Principle of rational prescribing and deprescribing in older adults with multiple chronic conditions

Abstract

Although the majority of older adults in the developed world live with multiple chronic conditions (MCCs), the task of selecting optimal treatment regimens is still fraught with difficulty. Older adults with MCCs may derive less benefit from prescribed medications than healthier patients as a result of the competing risk of several possible outcomes including, but not limited to, death before a benefit can be accrued. In addition, these patients may be at increased risk of medication-related harms in the form of adverse effects and significant burdens of treatment. At present, the balance of these benefits and harms is often uncertain, given that older adults with MCCs are often excluded from clinical trials. In this review, we propose a framework to consider patients’ own priorities to achieve optimal treatment regimens. To begin, the practicing clinician needs information on the patient’s goals, what the patient is willing and able to do to achieve these goals, an estimate of the patient’s clinical trajectory, and what the patient is actually taking. We then describe how to integrate this information to understand what matters most to the patient in the context of an array of potential tradeoffs. Finally, we propose conducting serial therapeutic trials of prescribing and deprescribing, with success measured as progress towards the patient’s own health outcome goals. The process described in this manuscript is truly an iterative process, which should be repeated regularly to account for changes in the patient’s priorities and clinical status. With this process, we aim to achieve optimal prescribing, that is, treatment regimens that maximize benefits that matter to the patient and minimize burdens and potential harms.

Keywords

chronic conditions deprescriptions health priorities medications multimorbidity polypharmacy prescriptions

Introduction

At the core of prescribing decisions is the clinician’s attempt to provide health benefits to the patient. While laudable in motivation, the narrow focus on potential benefits leaves out additional critical pieces of decision-making for older adults with multiple chronic conditions (MCCs) such as potential harms, treatment burdens, and patients’ health priorities. To illustrate the potential pitfalls of prescribing in these complex individuals, consider the following clinical decision, that is, whether to add an antihypertensive medication, in two patients: patient 1 is a 50-year-old woman with longstanding hypertension and no other conditions; patient 2 is an 85-year-old woman, also with longstanding hypertension, but also living with atrial fibrillation, diabetes, congestive heart failure, arthritis, and depression, and taking medications daily for each condition.

For patient 1, the clinician considers the benefits of treatment, here, reductions in the risk of future cardiovascular events. The clinician may consider treatment-related harms but these are likely to be mild. Burden of treatment is small and unlikely to change decision-making. For patient 2, the benefit of increasing antihypertensive intensity in reducing cardiovascular risk is relevant. However, the clinician must also consider other important questions. Will the patient live long enough to benefit? Will additional medications make any of the patient’s conditions worse or interact with her other medications? Will the treatment be so burdensome given her extensive list of medications that the patient will say, “It’s not worth the trouble”?

Here we seek to provide a reasoned approach to medication prescribing and deprescribing decisions for older adults with multiple chronic conditions, which aims to achieve clinical outcomes that matter most to each individual patient.

Rationale

Determining optimal treatment regimens for older adults with multiple coexisting chronic conditions is a common, yet challenging, problem. Among those 65 years and older in the United States, approximately 60% have at least two chronic conditions and 25% have at least four.¹ The prevalence of multiple chronic conditions is similarly high in the European Union.² Patients with MCCs have higher mortality, more impaired function, and lower quality of life than age-matched persons with fewer than three conditions.^3–5 The presence of MCCs also burdens individuals with medical visits, testing and monitoring, self-management tasks, as well as complex multidrug treatment regimens.⁶ Such healthcare utilization results in high costs to both individual patients and to society. In the United States, individuals with MCCs accounted for 71% of total healthcare spending in 2010 and also face out-of-pocket costs that are significantly greater than those with fewer conditions.⁷ Despite these large costs, it is often not clear whether prescribed medications provide net benefits or whether these potential benefits are aligned with what matters most to older adults with MCCs.

The relative benefits and harms of treatments for older adults with MCCs are often uncertain. Clinicians frequently turn to the panoply of evidence-based guidelines to assist with clinical decision-making, yet extrapolation of data from healthier patients to older adults with MCCs is not appropriate for several reasons. First, the clinical trials on which guidelines are based largely exclude or significantly underrepresent older adults with MCCs.^8,9 Few studies have investigated whether the treatment effects are modified in the presence of MCCs.¹⁰ In addition, medications prescribed for chronic diseases often require years of consistent treatment to achieve a clinical benefit. In individuals with significant mortality risk from other coexisting conditions, this lag time to benefit may be longer than their life expectancy.¹¹

On the harms side, adhering to guidelines for individual chronic diseases frequently leads to burdensome monitoring regimens and significant polypharmacy.¹² Nearly 50% of Medicare beneficiaries reported taking five or more medications;¹³ this polypharmacy places patients with MCCs at increased risk of adverse drug events, drug–drug interactions, and drug–disease interactions.¹⁴ Polypharmacy has been specifically associated with such undesirable outcomes as increased falls, hospitalizations, decreased function, and decreased cognition.¹⁴

Most guidelines recommend treatments with the goal of improving a disease-specific outcome or overall mortality.¹² This narrowly focused decision-making paradigm, however, paints an incomplete picture for those with MCCs. Older adults with MCCs face a number of potential undesirable health outcomes which are not limited to death; other common and important outcomes in this population include bothersome symptoms with multifactorial etiologies and declines in cognitive and physical function.¹⁵

There is also significant variation among these individuals as to which outcomes matter the most. Some patients value longevity at almost any cost, while others prefer maintenance of function or symptom control when faced with tradeoffs among these outcomes.^16–18 To provide truly patient-centered care, that is, to achieve outcomes that matter to patients, it is critical that clinicians consider patients’ values and preferences alongside benefits and harms.

