Utilization of Patient Electronic Messaging to Promote Advance Care Planning in the Primary Care Setting

Abstract

Background:

Advance care planning (ACP) is an instrumental mechanism aimed at preserving patient autonomy. Numerous interventions have been proposed to facilitate the implementation of ACP; however, rates of completed advance directives (ADs) are universally low. Patient electronic portal messaging is a newer tool in patient–provider communication which has not been studied as a method to promote ACP. In this study, we hypothesized that the use of ACP-specific patient electronic messages would increase rates of AD completion in patients aged 65 years and older in an academic primary care practice.

Methods:

All primary care patients, aged 65+, who had previously enrolled in a patient electronic messaging system, within an academic primary care practice, were included for randomization. Two hundred patients were randomized to receive an electronic message. The primary outcome was the proportion of patients in each group who completed an AD, 3 months after intervention. Secondary outcomes included clinical utility of the completed ADs and proportion of patients who viewed their electronic messages.

Results:

The intervention group completed an AD 5.5% of the time when compared to 2% in the control group (odds ratio 3.2 [1.6-6.3]). Up to 74.5% of patients opened their electronic messages.

Conclusion:

Among primary care patients aged 65 years and older, use of AD-specific electronic messaging statistically significantly increased the rate of AD completion, but the absolute number of completed AD remained relatively low. These data suggest that this valuable communication tool holds opportunities for further improvement. Older, frailer adults were more likely to complete an AD, and prompted directives were more likely to include a written expression of the individual’s health-care values and preference.

Keywords

advance care planning advance directive patient electronic messaging

Background

Advance care planning (ACP) is a dynamic, iterative process for patients to communicate their preferences for medical care should they ever lose the capacity to express their wishes or the ability to make medical decisions.¹ Although the efficacy of ACP has been debated in the past, numerous studies have shown that the process promotes patient autonomy and enhances communication between providers, patients, and their families.^2
–4 Now, over 20 years since the Patient Self-Determination Act, ACP is emerging as a reportable quality metric in some areas and has been endorsed by the National Quality Forum.⁵ Documentation of ACP within the electronic health record (EHR) can take many forms including the designation of code status, health-care proxy, discussions between a patient and a provider, or most commonly via an advance directive (AD).

Despite the aforementioned drivers of ACP, AD completion rates remain universally low, at approximately 30% among patients aged 18 years and older, in the United States.⁶ Numerous barriers to successful completion of an AD have been identified, and wide-ranging solutions have been proposed including physician training sessions, computer-generated reminders for providers, appointments targeted toward ACP, interactive patient education seminars, and prolongation of the health maintenance examination visit.^4,7,8 Many of these interventions have been temporarily successful in improving AD completion but can be difficult to sustain long term due to financial and time constraints.^9
–11Few studies have taken into account the changing technological landscape within medicine, specifically, electronic patient messaging, as a means of reaching a wider patient population in a cost-effective manner.¹²

Secure patient electronic messaging (PEM) is rapidly being adopted by many medical institutions. Patient electronic messaging is typically a feature of the institution’s EHR, which allows patients to interface with their medical record, check test results, and securely send and receive electronic messages from care providers. The Markle Survey on Health in a Networked Life identified that 9% of patients in the United States were enrolled in a PEM system in 2010.¹³ This survey was published only 1 year after the initiation of the Health Information Technology for Economic and Clinical Health Act in 2009, which has since increased EHR adoption by as much as 7% and likely spurred increases in PEM in the process.¹⁴ Currently, multiple medical centers throughout the United States actively employ patient portals, including the Veterans Administration, Kaiser Permanente, Cleveland Clinic, MD Anderson Cancer Center, Vanderbilt, Northwestern Medicine, and Mayo Clinic.^15

–18

Since the emergence of PEM, multiple groups have studied its effect on various patient populations. In multiple studies, patients older than 65+ years of age constituted approximately 15% to 25% of active PEM users.^19,20 There are several recent studies and systematic reviews that have shown overall positive results with respect to patient satisfaction, improved patient–provider communication, and better self-management of chronic conditions through the use of a messaging system.^19,21
–23 Given the widespread use of portal technology in chronic disease management, experts in this field have called for trials examining their utility in promotion of ACP.¹²

The primary aim of this study was to measure the impact of a personalized ACP electronic message on AD completion in the primary care setting. We hypothesized that a nonintrusive motivational electronic message linked to appropriate resources would improve AD completion rates among senior adults.

