End-of-Life Care in Cystic Fibrosis: Comparing Provider Practices Based on Lung Transplant Candidacy

Abstract

Background:

The optimal timing to introduce palliative care (PC) and end-of-life (EOL) conversations into the lives of people with cystic fibrosis (CF) has not been established.

Objective:

Compare EOL care practices for people with CF who died without a lung transplant (LT), are living without an LT, and those who received an LT.

Design:

Retrospective chart review.

Setting/Subjects:

People with CF who received care from 2012 to 2017 at the University of Texas Southwestern Medical Center.

Measurements:

Primary outcomes were (1) EOL discussion with a pulmonologist, (2) time of EOL discussion before death or LT, (3) evaluation by PC, and (4) documentation of advanced directive or medical power of attorney.

Results:

Twenty-three patients died without LT, 40 patients received an LT, and 222 were living without an LT. Among LT recipients, 10% had EOL conversations compared with 74% of deceased patients and 5% of living patients without LT (p = 0.001). Among deceased patients, 39% had EOL conversations more than six months before death, while 5% of transplanted patients had EOL conversation more than six months before LT (p < 0.001). Deceased patients were more likely to have seen PC (57%) than either patients who received LT (2%) or those living without LT (3%, p = 0.0001).

Conclusions:

Patients who died without LT were more likely to have seen PC and had an EOL conversation than patients who received LT or who are living without LT. Further research should explore the optimal timing to discuss EOL care and the best timing to involve PC.

Introduction

Cystic Fibrosis (CF) is a genetic disease in which the median predicted survival is improving due to recent advances in treatment, from 31 years in 1993 to 47.4 years in 2018.¹ However, people with CF still experience a shortened life expectancy and the most common cause of death is respiratory failure.¹ Lung transplant (LT) improves survival of people with CF^2,3 and is offered at a point in which one is at a high risk of dying from lung disease in the next two years.^4,5

Due to the shortened life expectancy and large symptom burden of people with CF,^6,7 it has been recommended that palliative care (PC) be introduced early into treatment planning discussions.⁸ The current guidelines for LT and CF recommend discussion of values and goals of care when FEV1 is <50% predicted,⁵ although it is not required by LT guidelines or consensus statements.⁴ However, the optimal timing for PC involvement and end-of-life (EOL) discussions has not been established. Due to high mortality at the time of consideration for LT, this may be an important time point for involvement of PC for people with CF.

It is well documented that people with CF who receive an LT are more likely to pursue aggressive care, die in the intensive care unit (ICU),^9–11 initiate do-not-resuscitate (DNR) orders later, and are less likely to participate in EOL discussions.¹⁰ Studies evaluating the involvement of PC in patients who are being considered for LT are limited in number^12,13 and have not investigated how LT candidacy affects EOL conversations; none of these studies focused on people with CF but rather included people with all end-stage lung diseases. The risks associated with lung transplantation, high symptom burden in people with end-stage CF, and the mortality associated with disease progression make the time of transplant referral a potentially advantageous time to discuss EOL preferences for some patients.

More information is needed regarding when and how to introduce PC to people with CF. Our objective is to describe and compare EOL care practices for people with CF who have died without an LT, those who received an LT, and those who are living without an LT at one academic medical center, in particular focusing on EOL conversations, involvement of PC, and documentation of advanced care planning.

Methods

Study design overview

We conducted a retrospective chart review of all people with CF who received care at the William P. Clements Jr. Hospital at the University of Texas (UT) Southwestern from August 1, 2012, to December 31, 2017. Patients were identified using the CF Foundation Patient Registry. Manual chart review was conducted to extract all data from the electronic health record (EHR). All chart review data were conducted by three authors A.V.L., A.R., and J.D.F.

