Outcome Goals and Health Care Preferences of Older Adults With Multiple Chronic Conditions

Key Points Question What are the common goals and health care preferences of older adults with multiple health conditions? Findings In this cross-sectional study, 163 older adults reported 459 goals, most commonly encompassing activities with family and friends (24.2%), shopping (6.1%), exercising (4.6%), and living independently (4.4%). Medications most commonly cited as helpful were nonopioid pain medications (65.5% of users), sleep medications (52.9% of users), and inhalants (42.2% of users), whereas statins (25.8% of users) and antidepressants (32.5% of users) were the most commonly reported bothersome medications. Meaning Participants identified realistic and doable goals and health care preferences; this information can inform decision-making.


Introduction
Clinical decision-making can be difficult for the two-thirds of older adults who have multiple chronic conditions (MCCs). 1 These individuals are excluded from many clinical trials and may accrue less benefit or greater harm than suggested by disease guidelines because of coexisting conditions. [2][3][4][5][6] Most treatments are recommended to address disease-specific outcomes or survival, whereas older adults with MCCs vary in the outcome goals that they most desire. [7][8][9] Furthermore, the medications, health care visits, testing, procedures, and self-management tasks entailed in treating MCCs require investments of time and effort that may be burdensome and conflict with what patients are willing and able to do. [10][11][12] There is growing awareness of the need to transition health care, particularly for persons with MCCs, from treating single diseases in isolation to health care that is aligned with patients' priorities (ie, values, goals, and preferences). 5,9,[13][14][15][16][17] Most tools available for eliciting patients' goals and preferences were developed in the context of advanced illness or specific conditions or populations. [18][19][20][21][22][23][24][25][26][27][28][29] Some of these approaches may be less useful to older adults weighing the benefits, burdens, and tradeoffs associated with long-term management of several chronic conditions. Goal attainment scaling is appropriate for this latter population, although some applications include goal categories representing medical and supportive care, which are better considered care preferences that support attainment of desired life goals. [22][23][24][25][26][27][28][29] In response to the need for approaches that support the alignment of clinical decision-making with the priorities of older adults with MCCs, a diverse group of patients, health care systems experts, and clinicians developed Patient Priorities Care (PPC). 13,14 Aimed at both patients and clinicians, PPC begins with a structured process in which patients work with a member of the health care team to identify the outcome goals that they most desire to achieve based on what matters most to them (ie, their values-based outcome goals) and to specify what they are willing (or unwilling) to do to achieve those outcomes (ie, their health care preferences). [30][31][32] Clinicians then use the results of these facilitated discussions to align their decisions and care with patients' priorities. [31][32][33][34] Evidence from a recent study 31 indicated that the PPC framework was feasible to patients and clinicians and was associated with increased care aligned with patients' priorities and decreased treatment burden and unwanted health care. The aim of the present study is to describe the outcome goals and health care preferences identified by older adults in this study. 26

Study Design and Setting
Details of the PPC study have been described elsewhere. 30,31,33 Briefly, 10 primary care clinicians from a large multisite primary care practice in Connecticut invited their patients to participate in the PPC study during a routine visit. Relevant to the present cross-sectional study, patients who agreed to participate met with 1 of 2 trained members (K.H.-B. and an advanced practice registered nurse) of the practice to identify their priorities. The institutional review board at the Yale University School of Medicine, New Haven, Connecticut, approved this study; oral consent was obtained for all participants. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cross-sectional studies. 35

Participants
Inclusion criteria consisted of being 65 years or older, the ability to speak English, and having at least 3 chronic conditions; in addition, participants used at least 10 prescription medications, were seen by 2 or more specialists, or had at least 2 emergency department visits or 1 hospitalization during the past year. Exclusion criteria included advanced dementia, hospice eligibility, receiving dialysis, or nursing home residence. Potentially eligible participants were identified through electronic health record data. Of the 236 eligible patients treated by clinicians at the PPC practice, 163 (69.1%) agreed to participate in the PPC study and were included in the present analysis.

Patient Priorities Identification Process
Development of the process to identify patients' priorities was described previously. 30 Details of the identification process are included in the facilitator manual. 36  something you would like to be able to do more of. Goals need to be specific and realistic so that your doctor can work with you on making sure your health care is focused on achieving that goal." Prompts and probes were used to get the goals as specific, actionable, and realistic as possible. Participants then were asked what health problems or symptoms they believed most interfered with or were barriers to their identified outcome goals. Participants next identified as many as 3 health care activities (ie, medications, health care visits, procedures, diagnostic tests, and self-management tasks) that they thought would help most with their outcome goals, were doable, and were not too difficult or bothersome and as many as 3 health care activities in these categories that they found burdensome, bothersome, unhelpful, or that they wanted to stop if possible. Together these were considered health care preferences.
Facilitators recorded the results in a template that included participants' key values, outcome goals, and any barriers to achieving them and their health care preferences. Completed templates were transmitted to the electronic health record. A sample completed template is in the eFigure in the Supplement.