Methods

Setting

Mayo Clinic’s Division of Primary Care Internal Medicine (PCIM) Clinic in Rochester, Minnesota, located in Olmsted County, provides primary care to over 35 000 patients. At the time of this intervention, the division was composed of 48 internists, 96 resident physicians, and 29 advance practice providers. This division provides primary care to the local community, to Mayo Clinic employees, and their dependents. Most patient visits occur on the downtown campus. Olmsted County has a population of over 150 000 people, with 13.6% of the population comprised of those aged 65 and older.⁵

The Mayo Clinic Patient Online Services system is a secure EHR and PEM system that has been offered without cost to all patients since 2009. Registration is patient initiated and not mandatory in order to receive care at the institution. Registration allows patients secure access to physician notes, laboratory test results, imaging studies, and appointment details. In addition, there is a messaging system that allows patients and their providers to communicate electronically. As of 2015, there are over 700 000 patients enrolled, with an adoption rate of approximately 45%.

Study Design

We conducted a randomized controlled intervention within the Division of Primary Care Internal Medicine. The study was conducted from May 2015 through July 2015.

Participants

Patients empaneled to a PCIM provider were evaluated for inclusion. Patients were included if they were aged 65 or older, did not have an AD on file within their EHR, and had access to the Mayo Clinic Patient Online Services system.

Outcomes

The primary outcome was the completion and return of an advance care plan within 3 months of receiving a personalized message prompt. The secondary outcomes included clinical utility of the completed AD based on (1) presence or absence of a health-care proxy, (2) designation of code status, and (3) personal description of health-care wishes and values.

The AMALGA Hospital Information System IT platform was used to obtain baseline patient data for all patients participating in the study. Baseline data included demographic information, Johns Hopkins Adjusted Clinical Group medical complexity tier, and Elder Risk Assessment (ERA)²⁴ score. The Johns Hopkins Adjusted Clinical Group (ACG) model is a proprietary model that characterizes patient morbidity. Higher medical tier is associated with higher morbidity.²⁴ The ERA index is a risk model that has been validated to identify elderly patients at risk of nursing home placement and mortality within 2 years.²⁵ Patients are assigned points based on numerous factors, including age, race, and presence of specific chronic medical conditions. High ERA scores are associated with high risk of mortality and morbidity.

Sample Size

It was determined that in order to detect a 10% difference with 90% power, assuming a 5% AD completion rate in the usual care group, the study would require 200 patients in each arm. Thus, 200 patients were individually, electronically randomized to the intervention arm, and the remainder of the population comprised the control arm, which received usual care.

Intervention

Patients randomized to the intervention arm received a personalized PEM explaining the ACP process and encouraging them to complete an AD. The electronic message also included a link to an ACP workbook that guided the individual and their loved ones through each step of the ACP process. This workbook included a Minnesota-specific AD form with instructions on how to return the completed document to the patient’s primary care provider. Eight weeks after the initial message was sent out, a reminder message was delivered to those individuals who had not yet returned an advance care plan. Final outcomes data were collected at 12 weeks. This quality improvement intervention was reviewed and deemed exempt by the Mayo Clinic institutional review board.

