Are We Getting What We Really Want? A Systematic Review of Concordance Between Physician Orders for Life-Sustaining Treatment (POLST) Documentation and Subsequent Care Delivered at End-of-Life

Abstract

Background:

Physician Orders for Life-Sustaining Treatments (POLST) is an advance care planning (ACP) tool that is designed to facilitate End-of-Life (EoL) care discussions between a medical provider and a terminally ill patient. It is often used as a tool to translate care wishes into a medical order, which can be honored across healthcare settings. With an increased utilization of the POLST paradigm in various healthcare settings along with continued dissemination across the nation, it is critical to examine whether documented wishes on POLST are concordant with subsequent care delivered. Purpose of this article was to examine concordance rate between POLST and subsequent care delivered in any care settings and communities.

Design:

Systematic review.

Results:

Of 1,406 articles identified, 10 articles met inclusion criteria. Together, included studies represent 5,688 POLST forms reviewed from individuals residing in a total of 126 nursing care facilities, 9 elderly care centers, 4 community settings, and 2 hospitals. Preference for cardiopulmonary resuscitation and actual delivery/ withholding of resuscitation was the most observed intervention in study of concordance (n = 8). It is also where highest concordance rate (97.5%) was reported. Seven studies compared care provided during EoL and the level of medical intervention requested on POLST forms (91.17% concordance). Preference to use artificial nutrition/ hydration, and actual delivery was 93.0% (n = 4 studies), and antibiotics use preference and delivery was 96.5% (reported in 4 studies).

Conclusion:

Published literature evidence suggests that EoL care wishes documented on POLST forms were largely concordant with subsequent care delivered. Additional research is needed to evaluate concordance between POLST documentation and care received among POLST users, who experienced multiple care transitions across healthcare settings, or across state during EoL care journey.

Keywords

POLST end-of-life advance care planning goal-concordant care

Introduction

In 2014, the Institution of Medicine (now the National Academy of Medicine) released a consensus report, Dying in America, which raised an awareness on the importance of providing high quality End-of-Life (EoL) care for those who are living with serious illnesses. In this report, appropriate EoL care was described as a vehicle that enhances quality of remaining life, and an essential factor to a more sustainable care system.¹ One of the conclusions called for an increased need for a state-wide adoption and active implementation of the patient-centered and communication-based advance care planning program, Physician Orders for Life-Sustaining Treatment (POLST).¹

The POLST program was first developed in early 1990s, to enhance the delivery of goal-concordant care among those who are living with frailty, and life-limiting illnesses.² The POLST form was specifically designed to facilitate EoL care discussions between a provider and a patient, and to encourage documentation of specific care wishes that align with one’s values, beliefs and care goals. What differentiated the POLST from other advance care planning tools (e.g., living wills or advance directives) was that documented care wishes on POLST forms become pre-signed actionable, and transferrable medical orders that can be honored across healthcare settings.³

Often printed in a distinctive bright pink background, POLST form contains 3 key sections; Section A: Cardiopulmonary Resuscitation (CPR), Section B: Medical Interventions (Comfort Measures Only: primary goal of maximizing comfort; Limited Treatment : goal of treating medical condition with basic medical treatments, but no intubation or ventilation; and Full Treatment: all medical treatments including breathing machine with goal to prolong life by all medically effective means), Section C: Artificially Administered Nutrition. It is noteworthy that medical interventions pertinent to EoL care (e.g., use of Intensive Care Unit (ICU), transfer to hospital) are listed under section B. Although some regional POLST forms contain additional sections (e.g., Antibiotics Preference for New York MOLST (Medical Orders for Life-Sustaining Treatment); and Blood Products Preference section for Idaho POST (Physician Orders for Scope of Treatment)), the National POLST Paradigm Task Force acknowledges sections A through C to be required elements of the official POLST form (see Figure 1).⁴

Figure 1.

Sample POLST form.

Nearly 3 decades since the POLST program was first introduced, it has gotten increased attention across the nation. As of 2020, a majority of U.S. states have, or are developing a state-wide POLST program (n = 46, including Washington D.C.) and regional data from California, Oregon and West Virginia showed that the POLST program is actively being used in more than half of its regional medical facilities (e.g., hospices, nursing homes, and hospitals).⁵ Moreover, scientific evidence indicating positive impacts of the POLST program (e.g., higher care satisfaction) on one’s EoL care journey continues to accumulate. For instance, POLST program usage was correlated with higher hospice service utilization,^6
-8 out-of-hospital deaths,^6,9 and higher likelihood of having treatment preferences documented as medical orders among older adults living with serious and/or terminal illnesses.¹⁰ While current research evidence supports the association between POLST program and care outcomes that are indicative of high quality EoL care, there exists a gap in knowledge on whether the POLST documentation achieves what it was designed to achieve; deliver EoL care that are in concordance with documented care preferences.

In EoL care delivery, where honoring patient’s care wishes remain as a top priority, identifying concordance between POLST documentation and actual care delivered during EoL and/or at the time of patient death is highly important. Therefore, the objective of this systematic review is to examine published evidence to identify concordance between documented care wishes on the POLST form and actual care delivered to POLST users.

Methods

This systematic review followed the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).

Eligibility Criteria

Articles were eligible for inclusion if the authors examined the concordance between documented care wishes on POLST forms and subsequent/actual care delivered to patients. Other eligibility criteria included: studies that were written in the English; conducted in U.S.; used quantitative methods; and published in a peer reviewed journal. Studies were excluded if the authors examined: the patient and/or family satisfaction on POLST form usage and/or care delivered; the health care providers’ perspectives on the ease of POLST use; the legal or ethical issues surrounding advance directives; the effects of advanced care planning tools other than POLST (e.g., living wills and/or advance directives); or, the quality of POLST form completion. Editorials and qualitative studies were also excluded.

Search Strategy

The literature search was conducted using 4 electronic databases: PubMed, PsycINFO, Embase, and Cumulative Index of Nursing and Allied Health Literature (CINAHL). A manual search of the reference lists to screen additional articles for eligibility was also conducted. After a consultation with a library informationist, a comprehensive search methodology was developed using free-text words and Medical Subject Headings (MeSH). Search terms included the following keywords: “Physician Orders for Life-Sustaining Treatment”, “POLST”, “MOLST”, “Advance Directive Adherence/Utilization”, “Advance Care Planning”, “Treatment Adherence and Compliance”, “End of Life Care Planning”, “Consistency”, “Congruence” and “Concordance”. Searches were not restricted by patient age, publication date, or care settings where the POLST user resided (e.g., home, nursing home, acute care facility, or hospice). This was to allow a comprehensive screening of articles that may meet our eligibility criteria. Table 1 contains a full list of search terms used for literature search. The initial search was conducted in February of 2019, then repeated again in August 2020 to ensure that all potentially eligible, and up-to-date articles are included.

