Patients from both the patient priorities care (PPC) and usual care (UC) sites were prescreened using administrative data based on the eligibility criteria listed in the Methods section. Primary care clinicians at the PPC site decided which of the prescreened patients they would invite to participate. The UC site did not include the step of recruitment. Therefore, all potentially eligible participants were recruited from the UC site. Usual care clinicians did not review their prescreened patients; thus, all UC patients who met the administrative data screen were assumed to be eligible.
aMember of the health care team who helps patients identify their health priorities.
bHealth outcome goals are the health and life outcomes that patients desire from their health care. Health care preferences refers to the health care activities (eg, medications, self-management tasks, health care visits, testing, and procedures) that patients find helpful and doable or bothersome, unhelpful, or unwanted. Health priorities refers to the specific, actionable, and reliable health outcome goals that are consistent with their health care preferences.
eAppendix. Patient Priorities Care: Health Priorities Template
eTable 1. Tasks and Times Associated With Patient Priorities Care
eTable 2. Baseline Comparisons of Participants who Completed Follow-up Interview and Participants Lost to Follow-up by Group
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Tinetti ME, Naik AD, Dindo L, et al. Association of Patient Priorities–Aligned Decision-Making With Patient Outcomes and Ambulatory Health Care Burden Among Older Adults With Multiple Chronic Conditions: A Nonrandomized Clinical Trial. JAMA Intern Med. 2019;179(12):1688–1697. doi:10.1001/jamainternmed.2019.4235
Is care for older adults with multiple chronic conditions that is aligned with their health priorities associated with improved patient-reported outcomes and reduced unwanted care?
This nonrandomized clinical trial of 366 adults 65 years or older with multiple chronic conditions found that, although there was no difference in perception of whether their care was goal-directed or coordinated, participants receiving patient priorities care vs usual care reported a greater reduction in treatment burden, and their health records reflected more medications stopped and fewer self-management tasks and diagnostic tests added.
This study’s findings suggest that aligning care with patients’ priorities may improve outcomes for patients with multiple chronic conditions.
Health care may be burdensome and of uncertain benefit for older adults with multiple chronic conditions (MCCs). Aligning health care with an individual’s health priorities may improve outcomes and reduce burden.
To evaluate whether patient priorities care (PPC) is associated with a perception of more goal-directed and less burdensome care compared with usual care (UC).
Design, Setting, and Participants
Nonrandomized clinical trial with propensity adjustment conducted at 1 PPC and 1 UC site of a Connecticut multisite primary care practice that provides care to almost 15% of the state’s residents. Participants included 163 adults aged 65 years or older who had 3 or more chronic conditions cared for by 10 primary care practitioners (PCPs) trained in PPC and 203 similar patients who received UC from 7 PCPs not trained in PPC. Participant enrollment occurred between February 1, 2017, and March 31, 2018; follow-up extended for up to 9 months (ended September 30, 2018).
Patient priorities care, an approach to decision-making that includes patients’ identifying their health priorities (ie, specific health outcome goals and health care preferences) and clinicians aligning their decision-making to achieve these health priorities.
Main Outcomes and Measures
Primary outcomes included change in patients’ Older Patient Assessment of Chronic Illness Care (O-PACIC), CollaboRATE, and Treatment Burden Questionnaire (TBQ) scores; electronic health record documentation of decision-making based on patients’ health priorities; medications and self-management tasks added or stopped; and diagnostic tests, referrals, and procedures ordered or avoided.
Of the 366 patients, 235 (64.2%) were female and 350 (95.6%) were white. Compared with the UC group, the PPC group was older (mean [SD] age, 74.7 [6.6] vs 77.6 [7.6] years) and had lower physical and mental health scores. At follow-up, PPC participants reported a 5-point greater decrease in TBQ score than those who received UC (ß [SE], –5.0 [2.04]; P = .01) using a weighted regression model with inverse probability of PCP assignment weights; no differences were seen in O-PACIC or CollaboRATE scores. Health priorities–based decisions were mentioned in clinical visit notes for 108 of 163 (66.3%) PPC vs 0 of 203 (0%) UC participants. Compared with UC patients, PPC patients were more likely to have medications stopped (weighted comparison, 52.0% vs 33.8%; adjusted odds ratio [AOR], 2.05; 95% CI, 1.43-2.95) and less likely to have self-management tasks (57.5% vs 62.1%; AOR, 0.59; 95% CI, 0.41-0.84) and diagnostic tests (80.8% vs 86.4%; AOR, 0.22; 95% CI, 0.12-0.40) ordered.