Statistical Analysis
Demographic data were ascertained during the baseline telephone interview and verified by electronic health record review. Chronic conditions were determined from the problem list and medications from the medication list of the primary care practice's electronic health records at the time of study enrollment. Medications included those prescribed for chronic conditions. Medications used for time-limited conditions (eg, antibiotics) were not included.
Participant characteristics were summarized with descriptive statistics. The outcome goals and health care preferences were obtained from the deidentified priorities templates. One author (M.E.T.) initially combined goals that reflected similar activities. The other authors reviewed and reached consensus on the goal groupings listed. The health care preferences were aggregated into the predetermined categories of medications, health care visits, procedures and tests, and selfmanagement tasks. Frequencies were ascertained for the outcome goals and health care preferences.

Results
Participant baseline characteristics are shown in Table 1 reported at least some college. Participants had a median of 4 (interquartile range, 3-5) chronic conditions.

Outcome Goals
The 163 participants identified 459 outcome goals ( Goals could be specific, such as "I want to go down to the dining room to eat and socialize each day." In other cases, they were less specific, such as, "I want to help my sister every day." Twenty individuals (12.3% of participants and 4.4% of goals) noted the desire to live independently or stay in their home but did not identify specific activities needed to ensure this could happen. Some outcome goals expressed a desire to continue a current activity (eg, "I want to continue to babysit my grandchildren every day"), whereas others reflected the aspiration to do something they were unable to do for health reasons, such as "I would like to be able to play golf again." Many health goals addressed more than 1 value. Most often this involved a function, such as personal mobility, cooking, driving, or traveling, that supported a goal related to relationships, enjoying life, or productivity (eg, "I want to cook and host my children for dinner each month" or "I want to continue to drive to the opera house 4 times a year with my friends").

Barriers to Achieving Outcome Goals
Although only 4 individuals reported outcome goals related to managing health or living longer-all of whom linked their health to meaningful events-participants did identify a health-related barrier to

Health Care Preferences
Health care activities that participants perceived as helpful and doable are shown in Table 3, whereas those activites that they believed were burdensome or bothersome are listed in Table 4

JAMA Network Open | Geriatrics
Outcome Goals and Health Care Preferences of Older Adults With Multiple Chronic Conditions shoulder pain" (Table 3). Fifteen participants (9.2%) reported having too many visits or clinicians: "I'm tired of going to so many doctors" (Table 4).
Procedures and tests were mentioned as acceptable and helpful by 38 participants (23.3%).
They mentioned both past procedures ("I'm very pleased with my recent knee surgery") and ones they hoped to have ("I would like to try hearing implant surgery") ( and too bothersome to 9 others (18.8%) ( Table 4). Other commonly cited activities included continuous positive airway pressure, which was helpful to 14 and bothersome to 9 participants and assistive devices (helpful to 21 and bothersome to 6 participants). The total numbers of participants using continuous positive airway pressure or assistive devices were unavailable.

Discussion
Identifying outcome goals that are realistic and actionable appears to be feasible among older adults with MCCs. No participant-selected goals were grandiose or unrealistic, likely because arriving at realistic and actionable goals given each person's health status was part of the facilitation process.
Goals often were linked to multiple values as expected, and reported previously, for meaningful human activities. 39    Approaches to identifying patients' goals have been reported for specific health problems or patient populations. [18][19][20][21][22][23][24][25][26][27][28][29] A goal taxonomy recently was developed for persons with functional limitations and complex care needs. 26,27 The patient priorities identification process builds on these efforts, particularly goal attainment scaling, by implementing a process that facilitates the separate identification of value-based outcome goals and health care preferences for persons with MCCs. 30 This separation supports clinical decision-making by focusing decision-making on the achievement of each person's outcome goals within the context of the health care the individual is willing and able to receive.
Participants varied in their health care preferences, the aspects of their current health care that they believed were acceptable and helpful, and the aspects that were burdensome, bothersome, or unhelpful. Although instructed to define helpful care in relation to their outcome goals, persons usually mentioned treatments that were helpful for individual conditions, such as hypertension or diabetes.
Not surprisingly, medications were the most frequently cited health care activity. Participants mentioned preventive medications, particularly statins, more often in the context of burden than benefit. Whether they were correct in ascribing bothersome effects to these medications is uncertain, but many participants found these medications to be burdensome. 41,42 This finding is consistent with previous work showing that older adults are reticent to take preventive medications associated with adverse symptoms that they consider to be health outcomes that need to be balanced against future benefit. 43 By comparison, medications that address current symptoms, such as pain or sleep, were more likely to be perceived as helpful than bothersome, although clinicians and regulators try to limit use of these medications because of adverse effects. 44 The differing perceptions of some patients and clinicians on the benefits and harms of preventive medications vs medications that alleviate symptoms needs to be acknowledged. Decision-making guided by patients' outcome goals may help address these discordant perspectives. h All 163 participants used to calculate denominator because most received at least 1, and other than a single mention of atenolol, participants did not name specific medications but rather the conditions for which medications were prescribed (eg, blood pressure, atrial fibrillation). Seven participants mentioned that blood pressure medications were not causing adverse effects; 6 mentioned a benefit such as blood pressure or heart rate control.
i Includes the 16 participants who received insulin.
k Includes the 48 participants who had a diagnosis of diabetes. • Will not take a prescription medication for sleep because it will "knock me out for 2 days" • "The sleeping pills make me too groggy in the morning" • "I stopped clonazepam because of my breathing, it caused the need for the CPAP" Diuretics k 7 (14.9) • Furosemide is helping but "I am peeing so much" • "They want me to take the water pill, I am going all afternoon and all night" Other medications mentioned as burdensome l 18 (11.0) • "I can't take PO meds for my osteoporosis because they caused chest pain" Health care visits Too many clinician visits or specialists 15 (9.2) • Would like less frequent clinician visits • Wants to limit number of specialists seen • "I'm tired of going to so many doctors" • "I know that my care is specialized but I want 1 person to look over everything because all the specialists tell me different things" Rehabilitation (physical therapy; pain management) 8 (4.9) • Did not do the physical therapy this round, "I don't think it helped" • The exercises put pressure on right shoulder • Went to pain management; the injections did not work • "I know they want me to go for the scans and the tests but I'm not going through that" • "I really want to talk about my blood, to see if I don't have to go that blood work anymore" • Mammogram and Papanicolaou smears: "I am trying to cut back on things because I am 90"