Due to uncertainty about outcomes and variability in individual patient’s preferences, primary care clinicians describe significant difficulty using guidelines to select care options for patients with MCCs.^19,20 Beyond the guidelines, a variety of multidisciplinary interventions have been developed to reduce potentially inappropriate prescribing for older adults. For example, pharmacist-directed interventions have provided both education to clinicians and patients, as well as real-time collaboration to identify potentially inappropriate medications.^21–24 Other interventions have used automated computerized decision support to provide similar information.^25,26 To capture more nuanced information to guide prescribing practices, some have used comprehensive geriatric assessment to better understand prognosis and the potential contributions of medications to multifactorial geriatric syndromes such as urinary incontinence, falls, and delirium.²⁷ Certainly, comprehensive geriatric assessment, here, provides the clinician with necessary information about patients’ clinical and functional statuses as well as psychosocial supports, all factors which are important to the decision-making process in older adults.²⁸

More nuanced frameworks for deprescribing medications in older adults have addressed the need to integrate potential benefits and harms, patients’ prognoses, and patients’ goals of care into decision-making.^29–31 In this paper, we propose the development of specific, measurable goals to facilitate continual optimization of a medical regimen over time and consideration of treatment burden in decision-making. We also build on the concepts defined in these frameworks by using patients’ clinical goals, what they are willing to do to achieve these goals, clinical trajectory and prognosis, and the benefit/harms of each medication to construct easily understandable key tradeoffs to facilitate shared decision-making.

To guide clinicians through this uncertainty and difficulty, we propose three practical steps to decision-making about medications for older adults with MCCs: gather information, identify key tradeoffs, and conduct ‘therapeutic trials’ (Figure 1). These steps have been informed by the Patient Priorities Care approach, which seeks to align medical care provided by all of a patient’s clinicians with patients’ own health priorities.^32,33 With Patient Priorities Care, optimal medication decision-making maximizes benefits, minimizes harms and treatment burdens, and addresses a range of potential tradeoffs. A typical patient vignette illustrates the three steps.

Figure 1.

Iterative three step approach to medication decisionmaking in older adults with MCCs.

The case

Mr T is an 83-year-old man with stable heart failure with reduced ejection fraction (EF 30%), hypertension, and type II diabetes mellitus who presents for a primary care visit today. He lives with his wife of 60 years and remains independent in his activities of daily living, that is, he can bathe, dress, eat, use the toilet, and transfer unassisted. He requires help from his wife for housekeeping, driving, and meal preparation. Today, you plan on focusing on the management of his hypertension and diabetes.

Mr T brings in a log of weekly blood pressures taken with an automatic cuff at home with most readings measuring 150–160 mmHg systolic. His blood today in the office is 156/92. His hemoglobin A1c measured 1 week prior to this visit was 7.6 mg/dl. His current medications include metoprolol 100 mg twice daily, lisinopril 20 mg daily, furosemide 40 mg daily, and insulin glargine 50 U daily. Mr T and his wife want to know what should be done moving forward.

Step 1: gather information

At this point, the clinician may turn to the guidelines and add additional antihypertensive medications. In recent years, several guidelines acknowledge the uncertainty in evidence for those with MCCs, yet have given little practical guidance. Relevant to our case patient, the 2017 American College of Cardiology/American Heart Association hypertension guidelines recommend considering risks, benefits, and patient preferences when prescribing antihypertensives to older adults with limited life expectancy and high levels of MCC burden.³⁴ The American Diabetes Association goes a step further and gives recommendations for relaxed glycemic control goals when patients have MCCs or functional impairment. For a patient like Mr T, the guideline suggests a hemoglobin A1c goal of under 8 mg/dl and acknowledges that evidence is quite limited to support this recommendation.³⁵ While the acknowledgement of uncertainty is certainly an improvement over the one-size-fits-all approach of the past, the ultimate goals of guideline recommendations are still reductions in mortality or disease-specific outcomes. These outcomes may, or may not, be what matters most to older patients with MCCs who face several different health outcomes.

Given the substantial uncertainty in the evidence and varying preferences of patients, the clinician should use this opportunity to determine four key factors that will help future decisions: What matters most to the patient? What is the patient willing and able to do to achieve these outcomes? What is the patient’s likely health trajectory? and What is the patient actually taking?

What matters most?

In 2001, the Institute of Medicine advocated for improving healthcare quality by providing ‘care that is respectful of and responsive to individual patient preferences, needs, and values and ensures that patient values guide all clinical decisions’.³⁶ Prior studies demonstrated that older adults with MCCs are willing to engage with clinicians to set clinical goals that are consistent with their own priorities.³⁷ They are also able to understand potential tradeoffs that may occur with any of their health care decisions.^16,18,38,39 When discussing potential tradeoffs, there is significant variation in which outcomes matter most,¹⁵ dispelling the problematic perception among clinicians that all patients have the same goal, that is, ‘to be healthy’.⁴⁰ The process of eliciting patients’ priorities prior to decision-making also promotes patient engagement and improves treatment adherence.^39,41