Conclusions and Relevance
This study’s findings suggest that patient priorities care may be associated with reduced treatment burden and unwanted health care. Care aligned with patients’ priorities may be feasible and effective for older adults with MCCs.
ClinicalTrials.gov identifier: NCT03600389
Older adults with multiple chronic conditions (MCCs) receive large amounts of health care,1 much of which is of uncertain benefit.2,3 The observed benefits and harms generated in clinical trials involving populations with few health conditions may not apply to persons with MCCs.4-7 Furthermore, benefits have primarily been defined by disease-specific outcomes or survival, which may not be what matters most for older adults with MCCs.8-10 Concomitantly, practitioners and the public are increasingly aware of treatment burden for persons with MCCs.2,11-13 These patients and their caregivers spend an average of 2 hours a day on health care–related activities.14 Consequently, appropriate health care for individuals with MCCs is uncertain, as are the outcomes that define its effectiveness.
The National Academy of Medicine (formerly the Institute of Medicine) defines patient-centered care as care “respectful of, and responsive to, individual patient preferences, needs, and values and ensuring that patient values guide all clinical decisions.”15(p 3) In other words, appropriate care is care aligned with the individual’s specific health outcome goals and health care preferences.16 This definition of appropriate care is particularly important for older adults with MCCs for whom evidence of treatment benefit is uncertain and outcome goals vary.2-10 Although methods for ascertaining and acting on patients’ goals and preferences exist for persons with advanced illness,17-20 less is known for the large number of older adults with MCCs.
To fill this gap, we instituted a multistakeholder initiative to define an appropriate approach to health care decision-making for adults aged 65 years or older who had MCCs.21-25 With input from patients, caregivers, clinicians, health system leaders, payers, and health care design experts, we developed patient priorities care (PPC), which involves identifying, communicating, and providing care consistent with patients’ health priorities.21,22 Patients’ health priorities include the health outcome goals they most desire given the health care they are willing and able to receive (ie, health care preferences). Previous studies reported on the development of a values-based approach to helping older adults identify their health priorities23 and on the feasibility of incorporating PPC into clinical practice.24 Challenges faced in aligning decision-making with patients’ health priorities and strategies for addressing these challenges were identified through experience in the pilot PPC practice25 and through action steps developed for the American Geriatrics Society’s Guiding Principles for the Care of Older Adults With Multimorbidity.2,26
The first objective of this study was to evaluate the association between participation in PPC and patients’ perception of whether their care was more focused on their health goals and less burdensome compared with care before PPC and with the perceptions of patients who received usual care (UC). The second objective was to assess changes in ambulatory health care in patients receiving PPC and UC.
The development, implementation, and feasibility phases of this nonrandomized clinical trial have been described previously.23,24,27,28 Participant enrollment occurred between February 1, 2017, and March 31, 2018. Participant follow-up ended September 30, 2018. The study was approved by the Yale University Institutional Review Board, and the complete trial protocol is available in Supplement 1. Oral informed consent was obtained from study participants.
Patient priorities care was implemented in a multisite primary care practice that provides care to almost 15% of Connecticut residents.29 Two comparable sites served as the PPC and UC site. Both sites participate in improvement and research projects, and more than 15% of their patients are Medicare beneficiaries. All clinicians at the sites were invited to participate as were their patients who met eligibility criteria. The selection and recruitment of the practice and participating clinicians have been described previously.24
The leadership of the practice chose 1 site to implement PPC. All 10 primary care clinicians (2 advanced practice nurses, 1 physician assistant, and 7 physicians) at the PPC site agreed to participate. Because a key PPC principle is alignment between primary and specialty care, the 5-member cardiology practice that provides cardiac care to the greatest number of PPC patients was selected as the specialty practice. All 5 cardiologists agreed to participate. Cardiology was chosen because it is the medical specialty that provides the greatest amount of care to older adults with MCCs. The primary care clinicians and cardiologists received modest stipends for participating in PPC. A clinical “champion” was identified from the primary care practice and the cardiology practice to support their colleagues and serve as a liaison between their practices and the PPC implementation team.
