Advance Care Planning Practice Patterns Before and During a Pandemic

Abstract

Introduction:

The COVID-19 pandemic resulted in introspection of the United States health care infrastructure, especially with advance care planning (ACP).

Methods:

This is a retrospective chart review assessing the frequency of ACP discussions and formal documents reflecting ACP wishes in an outpatient palliative medicine (PM) practice. The study site was at University Medical Center New Orleans from pre-COVID-19 surge (November 2019–February 2020) to months during and post-COVID-19 surge (March–April 2020).

Results:

Results showed an increase in ACP discussions during the post-COVID-19 months. Patients seen during the surge and post-COVID-19 months were more likely to discuss medical power of attorney (odds ratio [OR] = 1.78, p = 0.045) and preferred code status (OR = 2.82, p < 0.001).

Conclusion:

Our study showed that more ACP discussions were conducted post-COVID-19 versus pre-COVID-19. However, formal documents reflecting these wishes were lacking. These results can help guide ACP use in crisis periods and improve understanding of ACP discussions in an outpatient PM clinic.

Introduction

The COVID-19 pandemic forced reflection and evaluation of health care infrastructure. COVID-19, caused by the SARS-CoV-2 virus, resulted in significant morbidity and mortality.^1,2 Palliative medicine (PM) took a major role in the pandemic by focusing on caring for people living with serious illnesses and facilitating communication around goals of care (GOC) and quality of life (QOL).³

Advance care planning (ACP) is a process wherein patients discuss health care goals with their health care providers.⁴ Studies have shown that ACP positively impacts QOL in end-of-life care settings.^5–10 There are several ways to document ACP discussions, including completion of medical power of attorney (mPOA) forms, living wills, and/or physician orders for life-sustaining treatment (POLST) forms. This study assessed the frequency of ACP discussions and documentation in an outpatient (PM) clinic before the onset of the COVID-19 pandemic. We examined how the initial COVID-19 surge—in one of the hardest hit metro regions in the United States—impacted this.^11–13

Methods

Participants and study design

A retrospective chart review was performed on 309 patients seen in a free-standing PM clinic at an academic safety-net medical center in New Orleans, Louisiana, between November 2019 and April 2020. The first COVID-19 surge was experienced between March and April 2020. This study examined differences in ACP documentation during PM encounters occurring prepandemic (November 1, 2019, to March 8, 2020), and during the COVID-19 surge (March 9, 2020, to April 30, 2020).

The first confirmed case of COVID-19 in the region was in March 9; this date was used to distinguish the two data sets. Inclusion criteria were 18 years or older, and patients who attended the PM clinic during this period. Many patients were seen frequently due to high symptom burden; therefore, the study team used narrative data from the patient's most recent visit only.

Data collection

Data were collected by a team of individuals through review of demographic information listed in the electronic medical record (EMR), presence of an advance directive under the ACP tab, and review of the clinician's narrative in the most recent PM note. Data were coded charts for the following characteristics: age, gender, race/ethnicity (Black, Caucasian, Hispanic, or Asian), and diagnosis (cancer vs. noncancer). Additional coding included visits occurring during the COVID-19 surge (after March 9, 2020), telemedicine appointments, use of an interpreter, and whether any indicators of ACP (listed below) occurred during the visit or at any prior visit, as listed under the ACP tab.

Data were collected on three indicators of physician–patient communication relating to ACP (coded as 1 = occurred during or before visit; 0 = did not occur): 1.

GOC discussion

mPOA discussion

Preferred code status discussion

Data were also collected on four indicators of ACP documentation (coded as 1 = occurred during or before visit; 0 = did not occur):

Completed mPOA document

Completed POLST document

Living will documentation

Composite variable, indicating whether at least one of the three specific indicators above occurred

Statistical analyses

Statistical analysis was conducted using SPSS version 27 data analysis software. Descriptive statistics were examined in the overall patient cohort and within two subgroups: visits occurring prepandemic and during the COVID-19 surge.

