Palliative Care Consults Among Critically Ill Patients: Does Timing Matter?

Introduction

Palliative care has become an integral part of providing quality care to patients in the intensive care unit (ICU).^1,2 Critically ill patients often have a high symptom burden, require complex care coordination, and can benefit from early patient and family support. ³ Palliative care involvement has led to benefits of earlier goals of care meetings that allow for rapport building and expectation setting with patients and their families, resulting in shorter hospital length of stay (LOS). ⁴ Although benefits of palliative care are clear, workforce limitations of specialty-trained palliative care clinicians restrict our ability to meet the needs of these critically ill patients.^5,6,

To increase access to palliative care, we embedded a specialty palliative care team of a physician and a registered nurse into the medical intensive care unit (MICU) to deliver palliative care to patients admitted to the MICU.^7-10 The team structure was designed to best meet the needs of patients and clinicians in the MICU, and was developed on published research, palliative care clinician observation, and feedback from the MICU team. The embedded team screened the EMR of each MICU patient to assess the timing and urgency of seeing the MICU patients, based on concurrent serious illness, LOS, symptoms, and current clinical status. Next, during interdisciplinary rounds, the embedded palliative care team reviewed each patient currently admitted to the MICU with the MICU interdisciplinary team (IDT) (e.g., physician, social worker, nurse manager, and physical therapy). During these rounds, highest priority patients for new consult and highest priority family meetings would be flagged. Third, following rounds, the palliative care RN and physician determined which clinician would be most appropriate to complete the consults (e.g. physician for symptom management or RN for identification of surrogate decision-maker, in the case of a patient without capacity due to intubation and sedation). Finally, patients were seen for symptom management, psychosocial support, or goals of care meetings. Nonetheless, many patients did not receive a palliative care consult during their MICU admission. In this manuscript, we compared patients seen by the embedded palliative care team during the MICU stay and those seen after the MICU stay and to identify sociodemographic or clinical characteristics to determine which patients should be prioritized to receive an in-ICU consultation.

Methods

Results

Of the 258 patients, 195 (76%) received a palliative care consult during their MICU stay (in-ICU consult) by the embedded team and 63 (24%) patients received a post-ICU consult by other members of the specialty palliative care team. There were no statistically significant differences in sociodemographic characteristics between the 2 groups (Table 1). There was no statistical difference in the prevalence of the five most common serious illness diagnoses between groups (Table 1). For both groups, the median Karnofsky performance status (KPS) at the time of palliative care consultation was 20%, which represents “patients who require extensive nursing care by professionals”. In both groups, the median Elixhauser comorbidity index was six.

OPEN IN VIEWER

Table 1.

Comparison of Sociodemographic and Clinical Characteristics of Patients Who Received an In-ICU Palliative Care Consultation and Those Who Received a Post-ICU Palliative Care Consultation

	In-ICU (n = 195)	Post-ICU (n = 63)	P-value
Mean age, years (median)	66.4 (68)	67.1 (69)	0.793
Female (%)	51.8	61.9	0.21
Race (%)			0.38
White	23.6	23.8
African-American	25.1	17.5
Hispanic	29.2	39.7
Other	13.8	6.3
Primary language (%)			0.05
English	72.8	69.8
Spanish	13.3	23.8
Other	13.8	6.3
Insurance (%)			0.71
Medicare	30.4	28.3
Medicaid	37.2	42.4
Elixhauser comorbidity index, mean (median)	6	6	0.880
Serious illness diagnoses, 5 most common (%)
Sepsis	62.3	63.3	0.99
Cancer	20.6	20.2	0.99
Liver disease	18.5	25.3	0.36
Renal failure	15.4	19	0.67
Dementia	11.7	10.1	0.85
Mean hospital LOS, days (median)	32.4 (20.4)	38 (30)	0.15
Mean ICU LOS, days (median)	13.4 (10)	10.6 (7)	0.03 (<0.01)
Mean KPS at the time of consult, % (median)	22.3 (20)	23.5 (20)	0.084
Mean number of days from palliative care consult to hospital discharge, (median)	21 (14)	14 (6)	<0.01

There were statistically significant differences (P < .01) in primary reason for palliative care consultation among those who received an in-ICU consult and those who received a post-ICU consult (goals of care: 63.6% (in-ICU) vs 44.4% (post-ICU); transfer to in-hospital palliative care unit: 1% (in-ICU) vs 14.3% (post-ICU); physical symptoms 27.7% (in-ICU) vs 39.7% (post-ICU); and emotional symptoms: 7.2% (in-ICU) vs 7.9% (post-ICU)).

