McAvay

et al. A multifactorial intervention to reduce the risk of falling among elderly people living in the community. N Engl J Med, 1994; 331:821–827.

13.

Greenberg

, Haviland

, Yu

, Farley

. Safety outcomes in the United States: trends and challenges in measurement. Health Services Res, 2009; 44:739–755.

14.

National Hospice and Palliative Care Organization Facts and Figures. Hospice Care in America. 2010. http://www.nhpco.org/files/public/Statistics_Research/Hospice_Facts_Figures_Oct-2010.pdf.

15.

Clinical Excellence and Safety Standards. National Hospice and Palliative Care Organization: Alexandria, VA, 2010.

16.

National Voluntary Consensus Standards for Public Reporting of Patient Safety Event Information. National Quality Forum, February 2011. http://www.qualityforum.org/Publications/2011/02/National_Voluntary_Consensus_Standards_for_Public_Reporting_of_Patient_Safety_Event_Information.aspx.

17.

Fried

, Bradley

, Towle

, Allore

. Understanding the treatment preferences of seriously ill patients. N Engl J Med, 2002; 346:1061–1066.

18.

McQuellon

, Muss

, Hoffman

, Russell

, Craven

, Yellen

. Patient preferences for treatment of metastatic breast cancer: a study of women with early-stage breast cancer. J Clin Oncol, 1995; 13:858–868.

19.

Medicare Hospice Regulations. 42 Code of Federal Regulations, Part 418.22. 1996.

20.

Casarett

. Rethinking hospice eligibility criteria. JAMA, 2011; 305:1031–1032.

21.

Briesacher

, Limcangco

, Simoni-Wastila

, Doshi

, Gurwitz

. Evaluation of nationally mandated drug use reviews to improve patient safety in nursing homes: a natural experiment. J Am Geriatr Soc, 2005; 53:991–996.

22.

Fick

, Cooper

, Wade

, Waller

, Maclean

, Beers

. Updating the Beers criteria for potentially inappropriate medication use in older adults: results of a US consensus panel of experts. Arch Intern Med, 2003; 163:2716–2724.

23.

Steele

, Fitch

. Needs of family caregivers of patients receiving home hospice care for cancer. Oncology Nurs Forum, 1996; 23:823–828.

24.

Haley

, LaMonde

, Han

, Narramore

, Schonwetter

. Family caregiving in hospice: effects on psychological and health functioning among spousal caregivers of hospice patients with lung cancer or dementia. Hospice J, 2001; 15:1–18.