(continued)
A quarter of participants mentioned that visits or clinicians were helpful; 9.2% reported desiring fewer visits and clinicians. Because participants were not required to comment, results do not imply level of satisfaction with visits or clinicians. Some participants wished that 1 person could oversee their care or expressed frustration when clinicians' recommendations conflicted. Aligning decisions with patients' priorities is an effective strategy for coordinating care and avoiding conflicting recommendations. 33,34 Participants typically believed that the procedures they received were helpful. The 15% who raised concerns did so in the context of prospective procedures. More in-depth discussions are necessary to explore individuals' desires about future procedures. Having gone through the process of identifying their priorities may help guide these discussions, particularly if the tradeoffs are framed within patients' priorities.
As for other health care activities, individuals varied in their acceptance of self-management tasks, such as monitoring of glucose levels and use of continuous positive airway pressure, diets, assistive devices, and hearing aids. Of note, even among the 37.5% of patients with diabetes who reported that monitoring glucose levels was acceptable, most acknowledged that they did not check their glucose levels as often as recommended.

Limitations
Participants were drawn from a single practice with a homogeneous patient population; results may not generalize to other populations. Identifying the priorities of diverse groups is essential. It remains to be determined whether clinicians can use patients' goals to guide decision-making; preliminary evidence suggests they can. 31,32 Future work requiring larger samples from diverse populations includes categorizing goals based on agreed upon characteristics and ascertaining the abilities • Seeing the nutrition person was not helpful; "she told me to add up all the grams for every single thing that I buy, I can't do that, that is too much, I feel like she didn't help me" • Not motivated to diet or make healthier food choices at this time Assistive device 6 (3.7) • Quad cane catches left foot • "I won't use a cane because that is embarrassing" • "The walker helps me but it doesn't let me go into places and be safe" Hearing aids 5 (3.1) • Bilateral hearing aids, I need them but sometimes they pop out" • "I finally got new hearing aids, when I closed the fridge it sounded like a bomb went off" Other self-management tasks 4 (2.5) • "I want to continue to be a caregiver for my daughter each day. Health care tasks make this difficult, burdensome" Abbreviations: CPAP, continuous positive airway pressure; PO, oral. a The 163 participants were able to identify as many as 3 health care activities they considered burdensome, unhelpful, or unwanted.

JAMA Network Open | Geriatrics
Outcome Goals and Health Care Preferences of Older Adults With Multiple Chronic Conditions required to achieve specific goals, and conversely, the conditions, symptoms, and impairments impeding their achievement. Ultimately, methods will be needed for determining which interventions most effectively support achievement of specific goals.

Conclusions
Although further research is needed, this study suggests that asking people about their goals and preferences and getting responses that can inform decision-making is feasible. Combining patients' health conditions, function, and health trajectory with these goals and preferences should focus care for older adults with MCCs. Because various combinations of conditions and impairments as well as social determinants may affect goal achievement, aligning care with patients' priorities will require input from many health care professionals as well as community and other services. The variability in goals and preferences supports patients' priorities as the targets toward which to aim all health and support services.
In related work, strategies for aligning decision-making with patients' outcome goals and health care preferences have been developed and tested. 33,34 Tips and scripts for addressing challenges, such as nonactionable or nonspecific goals or goals that are not achievable because of health and functional status or because patients are unwilling to receive the health care necessary to achieve their goals, have also been created. 34 The eventual objective of this line of investigation is an approach to decision-making and care that helps older adults with MCCs achieve what matters most to them while minimizing harm and treatment burden.