Clinicians may have concerns about how to best discuss patients’ priorities. Several methods for eliciting priorities based on decision analysis have been developed,^42,43 although they have not yet entered widespread clinical use due to complexity and time constraints. The simplest and most feasible methods for clinical use focus on prioritization among a shared set of health domains affected by all chronic conditions, that is, universal health outcomes, namely mortality, function, and symptom burden.^44,45 In order to more clearly determine whether prescribing or deprescribing accomplishes what matters most to the patient, it is helpful to determine specific, measureable, actionable, realistic, and time-limited goals (SMART).⁴¹ This specific, measurable, actionable, realistic, and time-limited (SMART) paradigm, derived from the business literature,⁴⁶ has been used successfully to facilitate goal attainment scaling in the rehabilitation community.^47,48 A recent study from the Netherlands found setting SMART goals particularly helpful for medical students learning to design medication treatment regimens and follow-up plans.⁴⁹ The creators of the Patient Priorities Care approach, which informed the process described in this paper, developed a simple approach to identify and integrate patients’ SMART health outcome goals and their treatment preferences.⁵⁰ In the United Kingdom, a current cluster randomized trial seeks to understand the feasibility of goal setting in older adults with MCCs.⁵¹ In this intervention, physicians conduct the goal setting intervention with patients. With the Patient Priorities Care approach, the goal setting is done by trained nonphysician members of the clinical team.³³ An interactive, patient-facing website, Prepareforyourcare.org, designed for Advanced Care Planning,⁵² is another potentially useful alternative to prime patients for discussion of health priorities and potential tradeoffs. While these studies provide some initial strategies for integrating patient preferences in care provided to patients with MCCs, more work is needed to understand how to tailor the process of eliciting patient preferences to different practice settings with unique logistical constraints.

You return to Mr T and begin to explore his health priorities. Mr T reports that he is very concerned about becoming more sedentary and isolated. Six months ago, Mr T and his wife were participating in activities at their local senior center several times a week. Now, he no longer enjoys going to the senior center or to family gatherings due to fatigue and weakness. Mr T tells you that he has had several episodes over the last 6 months where he feels that his knees are going to buckle. When this has happened in public, Mr T had to quickly find somewhere to sit and wait until his wife could assist him.

You ask Mr T, ‘What matters most to you?’ He reports remaining as independent as possible is most important, especially because he does not want to be a burden to his wife. In order to accurately assess the effect of any intervention in the future, you work with Mr T to develop a SMART goal. Mr T wants to feel strong enough to play cards at the senior center at least once per week over the next 6 months. Importantly, this goal is currently out of his reach but is not unreasonable. Mr T identifies feelings of weakness and fatigue as the major barriers to accomplishing his goal.

What is the patient willing and able to do?

With the current disease-specific paradigm of treating chronic disease, older adults with MCC often face a complex array of fragmented health care tasks, each recommended by siloed clinicians.⁵³ In the US, the average older adult has visits with seven different clinicians per year.⁵⁴ In recognizing that a mounting treatment burden results in poor adherence and poor outcomes for patients with MCC, the minimally disruptive medicine approach calls for providing care that advances patients’ goals while minimizing the burden of treatment.⁵⁵ This approach relies on several related conceptual models, including the Cumulative Complexity Model and the Burden of Treatment Theory, which describe the balance of patient workload and patient capacity and how this balance affects patients’ quality of life.^56,57 The work of being a patient includes a wide array of health-related tasks, including but not limited to organizing and taking medications, lab testing, in-home monitoring and other self-management tasks, health care visits, coordinating care and findings between multiple clinicians, and coordinating the financing of care. Capacity includes the sum of the physical, cognitive, and social abilities of the patient and their available social networks that enable performance of health-related tasks.

High levels of treatment burden occur when patient workload eclipses capacity. Resultant nonadherence wastes health care resources and strains clinician–patient relationships.^55,58 More importantly, patient quality of life suffers with high treatment burden.^59,60 Several scales have been developed for research purposes to quantify levels of burden,^59–61 and the recently developed Revised Patients’ Attitudes Towards Deprescribing Questionnaire integrates assessment of burden with willingness to consider deprescription.⁶² One significant challenge to the assessment of burden in busy clinical practice is that administration of the above scales may be overly time consuming. As such, assessment of burden may be obtained by trained nonphysician members of the clinical team prior to a visit with the primary clinician. At present, efficient but comprehensive means of obtaining this information has been understudied. Experiential learning during the pilot of the Patient Priorities Care approach at a community primary care practice has identified questions such as ‘Are there aspects of your care that you find difficult or burdensome?’ or ‘If we could change one aspect about your healthcare, what would it be?’ as simple, efficient ways of evaluating burden in clinical practice.⁶³

You ask Mr T and his wife whether there is any part of his care that they find difficult. At this point, his wife mentions that they are starting to feel overwhelmed with his medical care and Mr T nods his head in agreement. ‘I have so many doctors’ appointments and I really hate pricking my finger once a day to check my sugar.’

What is the patient’s likely health trajectory?

Understanding the patient’s likely health trajectory is critical to making informed clinical decisions. Health trajectory includes life expectancy, but is a broader concept which also encompasses likely changes in cognition, function, symptoms, and clinical stability. The simplest case of applying knowledge of a patient’s likely trajectory to decision-making occurs when a patient’s life expectancy is shorter than the lag time to benefit of a particular medication. In this case, the medication is unlikely to provide benefit.¹¹ Trajectories that include outcomes other than mortality are also important; patients may prioritize possible clinical outcomes differently in differing clinical contexts.^41,64 For example, a patient who previously prioritized longevity may prefer symptom control if their likely trajectory involved substantial functional loss and need for nursing home placement.

Primary care clinicians often recognize the value of estimating a patient’s prognosis but there remains significant uncertainty about how to most accurately do so and how best to incorporate that information in clinical decision-making. In two qualitative studies, primary care clinicians described several factors that influence their own perception of patients’ trajectories, but expressed little familiarity or comfort with using validated prognostic indices.^65,66 Evidence suggests that clinicians’ ability to prognosticate is poor; estimates vary particularly when patients have longer predicted life expectancies.^67,68 Several prognostic indices have been developed and validated for predicting mortality in older adults.^69–72 The website EPrognosis provides an interactive tool to allow for easy calculation of mortality risk in the short term (1 year) and long term (5 and 10 years) using these indices.^73,74 Notably these indices are not disease specific and likely provide more appropriate estimates for patients with MCCs compared with the multiple of disease-specific indices.