Table 1.

Electronic Search Strategy and Keywords.

PubMed	1	((((((((((((“polst”) OR “molst”) OR “colst”) OR “physician orders for life sustaining treatment”) OR “Portable orders for life sustaining treatment”) OR “Medical orders for life-sustaining treatment”) OR “summary of physician orders for scope of treatment”) OR “clinician orders for life sustaining treatment”) OR “physician orders for scope of treatment”) OR “life sustaining treatment”))) AND (((((((((“Terminal Care/methods”[Mesh] OR “Terminal Care/nursing”[Mesh] OR “Terminal Care/organization and administration”[Mesh] OR “Terminal Care/standards”[Mesh] OR “Terminal Care/statistics and numerical data”[Mesh])) OR (“Hospice Care/classification”[Mesh] OR “Hospice Care/methods”[Mesh] OR “Hospice Care/organization and administration”[Mesh] OR “Hospice Care/standards”[Mesh] OR “Hospice Care/trends”[Mesh])) OR “Palliative Medicine”[Mesh]) OR “Hospice Care”[Mesh])) OR (((“end of life”) OR “end of life care”) OR “eol”))) AND ((((((((“Living Wills”[Mesh]) OR “Advance Directives”[Mesh]) OR “Advance Care Planning”[Mesh]) OR “Patient Care Planning”[Mesh])) OR “healthcare proxy”) OR (((“care planning”) OR “end of life care discussion”) OR “EoL care discussion”)) OR “life care planning”))	270
PsycInfo	1	“living wills”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	911
	2	“advance care planning”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	1,343
	3	“care planning”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	5,478
	4	“end of life”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	11,491
	5	“eol”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	669
	6	“end of life discussion”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	253
	7	“eol care discussion”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	1,153
	8	1 or 2 or 3 or 4 or 5 or 6 or 7	16,669
	9	“terminal care”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	4,534
	10	“hospice care”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	3,134
	11	Palliative Care/	18,153
	12	Palliative Care/ or Hospice/ or Terminally Ill Patients/	36,224
	13	9 or 10 or 11 or 12	36,949
	14	8 and 13	8,363
	15	polst.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	40
	16	“ physician orders for life sustaining treatment “.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	41
	17	“medical orders for life-sustaining treatment”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	15
	18	“life sustaining treatment”.mp. [mp=title, abstract, heading word, table of contents, key concepts, original title, tests & measures, mesh]	1,022
	19	15 or or 16 or 17 or 18	1,024
	20	14 and 19	375
CINAHL	1	living wills	11,001
	2	(MH “Advance Care Planning”)	3,519
	3	(MH “Living Wills”) OR (MH “Resuscitation Orders”)	3,864
	4	“health care proxy”	101
	5	“care planning”	7,523
	6	“end of life” OR “eol” OR “end-of-life” OR “end of life care discussion” OR “eol care discussion”	29,492
	7	1 OR 2 OR 3 OR 4 OR 5 OR 6	47,542
	8	(MH “Terminal Care”) OR (MH “Hospice Care”) OR (MH “Palliative Care”)	55,743
	9	7 AND 8	16,949
	10	“polst”	191
	11	“colst” OR “physician orders for life sustaining treatment” OR “medical orders for life-sustaining treatment” OR “summary of physician orders for scope of treatment” OR “physician orders for scope of treatment” OR “life sustaining treatment”	2,620
	12	10 OR 11	2,671
	13	9 AND 12	488
Embase	1	(“advance care planning”/exp OR “advance care planning”) AND (“living will”/exp OR “living will”) OR “care discussion” OR “end of life” OR “end of life care discussion” OR “care planning” OR “end of life care planning” OR “eol care planning”	175,307
	2	“terminal care” OR “palliative therapy” OR “palliative nursing” OR “hospice nursing” OR “hospice” OR “end of life”	165,056
	3	1 AND 2	40,075
	4	“POLST” OR “MOLST” OR “COLST” OR “portable orders for life sustaining treatment” OR “physician orders for life sustaining treatment” OR “medical orders for life-sustaining treatment” OR “summary of physician orders for scope of treatment” OR “clinician orders for life sustaining treatment” OR “clinical orders for life sustaining treatment” OR “physician orders for scope of treatment”	474
	5	3 AND 4	269

Study Selection

After removing duplicate articles, titles and abstracts were screened by 2 independent reviewers (AT, JS). When the article was deemed relevant, full text was obtained for further screening. Any discrepancies were resolved through a consensus discussion with a third reviewer (JP).

Data Collection Process

A standardized data abstraction tool included the following information about the manuscript demographics: title, last name of first author, and year study was published. Data audited on the study itself included: study design, data collection method, study setting, study region(s), sample number, sample characteristics, care intervention(s) examined for concordance (e.g., CPR, hospital transfer, Intensive Care Unit (ICU) admission, level of medical interventions (e.g., comfort-measures only, limited interventions, full treatment/life-preserving interventions), antibiotics and artificially administered nutrition. An evidence base table was developed to present the study characteristics and findings. Table 2 depicts overview of audited studies (i.e., study aim, design/ method, setting and important findings), and Table 3 contains detailed information on study characteristics (e.g., study region, sample number, demographics, participant criteria and outcome measured per each study).

Table 2.

Overview of Audited Studies.