Inclusion and exclusion criteria

During our study period, there were 348 adults who received care at the Adult CF Clinic and inpatient care at William P. Clements Jr. Hospital at UT Southwestern. All patients included in our study fit one of the following criteria: (1) died of any cause without LT, (2) received an LT for end-stage CF, or (3) were alive and receiving care without an LT between August 1, 2012, and December 31, 2017. All patients were at least 18 years of age at the time of death, transplant, or by August 1, 2012. The primary location of their CF care was UT Southwestern, which was determined by the fact that they had the majority of their visits at that institution. Among patients who died after receiving an LT in our time frame, we only included data leading up to their LT. We excluded patients who received the majority of their care at another institution due to limited data. During this time, there were 413 LTs at UT Southwestern. Of those 413, 12% were done for CF.

To better understand when EOL conversations occurred in the care of a patient, we collected data on FEV1. For deceased patients without an LT, we recorded the FEV1 nearest to the time of the EOL conversation. For patients who received an LT and those living without an LT, we also recorded the FEV1 closest to the time of their EOL discussion. If the patient did not have an EOL discussion, we recorded the FEV1 at the time of transplant referral (if this was made) or LT surgery. If neither of these occurred, we did not collect information regarding their FEV1.

Primary outcome variables

The primary outcomes for this study were (1) documentation of an EOL conversation with a pulmonologist, (2) timing of EOL discussion relative to death or LT, (3) involvement of PC, and (4) documentation of advanced directive or Medical Power of Attorney (MPOA). EOL conversation was defined by documentation in the EHR by a pulmonologist that code status, EOL preferences, worsening illness trajectory, goals of care, or limitations of the patient's care were discussed. If several EOL conversations were conducted, we chose the date of the first conversation. Patients deciding to pursue LT receive extensive education on the risks and benefits of LT, and choosing an LT could be viewed as a clear statement of a patient's goals of care. If a discussion of transplant and the alternatives occurred and were documented, this was considered an EOL conversation. However, a discussion of only LT does not ensure that this goal will be met, and if a provider did not document that the alternative of not receiving a transplant was discussed, this was not considered an EOL conversation. Furthermore, if the discussion of the decision to pursue LT is not explicitly documented and discussed in the EHR, the decision to pursue LT on its own was not considered an EOL conversation. We are interested in determining how often these conversations are held by pulmonologists who care for CF patients and how these conversations differ for patients who receive transplant and those who do not. We used manual chart review to determine if an EOL conversation was documented, whether the patient was evaluated by PC, and if an advanced directive or MPOA was documented. The timing of EOL discussion relative to death or LT was determined using the date of their death or LT from manual chart review. Evaluation by PC was defined as having had at least one visit, inpatient or outpatient, with a PC physician. During the study period, the PC team at UT Southwestern expanded to include advanced practice providers, social workers, pharmacists, and a chaplain; however, for the majority of the study period, the PC team included only physicians, and therefore, for the purpose of our study, the physicians' involvement was required to be considered as evaluation by PC. The presence of advanced directive or MPOA was determined by EHR review, not paper chart review. Paper documents are scanned into the EHR after hospital discharge, although we do not have data on the accuracy of this process.

Secondary outcome variables

Secondary outcomes included the presence of a DNR order. This was done by manual chart review and was defined by an order in the chart at the time of death or end of study period.

Statistical analysis

Various characteristics of subjects who did or did not receive a transplant with regard to EOL discussions were examined. Frequency counts and percentages were determined for the categorical variables for the entire group and for the transplant groups (received transplant yes or no). Chi-square contingency table analyses or Fisher's exact tests were utilized to compare the transplant groups. Means, medians, standard deviations, and ranges were determined for the numerical measures, including age, percent predicted FEV1, and time from DNR to death. Fisher exact tests for three independent groups were utilized to compare the groups on each of the categorical measures. Version 9.3 of SAS software was used for the analyses.

Results

Out of the 348 adults with CF who received care at UT Southwestern from August 1, 2012, to December 31, 2017, we identified 287 patients who met the study criteria, 42 patients who received an LT, 23 who died without LT, and 222 who were living without an LT (Fig. 1). Two patients who died were listed but died before LT, 9 patients were referred but were not listed, 10 patients were not referred for LT, and 2 patients declined transplant. Reasons for no referral for LT or not being listed for transplant after an evaluation included unexpected death, chronic infections such as Burkholderia or Mycobacterium abscessus, poor social support, and complex medical illness (congenital heart disease, thalassemia, requirement of simultaneously performing lung/liver/renal transplants). Reasons for exclusion included the following: age younger than 18 years at the time of death, time of LT, or at the beginning of study period, UT Southwestern was not the primary site of their CF care, LT occurred before the study window, or the patients died after an LT and the LT did not fall within the study window.