Potentially eligible patients cared for by the PPC site were identified by electronic health record (EHR) data. Criteria included more than 3 chronic conditions and either 10 medications or visits to more than 2 specialists in the past year. Exclusion criteria included hospice eligibility, receiving dialysis, advanced dementia, or nursing home residence. Clinicians reviewed lists of eligible patients the week before scheduled primary care practitioner (PCP) visits; the final decision about inviting patients was left to clinician discretion. The flow diagram of patient participants is shown in Figure 1.
We performed 2 levels of matching to identify the UC practice. We first used commercial market segmentation data to identify census block groups with demographic and socioeconomic characteristics similar to the geographic catchment area of the PPC site.30 Second, among the 8 practice sites located in areas with demographic and socioeconomic characteristics similar to the PPC site, we selected the practice most similar to the PPC site in types of PCPs and percentage of the patient population who were Medicare beneficiaries. The 7 PCPs (2 advanced practice nurses, 2 physician assistants, and 3 physicians) at the UC site all agreed to participate. Their participation entailed sending a prepared letter to their eligible patients stating that they would be contacted about the study. All patients cared for by the UC clinicians who met the same eligibility criteria (without prereview by their clinicians) were invited to participate in the study.
The development and feasibility of implementing PPC has been described previously.21,22,24 The training and implementation of PPC was led by a multidisciplinary team of 4 geriatricians, 1 general internist expert in clinician training, 1 expert in behavioral medicine, and 2 project managers. The PPC approach is defined in detail elsewhere and summarized here.23,24
Patients were invited by their PCP to participate in PPC during routine visits. Those who accepted met in person or by telephone with an advanced practice nurse or a care manager employed by the primary care practice who served as facilitator (step 1 in Figure 2). The facilitators followed a clinically feasible process to help patients identify their values-based health priorities and interact with their clinicians concerning their health priorities.23 Facilitators completed 1-page health priorities templates that included brief summaries of function and support, key values, up to 3 health outcome goals, 3 helpful and doable and 3 unhelpful or difficult health care activities (health care preferences), and the one thing (eg, symptom, health condition) or health care activities (eg, medication, self-management task) on which they most wanted to focus to achieve their health outcome goals. Facilitator training included review of training materials, role playing, observation, and feedback.23 The patient priorities conversation guides for the facilitator and patients and a link to online PPC training are available elsewhere.31
Facilitators transmitted patients’ health priorities templates (eAppendix in Supplement 2) to the participating PPC clinicians (via a scanned EHR document to the PPC primary care clinicians and via fax to PPC cardiologists). Patient priorities care clinicians, who were reminded by office staff to review the templates the day of the first visit after patients identified their priorities, used this information in their communication and decision-making with patients and other clinicians (steps 2 and 3, Figure 2). Decisional strategies that support patient priorities–aligned decision-making and care were identified through emergent learning among the PPC implementation team and the participating clinicians.25 The recommended strategies included starting with a single thing that mattered most to the patient (Specific Ask in the Health Priorities template in the eAppendix of Supplement 2); conducting serial trials of starting, stopping, or continuing interventions; focusing on achieving the patient’s desired activities rather than completely eliminating symptoms, which is often not possible in persons with multiple conditions and symptoms; basing communications, decision-making, and measures of success or failure on patients’ priorities, not solely on diseases; and using collaborative negotiations when there are differences in perspectives between patient and clinician or among clinicians to arrive at agreed-upon decisions.25 An example of how these strategies may be used is provided in the eAppendix in Supplement 2. The clinical workflow, preparation of participating clinicians, and practice change framework used to encourage adoption of PPC has been reported.24,25
To estimate the time required to implement PPC, we worked with a process improvement staff member of the participating primary care practice to build workflows for the tasks needed to carry out PPC by staff and clinicians. We calculated the time spent on each task based on interviews with the relevant staff and clinicians. The tasks are included in eTable 1 in Supplement 2.
Patients’ perceptions as to whether their health care was collaborative and focused on their goals were measured by using the Older Patient Assessment of Chronic Illness Care (O-PACIC) score (11 items; possible range, 1-5; Cronbach α = 0.87; higher score indicates better perceived experience of chronic disease care) and CollaboRATE (3 items; possible range, 0-100; Cronbach α = 0.89; higher score indicates greater perceived shared decision-making and goal ascertainment), and perceptions of burdensomeness were measured by the Treatment Burden Questionnaire (15 items; possible range, 0-150; Cronbach α = 0.90; higher score indicates greater perceived burden of treatment).11,32,33 These patient-reported outcomes were ascertained by telephone at baseline and after approximately 6 months’ follow-up.