First Author (Year)	Study Aim	Study Design / Method	Study settings	Findings
Tolle (1998)¹⁵	Examine the extent to which the POLST form ensured that nursing home residents’ wishes were honored for DNR and transfer request only if comfort measures fail	Prospective/ Chart Review	Long-term, adult care facilities(n = 8)	Section A: Cardiopulmonary Resuscitation (CPR) Twenty-six hospitalizations occurred; no subject received CPR in the hospital or at the care center (100% concordance) Section B: Medical Intervention Hospitalization: Total hospitalization (n = 26) among 24 residents Twenty-two cases were for comfort enhancement (i.e., 85% of concordance), while 15% of hospitalizations were to extend life Intensive Care Unit care: Orders to focus on comfort-oriented care were universally respected (i.e., no study subject received ICU care (100% of concordance)
Lee (2000)¹⁶	To evaluate whether terminal care was consistent with POLST	Retrospective/ Chart Review	ElderPlace, a Program of All-inclusive Care for the Elderly (PACE) site	Section A: CPR For those who opted DNR: CPR category was consistent with POLST instructions in 49/54 cases (90.7% of concordance) Of remaining 5/54 participants (9% of discordance); 3 participants received CPR, but had DNR on the POLST; 2 participants did not receive CPR although POLST indicated that they were to receive it Section B: Medical Intervention Level of Medical Intervention: The treatment received was at the medical level ordered in 25/54 cases (46.3% of concordance), while patients received at a less invasive level in 18/54 (33.3%), and at a more invasive level in 11/54 (20%) Section C: Artificially administered Nutrition Nutrition: Thirty-four residents had feeding tube preference documented. Care provided matched POLST instructions for 32 patients (94.1% concordance) Hydration: Of 38 participants who were not drinking adequate fluids and specified IV fluid instruction in POLST, 32/38 participants’ care matched POLST instruction (84.2% of concordance) Other section: Antibiotics Antibiotics usage complied with POLST instructions for 86% of the participants for whom antibiotics were indicated
Hickman (2011)¹⁷	To evaluate the consistency between treatments provided and POLST orders	Retrospective/ Chart Review	Random sample of nursing facilities(n = 90)	Section A: CPR Provision or withholding of CPR was consistent with POLST orders regarding resuscitation for 300/306 (98.0% of concordance) Of 299 deceased residents with a DNR order, none received unwanted CPR (100% of concordance) Section B: Medical Interventions Level of Medical Intervention: The overall concordance between treatments provided and level of medical interventions was 91.1% (102/112) “Comfort care only” wishes were concordant in 26/35 residents (74.3% of concordance) “Limited treatment” were concordant with POLST documentation in for 57/58 (concordance of 98.3%) “Full treatment” wishes were concordant 100% (19/19) Section C: Artificially administered Nutrition Nutrition: Of twenty-two residents requested to have feeding tube on POLST documentation, and 14 residents received feeding tubes (63.6% concordant) Other section: Antibiotics Antibiotic preference documented and care delivered were concordant 92.9% (224/241)
Hammes (2012)¹⁸	To determine demographics of individuals with a POLST form, the variety of orders that exist on the POLST form at the time of death and overall relationship between POLST and use of medical treatments and to rate discrepancy between POLST orders and treatments provided	Retrospective/ Chart Review	Community	Section A: CPR Of 255 deceased residents with a DNR order, none received unwanted CPR (100% concordance) Five residents requested that CPR be attempted received CPR (100% concordant) Section B: Medical Intervention Hospitalization: Overall, among hospitalized patients, wishes were honored in 49/51 (96.1%). In specific, 157 decedents with Comfort Measures Only, 15 were hospitalized. Thirteen cases were to enhance comfort, 2 did not have their preference to avoid hospitalized honored. ICU stay: Of the 157 decedents with POLST orders, no one intubated or received care in an ICU (100% of concordance) Section C: Artificially administered Fluids and Nutrition Nutrition: 221 POLST documentation, only 2 case were incongruency (219/221, 99.1% of concordance) The majority (148 cases, 58%) did not wish to have a feeding tube to sustain life and universally respected (100% of concordance) Four who had orders for feeding tube use, 2 received feeding tube and 2 did not as treatment was not indicated (4/4, 100% of concordance) Other section: Antibiotics Of the 255 decedents who opted for antibiotics treatment, 115/255 (45.1%) were for comfort only, 80/255 (31.4%) were for aggressive antibiotic use Although 20 decedents with POLST orders received antibiotics, the decision-maker authorized antibiotics use despite the order. Ten decedents with POLST orders for No Intravenous/Intramuscular antibiotics received antibiotics, but it was administered orally in all cases (overall, 100% of concordance)
Schmidt (2013)¹⁹	Determine if new POLST registry allows Emergency Medical Technician to honor patient preferences at the time of an EMS call	Retrospective/ Chart Review	Community	Section A: CPR Of 7 chart reviews, one patient was dead at the scene and no medical treatment was provided. CPR performed were concordant with POLST documentation in 5 cases (83.3% of concordance).
Richardson (2014)²⁰	Evaluate concordance of care provided by emergency medical services in out-of-hospital cardiac arrest cases	Retrospective/ Chart Review	Community	Section A: CPR Overall, provision or withholding of CPR was consistent with POLST orders regarding resuscitation for 66/82 (80.5% of concordance) Of the 50 patients with a DNR order on POLST documentation, wishes were honored in 39/50 (78.0% of concordance). Of 32 patients with a POLST form specifying attempt resuscitation, 27/32 had resuscitation attempted (84.4% of concordance).
Lum (2017)¹²	To describe care preferences of patients with heart failure admitted to skilled nursing facilities for rehabilitation based on MOST documentation, and evaluate goal-concordant care based on MOST documentation	Retrospective/ Chart Review	Skilled Nursing Facilities(n = 35)	Section B: Medical Intervention Hospitalization: Of 66/182 (36.3%) who had an Emergency Department or hospital visits, 63/66 patients received goal-concordant care during the event (95% of concordance)
Reinhardt (2017)¹⁴	Identify specific care indicators that are significantly associated with having each of the 6 medical orders on the MOLST in place	Retrospective/ Chart Review	Long Term Care facility (n = 1)	Section A: CPR No incongruence found (the specific number was not presented) Section B: Medical Intervention Hospitalization: Of 160 residents with a “Do not hospitalization” 6 were hospitalized. However, all hospitalizations (n = 6) were appropriate (i.e., reason for stay was to enhance comfort that cannot be managed at the care facility) (100% of concordance) Section C: Artificially administered Nutrition Hydration: Of the 47 residents with an order for “no artificial hydration,” 12 received hydration. However, 3 residents received hydration with family agreement. And for 9 patients, the treatment came before the order was put in place (100% concordant) Other section: Antibiotics Eight of 17 residents with a “No Antibiotics” order in place received antibiotics, but the treatment preceded the “No Antibiotics order” (100% concordance)
Hopping-Winn, (2018)¹¹	Review of wishes of patients who participated in a POLST Paradigm and whether those wishes were respected at End of Life	Retrospective/ Chart Review	Community	Section A: CPR Provision or withholding of CPR was concordant in 290 patients (96.7% concordant) For CPR/Full 100% concordant (19/19), DNR/Full 87.5% (7/8) concordant, DNR/select 95.9% (93/97) concordant, DNR/comfort 97.18% (138/142) concordant Incomplete POLST 33% (97.06)
Lee (2020)¹³	To evaluate the associations between POLST for medical interventions and Intensive Care Unit admission for patients hospitalized near EoL	Retrospective/ Chart Review	Hospital(n = 2)	Section B: Medical Intervention Level of Medical Intervention The incidence of POLST-discordant delivery of life-sustaining treatments was 14% in the comfort-only group and 20% in the limited interventions group, for a combined incidence of 18% ICU admission Overall, ICU admission was concordant in 726/1,162 patients (62.5%): 280 of 401 patients who opted for comfort measures only orders (69.8% concordant); 446 of 761 patients who opted limited-interventions order (58.6% concordant)