Statistical Analysis

Data were summarized using mean ± standard deviation for continuous variables and frequency counts and percentages for categorical variables. The percentage of patients who completed an AD was compared between study groups (electronic message intervention vs usual care) using the χ² test. To assess other characteristics potentially associated with the completion of an AD, a series of logistic regression analyses were performed. Separate analyses were performed for each of the patient characteristics listed in Table 1. For these analyses, completion of AD was the dependent variable, and the explanatory variables included the characteristic of interest and study group (electronic message vs control). The 2-way interaction between the given characteristic and study group was included in an initial analysis to assess whether the effect of electronic message was dependent on the given characteristic. In addition, to assess each characteristic individually, 2 preplanned multivariable analyses were performed. Age, sex, and study group as well as a multivariable model included the Johns Hopkins ACG medical tier, and the other model included ERA score.

Table 1.

Baseline Characteristics.

Characteristic	Control (N = 2294^a)	Electronic Message (N = 199^a)	P value
Age, years			.626
Mean ± SD	71.7 ± 5.8	71.9 ± 5.9
(Min, max)	(65, 96)	(65, 93)
Sex, n (%)			.140
Female %)	1197 (52)	93 (47)
Male, %)	1097 (48)	106 (53)
Need for interpreter, n (%)^b	36 (2)	4 (2)	.635
Chronic medical conditions, n (%)
Diabetes mellitus (%)	407 (18)	38 (19)	.634
Asthma (%)	196 (9)	18 (9)	.809
Hypertension (%)	1372 (60)	122 (61)	.679
Vascular disease (%)	538 (23)	51 (26)	.488
Obesity (%)	171 (7)	15 (8)	.966
Depression (%)	327 (14)	28 (14)	.943
Current tobacco use (%)	129 (6)	5 (3)	.061
Medical tier, n (%)			.189
−1, 0	237 (10)	17 (9)
1	1146 (51)	115 (58)
2	608 (27)	47 (24)
3	196 (9)	10 (5)
4	84 (4)	8 (4)
ERA score, n (%)			.299
7 to −1	526 (23)	48 (24)
0 to 3	1074 (47)	91 (46)
4 to 8	467 (20)	38 (19)
9 to 15	177 (8)	13 (7)
16+	50 (2)	9 (5)

Abbreviations: ERA, Elder Risk Assessment; SD, standard deviation.

^aAll patients had complete data with the exception of 25 (23 control, 2 intervention) who were missing information on medical tier.

^bInterpreter languages included Arabic, Bosnian, Cambodian, Chinese, Lao, Somali, Spanish, and Vietnamese.

In analyses restricted to patients in the electronic message group, exploratory analyses were performed using the 2-sample t test or χ² test to compare characteristics between those who opened the electronic message versus not. In all cases, 2-tailed P values of ≤.05 were considered statistically significant. Data related to this study may be obtained by e-mailing the corresponding author at tieu.christina@mayo.edu.

Results

Baseline Characteristics

A total of 2526 patients within the Division of PCIM, aged 65 and older, with access to the secure PEM system and without an AD on file, were included for randomization. Two hundred patients were individually randomized to the intervention group. The remaining 2326 patients comprised the control group. Of these, 32 patients left the practice during the time of our study and were not included in the final statistical analysis. Clinical characteristics and demographics were balanced between the group randomized to receive the motivational message and those in the control group (Table 1.)

Primary Outcome: AD Completion and Return

Among the intervention group, 11 (5.5%) completed and returned an AD during the study period compared to 46 (2%) in the usual care group (odds ratio [OR] 3.2, confidence interval = [1.6-6.3]). Additional factors associated with completion of an AD included higher medical complexity (P = .017) and older age (OR per 5 year age increase 1.5, P < .001)

Secondary Outcomes

Of the 200 patients randomized to receive the intervention, 149 (74.5%) opened their electronic message.

Among the returned ADs, those directives completed by individuals receiving the personalized message and ACP workbook were more likely to include a personal expression of their health values (P < .02; Table 2)

Table 2.

Comparison of Advance Directive Content.