FIG. 1.

Consort diagram for patient inclusion in chart review.

Of the 42 patients who received an LT, median age at transplant was 29 years, among those who died without LT, median age at death was 27 years, and among those living without LT, median age at the end of the study period was 29 (Table 1). Patients who received an LT were 55% female, those who died without LT were 39% female, and those who were living without LT were 50% female. In patients who received both an LT and had an EOL discussion, the median FEV1 at the time of an EOL discussion was 23% predicted, and the median FEV1 at the time of the transplant discussion was also 23% predicted (Table 1). In patients who died without an LT, the median FEV1 at time of the EOL conversation was 32% predicted (Table 1). For those patients who were living without an LT, data are displayed similarly, and the median FEV1 was 32% predicted at time of the transplant discussion and 28% predicted at time of the EOL discussion.

Table 1.

Demographic Information for People with Cystic Fibrosis 2012–2017

Characteristic	Received LT (42) Median (range) or N (%)	Deceased without LT (23) Median (range) or N (%)	Living without LT (222) Median (range) or N (%)	CF annual report data for comparison median or %
Age in years at death, time of LT, or endpoint of study	29 (18–57)	27 (21–72)	29 (18–67)	30.8^a
Female gender	23 (55)	9 (39)	110 (50)	48.2^b
Genotype homozygous for DF 508	15 (36)	16 (70)	86 (39)	44.2^b
FEV1 at time of transplant discussion	23 (14–43) N = 38		32 (21–47) N = 24
FEV1 at time of EOL discussion	23 (19–26) N = 4	32 (18–75) N = 16	28 (17–57) N = 13
Location of death if deceased		ICU 10 (44)
		Ward 5 (22)
		Home without hospice 3 (13)
		Home with hospice 4 (17)
		Hospice facility 1 (4)

Median age at death, from CF Foundation Annual report 2018.

From CF Foundation Annual report 2018.

CF, cystic fibrosis; FEV1, forced expiratory volume in one second; LT, lung transplant; EOL, end of life; ICU, intensive care unit.

Of the deceased patients without LT, 74% had an EOL conversation with a pulmonologist, compared with 10% who received LT and 5% of the living without transplant group (p = 0.001, Table 2). Of the deceased patients, 39% had an EOL conversation more than six months before death; in contrast, among the patients who received LT, 5% had an EOL conversation more than six months before LT (p = 0.001). Of the patients who had an EOL conversation, the majority of these EOL conversations occurred in the age group 18–30, compared with patients who were 31–40 years old or >41 years old (Table 3). The triggering factors for EOL conversations are listed for the three groups, the most common of which is a CF exacerbation (Table 4). The majority of the EOL conversations for patients who died without an LT occurred in the inpatient setting, and approximately evenly split between the hospital ward and the ICU (Table 2).

Table 2.

Comparison of End-of-Life Conversations, Involvement of Palliative Care, Documentation of Advanced Directives in People with Cystic Fibrosis Who Received Lung Transplant or Died without Lung Transplant

Characteristic	Received LT (42) N (%)	Deceased (23) N (%)	Living without LT (222) N (%)	p-Value
EOL conversation with pulmonologist	4 (10.0)	17 (74)	10 (5)	0.001
Location of EOL conversation				0.215
Clinic	3 (7)	5 (22)	6 (3)
Hospital ward	1 (2)	7 (30)	4 (2)
Hospital ICU	0 (0)	5 (22)	0 (0)
Palliative care involved (at least one visit with attending physician)	1 (2)	13 (57)	6 (3)	0.0001
EOL conversation that occurred >6 months before death or LT	2 (5)	9 (39)	NA	0.001
Advanced directive	7 (17)	4 (17)	8 (4)	0.001
Medical power of attorney	11 (26)	9 (39)	18 (8)	0.0001
Intubation at time of death or transplant	4 (10)	7 (30)	NA	0.043
ECMO at time of death or transplant	4 (10)	3 (13)	NA	0.690
DNR order documented in chart at time of death or transplant	0 (0)	18 (78)	NA	0.001
Time of DNR order before death
<1 month	NA	15 (65)	NA
1–6 months		3 (13)
>6 months		0 (0)