Footnote: CPR: Cardiopulmonary Resuscitation; DNR: Do Not Resuscitate, EoL: End-of-Life; ICU: Intensive Care Unit; POLST: Physician Orders for Life-Sustaining Treatment.

Table 3.

Study Characteristics.

First author (year)	Study region	Sample number/ Demographics	Study participants criteria	Measurements
Tolle (1998)	Oregon	Sample (n) = 180 Mean age (y): 83.2 Female: 69%, White: 98%	NH residents, who had a POLST recording DNR designation	All subjects: treatment and disposition after significant health status changes; orders for narcotics for provision or limitation of aggressive interventions. For hospitalized subject: diagnosis, medical interventions, and DNR orders. Those who died: cause and location of death, life-sustaining treatments attempted, and comfort measures provided
Lee (2000)	Oregon	Sample (n) = 54 Mean age (y): 84 Female: 70%, White: 91%	Elderly care site residents who died in 1997 and completed POLST	POLST instructions for each participant and whether or not each of treatments addressed by the POLST was administered in the final 2 weeks of life
Hickman (2011)	Oregon, West Virginia and Wisconsin	Sample (n) = 870 Mean age (y): 84.1 Female: 69%, White: 88%	NH residents aged 65 and older with a minimum 60-day stay, who died between 2007-2008	POLST form orders and related treatments over a 60-day period
Hammes (2012)	Wisconsin	Sample (n) = 400 Mean age (y): 80 Female: 55%, White: N/S	Adults who died between2007-2008	Presence of a POLST form, medical treatment provided in the last 30 days of life
Schmidt (2013)	Oregon	Sample (n) = 7 Mean age (y): N/S Female: N/S, White: N/S	Both EMTs who contacted the Oregon POLST Registry and received orders from registered POLST forms and the patients (or surrogates) for whom the calls were made between 2010-2011	Change of care preference documented on POLST form Intervention delivered at EoL matching the expressed preferences on POLST
Richardson (2014)	Oregon	Sample (n) = 1,577 Mean age (y): 60 (No POLST, No DNR), 69 (Attempt CPR), 83 (DNR) Female: 34% (No POLST, No DNR) 56% (Attempt CPR) 55% (DNR), White: N/S	Patients in out-of-hospital cardiac arrest who were evaluated by EMS in the 5 counties in 2010	Concordance for resuscitation preference/ delivery of resuscitation order indicated on POLST Concordance for delivery of medical interventions (intravenous access, intraosseous access, epinephrine, cardioversion attempted, and advanced airway placement)
Lum (2017)	Colorado	Sample (n) = 182 Mean age (y): 78.6 Female: 58%, White: 93%	Patients with HF admitted in SNFs between 2014 and 2016.	Validity of MOST forms, care preference documentation, and ED visits/hospitalizations within 60 days of SNF admission.
Reinhardt (2017)	New York	Sample (n) = 300 Mean age (y): 87 Female: N/S, White: 47%	Long-stay nursing home decedents over a 1-year period (beginning August 2013)	The decision to “not” conduct a particular life sustaining treatment; interventions provided (resuscitation, intubation, hospitalization, artificial hydration, feeding tube, and antibiotics)
Hoppin-Winn (2018)	California	Sample (n) = 300 Mean age (y): 70.8 Female: 41%, White: 76%	Patients who died in 2015 and had participated in the Advanced Steps program	Patients’ wishes, any changes of those wishes before death Care that patients received in the last care setting from patients’ EHR Care received in setting before death and concordance with documented wishes on patient’s POLST form
Lee (2020)	Washington	Sample (n) = 1,818 Mean age (y): N/S Female: N/S, White: N/S	Patients with POLSTs andwith chronic illness who died between January 1, 2010, and December 31, 2017, and were hospitalized within 6 months or less before death	Association between POLST order and ICU admission during the last hospitalization Receipt of a composite of 4 life-sustaining treatments: mechanical ventilation, vasopressors, dialysis, and CPR

Footnote: DNR: Do-Not-Resuscitate; ED: Emergency Department; EHR, Electronic Health Records; EMS: Emergency Medical Services; EMT: Emergency Medical Technicians; ICU: Intensive Care Unit; MOST: Medical Orders for Scope of Treatment; N/S: Not specified; POLST: Physician Orders for Life-Sustaining Treatment; SNF: Skilled Nursing Facility.

Results

The search strategy yielded a total 1,406 results. After removing duplicated studies (n = 380), titles and abstracts reviews for 1,026 studies. One hundred and thirty-nine articles were deemed relevant to our study aim and were subjected to further review. After reviewing full-text of 139 studies, we found 10 studies that met eligibility criteria. PRISMA flow diagram is depicted in Figure 2.

Figure 2.

PRISMA flow diagram.

Overview of Included Studies

Four studies were published within the last 5 years,^11

-14 and the remaining 6 were published between 1998 and 2014.^15

-20 Most studies (n = 4) were conducted in Oregon,^15,16,19,20 and one study per California,¹¹ Colorado,¹² New York,¹⁴ Washington,¹³ and Wisconsin.¹⁸ One study was conducted in multi-state setting (i.e., Oregon, Wisconsin and West Virginia).¹⁷ While most study designs were retrospective cohort studies,^{11

-14,16

-21} Tolle’s study was the only prospective study that collected data in real time and conducted chart reviews.¹⁵

The combined study sample size was 5,688 individuals who documented EoL care wishes on POLST forms, recruited from a various care settings; nursing homes,^12,14,15,17 elderly care facilities,¹⁶ community settings,^11,18
-20 and academic hospitals.¹³ Majority of study samples were White (78.8%), female (54.3%), with mean age of 79.8 years.