	Intervention group (% of returned AD)	Control group (% of returned AD)	P value
Health-care proxy	10 (91%)	39 (87%)	.596
Code status	8 (73%)	24 (53%)	.218
Expression of health values	8 (73%)	16 (36%)	.022

Abbreviation: AD, advance directive.

Discussion

This study revealed that individuals receiving personalized, electronic motivational prompts and electronic resources were statistically significantly more likely to complete and return an AD document when compared to usual care. Additionally, the content of the ADs that were completed as a result of this minimally intrusive intervention was both comprehensive and clinically useful. Although the absolute number of completed AD remained low (11 of 199), the trend toward a positive association between PEM and AD completion is important to recognize. To the best of our knowledge, this project reports the only randomized trial utilizing electronic messaging as a motivational tool for ACP. This demonstrates both the feasibility and the utility of electronic messaging systems in facilitating ACP in some primary care patients.

Motivational portal messaging has established utility as a low touch, inexpensive intervention that has increased patients’ willingness to participate in various aspects of health care from diabetes management to medication reconciliation.^26,27 Utilization of health messaging allows patients to consider important health data, medical recommendations, and view patient education during their own time in the privacy of their own home. This modality also allows patients to share information with loved ones and decision makers who may not be able to attend office visits.

Other authors have successfully utilized nonelectronic messaging to prompt ACP. These interventions include ACP counseling offered via face-to-face visits with primary care providers or telephonic-based visits.^28
–30 Our group demonstrated that mailing ACP materials prior to a health maintenance visit also increased rates of AD completion.³¹ Several state-wide initiatives to improve ACP utilizing trained volunteers have also proven feasible.^32,33 However, each of these ACP enhancement modalities requires substantial time and resource utilization for both the institution and the patient and requires ACP to occur during a time frame that is convenient for the health-care institution but not necessarily convenient or acceptable for the patient. Few studies have employed electronic tools to aid in ACP. One study investigated the use of electronic reminders to physicians, urging providers to discuss ACP with their patients.³⁴ Other authors have studied electronic decision aids to facilitate ACP conversations.³⁵ However, these modalities are dependent on an office visit which are challenging to sustain in this era of increasing nonvisit care and limited face-to-face visits. Additionally, existing limitations in primary care access for older adults minimizes the feasibility of dedicated ACP visits.¹³

We view PEM as a feasible and emerging vehicle for delivering important ACP tools and a means of normalizing discussions about this sensitive topic. This modality also allows providers to tailor PEM to target specific, at-risk populations within their practices. For example, materials may be translated to reach non–English-speaking patients who historically have lower rates of ACP.³⁶ Additionally, inclusion of video technology may allow providers to better reach lower rates of health-care literacy.³⁷ Messages may also be sent at multiple time points, allowing for evolution of material as a patient’s understanding of ACP grows. This may be crucial to success because the process of ACP has been conceptually likened to the behavior change model.¹ Thus, use of PEM would allow for facile communication of the right information at the right time.

For health-care institutions with electronic messaging capabilities already in place, this intervention is both cost efficient and sustainable once automated. These results are generalizable regardless of the PEM system in place. A relatively simple, text-based message was utilized in this study, which can be performed by most messaging platforms.

This study is not without limitations. First, the authors recognize that ACP needs to vary at different points in an individual’s life. At times, planning may require intensive face-to-face communication between the individual, his or her loved ones, and their health-care team. While our intervention does not replace this valuable longitudinal communication between the patient and their health-care team, it does represent a starting point for ACP and normalizes dialogue regarding the topic. Second, patients targeted in this study must have access to a computer and have registered for electronic messaging. This confers selection bias for individuals with greater socioeconomic security and higher degrees of health literacy. This intervention also poses a problem with those who are not literate in English, although translation of culturally sensitive ACP materials is feasible. Third, the possibility exists that patients could have been exposed to other measures designed to increase ACP awareness which were in effect at the same time as the study; however, our randomization process minimized the impact of confounding factors.