Patients who were seen in-MICU had a longer ICU LOS (10 (in-ICU) vs 7 days (post-ICU), P < 0.01). These patients also had more days between palliative care consultation and hospital discharge (median 14 (in-ICU) vs 6 days (post-ICU), P < .01) and were more likely to have life-prolonging goals at the time of consultation (40% (in-ICU) vs 25.4% (post-ICU), P < .01). There was no significant difference between in-hospital mortality rate among those who received an in-ICU consult and those who received a post-ICU consult (60% (in-ICU) vs 57.1% (post-ICU), P = .73). (Figure 1) There were also no significant differences in discharge disposition between in-ICU and post-ICU consults (facility: 22.1% (in-ICU) vs 23.8% (post-ICU); home: 14.4% (in-ICU) vs 14.3% (post-ICU); hospice 3.1% (in-ICU) vs 3.2% (post-ICU)).

OPEN IN VIEWER

Figure 1.

Comparison of Percentage of Life-Prolonging Goals at the Time of Palliative Care Consultation and In-Hospital Mortality Among Patients Who Received an In-ICU Palliative Care Consultation and Those Who Received a Post-ICU Consult

Discussion

In order to increase access of palliative care to critically ill patients, it is imperative to consider innovative models of palliative care delivery within the ICUs. Even with an embedded, dedicated MICU model, which continued to be prioritized by the healthcare system, it was not possible to see all patients within the ICU due to workforce limitations, short MICU stays, or clinical uncertainty, among other factors. Our findings suggest that there were no differences in sociodemographic and clinical characteristics between those who received an in-ICU palliative care consult and those who received a post-ICU consult. Furthermore, markers of severity of illness, including functional status and prevalence of comorbidities were not significantly different between the two groups.

Yet, there were differences in the acute healthcare utilization markers of patients who received an in-ICU and post-ICU consult. Specifically, patients seen in-ICU had a longer ICU LOS, more days between palliative care consultation and discharge, and had a higher rate of life-prolonging goals at the time of consultation, as compared to those who received a post-ICU consult. Of note, patients who received a post-ICU consult were seen by palliative care much closer to death. A higher percentage of post-ICU consults were for transfer to the palliative care unit, which is typically meant as a short-stay unit and often for patients at the end of their life. This suggests that these post-ICU consults were for patients later in their disease trajectory who were too ill to leave the hospital and more likely to shift their goals to comfort-focused care. Regardless of the timing of the palliative care consult, the in-hospital mortality was high and did not differ significantly. This suggests that patients not seen while in the ICU should receive access to palliative care soon after ICU discharge.

There are limitations to our study. This study was conducted in an urban quaternary academic hospital with an established, well-resourced program and therefore may not be generalizable to other health systems. We also only included patients who received palliative care consultation and not those who never received a palliative care consultation. Post-ICU consultations were completed after receiving a consultation request from the primary team (i.e., hospital medicine, oncology) rather than the ICU team directly working with the embedded team. As a result, there may have been other factors that we did not capture which impacted the decision to consult palliative care. Of those that received palliative care consultation, we did not capture the number of encounters with each individual patient. Similarly, though we collected occurrence of advance care planning documentation, we were not able to evaluate the quality of advance care planning. Of note, we included only patients who had a MICU stay, which may not allow this data to be generalizable to other types of ICUs, such as cardiac or neurologic, due to different needs of these distinct clinical populations.

Although growing evidence supports delivery of palliative care at all ages and stages, regardless of risk of in-hospital mortality, efforts to identify which ICU patients need in-ICU consult are limited. Our findings suggest that sociodemographic and clinical characteristics are not used to determine who accesses palliative care among patients who spend at least 72 hours in the ICU. Furthermore, ICU clinicians may not be able to identify patients upstream given they have so many responsibilities and considerations to manage that palliative care may not be high priority. Many patients who are discharged from the ICU without palliative care die in the hospital. Overall, critically ill patients seen by palliative care in-ICU and those seen post-ICU have similar in-hospital mortality, which serves as a proxy marker for unmet palliative care needs as evidence has shown patients at the end of life often benefit from palliative care.^14,15 When using in-hospital mortality as a proxy for palliative care needs, we determined the embedded model is resource intensive and remains insufficient to meet the requirements of critically ill patients. Therefore, innovative models to ensure more seriously ill patients receive palliative care during or soon after their ICU stay, in spite of the limited specialty palliative care workforce are required. Potential solutions may include a formal check-in for assessment of palliative care needs prior to ICU discharge, an automatic trigger for patients to be seen based on their ICU LOS, and a primary palliative care curriculum to educate all providers about serious illness communication skills and symptom management. As a result of this work, we have standardized timing of in-ICU palliative care consultation to all patients with an ICU LOS of at least 5 days. In addition, we have also prioritized seeing patients as they are transferred out of the ICU to the floor, with the understanding that these transition points can be critical to their clinical trajectory.