Concordance for section A: Cardiopulmonary resuscitation (CPR)

Eight studies reported concordance between documented care wishes section A: CPR and whether the subsequent delivery/ withholding of CPR was concordant with written wishes.^{11,14

-20} Overall, resuscitation efforts delivered or withheld were highly concordant with POLST documentation (mean 93.6%, median 97.5%, range 80-100%).

Hammes and colleagues conducted a retrospective review of medical records of community dwelling adults who completed POLST forms prior to their deaths.¹⁸ Authors found that of 300 individuals, 250 (98%) opted for Do-Not-Resuscitate (DNR) order under Section A, and none received CPR in the last 30 days of life. Five individuals (2%) requested resuscitations be performed, and their medical record indicated that all 5 were hospitalized at EoL, with admitting orders for full-code status. In Reinhardt and colleagues’ study, which was a review of 300 medical charts of long-stay nursing home (NH) residents, the researchers found that CPR wishes were honored at all times (100% of concordance).¹⁴ Tolle and colleagues conducted a prospective study of 24 adults who opted for DNR option under POLST form Section A. The authors found that only 2 participants died in the hospital setting, and neither received CPR in the hospital or nursing home (2/2, 100% of concordance).¹⁵

Hickman and colleagues also reported a highly concordant findings between POLST documentation and provision or withholding of CPR (300/306, 98% of concordance).¹⁷ When the authors examined the cases where POLST documentation indicated DNR order, none of the 299 deceased residents with a DNR order received unwanted or erroneous resuscitation (100% of congruence). In a chart review of 54 residents from a single NH, Lee and colleagues investigated whether resuscitation performed were concordant with POLST documentation.¹⁶ It was found that 49 residents (91%) received concordant care. One study, in which researchers randomly selected 300 community-dwelling adults who participated in advance care planning program, to identify whether CPR provided to POLST users were concordant with documented care wishes. It was found that wishes were honored in 257 individuals (97%).¹¹

Two studies showed concordance rates that were little over 80% of the time. Schmidt and colleagues conducted a study where authors examined accessibility of POLST registry for Emergency Medical Services (EMS) personnel working at the scene.¹⁹ The authors noted an overall provision or withholding of CPR was concordant with patient’s POLST documentation 83.3% (5/6 cardiac arrests) of the time. It was noteworthy that one patient was dead at the scene and did not receive any medical interventions at the scene.

The lowest rate of concordance was found in Richardson and colleagues’ studies.²⁰ The authors also examined concordance of provision/ withholding of resuscitation delivered by EMS in out-of-hospital cardiac arrest cases. Of total 82 cases of cardiac arrests where EMS calls were made, only 66 (80.6%) received care that were aligned with POLST documentation.

Concordance for section B: Medical interventions

Seven studies examined if the care provided during EoL was concordant with written care wishes on POLST form, section B. Specifically, 3 studies compared the preferred level of care (e.g., comfort only vs. limited vs. full treatment) indicated on POLST forms and whether wishes were respected.^13,16,17 Four studies examined the POLST wishes for hospitalization,^12,14,15,18 and 3 studies examined the POLST wishes for the receipt of care in an ICU setting.^15,16,18

Overall, preferred level of medical interventions showed concordance of 68.5% (range: 46%-91.1%), and the wishes to avoid/ transfer to hospital or ICU settings were concordant wish documented wishes 94% of the time (median 95.5%, range 85-100%).

Level of Medical Intervention

The concordance rate reported for preferred level of medical intervention and the actual level of care delivered varied greatly, ranging between 46.0%-91.1%.

In a study where authors evaluated the concordance between treatment provided to patients who are dying and their written wishes for the level of medical care (i.e., full vs. limited vs. comfort-only levels of care),¹⁷ authors noted an overall concordance of 91.1%. Specifically, of all patients who opted for full treatment level of medical care (n = 19) always had their wishes honored (100%). Similarly, those who opted for limited treatment level of medical intervention (n = 58) almost always (98.3%) had their wishes honored. However, when 35 POLST forms with comfort measures only level of medical care was reviewed, only 26 patients (74.3%) received concordant care. In the most recent study where authors evaluated association between POLST documentation for medical intervention,¹³ it was found that one or more life-sustaining/ prolonging treatments were delivered to 86.0% of patients who had comfort measures-only and 80.0% of patients who opted for limited treatment as their preferred level of medical care (i.e., combined 82.0% concordance) .

In contrast, a retrospective chart review of 54 residents, conducted by Lee and colleagues, found that less than half of residents (46.0%) received EoL care at the level they requested; 33 percent (n = 18) received care at a less invasive level, while 20% (n = 11) received care at a more invasive level.¹⁶

Hospitalization / ICU admission

Five studies examined concordance between documented wishes to be transferred and hospitalized for further management and whether such wishes were honored.^12

-15,18 When Tolle and colleagues examined 26 hospitalizations that occurred among 24 elderly residents, for whom POLST wish indicated transfer to hospital only for comfort enhancement, 22 hospitalizations were found to be for the management of pain and sufferings that were not controlled in their current care facilities (84.6% concordance).¹⁵ The rest (15.4%) were for the life-sustaining purposes, contrary to residents’ POLST documentation. Authors also added that of all hospitalized residents, their wish to receive comfort-care were universally respected (i.e., none received aggressive treatments, or were admitted to ICU settings).

Hammes and colleagues examined the medical records and POLST documentation of community dwelling individuals to examine if preference for hospital admission and ICU utilization were honored.¹⁸ Authors noted that of 157 decedents opted for comfort-measures only, and none received aggressive treatments (e.g., intubation or ICU admission). There were 15 hospitalization that were noted, and further chart review showed 13 cases were congruent with POLST documentation. That is, 13 decedents specified wishes to be transferred to hospital for comfort enhancement purposes only. And all 13 transfers were attributable to need for higher level of pain management. However, 2 cases were discordant as those individuals specified specific wish to avoid hospitalization.

Lum and colleagues examined medical charts of elderly patients living with heart failure, admitted to skilled nursing facilities.¹² Of 182 patients who had Medical Orders for Scope of Treatment (MOST) forms, 66 patients had hospital visits or were hospitalized. Although the authors stated that the majority received goal-concordant care during the hospitalization (95%),¹² no further information (e.g., specific medical intervention provided) was provided.