Given the priority many older adults place on outcomes other than death, prognostication of other outcomes is also important in clinical decision-making.⁶⁶ However, there are few tools to predict the probability of multiple competing outcomes. A promising methodology has recently been proposed by Allore and colleagues to calculate the absolute risk of several competing outcomes in older adults with MCCs.⁷⁵ In clinical practice, comprehensive geriatric assessment can add important information about function, cognition, and the presence of geriatric syndromes, which are not captured by a disease-centric approach to care. These elements together influence patients’ mortality risk but are also important outcomes in and of themselves.

There remains uncertainty about how to best discuss estimated trajectories with patients. Qualitative studies of older adults demonstrated wide variation in patients’ willingness to discuss life expectancy,^76–78 as well as the optimal timing and content for these discussions.⁷⁹ Given this heterogeneity, clinicians should ask how much information about prognosis the patient would like to know and why that information may be relevant to clinical decision-making.⁷⁸ Furthermore, some methods of presenting prognostic information are better received by patients than others. In a recent qualitative analysis of older adults’ perceptions of stopping cancer screening, patients preferred the positively framed ‘This intervention will not help you live longer’ to the more negatively framed ‘You will not live long enough to benefit from this intervention’.⁷⁷ This positive framing can be easily applied to discussing medication-related decisions with patients with limited life expectancy.

Based on your knowledge of Mr T’s chronic conditions, functional impairments, and age, you estimate that Mr T is unlikely to live 10 years and is probably unlikely to live 5 years. You recall that clinician estimates are often crude, and plug in Mr T’s information into the calculator on EPrognosis. Based on the Lee and Schonberg Indices, Mr T’s predicted mortality at 5 years is 63–70% and 92–93% at 10 years (in fact, not very different from your clinical gestalt in this case). You also estimate that Mr T is very likely to be become more dependent in the next several years, knowing that his chronic conditions and their treatments have already affected his function. You ask him, ‘What do you think the next few years will look like for you?’ Mr T replies, ‘I know I’m probably going to get worse, but I’d like to stay independent as long as I can. I’m 83. I know I won’t live forever.’

What is the patient actually taking?

Optimizing a patient’s medical management is predicated on having an accurate understanding of the medications that patient is actually taking. This can be easier said than done, with multiple potential specialist providers, transitions of care in which the medication list may be adjusted, and potential use of over-the-counter drugs or supplements. With these potential complications, patient-reported medication lists are often inaccurate.⁸⁰ The simple brown bag approach in which a patient is asked to bring all of their medications into a clinic visit has promise to improve accuracy of the medication list, yet may be limited by the clarity of instructions provided to patients.⁸¹ A ‘Brown Bag Medication Review’ tool, developed by the Agency for Healthcare Research and Quality as a portion of its Health Literacy Universal Precautions Toolkit, holds significant promise to improve the quality of medication review.⁸² Potential other sources of information about prescriptions in older adults include dispensing pharmacies and caregivers. With attention to the time necessary to perform a thorough medication review, this process may be best accomplished by other members of the clinical team, including pharmacists, nurses, and medical assistants.

Step 2: identify key tradeoffs

After gathering information about the patient’s health priorities, what they find burdensome, and estimating their clinical trajectory, the clinician considers how these factors result in meaningful tradeoffs. Addressing potential tradeoffs of drug treatment is at the core of effective decision-making for older adults with MCCs. Several prior studies have demonstrated that patients with MCCs are able to articulate their priorities in the face of tradeoffs.^18,38,83 In Table 1, we listed common types of tradeoffs, illustrative clinical examples, and strategies for clinical management.

Table 1.

Tradeoffs encountered by older adults with MCCs.

Type of tradeoff	Clinical examples	Strategies for clinical management
Future benefit versus current harm		Consider patient’s likely trajectory and estimate likelihood of benefitAssess current harms, that is, adverse effects, worsening of other conditions, significant treatment burdenPresent estimates of future benefits in the context of patient’s SMART goals and treatment preferences.Trial of reducing/stopping medication with goal of improving the current harmTrial of different medication with different adverse effect profile with goal of improving the current harmMeasure success by patient’s ability to perform desired activity
Adverse effect	Patient with hypertension on multiple medications with goal of reducing risk of stroke but has orthostasis and dizziness
Therapeutic competition	Patient with heart failure and chronic obstructive pulmonary disease (COPD). Metoprolol prescribed to reduce mortality but makes COPD symptoms worse
High treatment burden	Patient with diabetes is prescribed insulin to reduce the risk of diabetes-related complications. Although he does not have adverse effects, he feels burdened by fingersticks to check his glucose and insulin injections
Future benefit versus future harm	Patient with atrial fibrillation and prior gastrointestinal bleeding is prescribed warfarin for reduction in the risk of stroke but has increased bleeding risk	Consider patient’s likely trajectory and estimate likelihood of benefitConsider treatment burdenPresent estimates of future benefits and harms of potential strategies in the context of patient’s SMART goals and treatment preferences
Current benefit versus current harm	Patient with peripheral neuropathy is prescribed gabapentin, which does improve her neuropathic pain but also makes her feel sedated	Identify hierarchy of current outcomes given specific SMART goals and treatment preferencesTrial of decreasing or stopping medication to determine if tradeoff improvesTrial of other medications with different adverse effect profileMeasure success by patient’s ability to perform desired activity
Current benefit versus future harm	Patient with dementia with violent behavioral disturbance is prescribed risperidone, which improves his aggression but increases his risk of death and stroke	Identify hierarchy of outcomes, in this case, with the patient’s caregiverTrial of decreasing or stopping medication to determine if the longer-term risk is necessary to achieve the current benefitTrial other medications with less harms

MCCs, multiple chronic conditions.