In this era of burgeoning use of health information technology, PEM represents an inexpensive and feasible adjunct to many of the measures currently in place throughout health-care institutions. We have demonstrated that motivational electronic messaging allows patients to participate in ACP and complete an advance care directive on their own terms independent of office visits. Further studies of electronic ACP dialogue are needed to determine patient acceptability and satisfaction with the use of this technology.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Sudore

Schickedanz

Landefeld

. Engagement in multiple steps of the advance care planning process: a descriptive study of diverse older adults. J Am Geriatr Soc. 2008;56(6):1006–1013.

Brinkman-Stoppelenburg

Rietjens

van der Heide

. The effects of advance care planning on end-of-life care: a systematic review. Palliat Med. 2014;28(8):1000–1025.

Dixon

Matosevic

Knapp

. The economic evidence for advance care planning: systematic review of evidence. Palliat Med. 2015;29(10):869–884.

Kirchhoff

Hammes

Kehl

Briggs

Brown

. Effect of a disease-specific advance care planning intervention on end-of-life care. J Am Geriatr Soc. 2012;60(5):946–950.

Blaum

C MC

Alexander

Bangasser

. Geriatrics: Percentage of Patients Aged 65 Years and Older Who Have an Advance Care Plan or Surrogate Decision Maker Documented in the Medical Record or Documentation in the Medical Record That an Advance Care Plan was Discussed but the Patient Did Not Wish or Was Not Able to Name a Surrogate Decision Maker or Provide an Advance Care Plan. Washington, DC: National Quality Measures Clearinghouse, American Medical Association; 2009.

Morhaim

Pollack

. End-of-life care issues: a personal, economic, public policy, and public health crisis. Am J Public Health. 2013;103(6):e8–e10.

Jain

Corriveau

Quinn

Gardhouse

Vegas

You

JJ.

Video decision aids to assist with advance care planning: a systematic review and meta-analysis. BMJ Open. 2015;5(6):e007491.

Hohman

Lee

Ngo

Woodruff

. Feasibility and acceptability of a brief motivational stage-tailored intervention to advance care planning: a pilot study. Am J Hosp Palliat Care. 2015;33(9);834–842.

Bravo

Dubois

Wagneur

. Assessing the effectiveness of interventions to promote advance directives among older adults: a systematic review and multi-level analysis. Soc Sci Med. 2008;67(7):1122–1132.

10.

Heiman

Bates

Fairchild

Shaykevich

Lehmann

. Improving completion of advance directives in the primary care setting: a randomized controlled trial. Am J Med. 2004;117(5):318–324.

11.

Pearlman

Starks

Cain

Cole

. Improvements in advance care planning in the Veterans Affairs System: results of a multifaceted intervention. Arch Intern Med. 2005;165(6):667–674.

12.

Bose-Brill

Pressler

. Commentary: opportunities for innovation and improvement in advance care planning using a tethered patient portal in the electronic health record. J Prim Care Community Health. 2012;3(4):285–288.

13.

Markle. Markle Survey on Health in a Networked Life 2010. New York, NY: Markle Foundation; 2011.

14.

Mennemeyer

Menachemi

Rahurkar

Ford

. Impact of the HITECH act on physicians’ adoption of electronic health records. J Am Med Inform Assoc. 2016;23(2):375–379.

15.

Osborn

Rosenbloom

Stenner

. MyHealthAtVanderbilt: policies and procedures governing patient portal functionality. J Am Med Inform Assoc. 2011;18(suppl 1):i18–i23.

16.

Turvey

Zulman

Nazi

. Transfer of information from personal health records: a survey of veterans using My HealtheVet. Telemed J E Health. 2012;18(2):109–114.

17.

Wald

. Variations in patient portal adoption in four primary care practices. AMIA Annu Symp Proc. 2010;2010:837–841.

18.