The chi-square test of independence described unadjusted differences between these two subgroups on dependent variables of interest. The dependent variables included the three dichotomous indicators of physician–patient communication (GOC, mPOA, and preferred code status discussion), and three of the four dichotomous indicators of ACP documentation (mPOA document, POLST document, composite indicator).

Primary analyses included six adjusted binary logistic regression models predicting each of the six dependent variables of interest. Models tested whether the COVID-19 surge was associated with changes in physician–patient communication and in ACP documentation. The independent variable was the COVID-19 surge, and we adjusted for the following covariates: age (coded in decade units), female gender, race/ethnicity (non-Caucasian vs. Caucasian), cancer diagnosis (vs. noncancer diagnosis), and whether an interpreter was used.

Results

Sample characteristics and unadjusted analyses

An analysis of 309 male (51.1%) and female (48.9%) patients occurred. The average age was 56.60 years old (standard deviation = 11.54). The majority were Black (n = 186, 60.2%) or Caucasian (n = 89, 28.8%) and had a cancer diagnosis (n = 263, 85.1%). Almost half of the patients were seen during the COVID-19 surge (n = 144, 46.6%), and all COVID-19 surge visits were held through telemedicine. Table 1 summarizes descriptive statistics in the overall sample and across subgroups.

Table 1.

Sample Characteristics

Characteristic	n (%)
Characteristic	Overall sample (N = 309)	Prepandemic (n = 165)	COVID-19 surge (n = 144)
Age, mean (SD)	56.60 (11.54)	57.75 (11.84)	55.27 (11.07)
Gender, female	151 (48.9)	81 (49.1)	70 (48.6)
Race/ethnicity
Black	186 (60.2)	112 (67.9)	74 (51.4)
Caucasian	89 (28.8)	36 (21.8)	53 (36.8)
Hispanic	26 (8.4)	14 (8.5)	12 (8.3)
Asian	8 (2.6)	3 (1.8)	5 (3.5)
Diagnosis
Cancer	263 (85.1)	132 (80.0)	131 (91.0)
Noncancer	46 (14.9)	33 (20.0)	13 (9.0)
Telehealth visit^a	144 (46.6)	0 (0.0)	144 (100.0)
Use of interpreter	25 (8.1)	14 (8.5)	11 (7.6)
Physician–patient communication
Goals-of-care discussed	178 (57.6)	99 (60.0)	79 (54.9)
mPOA discussed	229 (74.1)	113 (68.5)	116 (80.6)
Preferred code status discussed	134 (43.4)	56 (33.9)	78 (54.2)
Advance care planning documentation
mPOA	88 (28.5)	58 (35.2)	30 (20.8)
POLST	28 (9.1)	19 (11.5)	9 (6.3)
Living will	3 (1.0)	2 (1.2)	1 (0.7)
Any completed advance care planning document^b	104 (33.7)	68 (41.2)	36 (25.0)

Note: N = 309. Visits during the COVID-19 surge occurred on March 9, 2020, to April, 2020. Prepandemic visits occurred from November, 2019, to March 8, 2020.

Perfect correspondence between telehealth visits and COVID-19 surge visits.

Composite variable indicating presence of at least one of the following: mPOA, POLST, living will.

POLST, Physician Orders for Life Sustaining Treatment; mPOA, Medical Power of Attorney; SD, standard deviation.

Results from unadjusted analyses found between-group differences in ACP based on the visit date (Table 1). Compared with patients seen before the pandemic, a higher proportion of visits during the COVID-19 surge included discussions regarding mPOA (80.6% vs. 68.5%, p = 0.016) and preferred code status (54.2% vs. 33.9%, p < 0.001). However, a smaller proportion of patients seen during the COVID-19 surge completed any ACP documents (25.0% vs. 41.2%, p = 0.003), particularly mPOA documents (20.8% vs. 35.2%, p = 0.005).