One common tradeoff for older adults with MCCs is between a future benefit and a current harm. Current harms may be intolerable adverse effects or treatment burden. In addition, in patients with MCCs, current harms can take the form of therapeutic competition in which the treatment for one condition exacerbates the symptoms of another.⁸⁴ By their nature, tradeoffs are context specific in that one outcome is not evaluated by itself but rather in relation to another. Here, a concrete example illustrates this important point. A patient with diabetes may experience only a small amount of burden from daily fingerstick glucose measurements and be mildly bothered by the loose stools as a result of metformin. If the potential benefit of tight glucose control is large, the patient may be willing to deal with these current harms. On the other hand, if the benefit is small or uncertain, these treatment activities may be considered ‘too much’. Other important categories of tradeoffs are between future benefits and future harms, between current benefits and current harms, and current benefits and future harms. Often clinical decision-making about one drug can encompass more than one type of tradeoff.

For Mr T, you identify the following key tradeoffs. Mr T’s metoprolol and lisinopril may have benefit in terms of reducing his mortality and exacerbations (a future benefit); however, it may also be a contributor to his fatigue (a current harm). His insulin is unlikely to reduce his mortality given his overall health trajectory and the adverse effect of hypoglycemia may be contributing to current symptoms of weakness and fatigue.

Mr T understands that that there is uncertainty about the potential mortality benefits of his antihypertensives and his diabetes medications. When weighing a potential small benefit against his current fatigue, he states, ‘I’ve got to be able to move, doc. Even if it means I don’t live as long, I hate having so much difficulty moving around.’

Step 3: serial coordinated ‘therapeutic trials’

Armed with the necessary information about the patient’s estimated clinical trajectory, priorities among multiple possible health outcomes, and which healthcare tasks are burdensome, the clinician is poised to discuss potential tradeoffs with patients and any other persons that the patient would like involved (e.g. family members, friends, or caregivers). In this review, we focus on decision-making for older adults who still have the capacity to make their own decisions. Certainly, similar albeit more complicated methods may be helpful to navigate shared decision-making that occurs primarily with caregivers in the setting of advanced cognitive impairment.

In patients with MCCs, it is often difficult to determine where to start. There may be a multitude of potential tradeoffs which need to be addressed. It is often unclear which conditions and treatments contribute to the outcomes that matter to patients; survival, function, symptoms, and quality of life are likely to be multifactorial outcomes in these patients.⁸⁵ In 2010, the American Geriatrics Society convened an expert panel to address the provision of optimal care for older adults with MCCs, resulting in the development of the MCC Guiding Principles.⁸⁶ The panel has now translated these principles into a series of practical action steps for practicing clinicians. One practical recommended strategy is to perform serial trials of starting or stopping medications (manuscript under review).⁸⁷ Experiential learning via implementation of the Patient Priorities Care approach in a community primary care practice has affirmed the feasibility and utility of this strategy.⁶³

Initial targets for deprescribing are those medications in which the harm of treatment greatly outweighs the potential benefits. Several evidence-based criteria have been developed to identify these high-risk medications for older adults, including STOPP-START and the Beers List of Potentially Inappropriate Medications.^63,65 Examples of medications with high risk of current harms include benzodiazepine and anticholinergic agents. Clinicians should discuss with patients that alternative agents or nonpharmacologic treatments may offer similar potential benefits with much less harm (for example, timed voiding rather than anticholinergics for urinary incontinence). In most cases, these alternatives exist. In a case in which other alternatives have been explored and ineffectual, a competent patient may choose to remain on a high-risk medication for its current benefit. This does not apply to patients in whom there is suspicion of drug seeking and addiction, a discussion of which is outside the scope of this review.

Next, the clinician and patient should discuss and prioritize addressing medications that offer little benefit given that the patient’s predicted trajectory and medications are not consistent with the patient’s outcome goals given their care preferences. Deprescribing of statins, an example of the former category, has been demonstrated to be safe and improve quality of life in patients with advanced life-limiting illness.⁸⁸ The Garfinkel method of deprescribing multiple medications has demonstrated that several unnecessary medications can be safely deprescribed with improvements in a variety of outcomes, including general satisfaction, cognition, and function, and no increases in hospitalizations or mortality.⁸⁹ Clinicians should also give consideration to starting or continuing medications that are likely to be beneficial given the patient’s trajectory and are consistent with the patient’s goals and preferences. An example, here, may be continuing clopidogrel in a patient with a very recent myocardial infarction if he or she is not at increased risk of bleeding and is not having adverse effects.

Since Mr T’s goal was playing cards at the senior center a few times each week and fatigue is his biggest barrier to that, you decide to explore possible causes of his fatigue. There are many potential etiologies including hypotension, hypoglycemia, bradycardia, adverse effect of his metoprolol (independent of blood pressure effects), worsening of his heart failure, or a combination of several of these. Given that he describes burden from his insulin injections and checking his blood sugar, you decide to begin by stopping his insulin gradually.

While they can be quite simple in practice, there are a few key factors which must be addressed to optimize the ability of a therapeutic trial to provide useful information. First, it is critical that both the patient and clinician can measure the success of a given treatment change. For this reason, it is important to make specific, measurable, actionable, realistic, and time-limited (SMART) goals as shared measuring sticks. As patients with MCCs often see several clinicians, it is important that all clinicians communicate clearly with each other and coordinate care effectively. One can imagine a situation in which Mr T agrees to stop insulin, only to be placed on a sulfonylurea by another clinician. Such fragmentation is frustrating and confusing to patients and clinicians alike, and can complicate the interpretation of a therapeutic trial. At the end of a therapeutic trial, whether or not progress was made towards the patient’s goals, there will be more information to guide discussion of tradeoffs.