Reinhardt and colleagues examined 160 residents from a single care facility and had “do not hospitalize” order indicated on POLST forms.¹⁴ Although medical chart review revealed 6 hospitalizations had occurred, all (100%) were deemed “appropriate” as all 6 hospitalizations were needed to enhance comfort (without using any aggressive measures while hospitalized).

The most discordant cases were found in a recent study, conducted by Lee and colleagues.¹³ Authors examined a total of 1,818 patients who were hospitalized in 2 different hospitals and had ICU admissions near EoL. Of 401 patients who opted for “comfort measures only” level of intervention, 123 patients were admitted to ICU setting (69.3% concordance). For those who opted “limited treatment” level of intervention (n = 761), nearly half (n = 349) were admitted to ICUs, showing 54.1% concordance between POLST documentation and medical interventions delivered. Therefore, overall 60.9% concordance was achieved.

Concordance for section C: Artificially administered nutrition / hydration

Overall, documented care wishes to receive artificially administered nutrition (e.g., feeding tube) and/or intravenous hydration were generally concordant with the actual care delivered (mean 88.3%, median: 94.0%, range: 63.6%-100%).

A review of 221 POLST documentation showed only 2 cases (1%) were discordant. Little over half (58%) of patients indicated wishes to avoid feeding tube, and those wishes were universally respected (100% concordant). Four patients requested feeding tube to be used to sustain life; 2 received feeding tube at EoL, and 2 did not meet treatment indication (100% concordant when feeding tube was opted).¹⁸ When Hickman and colleagues reviewed 22 medical charts of older adults who opted to receive feeding tube, only 14 received feeding tubes (63.6% of concordance),¹⁷ and the rationale for those who did not receive feeding tube was not provided.

Reinhardt and colleagues reviewed 47 POLST forms that indicated patient wishes of “no artificial hydration” at EoL. Medical chart review revealed that despite “no artificial hydration” order in place, 12 patients (25.5%) received intravenously administered hydration.¹⁴ However, authors explained that 3 residents received hydration with family agreement/ consents, and rest received hydration prior to the preparation of the POLST documentation. Therefore, the concordance was reported as 100%. In the Lee and colleagues’ study,¹⁶ authors examined and reported concordance for both artificial nutrition and hydration. Of 37 participants who stopped eating sufficient calories to sustain life at EoL and had requested tube feeding, 34/37 patients received care that matched the POLST instructions (91.9% or concordance). Thirty-eight patients who were not drinking adequate fluids at EoL and specified preference to receive IV hydration on POLST forms, 32/38 participants received concordant care (84.2% of concordance).¹⁶

Other section: Concordance for antibiotic treatment

Four studies reported whether the use of antibiotics treatment at EoL among POLST users were concordant with POLST documentation.^14,16
-18 In general, use of antibiotic treatments were highly concordant with POLST documentation (mean: 94.7%, median: 96.5%, range: 86.0%-100%).

Reinhardt and colleagues noted that there were 17 residents who had “no antibiotics” order indicated on POLST forms.¹⁴ Chart review showed that 8 residents received antibiotics as treatment at EoL. However, further investigation of charts revealed that antibiotic treatment preceded the POLST order, in all cases (100% concordant). Similarly, Hammes and colleagues reviewed a total POLST forms to identify decedents’ antibiotics preference.¹⁸ Authors found that the majority (n = 115, 45%) opted for antibiotics for comfort measures only, followed by antibiotics for aggressive treatments (n = 80, 31%). The chart review revealed that 20 decedents with orders for antibiotics for comfort measures only received antibiotics treatment despite the order. However, further investigation showed that in all 20 cases, the surrogate decision maker authorized antibiotics use which overrode patients’ wishes documented on POLST forms. In the same study, it was also noted that 10 decedents indicated “no IV or IM antibiotics” and in all 10 cases, the antibiotics used were administered via oral routes only. Therefore, all patients always had their wishes honored (100%). In the Hickman and colleagues’ study, antibiotic preference documented and care delivered were concordant 92.9% (224/241): 7 patients indicated wishes to avoid antibiotics but wishes were not respected at all (0% concordance); 60 patients opted for limited antibiotics and it was respected in 50 cases (83.3% concordance); all wishes for 174 patients who opted for full treatment antibiotics were fully respected. According to Lee and colleagues’ study,¹⁶ antibiotics usage complied with the POLST documentation for 86% of the study participants, for whom antibiotics were indicated.

Discussion

To the best of our knowledge, this is the first systematic review that synthesized scientific evidence for the concordance between the documented care wishes on POLST forms, and subsequent care delivered to POLST users during EoL in both clinical and non-clinical settings.

Most of the studies included in this review were retrospectively conducted cohort studies, in medical facilities (e.g., nursing homes). Section A: CPR, preferences for receipt of resuscitation in the event of cardiac arrest, and the actual delivery of or withheld of resuscitation efforts was the most studied intervention of all. It is also where the highest congruence rates were reported.

There were several discordant cases noted in the delivery of EoL care interventions, namely, hospital transfers, antibiotics use, and feeding tube use. Most common reason for unwanted hospital transfers, or the use of antibiotics among dying patients were new onset of infections.^12,18 Previous studies on the health status of NH residents revealed that infections are, in fact, the most commonly encountered health problems among elderly population.^22
-24 Also, it is one of the most common factors that causes unplanned changes in one’s EoL care processes.²⁵ Combined with the fact that not all NHs are adequately equipped to provide screening of, or management of infections that are prevalent in EoL, it appears to be a challenging task to bring immediate changes, or measurable outcomes in this aspect of EoL care.²⁶ However, continued efforts in infection surveillance, antibiotic stewardship and in-depth understanding of resident’s EoL care goals are one of few strategies that can be implemented to minimize over-utilization of potentially harmful treatments.