Primary care clinicians’ notes were reviewed across 9 months of follow-up for documentation of decision-making based on patients’ health goals, health care preferences, or health priorities (disease-specific goals, such as blood pressure target, were not included). Also abstracted from visit notes were the number (percentage) of participants with medications and self-management tasks added or stopped plus diagnostic tests, referrals, or procedures ordered or avoided (defined as mentioned in the EHR that they were decided against because they were not considered by clinicians to be beneficial or they were unwanted by the patient). A data dictionary guided uniform abstraction. Uncertain decisions were adjudicated by the study’s primary investigators (M.E.T. and C.B.); 30% of EHRs were doubly abstracted.
All outcomes were ascertained by assessors blinded to patient group assignment. Outcome assessors were located separately from other members of the research and clinical teams, were unaware of specific aims and intervention details, and entered data on forms that did not include group assignment.
Age, sex, race/ethnicity, educational level, marital status, and insurance coverage were ascertained during the baseline telephone interview or EHR review. During the baseline interview, physical and mental health were ascertained with PROMIS-10,34 and the Montreal Cognitive Assessment 5-word recall test35 was used to assess memory. Number and types of chronic conditions were determined from the problem list and medications from the medication list of the PCPs’ EHR.
We compared baseline demographic and clinical characteristics for the PPC and UC participants using independent samples t tests for continuous variables and χ2 tests for categorical variables. We estimated propensity score models with the PSMATCH procedure in a statistical software package (SAS, version 9.4; SAS Institute Inc) to achieve balance on baseline characteristics between the 2 groups using inverse probability of treatment weights.36,37 Balance was evaluated by comparing the distribution of the covariates using absolute standardized mean differences of 0.25 or less,38 variance ratios between 0.5 and 2.0,39 and side-by-side plots of the distribution of each covariate.37,40 Extreme weights were truncated at the 98th percentile value for 5 PPC and 2 UC participants.37 Missing baseline data were handled by using multiple imputation using the fully conditional specification method (M = 20) implemented by the multiple imputation procedure in SAS statistical software. Rubin formulas were used to combine model estimates into a single set of results using the MIANALYZE procedure in SAS.41
Change in patient-reported outcomes from baseline to follow-up was assessed by means of difference scores (ie, follow-up score minus baseline score) because the distribution of difference scores tends to be closer to normal (suitable for regression analysis) and offers a more intuitive interpretation than transformed variables.42 Health care activities were measured dichotomously. Both the patient-reported outcomes and health care activities models were adjusted for demographic and clinical characteristics in addition to propensity weights.43 The patient-reported outcome models also included length of follow-up. The health care models also included number of clinician visits. All analyses were intention to treat and performed using SAS statistical software; 2-sided P < .05 was considered statistically significant.
Of the 366 patients, 235 (64.2%) were female, 350 (95.6%) were white, and 7 (1.9%) were Hispanic. Compared with persons in the UC group, persons in the PPC group were older (mean [SD] age, 74.7 [6.6] vs 77.6 [7.6] years) and had lower physical and mental health scores (Table 1). As shown in Figure 1, 163 of 231 (70.6%) eligible patients at the PPC site agreed to participate, as did 203 of 295 (68.8%) at the UC site. Mean (SD) time between baseline interviews and the first PCP visit, which initiated PPC, was 11.2 (0.8) weeks in the PPC group and 7.7 (0.6) in the UC group. Characteristics of the 2 groups were similar after propensity score weighting. Follow-up O-PACIC, CollaboRATE, and Treatment Burden Questionnaire measurements were available for 145 (89.0%) of the PPC group and 191 (94.1%) of the UC group. Health care activities were available for 158 (96.9%) PPC and 187 (92.1%) UC participants (Figure 1). Baseline characteristics of participants who did and did not complete the follow-up are in eTable 2 in Supplement 2.
Documentation by the PCPs or cardiologists of discussion or decision-making concerning patient’s health priorities were noted in 108 of 163 (66.3%) of the PPC participants vs none of the UC participants. Changes in patients’ perceptions of their care are given in Table 2. Participants in the PPC group reported a 5-point greater decrease in treatment burden score (β [SE], –5.0 [2.04]; P = .01) from baseline to follow-up than UC participants. Changes in O-PACIC and CollaboRATE scores were similar between the 2 groups.