DNR, do-not-resuscitate; ECMO, extracorporeal membrane oxygenation; NA, not applicable.

Table 3.

Age Distribution of End-of-Life Conversations in People with Cystic Fibrosis 2012–2017

	Received LT			Deceased without LT			Living without LT
	EOL conversation documented	No EOL conversation documented	Total	EOL conversation documented	No EOL conversation documented	Total	EOL conversation documented	No EOL conversation documented	Total
	N (%)	N (%)	N (%)	N (%)	N (%)	N (%)	N (%)	N (%)	N (%)
18–30 Years old	3 (8)	19 (50)	22 (52)	9 (39)	5 (22)	14 (61)	5 (2)	118 (53)	123 (55)
31–40 Years old	0	9 (24)	9 (21)	5 (22)	0	5 (22)	3 (1)	57 (26)	60 (27)
>41 Years old	1 (2)	10 (26)	11(26)	3 (13)	1 (4)	4 (17)	2 (1)	37 (17)	39 (18)
Total	4	38	42	17	6	23	10	212	222

Table 4.

Factors Triggering End-of-Life Discussion among People with Cystic Fibrosis 2012–2017

Triggering factor	Deceased without LT who had EOL conversation (17)	Received LT and had EOL conversation (4)	Living without LT and had EOL conversation (10)
Triggering factor	N (%)	N (%)	N (%)
CF exacerbation	8 (47)	2 (50)	4 (40)
Simultaneous discussion of lung transplant	2 (12)	0	3 (30)
Patient determined not to be a lung transplant candidate	3 (18)	0	1 (10)
Patient critically ill with no other therapeutic options	4 (23)	0	0
Chronic decline in lung function, not associated with acute exacerbation	0	2 (50)	2 (20)

Of the 23 patients who died without LT, 57% had seen PC, compared with 2% of patients who received LT and 3% of the living without LT group (p = 0.0001, Table 2). Documentation of an advanced directive was also different among groups, with the LT and deceased groups both having 17% with advanced directives and only 4% of the living without LT group having an advanced directive (p = 0.001). MPOA was more common in the deceased group (39%) compared with the transplant group (26%) and the living without transplant group (8%, p = 0.0001). DNR orders were very different between the deceased group and the transplanted groups. No patient who received an LT had a DNR order before the LT, while 78% of the patients who died had a DNR order in place at time of death. The majority of DNR orders occurred less than one month before death and none occurred more than six months before death.

Discussion

This study was designed to compare EOL practices among pulmonologists for people with CF who died without an LT, received a transplant, or were living without an LT. Patients who died without an LT were more likely to have an EOL conversation and see PC than patients in the other two groups. Advanced care planning (either an advanced directive or MPOA) occurred most frequently in the deceased group compared with the other two groups, but rates of both advanced directives and MPOA were low. We also found a difference between groups in the number of DNR orders, including no DNR orders among patients who received an LT. The difference in DNR orders is a result of a practice preference at our center by the LT team to have all patients who are listed for an LT remain full code. More EOL conversations occurred in the younger age group than the older age groups for all three patient groups (Table 3). The greater number of younger patients who had an EOL conversation may be due to the fact that overall more people with CF are in the lower age groups. This finding is unexpected because it is perceived that having EOL conversations with older patients is easier than for younger patients. It is difficult to compare our results with those of other studies because other studies of people with CF and EOL conversations do not stratify their results by age. These results suggest that patients who receive an LT and those who are living without LT are less likely to have EOL conversations or have an evaluation by PC even though they also have high expected mortality due to end-stage CF. Before LT, evaluation may be an important time for people with CF to discuss EOL preferences with their CF providers.