Overall congruence between preferred use of feeding tube and actual care delivered revealed another area that needs improvement. Researchers pointed that discordant cases in feeding tube use were most commonly seen among those who opted to have feeding tubes only for a defined trial period.¹⁷ Some patients had feeding tube in place for longer than 30 days, raising concern for what constitutes as a trial period, and some even died with feeding tube in place. Researchers called for the need for clarification of the term defined trial period as it has never been defined by the National POLST Program Task Force (NPPTF) organization. Although it is difficult to conclude that discordant cases in feeding tube use were solely due to lack of consensus on definition, such finding raised an important issue that languages used in medical forms, especially in advance care planning document, should avoid any terms that can be misinterpreted or misunderstood.²⁷

One of notable study findings was the evidence of high congruence for a preferred route of medication administration. Among those who expressed EoL care preferences to receive antibiotics, but only via oral routes, all (100%) study participants received care that was congruent with their written wishes.¹⁸ This signifies a high specificity that can be achieved through use of POLST forms as an Advance Care Planning (ACP) tools. This finding resonates with previous research work, which noted an effectiveness of POLST forms in guiding EoL care.²⁸ Meier and Beresford explained that POLST provides a quick and clear guidance in the EoL care, through highly specific care preference documentation for common interventions used in EoL journey.²⁸ When examining concordance between POLST documentation and subsequent care delivered to seriously ill patients, it is important to understand that some discordant cases may arise from issues that are related to the POLST form itself (structural issues) or difference in familiarity of the POLST program among general population. While POLST is gaining increase attention across the nation, NPPTF allows each U.S. state to develop a state-specific POLST program and operate it independently.^27,29 This, in turn, leaves room for inherent differences to arise in the type of EoL care addressed and the options one can choose from. For example, the Wisconsin POLST form offers 4 different options under tube feeding preferences (i.e., no feeding tube, defining period usage, long-term use or determine the use when needed) while West Virginia only offers 2 (i.e., no feeding tube, or long-term feeding tube use).

The type of EoL care options captured on state forms also differ significantly. For example, antibiotics options are no longer being assessed through the POLST form in Oregon. In 2012, Oregon’s POLST coalition made an executive decision to remove antibiotics preference section from its POLST form, nearly 20 years after POLST program was developed.³⁰ Such a decision was due to a new research findings. That is, Hickman and colleagues found that the antibiotics section on POLST forms had little impact on the actual usage of antibiotics between 2 groups (patients who opted for antibiotics at EoL and patients who opted for no antibiotics at EoL). In this retrospective cohort study, the actual use of antibiotics at EoL were similar between 2 groups, regardless of written wishes indicated on the antibiotics section; 32.1% and 30.4% respectively.¹⁷ In addition, lack of research evidence that antibiotics use at EoL enhances survival outcomes or comfort for those who are near death supported Oregon’s decision to remove the antibiotics use section in its entirety.

Each state participating in the POLST program has different levels of program designation, classified as maturity status. From lowest to highest status, a state POLST program can move from developing, endorsed, and mature status when it meets certain milestones or key criteria that NPPTF requires.^31,32 When the program first begins a state-wide dissemination of POLST, it obtains a developing status and then moves toward endorsed when the state consensus is met for a single state-wide POLST form, and then to mature status when POLST becomes a part of standard of care for elderly and frail patients and is used as an advance care planning tool in more than half of all medical facilities (e.g., nursing homes, hospitals, hospice care settings) within its state.³² Depending on the maturity status of the POLST program, public awareness on the POLST program itself, and how it can be utilized to tailor one’s EoL care may vary significantly. For example, in an area where the POLST program is used in the majority of healthcare settings, medical providers may be more familiar with the fact that POLST forms should be re-visited when patient status improves or declines further, and patients be given many opportunities to reflect any changes in care preferences during the EoL care journey. Unfortunately, different levels of public or provider awareness and competency on how POLST can be utilized to tailor one’s EoL care remains unstudied, which could explain discordant cases observed in included studies.

Making the EoL care decision is a challenging process. It can also be emotionally draining for those who are faced with or involved in the decision processes. Informed ACP requires a utilization of a guiding tool that facilitates care discussions, while presenting treatment options that are relevant to EoL care. It should also be systematically monitored for continued quality improvement and be closely examined for the outcomes it brings to its target population. This systematic review synthesized published evidence on the congruence between POLST documentation and subsequent care delivered at EoL, and provides comprehensive resource for healthcare providers, health science researchers or policy makers who are seeking scientific evidence on POLST use and the impact it brings. It is clear that gathering additional scientific evidence on the use of POLST across different care settings or regions will help advance future EoL care practices.

Limitations

There are number of limitations that are worth mentioning. First, although we developed search strategy with the help of library informationist, it is possible that our search strategy may have not identified all relevant research articles on this topic. Second, while we limited our inclusion criteria to research articles published in peer reviewed journals, potentially important and relevant findings that addressed our research question could have been existed in grey literature, and/or unpublished articles. That is, publication bias may be present with studies not finding concordance not being published. Third, we found that several authors on the included research articles had affiliations with NPPTF; some served as consultants or board members. In addition, one of the authors in the included studies appeared as co-author on 2 other included studies.

Another significant limitation of this systematic review is that 5 of the included studies were conducted in Oregon, and 2 other multi-state studies included Oregon as one of their geographic study locations. Although this could be considered as a limitation, POLST was first developed, introduced, and disseminated in Oregon. Needless to say, as a birthplace of POLST program itself, the most number of resources and data are available from Oregon.

Conclusion

Based on our findings, POLST is an effective ACP tool that can be used to discuss, document, and to deliver EoL care that reflects patient’s care preferences. Provision of resuscitation at the time of cardiac arrest or time of death was universally respected, while there were mixed results in the use of feeding tube, and use of IV fluids use at EoL.

Additional research is recommended to identify congruence between documented care wishes and actual care delivered among patients who experienced multiple care transitions or have moved cross state near the time of death. Further studies should also compare congruence between POLST documentation and care delivered between groups of patients who may suffer from different medical conditions. Empirical studies are needed to ascertain the comprehensiveness of POLST items in capturing relevant and important EoL care interventions. There is a need to conduct prospective observational studies within various patient care settings, across all the life span.

Footnotes

Authors’ Note

A Tark: Analysis and interpretation of data, significant contribution to draft and revision of manuscript. J Song: Acquisition of articles, analysis, interpretation of data and contribution for manuscript revision. J Parajuli: Identify and acquisition of articles, contribution for manuscript revision. S Chae: Acquisition of articles, contribution for manuscript revision. P Stone: Overall concept, and design. Critical revision of the manuscript for important intellectual content. All the authors approved the final version of this manuscript.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Center for Improving Palliative Care for Vulnerable Adults with MCC (CIPC) National Institute of Nursing Research Grant (NINR) GRANT (P20NR018072-01); the Comparative and Cost-Effectiveness Research Training for Nurse Scientists (CER2) (NINR grant T32NR014205); Pain and Associated Symptoms Research Fellowship (NINR T32 (NR011147)

ORCID iDs

Aluem Tark

Jyotsana Parajuli

References

Dying in America: improving quality and honoring individual preferences near the end of life. Mil Med. 2015;180(4):365–367.