Changes in physician–patient communication

Consistent with unadjusted analyses, patients seen during the COVID-19 surge were more likely to discuss their mPOA (odds ratio [OR] = 1.78, p = 0.045) and preferred code status (OR = 2.82, p < 0.001). In addition, with each increasing decade of age, patients were more likely to discuss GOC (OR = 1.43, p = 0.002) and preferred code status (OR = 1.70, p < 0.001). Patients with a cancer diagnosis were more likely to discuss mPOA than those with a noncancer diagnosis (OR = 2.22, p = 0.025). Patients who required use of an interpreter were less likely to discuss their mPOA (OR = 0.36, p = 0.020). Table 2 demonstrates complete results from the adjusted binary logistic regression models assessing associations between the COVID-19 surge and each dependent variable indicating physician–patient communication.

Table 2.

Binary Logistic Regression With COVID-19 Surge Predicting Changes in Physician–Patient Communication

	Goals-of-care discussion			mPOA discussion			Code status discussion
	OR	95% CI	p	OR	95% CI	p	OR	95% CI	p
COVID-19 surge	0.89	0.55–1.43	0.620	1.78	1.01–3.11	0.045	2.82	1.71–4.68	<0.001
Covariates
Age	1.43	1.15–1.77	0.002	1.24	0.98–1.57	0.069	1.70	1.34–2.16	<0.001
Gender, female	1.34	0.84–2.12	0.220	0.75	0.44–1.28	0.287	0.78	0.48–1.25	0.298
Diverse race/ethnicity	0.85	0.50–1.44	0.548	0.71	0.37–1.35	0.291	0.69	0.40–1.17	0.170
Cancer diagnosis	0.72	0.36–1.45	0.358	2.22	1.11–4.45	0.025	0.60	0.29–1.23	0.162
Interpreter	1.09	0.45–2.64	0.856	0.36	0.15–0.85	0.020	1.01	0.40–2.52	0.990

Note: N = 309. Age is in decade units to show how a 10-year change in age is associated with differences in outcomes. Race is coded as non-Caucasian versus Caucasian.

CI, confidence interval; OR, odds ratio.

Changes in ACP documentation

Consistent with unadjusted analyses, patients seen during the COVID-19 surge were less likely to have an ACP document completed (OR = 0.48, p = 0.003), especially an mPOA (OR = 0.49, p = 0.006). With each increasing decade of age, patients were less likely to complete an mPOA (OR = 0.79, p = 0.049) and yet were more likely to complete a POLST document (OR = 2.16, p < 0.001). Table 3 displays results from the adjusted binary logistic regression models assessing associations between COVID-19 surge covariates and each dependent variable indicating ACP documentation.

Table 3.

Binary Logistic Regression With COVID-19 Surge Predicting Changes in Advance Care Planning Documentation

	mPOA document			POLST document			Any ACP document ^a
	OR	95% CI	p	OR	95% CI	p	OR	95% CI	p
COVID-19 surge	0.41	0.24–0.71	0.001	0.56	0.23–1.36	0.200	0.45	0.27–0.74	0.002
Covariates
Age	0.79	0.62–1.00	0.049	2.16	1.45–3.22	<0.001	0.98	0.79–1.22	0.867
Gender, female	0.75	0.45–1.26	0.277	1.77	0.77–4.06	0.181	1.01	0.63–1.64	0.959
Diverse race/ethnicity	0.95	0.53–1.68	0.850	0.71	0.28–1.82	0.480	0.84	0.49–1.45	0.537
Cancer diagnosis	1.93	0.87–4.28	0.105	1.32	0.35–4.98	0.685	1.25	0.61–2.53	0.546
Interpreter	0.31	0.09–1.11	0.073	1.62	0.45–5.87	0.463	0.61	0.23–1.62	0.322

Note: N = 309. Age is in decade units to show how a 10-year change in age is associated with differences in outcomes. Race is coded as non-Caucasian versus Caucasian.