Previous studies have shown that despite the recommendations⁸ for early introduction of PC and discussion of goals of care⁵ for people with CF, in practice this does not occur. The majority of physicians and their patients with CF do not discuss EOL preferences^14,15 despite the fact that most patients would be comfortable talking about it.^14,16,17 Furthermore, most people with CF still die in a hospital setting, frequently in the ICU,^11,18,19 and with late involvement of PC services,^18,19 often when the patient is too ill to participate in EOL discussions.²⁰ In contrast to practice patterns, surveys indicate that most patients and caregivers prefer that EOL discussions occur during a period of clinical stability and not during acute illness or directly before death.^14,15,20 Physician surveys have demonstrated that CF physicians are overall ambivalent about when to have a discussion about EOL care, and that logical times for this discussion may be at the transition time between pediatric and adult care or at the time of LT referral.²¹ In our study, the rates of discussion of EOL conversations were higher in the deceased group and the transplanted group than that reported by Linnemann et al., who report 5% of patients in a sample of 38 people with CF had discussed EOL preferences with a member of the CF team.¹⁴ This number is similar to our rate of EOL conversations for patients living without LT. The rate of EOL conversations occurring in the group that died without LT (74%) was also higher than the rates reported by Chen et al., although our results demonstrated less involvement of PC.¹¹ The location of EOL conversations in our study among patients who died was most commonly in the hospital and not in clinic, contrary to the studies of patient preferences.

A few of our results were different than what we expected to find. Interestingly, among the patients who received an LT and those who were living without an LT, the majority of EOL conversations occurred in the clinic setting. Furthermore, among patients who died, 39% of these patients had an EOL discussion more than six months before death. These results suggest continued barriers to EOL conversations with people with CF in the clinic setting, many EOL conversations are happening earlier than previously documented by other studies, and the importance of pulmonary physicians' development of a primary PC skill set. Finally, the median FEV1 at the time of LT discussion and EOL discussion for patients who received an LT was the same, suggesting that clinicians consider EOL discussion at the same time point that they consider LT in transplant candidates.

Our results also indicate that, despite the majority of patients who died without an LT having had an EOL conversation, 44% of these patients still died in the ICU, 30% while intubated, and 13% on extracorporeal membrane oxygenation. In addition, few patients completed advanced directives or MPOA documents. Taken together, we suspect this pattern of high numbers of EOL conversations, high utilization of aggressive care at EOL, and rare completion of advanced care planning documentation reflects the challenge of having EOL conversations with a relatively young chronically ill population, and barriers to which are discussed below.

There are many barriers to EOL discussions among people with CF and their providers. Patient attitudes, such as the notion that PC and LT are mutually exclusive,²² the importance of maintaining hope, unrealistic expectations regarding outcomes of cardiopulmonary resuscitation,^23,24 and unrealistic expectations about recovery from exacerbations,²⁵ are all possible contributing barriers to EOL conversations. Physician surveys have demonstrated that time constraints²⁶ and fear of taking away a patient's hope²¹ as barriers. In addition, although there have been several attempts to develop prognostication tools,^27,28 determining the life span remains a challenge in CF. The organ failure illness trajectory makes it difficult for physicians to determine which exacerbation will be the terminal one,²⁹ and the addition of CF transmembrane conductance regulator modulators slowing the progression of CF lung disease makes it more difficult to determine the proper time to have this discussion.³⁰ All of these factors create uncertainty for the physician regarding when the topic of EOL and PC should be introduced.²¹ These beliefs held by both patients and providers may have contributed to the different EOL practices seen in our study. One way to overcome the difficulty with prognostication as a barrier to EOL conversations and involvement of PC would be to choose a specific time point for these conversations, such as the time of LT evaluation.