National POLST Paradigm. About the National POLST Paradigm Web site. Published 2019. Updated February 2020. Accessed July 2, 2020. https://polst.org/about-the-national-polst-paradigm/?pro=1

Brugger

Breschi

Hart

, et al. The POLST paradigm and form: facts and analysis. Linacre Q. 2013;80(2):103–138.

National POLST Paradigm. Published 2019. Updated February 2020. Accessed July 3, 2020. https://polst.org/national-form/

Tark

Agarwal

Dick

Stone

. Variations in physician orders for life-sustaining treatment program across the nation: environmental scan. J Palliat Med. 2019;22(9):1032–1038.

Jennings

Zingmond

Louie

, et al. Use of the physician orders for life-sustaining treatment among California nursing home residents. J Gen Intern Med. 2016;31(10):1119–1126.

Nugent

Slatore

Ganzini

, et al. POLST registration and associated outcomes among veterans with advanced-stage lung cancer. Am J Hosp Palliat Care. 2019;36(7):564–570.

Schmidt

Weaner

Long

. The power of advance care planning in promoting hospice and out-of-hospital death in a dialysis unit. J Palliat Med. 2015;18(1):62–66.

Pedraza

Culp

Knestrick

Falkenstine

Moss

. Association of physician orders for life-sustaining treatment form use with end-of-life care quality metrics in patients with cancer. J Oncol Pract. 2017;13(10): e881–e888.

10.

Hickman

Nelson

Perrin

Moss

Hammes

Tolle

. A comparison of methods to communicate treatment preferences in nursing facilities: traditional practices versus the physician orders for life-sustaining treatment program. J Am Geriatr Soc. 2010;58(7):1241–1248.

11.

Hopping-Winn

Mullin

March

Caughey

Stern

Jarvie

. The progression of end-of-life wishes and concordance with end-of-life care. J Palliat Med. 2018;21(4):541–545.

12.

Lum

Obafemi

Dukes

Nowels

Samon

Boxer

. Use of medical orders for scope of treatment for heart failure patients during postacute care in skilled nursing facilities. J Am Med Dir Assoc. 2017;18(10):885–890.

13.

Lee

Brumback

Sathitratanacheewin

, et al. Association of physician orders for life-sustaining treatment with ICU admission among patients hospitalized near the end of life. JAMA. 2020;323(10):950–960.

14.

Reinhardt

Downes

Cimarolli

Bomba

. End-of-life conversations and hospice placement: association with less aggressive care desired in the nursing home. J Soc Work End Life Palliat Care. 2017;13(1):61–81.

15.

Tolle

Tilden

Nelson

Dunn

. A prospective study of the efficacy of the physician order form for life-sustaining treatment. J Am Geriatr Soc. 1998;46(9):1097–1102.

16.

Lee

Brummel-Smith

Meyer

Drew

London

. Physician orders for life-sustaining treatment (POLST): outcomes in a PACE program. Program of all-inclusive care for the elderly. J Am Geriatr Soc. 2000;48(10):1219–1225.

17.

Hickman

Nelson

Moss

Tolle

Perrin

Hammes

. The consistency between treatments provided to nursing facility residents and orders on the physician orders for life-sustaining treatment form. J Am Geriatr Soc. 2011;59(11):2091–2099.

18.

Hammes

Rooney

Gundrum

Hickman

Hager

. The POLST program: a retrospective review of the demographics of use and outcomes in one community where advance directives are prevalent. J Palliat Med. 2012;15(1):77–85.

19.

Schmidt

Olszewski

Zive

Fromme

Tolle

. The Oregon physician orders for life-sustaining treatment registry: a preliminary study of emergency medical services utilization. J Emerg Med. 2013;44(4):796–805.

20.

Richardson

Fromme

Zive

Newgard

. Concordance of out-of-hospital and emergency department cardiac arrest resuscitation with documented end-of-life choices in Oregon. Ann Emerg Med. 2014;63(4):375–383.

21.

Shawon

MSR

Hossain

Adhikary

, et al. Attitude towards diabetes and social and family support among type 2 diabetes patients attending a tertiary-care hospital in Bangladesh: a cross-sectional study. BMC Res Notes. 2016;9:286.

22.

Klein

Livergant

JPI

. Systematic review of factors preventing referral for palliative radiotherapy. Int J Radiat Oncol. 2017;99(2):E522–E523.

23.

Kline

Bowdish

. Infection in an aging population. Curr Opin Microbiol. 2016;29:63–67.

24.

Kuswardhani

RAT

Sugi

. Factors related to the severity of delirium in the elderly patients with infection. Gerontol Geriatr Med. 2017;3:2333721417739188.

25.

Bowman

Elford

Dovey

Campbell

Barrowclough

. Acute hospital admissions from nursing homes: some may be avoidable. Postgrad Med J. 2001;77(903):40–42.

26.

Montoya

Mody

. Common infections in nursing homes: a review of current issues and challenges. Aging Health. 2011;7(6):889–899.

27.

Tripken

Elrod

Bills

. Factors influencing advance care planning among older adults in two socioeconomically diverse living communities. Am J Hosp Palliat Care. 2018;35(1):69–74.

28.

Meier

Beresford

. POLST offers next stage in honoring patient preferences. J Palliat Med. 2009;12(4):291–295.

29.

U.S. Department of Health & Human Services. Electronic End-of-Life and Physician Orders for Life-Sustaining Treatment (POLST) Documentation Access Through Health Information Exchange. 2017. Updated July 2016. Accessed Augusted 6, 2019. https://www.healthit.gov/sites/default/files/topiclanding/2018-07/POLSTRegistryKnowledge.pdf

30.

Oregon POLST. Statement on Oregon POLST Separation from National POLST. Published 2018. Updated February 2020. Accessed April 6, 2018. https://static1.squarespace.com/static/52dc687be4b032209172e33e/t/5b05ff462b6a28e3e73cabac/1527119687088/2018.05.23+Statement+on+Oregon+POLST+separation+from+National+POLST.pdf

31.

Hemmati Maslakpak

Razmara

Niazkhani

. Effects of face-to-face and telephone-based family-oriented education on self-care behavior and patient outcomes in type 2 diabetes: a randomized controlled trial. J Diabetes Res. 2017;2017:8404328.

32.

National POLST Paradigm. POLST program status. Published 2018. Updated February 2020. Accessed May 21, 2018. http://polst.org/wp-content/uploads/2018/08/2018.04.11-POLST-Program-Status.pdf