Composite variable indicating presence of at least one of the following: mPOA document, POLST document, living will document.

ACP, advance care planning.

Discussion

Increased likelihood of ACP discussions, specifically mPOA and code status, was seen during the initial COVID-19 surge. As the pandemic continued, we hypothesize that patients and families were more reflective on medical decision making and use of life-sustaining medical technologies, particularly during the initial surge with the possible concern for ventilator shortages as publicized by media.^14,15 Another potential contributor included the increased use of telemedicine and patients having more comfort discussing ACP topics from their homes. In addition, PM clinicians in the study's clinic worked both inpatient and outpatient and may have been more motivated to discuss ACP in an outpatient setting after seeing the severity of COVID-19 in the inpatient setting.

Our study demonstrated an increase in ACP conversations, and less formal documentation reflecting these discussions. All encounters during the first COVID-19 surge were completed through telemedicine, limiting formal documents being signed and scanned into the EMR. The medical field needs to consider alternative electronic options for completing paperwork for patients and families, including the use of mobile notaries. Another consideration is the need for state Medicaid programs to reimburse ACP and prolonged care codes in line with Centers for Medicare Services to provide adequate reimbursement to hospitals caring for marginalized, under-resourced, and underinsured populations.

Increasing patient age and the presence of a cancer diagnosis were both independently associated with an increase in the likelihood of GOC, code status discussions, and POLST documentation. However, increasing age was associated with less mPOA documentation, reflecting a need to emphasize the importance of completing formal documentation. In addition, noncancer patients and patients requiring interpreters were less likely to discuss and identify an mPOA. More work is needed to explore ideal timeframes to discuss ACP as well as potential barriers for these patient populations.

Contrary to prior studies, there was no association between race/ethnicity and ACP documentation, engagement of discussions, or completion of documentation, especially in Black patient populations.^16–19 PM programs within a safety net hospital, such as the one we examined, often serve racially diverse populations. Our study was not specifically designed to examine this, so future studies are needed to understand how safety net PM programs impact ACP and racially diverse populations.

Conclusions

The COVID-19 pandemic impacted the way in which we provide patient care and highlighted the need for creative strategies. Our study shows ACP conversations occurred more frequently during the pandemic, yet without formal documentation. Further studies are needed to explore the use of technology when discussing GOC and ACP and improved ways to document them. Results from this study can help future work toward solutions engaging populations not as willing to engage in ACP conversations as well as how to better document these conversations formally in an outpatient setting.

Funding Information

This research was funded by the American Cancer Society under the grant number: 134579-RSG-20-058-01-PCSM, Patient-Centered Outcomes Research Institute under the grant number: EASC-COVID-00265, and the National Institute of Health under the grant number: T32CA19319.

Footnotes

Author Disclosure Statement

The authors declare that there are no conflicts of interest.

References

World Health Organization. Coronavirus. 2022. Available from: https://www.who.int/health-topics/coronavirus#tab=tab_1 [Last accessed: January 2, 2022].

Wang

, Prather

, Sznitman

, et al. Airborne transmission of respiratory viruses. Science, 2021; 373(6558):eabd9149; doi: 10.1126/science.abd9149.

Lopez

, Finuf

, Marziliano

, et al. Palliative care consultation in hospitalized patients with COVID-19: A retrospective study of characteristics, outcomes, and unmet needs. J Pain Symptom Manage, 2021; 62(2):267–276; doi: 10.1016/j.jpainsymman.2020.12.015.

Bender

, Huang

K-N

, Raetz

. Advance care planning during the COVID-19 pandemic. J Am Board Fam Med, 2021; 34(Suppl.):S16–S20; doi: 10.3122/jabfm.2021.s1.200233.

Singer

, Robertson

, Roy

. Bioethics for clinicians: 6. Advance care planning. Can Med Assoc J, 1996; 155(12):1689–1692.

