New obstacles to improving the quality of end-of-life care in ICU

Background 
Because of high mortality, end-of-life care is an important component of intensive care. 
 
Methods 
Objective 
To evaluate the effectiveness of a quality-improvement intervention to improve intensive care unit (ICU) end-of-life care. 
 
Design 
Cluster-randomized trial of a multi-component educational intervention targeted at bedside ICU caregivers from 2004 to 2008. 
 
Setting 
12 community hospitals in Washington State, United States. 
 
Subjects 
2,318 patients dying in the ICU or within 30 hours of ICU discharge, as well as their families and the nurses participating in their care. 
 
Intervention 
The intervention targeted clinicians with five components: clinician education, local champions, academic detailing, clinician feedback of quality data, and system supports. 
 
Outcomes 
Quality of death and dying (QODD) and satisfaction surveys. 
 
Results 
The primary outcome, family-QODD, showed no change with the intervention (P = 0.33). There was also no change in family satisfaction (P = 0.66) or nurse-QODD (P = 0.81). There was a non-significant increase in ICU days before death after the intervention (hazard ratio = 0.9; P = 0.07). Among patients undergoing withdrawal of mechanical ventilation, there was no change in time from admission to withdrawal (hazard ratio = 1.0; P = 0.81). 
 
Conclusions 
A multifaceted quality improvement intervention was associated with no improvement in quality of dying among ICU decedents, as well as no reduction in resource utilization prior to the withdrawal of life sustaining measures.


Results
Th e primary outcome, family-QODD, showed no change with the intervention (P = 0.33). Th ere was also no change in family satisfaction (P = 0.66) or nurse-QODD (P = 0.81). Th ere was a non-signifi cant increase in ICU days before death after the intervention (hazard ratio = 0.9; P = 0.07). Among patients undergoing withdrawal of mechanical ventilation, there was no change in time from admission to withdrawal (hazard ratio = 1.0; P = 0.81).

Conclusions
A multifaceted quality improvement intervention was asso ciated with no improvement in quality of dying among ICU decedents, as well as no reduction in resource utilization prior to the withdrawal of life sustain ing measures.

Commentary
Th e subject of end-of-life care in the intensive care unit (ICU) has been an area of increasing interest for the past two decades. Because of high mortality among patients with critical illness, end-of-life care is an important component of intensive care, and yet studies suggest the current quality of this care is often poor [1]. Ideally, highquality end-of-life care would refl ect each patient's values and wishes, promote compassionate communication and provide family-oriented care. And by avoiding care that is not in keeping with each patient's wishes, ideal end-oflife care would lead to reduction of non-benefi cial care and avoidance of prolongation of dying, thus preventing patient suff ering and allowing for more equitable use of scarce ICU resources.
Th e fi rst major eff ort to improve the quality of care of high-risk patients at the end of life was SUPPORT -a large multicenter trial to determine if a specially trained nurse could facilitate physicians' knowledge of patients' resuscitation preferences and improve the quality of endof-life care [2]. Physicians in the intervention group received estimates of the likelihood of 6-month survival for every day up to 6 months. In addition, a trained nurse made multiple contacts with the patient, family, physician and hospital staff to elicit treatment preferences, improve understanding of outcomes, encourage attention to pain control and facilitate advance care planning and patientphysician communication. Unfortunately, patients in the intervention arm in SUPPORT experienced no improvement in the fi ve targeted outcomes: the incidence or timing of written do-not-resuscitate orders, physician's knowledge of their patient's preferences not to be resuscitated, the number of days spent receiving intensive treatment before death, and the level of reported pain. Nor did the intervention reduce the use of hospital resources. Th e relative failure of SUPPORT suggested that eff orts to improve communication were by themselves insuffi cient to improve care, and that more targeted interventions directed at clinician behavior were needed.
Th e subject of study by Curtis and colleagues was to evaluate the eff ectiveness of just such an intervention [3].
In an un-blinded cluster-randomized trial (targeting clinicians and hospital, not individual patients or family members), the authors sought to test the eff ect of an interdisciplinary, multifaceted quality improvement inter vention to improve ICU clinicians' ability to provide end-of-life care to critically ill patients and their families. Of 16 eligible hospitals in the greater Seattle, Washington area, 15 agreed to participate. Th ree hospitals were designated as pilot sites and 12 were randomized to receive either the intervention or control. Th e quality-improvement intervention was based on self-effi cacy theory: changes in clinician performance are facilitated by increasing knowledge, enhancing attitudes, and modeling appropriate behaviors [4]. Th e fi ve components of intervention included clinician education about palliative care in the ICU using variety of educational approaches (grand rounds, workshops, video presentations), identification and training of ICU clinician local champions for palliative care, academic detailing of nurse and physician ICU directors to address individual ICU-specifi c barriers to improving end-of-life care (environmental scan), feedback of individual ICU-specifi c quality data, including family satisfaction (audit and feedback), and implemen tation of system supports, such as palliative care order forms and get-to-know-me posters. Th e primary outcome measures included family member's rating of Quality of Dying and Death (QODD) assessed by questionnaire, nurse assessed QODD questionnaire, number of days in the ICU, and the time from ICU admission to withdrawal of mechanical ventilation, as well as key elements of palliative care (occurrence of a family conference within 72 hours of ICU admission, consul tation with palliative care experts, involvement of a spiritual care provider, social worker, avoidance of cardiopulmonary resuscitation in the last hour of life, DNR orders in place at the time of death). Although the investigators successfully implemented each of the intervention's components in a way that was well received by clinicians, the intervention had no measurable eff ect family satisfaction or family and nurse ratings of the quality of dying. Time from admission to withdrawal of life support was unchanged and ICU length of stay was also unchanged.
Strengths of the study were that the investigators made a concerted, sophisticated eff ort to promote state-of-theart end-of-life care, and that clinicians highly rated the educational and training components of the intervention. Th e fact that the intervention was so powerful, however, makes the results of this study somewhat frustrating, as it suggests that any educational intervention is unlikely to improve outcomes. Limitations of the study were that it was confi ned to one region of the United States, and that the outcome measures may not be sensitive enough to capture meaningful changes in end-of-life quality. Th ere are many factors related to family experiences in the ICU, and it is possible that the clinician's skill in the dying process provides little signal amidst all the noise. It is also possible that a larger sample might be needed to defi nitively exclude an important but smaller improvement in quality of dying.

Recommendation
Based on the results of this study, targeted quality improve ment eff orts to improve end-of-life care at the ICU level are premature. Instead, research eff orts should be focused on interventions that target clinicians earlier in their training or interventions with more direct interaction with patients that can be customized to patient needs. In the meantime, intensivists should not use this study as a reason to avoid paying attention to the needs of dying patients and their families, as this and other studies still indicate that high-quality end-of-life care is valued by caregivers and families alike [5].