In contrast to LT patients, involvement of PC during the evaluation process for heart transplant candidates is recommended by guidelines and has resulted in improved patient and family experiences. The American College of Cardiology Foundation/American Heart Association guidelines for heart transplant recommend involvement of PC in all patients being considered for a transplant or mechanical circulatory support devices.³¹ This practice has been shown to be beneficial by helping families understand patient preferences related to quality of life, financial issues, and the use of invasive measures.³² These data from heart transplant patients demonstrate that before LT may be an advantageous time to have these conversations and introduce PC to people with CF. Involvement of PC at the time of consideration for LT may help patients and families plan for complications after LT and may also help with coping with the patient's ultimate decline should transplant not occur.

Our study had several potential limitations. First is the size of the study. Due to time and funding constraints, chart review was limited to one academic institution, which led to a small sample size and may not be widely generalizable. Our single-center design may introduce other biases such as our transplant team's requirement of a “full code” order for those people on the transplant list, and limited involvement of PC. Second, it is possible that not all EOL conversations were documented in the medical record due to time constraints of the providers and human error, and therefore, it is possible that there are fewer EOL conversations documented in our study than actually occurred. However, given that this would likely affect both groups equally, it is unlikely that this would bias the results of the study. Third, this is a retrospective chart review, and therefore, we cannot assume any cause and effect but only associations between variables. Fourth, two patients remained listed for LT just before death potentially delaying the EOL conversation that normally occurs in patient with end-stage lung disease. Finally, it is possible that the advanced care planning documents were not scanned into the EHR after hospital discharge due to human error, resulting in a lower than actual number of documents completed.

In conclusion, our study was designed to compare the EOL practices for people with CF who died without LT, received an LT, and were living without LT. We found that patients who died without an LT were more likely to see PC and have an EOL conversation than patients who received an LT or patients who were living without an LT. Future areas of research should investigate the optimal timing for involvement of PC and EOL conversations and development of a primary PC curriculum for pulmonary physicians.

Footnotes

Funding Information

No funding was received for this study.

Author Disclosure Statement

No competing financial interests exist.

References

Cystic Fibrosis Foundation Patient Registry 2018 Annual Data Report: https://cff.org/Research/Researcher-Resources/Patient-Registry/2018-Patient-Registry-Annual-Data-Report.pdf 2019. (Last accessed December 9, 2019).

Titman

, Rogers

, Bonser

, et al.: Disease-specific survival benefit of lung transplantation in adults: A national cohort study. Am J Transplant, 2009; 9:1640–1649.

Thabut

, Christie

, Mal

, et al.: Survival benefit of lung transplant for cystic fibrosis since lung allocation score implementation. Am J Respir Crit Care Med, 2013; 187:1335–1340.

Weill

, Benden

, Corris

, et al.: A consensus document for the selection of lung transplant candidates: 2014—an update from the Pulmonary Transplantation Council of the International Society for Heart and Lung Transplantation. J Heart Lung Transplant, 2015; 34:1–15.

Ramos

, Smith

, McKone

, et al.: Lung transplant referral for individuals with cystic fibrosis: Cystic Fibrosis Foundation consensus guidelines. J Cyst Fibros, 2019; 18:321–333.

Sawicki

, Sellers

, Robinson

: Self-reported physical and psychological symptom burden in adults with cystic fibrosis. J Pain Symptom Manage, 2008; 35:372–380.

Quittner

, Goldbeck

, Abbott

, et al.: Prevalence of depression and anxiety in patients with cystic fibrosis and parent caregivers: Results of The International Depression Epidemiological Study across nine countries. Thorax, 2014; 69:1090–1097.

Yankaskas

, Marshall

, Sufian

, et al.: Cystic fibrosis adult care: Consensus conference report. Chest, 2004; 125(1 Suppl):1S–39S.

Ford

, Flume

: Impact of lung transplantation on site of death in cystic fibrosis. J Cyst Fibros, 2007; 6:391–395.

10.

Dellon

, Leigh

, Yankaskas

, et al.: Effects of lung transplantation on inpatient end of life care in cystic fibrosis. J Cyst Fibros, 2007; 6:396–402.

11.

Chen

, Homa

, Goggin

, et al.: End-of-life practice patterns at U.S. adult cystic fibrosis care centers: A national retrospective chart review. J Cyst Fibros, 2017 [Epub ahead of print]; DOI: 10.1016/j.jcf.2017.08.010.