Miller

Vandenbosch

Ivanov

Black

. Determinants of personal health record use: a large population study at Cleveland Clinic. J Healthc Inf Manag. 2007;21(3):44–48.

19.

Ralston

Rutter

Carrell

Hecht

Rubanowice

Simon

. Patient use of secure electronic messaging within a shared medical record: a cross-sectional study. J Gen Intern Med. 2009;24(3):349–355.

20.

Makai

Perry

Robben

. Evaluation of an eHealth intervention in chronic care for frail older people: why adherence is the first target. J Med Internet Res. 2014;16(6):e156.

21.

Kruse

Bolton

Freriks

. The effect of patient portals on quality outcomes and its implications to meaningful use: a systematic review. J Med Internet Res. 2015;17(2):e44.

22.

Kruse

Argueta

Lopez

Nair

. Patient and provider attitudes toward the use of patient portals for the management of chronic disease: a systematic review. J Med Internet Res. 2015;17(2):e40.

23.

Riippa

Linna

Ronkko

Kroger

. Use of an electronic patient portal among the chronically ill: an observational study. J Med Internet Res. 2014;16(12):e275.

24.

Kinder

K AC

Sutch

. The John Hopkins ACG System. Baltimore, MD: John Hopkins University; 2015.

25.

Crane

Tung

Hanson

Cha

Chaudhry

Takahashi

. Use of an electronic administrative database to identify older community dwelling adults at high-risk for hospitalization or emergency department visits: the elders risk assessment index. BMC Health Serv Res. 2010;10:338.

26.

Raghu

Frey

Chang

Cheng

Freimund

Patel

. Using secure messaging to update medications list in ambulatory care setting. Int J Med Inform. 2015;84(10):754–762.

27.

Lau

Campbell

Tang

Thompson

Elliott

. Impact of patient use of an online patient portal on diabetes outcomes. Can J Diabetes. 2014;38(1):17–21.

28.

Boettcher

Turner

Briggs

. Telephonic advance care planning facilitated by health plan case managers. Palliat Support Care. 2015;13(3):795–800.

29.

Song

Ward

Fine

. Advance care planning and end-of-life decision making in dialysis: a randomized controlled trial targeting patients and their surrogates. Am J Kidney Dis. 2015;66(5):813–822.

30.

Detering

Hancock

Reade

Silvester

. The impact of advance care planning on end of life care in elderly patients: randomised controlled trial. BMJ. 2010;340:c1345.

31.

Tung

Vickers

Lackore

Cabanela

Hathaway

Chaudhry

. Clinical decision support technology to increase advance care planning in the primary care setting. Am J Hosp Palliat Care. 2011;28(4):230–235.

32.

Wilson

Kottke

Schettle

. Honoring Choices Minnesota: preliminary data from a community-wide advance care planning model. J Am Geriatr Soc. 2014;62(12):2420–2425.

33.

Pecanac

Repenshek

Tennenbaum

Hammes

. Respecting ChoicesÒ and advance directives in a diverse community. J Palliat Med. 2014;17(3):282–287.

34.

Dexter

Wolinsky

Gramelspacher

. Effectiveness of computer-generated reminders for increasing discussions about advance directives and completion of advance directive forms. A randomized, controlled trial. Ann Intern Med. 1998;128(2):102–110.

35.

Butler

Ratner

McCreedy

Shippee

Kane

. Decision aids for advance care planning: an overview of the state of the science. Ann Intern Med. 2014;161(6):408–418.

36.

Carrion

Nedjat-Haiem

Martinez-Tyson

Castaneda

. Advance care planning among Colombian, Mexican, and Puerto Rican women with a cancer diagnosis. Support Care Cancer. 2013;21(5):1233–1239.

37.

Volandes

Ariza

Abbo

Paasche-Orlow

. Overcoming educational barriers for advance care planning in Latinos with video images. J Palliat Med. 2008;11(5):700–706.