Mr T reduces the dose of his insulin over a week and then trials 1 month without insulin. He tests his blood glucose every other day at various times and his glucose measures about 150 mg/dl (8.3 mmol/liter) fasting and about 200 mg/dl (11.1 mmol/liter) postprandially. While he is happy to be free of daily injections, he still feels quite fatigued and has not been able to get to the senior center.

This therapeutic trial provided two valuable pieces of information to readdress tradeoffs with Mr T. Now, we know that hypoglycemia was not the only driver of his fatigue. However, his treatment burden is much lower without daily injections and monitoring. Knowing this, it is reasonable to readdress the decision of whether to resume insulin or remain off of it; the tradeoff now is simply between a small potential mortality benefit and treatment burden. Second, not achieving the patient’s goal of reducing his fatigue sufficiently enough to participate in his desired activities tells us that we need to continue serial trials to identify a potential contributor to his symptoms.

You revisit the tradeoff between possible life prolongation and treatment burden associated with insulin with Mr T. He feels quite strongly that he prefers staying off insulin as his quality of life is improved without the burden of injections. Given that he is still not meeting his SMART goal, you decide together that you will trial a lowering of his metoprolol dose.

Summary

Clinical decisions to start or stop medications for older adults with MCCs can be quite complicated. These patients may derive less benefits than healthier patients due to their shortened life expectancy and the competing risk of several outcomes. In this paper, we propose a rational framework for considering mediation-related decisions. In step 1, we suggest gathering information about patients’ goals (i.e. what health and life outcomes matter most) and care preferences (i.e. what health-related tasks they are willing and able to do to achieve their desired health outcomes), their current medication regimens, and their likely health trajectories. The basic information gathered in step 1 enables step 2, which is the mutual exploration of the wide array of potential tradeoffs when choosing one medication regimen versus another. In step 3, the clinician and patient begin a series of therapeutic trials with the goal of achieving the patient’s desired outcome. The entire process is meant to be iterative; symptoms, clinical trajectories, and life circumstances often change over time and may result in changes in patients’ priorities. Patients want clinicians to understand that these changes occur and readdress the priorities frequently.⁹⁰ The approach to prescribing and deprescribing detailed here strives to minimize burden and achieve outcomes that matter to older adults with MCCs.

Footnotes

Funding

Funding for this work was supported by grants from The John A. Hartford Foundation, Patient-Centered Outcomes Research Institute, The Gordon and Betty Moore Foundation and The Robert Wood Johnson Foundation. The authors received additional support and resources from the Claude D. Pepper Older Americans Independence Center at Yale University School of Medicine (P30AG021342 NIH/NIA). Dr Gregory Ouellet is supported by the Yale Postdoctoral Training Program in Geriatric Clinical Epidemiology and Aging-Related Research (T32AG19134 NIH/NIA).

Conflict of interest statement

Dr Tinetti is the primary investigator of Patient Priorities Care, which significantly influenced the development of the principles expounded in this manuscript. Drs Gregory and Jennifer Ouellet have also made scientific contributions to Patient Priorities Care. The authors have no financial conflicts of interest to disclose.

ORCID iD

Gregory M. Ouellet

References

Lochner

Goodman

Posner

et al . Multiple chronic conditions among Medicare beneficiaries: state-level variations in prevalence, utilization, and cost, 2011. Medicare Medicaid Res Rev 2013; 3: mmrr.003.03.b02.

Onder

Palmer

Navickas

et al . Time to face the challenge of multimorbidity. A European perspective from the joint action on chronic diseases and promoting healthy ageing across the life cycle (JA-CHRODIS). Eur J Intern Med 2015; 26: 157–159.

Newman

Boudreau

Naydeck

et al . A physiologic index of comorbidity: relationship to mortality and disability. J Gerontol A Biol Sci Med Sci 2008; 63: 603–609.

Marengoni

Von Strauss

Rizzuto

et al . The impact of chronic multimorbidity and disability on functional decline and survival in elderly persons. A community-based, longitudinal study. J Intern Med 2009; 265: 288–295.

Bayliss

Ellis

Steiner

JF.

Subjective assessments of comorbidity correlate with quality of life health outcomes: initial validation of a comorbidity assessment instrument. Health Qual Life Outcomes. 2005; 3: 51.

Spencer-Bonilla

Quiñones

Montori

VM.

Assessing the burden of treatment. J Gen Intern Med 2017; 32: 1141–1145.

Gerteis

Izrael

Deitz

et al . Multiple chronic conditions chartbook: 2010 Medical Expenditure Panel Survey data. Rockville, MD: Agency for Healthcare Research and Quality, 2014.

Boyd

Vollenweider

Puhan

MA.

Informing evidence-based decision-making for patients with comorbidity: availability of necessary information in clinical trials for chronic diseases. PloS One 2012; 7: e41601.

Jadad

Emara

et al . Consideration of multiple chronic diseases in randomized controlled trials. JAMA 2011; 306: 2670–2672.

10.

Fried

O’Leary

Towle

et al . The effects of comorbidity on the benefits and harms of treatment for chronic disease: a systematic review. PloS One 2014; 9: e112593.

11.

Lee

Leipzig

Walter

LC.

“When will it help?” incorporating lagtime to benefit into prevention decisions for older adults. JAMA 2013; 310: 2609–2610.

12.

Boyd

Darer

Boult

et al . Clinical practice guidelines and quality of care for older patients with multiple comorbid diseases: implications for pay for performance. JAMA 2005; 294: 716–724.

13.

Safran

Neuman

Schoen

et al . Prescription drug coverage and seniors: findings from a 2003 national survey. Health Aff 2005; Suppl Web Exclusives: W5-152-w155-166.

14.

Fried

O’Leary

Towle

et al . Health outcomes associated with polypharmacy in community-dwelling older adults: a systematic review. J Am Geriatr Soc 2014; 62: 2261–2272.