Brinkman-Stoppelenburg

, Rietjens

, Heide

. The effects of advance care planning on end-of-life care: A systematic review. Palliat Med, 2014; 28(8):1000–1025; doi: 10.1177/0269216314526272.

Detering

, Hancock

, Reade

, et al. The impact of advance care planning on end of life care in elderly patients: Randomised controlled trial. BMJ-Brit Med J (Clin Res ed.), 2010; 340:c1345; doi: 10.1136/bmj.c1345.

Lunder

, Červ

, Kodba-Čeh

. Impact of advance care planning on end-of-life management. Curr Opin Support Palliat Care, 2017; 11(4):293–298; doi: 10.1097/SPC.0000000000000306.

Jacobsen

, Tran

, Jackson

, et al. Case 19–2020: A 74-year-old man with acute respiratory failure and unclear goals of care. N Engl J Med, 2020; 382(25):2450–2457; doi: 10.1056/nejmcpc2002419.

10.

American Cancer Society. Types of advance directives. 2019. Available from: https://www.cancer.org/treatment/treatments-and-side-effects/planning-managing/advance-directives/types-of-advance-health-care-directives.html [Last accessed: January 23, 2022].

11.

Brooks

. Why is New Orleans' coronavirus death rate twice New York's? Obesity is a factor. Reuters; April 2, 2020. Available from: https://www.reuters.com/article/us-health-coronavirus-new-orleans/why-is-new-orleans-coronavirus-death-rate-twice-new-yorks-obesity-is-a-factor-idUSKBN21K1B0 [Last accessed: January 23, 2022].

12.

Calvert

. New Orleans Area Has Worst Coronavirus Death Rate in U.S. Wall Street Journal; April 4, 2020. Available from: https://www.wsj.com/articles/new-orleans-area-has-worst-coronavirus-death-rate-in-u-s-11586008800 [Last accessed: January 23, 2022].

13.

Stein

. Data Center report: New Orleans' demographics make city more vulnerable to coronavirus spread and mortality. March 26, 2020. Available from: https://thelensnola.org/2020/03/26/data-center-report-new-orleans-demographics-make-city-more-vulnerable-to-coronavirus-spread-and-mortality [Last accessed: August 25, 2020].

14.

Neighmond

. As The Pandemic Spreads, Will There Be Enough Ventilators. NPR; March 14, 2020. Available from: https://www.npr.org/sections/health-shots/2020/03/14/815675678/as-the-pandemic-spreads-will-there-be-enough-ventilators [Last accessed: September 12, 2023].

15.

Kliff

, Satariano

, Silver-Greenberg

, Kulish

. There Aren't Enough Ventilators to Cope With the Coronavirus. The New York Times; March 18, 2020. Available from: https://www.nytimes.com/2020/03/18/business/coronavirus-ventilator-shortage.html [Last accessed: September 12, 2023].

16.

Degenholtz

, Arnold

, Meisel

, et al. Persistence of racial disparities in advance care plan documents among nursing home residents. J Am Geriatr Soc, 2002; 50(2):378–381; doi: 10.1046/j.1532-5415.2002.50073.x.

17.

Morrison

, Zayas

, Mulvihill

, et al. Barriers to completion of health care proxies: An examination of ethnic differences. Arch Intern Med, 1998; 158(22):2493–2497; doi: 10.1001/archinte.158.22.2493.

18.

Smith

, McCarthy

, Paulk

, et al. (2008). Racial and ethnic differences in advance care planning among patients with cancer: Impact of terminal illness acknowledgment, religiousness, and treatment preferences. J Clin Oncol, 2008; 26(25):4131–4137; doi: 10.1200/JCO.2007.14.8452.

19.

U.S. Census Bureau QuickFacts. U.S. Census Bureau QuickFacts: New Orleans City, LA, USA; 2020. Available from: https://www.census.gov/quickfacts/fact/table/neworleanscitylouisiana/POP010220#POP010220 [Last accessed: October 3, 2021].