The number of PCP visits (weighted mean [SE]) was similar across 9 months of follow-up for PPC (4.1 [0.2]) and usual care (3.6 [0.2]) participants. Compared with UC participants, PPC participants were more likely to have any medications stopped (adjusted odds ratio [AOR], 2.05; 95% CI, 1.43-2.95), including cardiovascular medications (AOR, 3.43; 95% CI, 2.10-5.60) (Table 3). As noted in Table 3, PPC participants had moderately greater odds of having psychotropic medications both added and stopped; only the former reached statistical significance. Patient priorities care participants had lower odds of having diagnostic tests ordered (AOR, 0.22; 95% CI, 0.12-0.40) and self-management tasks added (AOR, 0.59; 95% CI, 0.41-0.84) than UC participants.
The time required for training in and implementing PPC, as estimated by the facilitators, is outlined in eTable 1 in Supplement 2. The typical time for the health care team member to identify patients’ priorities was estimated at approximately 20 to 30 minutes; the scheduling, reviewing EHRs, and documenting required an additional 15 to 40 minutes. Time commitment required of health information technology staff (1.0-1.5 days total) and office staff (approximately 10 minutes per patient across the 15 months) was modest.
Clinician training, including initial face-to-face sessions plus twice-monthly, 20-minute case-based discussions, required 8 hours across 15 months. Incorporating PPC into workflow required 30 to 33 minutes for PCPs, distributed across the invitation visit and 2 post–priorities identification visits, after which no additional time was required. Cardiologists required training and implementation time similar to that for the PCPs.
Patient priorities care was associated with a much greater frequency of clinicians documenting decisions based on the patient’s health priories, goals, or preferences compared with UC. Electronic health records also showed that more PPC than UC patients had medications and self-management tasks stopped, but fewer PPC patients had diagnostic tests ordered. Patients receiving PPC reported a modestly greater reduction in treatment burden than those receiving UC. The high baseline CollaboRATE and O-PACIC scores, which left little room for improvement, at least partially explain the lack of difference in change in these measures between groups. The latter finding is consistent with reports that older age may be associated with greater patient satisfaction.33,44-46 Results of this study suggest that PPC may offer a decisional framework for focusing care on patients’ health goals, lessening their burden of health care, and avoiding potentially unwanted care.
Approaches focused on achieving patients’ goals, preferences, and priorities have been investigated for specific conditions and patients with advanced illness.17-20,47-53 Some approaches included facilitators who helped patients identify their preferences and priorities; results have been mixed.17-20,47-53 Trials addressing older adults with MCCs have not focused on patients’ health priorities. A systematic review of 18 systems- or patient-based interventions revealed limited effect on patient-reported health outcomes or health care use.54 A recent multidimensional intervention was effective at improving patients’ experience of patient-centered care but not health-related quality of life or treatment burden.55
Staff and clinician time to implement PPC was modest. The time required for priorities identification was lowest if performed immediately after patients were invited by their PCP, likely because it avoided time spent contacting the patient and scheduling a visit. This study’s findings suggest that patient priorities identification could be incorporated into the clinical workflow of team members such as medical assistants, social workers, nurses, case managers, or primary care practitioners. After training, clinicians’ time spent on PPC was minimal, suggesting a change in approach rather than inclusion of additional tasks.
The study has limitations. First, the sites were not randomized. Although we chose a UC site based on characteristics similar to the PPC site and used propensity adjustment, there may be remaining confounders that could explain outcome differences observed. The greater number of deaths as shown in Figure 1 (5 vs 1) likely reflects the fact that the PPC group was slightly older and sicker than the UC group. This pilot involved a single practice with a relatively homogeneous patient population. We cannot comment on how feasible or effective PPC would be in other settings and with other populations. Our results can only be considered suggestive; results require replication in larger, more diverse populations, preferably in a randomized trial. We could not blind the clinicians to group assignment, although the outcome assessors were blinded. Primary care practitioners and cardiologists were the only clinicians involved. Assuming that the outcomes would have shown greater differences between the patient care groups had all clinicians been prepared and agreed to align their care with patients’ health priorities, our results may underestimate the changes associated with PPC. Although we required EHR documentation that the health care activities stopped or avoided were stopped or avoided based on patient or clinician preference, we cannot be sure that they were all “unwanted,” nor can we be sure that the EHR documentation captured all of the changes in health care activities at the PPC or UC sites.