15.

Fried

Tinetti

Agostini

et al . Health outcome prioritization to elicit preferences of older persons with multiple health conditions. Pat Ed Couns 2011; 83: 278–282.

16.

Fried

McGraw

Agostini

et al . Views of older persons with multiple morbidities on competing outcomes and clinical decision-making. J Am Geriatr Soc 2008; 56: 1839–1844.

17.

Fried

Tinetti

Iannone

et al . Health outcome prioritization as a tool for decision making among older persons with multiple chronic conditions. Arch Intern Med 2011; 171: 1854–1856.

18.

Tinetti

McAvay

Fried

et al . Health outcome priorities among competing cardiovascular, fall injury, and medication-related symptom outcomes. J Am Geriatr Soc 2008; 56: 1409–1416.

19.

Austad

Hetlevik

Mjolstad

et al . Applying clinical guidelines in general practice: a qualitative study of potential complications. BMC Fam Pract 2016; 17: 92.

20.

Fried

Tinetti

Iannone

Primary care clinicians’ experiences with treatment decision making for older persons with multiple conditions. Arch Intern Med 2011; 171: 75–80.

21.

Spinewine

Swine

Dhillon

et al . Effect of a collaborative approach on the quality of prescribing for geriatric inpatients: a randomized, controlled trial. J Am Geriatr Soc 2007; 55: 658–665.

22.

Gustafsson

Sandman

Karlsson

et al . Reduction in the use of potentially inappropriate drugs among old people living in geriatric care units between 2007 and 2013. Eur J Clin Pharmacol 2015; 71: 507–515.

23.

Moga

Abner

Rigsby

et al . Optimizing medication appropriateness in older adults: a randomized clinical interventional trial to decrease anticholinergic burden. Alzheimers Res Ther 2017; 9: 36.

24.

Nightingale

Hajjar

Swartz

et al . A pharmacist-led medication assessment to identify prevalence and associations with potentially inappropriate medications (PIMs) in ambulatory senior adults with cancer. J Clin Oncol 2014; 32(15 Suppl.): e20516.

25.

Marasinghe

KM.

Computerised clinical decision support systems to improve medication safety in long-term care homes: a systematic review. BMJ Open 2015; 5: e006539.

26.

Fried

Niehoff

Street

et al . Effect of the tool to reduce inappropriate medications on medication communication and deprescribing. J Am Geriatr Soc 2017; 65: 2265–2271.

27.

Avelino-Silva

Farfel

Curiati

et al . Comprehensive geriatric assessment predicts mortality and adverse outcomes in hospitalized older adults. BMC Geriatr 2014; 14: 129.

28.

Lea

Watts

Davis

et al . The potential clinical benefits of medicines optimisation through comprehensive geriatric assessment, carried out by secondary care geriatricians, in a general practice care setting in North Staffordshire, UK: a feasibility study. BMJ Open 2017; 7: e015278.

29.

Holmes

Hayley

Alexander

et al . Reconsidering medication appropriateness for patients late in life. Arch Intern Med 2006; 166: 605–609.

30.

Scott

Gray

Martin

et al . Minimizing inappropriate medications in older populations: a 10-step conceptual framework. Am J Med 2012; 125: 529–537.

31.

Jansen

Naganathan

Carter

et al . Too much medicine in older people? Deprescribing through shared decision making. BMJ 2016; 353: i2893.

32.

Tinetti

Naik

Dodson

JA.

Moving from disease-centered to patient goals-directed care for patients with multiple chronic conditions: patient value-based care. JAMA Cardiol 2016; 1: 9–10.

33.

Tinetti

Esterson

Ferris

et al . Patient priority-directed decision making and care for older adults with multiple chronic conditions. Clin Geriatr Med 2016; 32: 261–275.

34.

Whelton

Carey

Aronow

et al . ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2017: 24430.

35.

American Diabetes Association. Standards of medical care in diabetes—2018 abridged for primary care providers. Clin Diabetes 2018; 36: 14–37.

36.

Baker

Crossing the quality chasm: a new health system for the 21st century. BMJ 2001; 323: 1192.

37.

Naik

Schulman-Green

McCorkle

et al . Will older persons and their clinicians use a shared decision-making instrument? J Gen Intern Med 2005; 20: 640–643.

38.

Case

O’Leary

Kim

et al . Relationship between universal health outcome priorities and willingness to take medication for primary prevention of myocardial infarction. J Am Geriatr Soc 2014; 62: 1753–1758.

39.

Abraham

Naik

Street

Jr et al . Complex antithrombotic therapy: determinants of patient preference and impact on medication adherence. Patient Prefer Adherence 2015; 9: 1657–1668.

40.

Schulman-Green

Naik

Bradley

et al . Goal setting as a shared decision making strategy among clinicians and their older patients. Patient Educ Couns 2006; 63: 145–151.

41.

Naik

Martin

Moye

et al . Health values and treatment goals of older, multimorbid adults facing life-threatening illness. J Am Geriatr Soc 2016; 64: 625–631.

42.

Lenert

Kaplan

RM.

Validity and interpretation of preference-based measures of health-related quality of life. Med Care 2000; 38(9 Suppl.): II138–II150.

43.

Horsman

Furlong

Feeny

et al . The Health Utilities Index (HUI(®)): concepts, measurement properties and applications. Health Qual Life Outcomes 2003; 1: 54.

44.

Case

Fried

O’Leary

How to ask: older adults’ preferred tools in health outcome prioritization. Patient Educ Couns 2013; 91: 29–36.

45.

Uhlig

Leff

Kent

et al . A framework for crafting clinical practice guidelines that are relevant to the care and management of people with multimorbidity. J Gen Intern Med 2014; 29: 670–679.

46.

Doran

GT.