12.

Colman

, Singer

, Barua

, et al.: Outcomes of lung transplant candidates referred for co-management by palliative care: A retrospective case series. Palliat Med, 2015; 29:429–435.

13.

Freeman

, Le

, Singer

, et al.: Impact of a transplant palliative care clinic on symptoms for patients awaiting lung transplantation. J Heart Lung Transplant, 2016; 35:1037–1039.

14.

Linnemann

, Friedman

, Altstein

, et al.: Advance care planning experiences and preferences among people with cystic fibrosis. J Palliat Med, 2018; 22:138–144.

15.

Chen

, Killeen

, Peterson

, et al.: Evaluation of pain, dyspnea, and goals of care among adults with cystic fibrosis: A comprehensive palliative care survey. Am J Hosp Palliat Care, 2017; 34:347–352.

16.

Sawicki

, Dill

, Asher

, et al.: Advance care planning in adults with cystic fibrosis. J Palliat Med, 2008; 11:1135–1141.

17.

Sawicki

, Sellers

, McGuffie

, et al.: Adults with cystic fibrosis report important and unmet needs for disease information. J Cystic Fibros, 2007; 6:411–416.

18.

Mitchell

, Nakielna

, Tullis

, et al.: Cystic fibrosis. End-stage care in Canada. Chest, 2000; 118:80–84.

19.

Philip

, Gold

, Sutherland

, et al.: End-of-life care in adults with cystic fibrosis. J Palliat Med, 2008; 11:198–203.

20.

Dellon

, Shores

, Nelson

, et al.: Caregiver perspectives on discussions about the use of intensive treatments in cystic fibrosis. J Pain Symptom Manage, 2010; 40:821–828.

21.

Dellon

, Sawicki

, Shores

, et al.: Physician practices for communicating with patients with cystic fibrosis about the use of noninvasive and invasive mechanical ventilation. Chest, 2012; 141:1010–1017.

22.

Braithwaite

, Philip

, Tranberg

, et al.: End of life care in CF: Patients, families and staff experiences and unmet needs. J Cyst Fibros, 2011; 10:253–257.

23.

Diem

, Lantos

, Tulsky

: Cardiopulmonary resuscitation on television. Miracles and misinformation. N Engl J Med, 1996; 334:1578–1582.

24.

Harris

, Willoughby

: Resuscitation on television: Realistic or ridiculous? A quantitative observational analysis of the portrayal of cardiopulmonary resuscitation in television medical drama. Resuscitation, 2009; 80:1275–1279.

25.

Murtagh

, Preston

, Higginson

: Patterns of dying: Palliative care for non-malignant disease. Clin Med (Lond), 2004; 4:39–44.

26.

Colman

, Curtis

, Nelson

, et al.: Barriers to optimal palliative care of lung transplant candidates. Chest, 2013; 143:736–743.

27.

Kerem

, Reisman

, Corey

, et al.: Prediction of mortality in patients with cystic fibrosis. N Engl J Med, 1992; 326:1187–1191.

28.

Nkam

, Lambert

, Latouche

, et al.: A 3-year prognostic score for adults with cystic fibrosis. J Cyst Fibros, 2017; 16:702–708.

29.

Murray

, Kendall

, Boyd

, et al.: Illness trajectories and palliative care. BMJ, 2005; 330:1007–1011.

30.

Middleton

, Mall

, Drevinek

, et al.: Elexacaftor-tezacaftor-ivacaftor for cystic fibrosis with a single Phe508del allele. N Engl J Med, 2019; 381:1809–1819.

31.

Yancy

, Jessup

, Bozkurt

, et al.: 2013 ACCF/AHA guideline for the management of heart failure: A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol, 2013; 62:e147–e239.

32.

Swetz

, Freeman

, AbouEzzeddine

, et al.: Palliative medicine consultation for preparedness planning in patients receiving left ventricular assist devices as destination therapy. Mayo Clin Proc, 2011; 86:493–500.