We lacked the claims data necessary to calculate costs or revenues from increases or decreases in health care activities. We also could not estimate the potential revenue generation through use of Centers for Medicare & Medicaid Services billing codes, such as advance care planning and care coordination, because these were not used by the practice. The financial effects of changes in utilization because of PPC depends on the payment model of the practice or health system.
Despite these limitations, there are important implications for the care of older adults with MCCs. Patient priorities care proved feasible and acceptable to patients and clinicians. It was incorporated into workflow with modest start-up time. If shown effective in settings with more heterogeneous populations, PPC could inform treatment deintensification or deprescribing by tailoring these decisions to each patient’s health priorities.56 Although helpful in guiding treatments that can be weaned, finding that some health care activities increased in the PPC vs the UC site suggests that patient priorities–aligned decision-making can also identify interventions likely to help patients achieve their outcome goals, offering the potential to address both overuse and underuse.57 Patient priorities care offers a platform for shared decision-making for persons with MCCs for whom disease-by-disease decisions can be cumbersome, conflicting, and potentially harmful. This approach works for patients who prefer either an active or passive role in decision-making.58 Even persons who do not desire active participation in treatment selection may articulate the values and outcomes they prioritize. Adherence may improve when recommendations align with patients’ goals, preferences, and capabilities.59 Finally, normalizing the identification of and alignment of decisions with patients’ health priorities may make decision-making in advanced illness less difficult.
Patient priorities care is expanding, with implementation in additional practices and health systems and with development of tools that facilitate adoption. The goal is to disseminate an approach to decision-making and care that maximizes benefits that matter to individuals with MCCs while minimizing harm and burden. Patient priorities care operationalizes the core patient-centered tenet that care should be driven by patient’s values, goals, and preferences.15 Patient priorities–aligned care may promote true value-based care.60
Accepted for Publication: August 3, 2019.
Corresponding Author: Mary E. Tinetti, MD, Department of Medicine, Yale School of Medicine, 333 Cedar St, PO Box 208025, New Haven, CT 06520 (email@example.com).
Published Online: October 7, 2019. doi:10.1001/jamainternmed.2019.4235
Author Contributions: Drs Tinetti and Costello had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Tinetti, Naik, Dindo, Esterson, Rosen, Smith, Blaum.
Acquisition, analysis, or interpretation of data: Tinetti, Naik, Dindo, Costello, Geda, Rosen, Hernandez-Bigos, Ouellet, Kang, Lee, Blaum.
Drafting of the manuscript: Costello, Smith, Lee.
Critical revision of the manuscript for important intellectual content: Tinetti, Naik, Dindo, Esterson, Geda, Rosen, Hernandez-Bigos, Ouellet, Kang, Lee, Blaum.
Statistical analysis: Costello.
Obtained funding: Tinetti.
Administrative, technical, or material support: Naik, Dindo, Esterson, Geda, Rosen, Hernandez-Bigos, Blaum.
Supervision: Tinetti, Esterson, Rosen, Smith, Blaum.
Conflict of Interest Disclosures: Dr Smith reported receiving grants from Yale University, that her husband is employed by Merck & Co, and that she and her husband jointly own stock and stock options in Merck & Co. No other disclosures were reported.
Funding/Support: The present study was supported by grants to Dr Tinetti from The John A. Hartford Foundation, The Gordon and Betty Moore Foundation, and The Robert Wood Johnson Foundation. Dr Naik received support from the Houston VA Health Services Research and Development Center for Innovations grant (CIN13-413). Investigators received additional support and resources from the Claude D. Pepper Older Americans Independence Center Grant from the National Institutes of Health to Yale University School of Medicine.
Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Denise Acampora, MPH (Yale University), provided project management; Lauren Vo, APRN (Bristol Prohealth), facilitated patient priorities identification; and Eliza Kiwak, BA (Yale University), provided project support and manuscript preparation. They did not receive compensation above their usual compensation for their work on this study. The clinicians of Bristol Prohealth and Bristol Cardiology helped develop patient priorities care and provided patient priorities care to their patients. These clinics received stipends for their participation.