There’s a S.M.A.R.T. way to write management’s goals and objectives. Manage Rev 1981; 70: 35–36.

47.

Bovend’Eerdt

Botell

et al . Writing SMART rehabilitation goals and achieving goal attainment scaling: a practical guide. Clin Rehabil 2009; 23: 352–361.

48.

Hersh

Worrall

Howe

et al . SMARTER goal setting in aphasia rehabilitation. Aphasiology 2012; 26: 220–233.

49.

Tichelaar

Uil den

Antonini

et al . A ‘SMART’ way to determine treatment goals in pharmacotherapy education. Br J Clin Pharmacol 2016; 82: 280–284.

50.

Naik

Dindo

Van Liew

et al . Development of a clinically-feasible process for identifying patient health priorities. J Am Geriatr Soc 2018. (accepted).

51.

Steel

Clark

Lipp

et al . Goal-setting in care planning for people with multimorbidity: feasibility study and intervention refinement. Study Protocol, https://www.uea.ac.uk/medicine/research/primary-care-and-epidemiology/goalplan. (2016, accessed 20 May 2018).

52.

Sudore

Boscardin

Feuz

et al . Effect of the prepare website vs an easy-to-read advance directive on advance care planning documentation and engagement among veterans: a randomized clinical trial. JAMA Intern Med 2017; 177: 1102–1109.

53.

Bohlen

Scoville

Shippee

et al . Overwhelmed patients: a videographic analysis of how patients with type 2 diabetes and clinicians articulate and address treatment burden during clinical encounters. Diabetes Care 2012; 35: 47–49.

54.

Pham

Schrag

O’Malley

et al . Care patterns in Medicare and their implications for pay for performance. N Engl J Med 2007; 356: 1130–1139.

55.

May

Montori

Mair

FS.

We need minimally disruptive medicine. BMJ 2009; 339: b2803.

56.

Shippee

Shah

May

et al. , Cumulative complexity: a functional, patient-centered model of patient complexity can improve research and practice. J Clin Epidemiol 2012; 65: 1041–1051.

57.

May

Eton

Boehmer

et al . Rethinking the patient: using Burden of Treatment Theory to understand the changing dynamics of illness. BMC Health Serv Res 2014; 14: 281.

58.

Naik

Dyer

Kunik

et al . Patient autonomy for the management of chronic conditions: a two-component re-conceptualization. Am J Bioeth 2009; 9: 23–30.

59.

Tran

Montori

Eton

et al . Development and description of measurement properties of an instrument to assess treatment burden among patients with multiple chronic conditions. BMC Med 2012; 10: 68.

60.

Eton

Yost

Lai

et al . Development and validation of the Patient Experience with Treatment and Self-management (PETS): a patient-reported measure of treatment burden. Qual Life Res 2017; 26: 489–503.

61.

Eton

Elraiyah

Yost

et al . A systematic review of patient-reported measures of burden of treatment in three chronic diseases. Patient Relat Outcome Meas 2013; 4: 7–20.

62.

Reeve

Low

Shakib

et al . Development and validation of the revised patients’ attitudes towards deprescribing (rPATD) Questionnaire: Versions for Older Adults and Caregivers. Drugs Aging 2016; 33: 913–928.

63.

Tinetti

Dindo

Smith

et al . Strategies for aligning decision-making with the health priorities of older adults with multiple chronic conditions. J Am Geriatr Soc 2018. (accepted).

64.

Bradley

Bogardus

Jr Tinetti

et al . Goal-setting in clinical medicine. Soc Sci Med 1999; 49: 267–278.

65.

Schoenborn

Bowman

et al . Primary care practitioners’ views on incorporating long-term prognosis in the care of older adults. JAMA Intern Med 2016; 176: 671–678.

66.

Thomas

Fried

TR.

Defining the scope of prognosis: primary care clinicians’ perspectives on predicting the future health of older adults. J Pain Symptom Manage 2018; 55: 1269–1275.

67.

Christakis

Lamont

EB.

Extent and determinants of error in doctors’ prognoses in terminally ill patients: prospective cohort study. BMJ 2000; 320: 469–472.

68.

Wirth

Sieber

CC.

Health care professionals underestimate the mean life expectancy of older people. Gerontology 2012; 58: 56–59.

69.

Lee

Lindquist

Segal

et al . Development and validation of a prognostic index for 4-year mortality in older adults. JAMA 2006; 295: 801–808.

70.

Schonberg

Davis

McCarthy

et al . Index to predict 5-year mortality of community-dwelling adults aged 65 and older using data from the National Health Interview Survey. J Gen Intern Med 2009; 24: 1115–1122.

71.

Schonberg

Marcantonio

et al . Predicting mortality up to 14 years among community-dwelling adults aged 65 and older. J Am Geriatr Soc 2017; 65: 1310–1315.

72.

Cruz

Covinsky

Widera

et al . Predicting 10-year mortality for older adults. JAMA 2013; 309: 874–876.

73.

Yourman

Lee

Schonberg

et al . Prognostic indices for older adults: a systematic review. JAMA 2012; 307: 182–192.

74.

Lee

Smith

Widera

et al . ePrognosis, https://eprognosis.ucsf.edu/ (accessed 10 March 2018).

75.

Allore

McAvay

Vaz Fragoso

et al . Individualized absolute risk calculations for persons with multiple chronic conditions: embracing heterogeneity, causality, and competing events. Intern J Stat Med Res 2016; 5: 48–55.

76.

Fried

Bradley

O’Leary

Prognosis communication in serious illness: perceptions of older patients, care givers and clinicians. J Am Geriatr Soc 2003; 51: 1398–1403.

77.

Schoenborn

Lee

Pollack

et al . Older adults’ views and communication preferences about cancer screening cessation. JAMA Intern Med 2017; 177: 1121–1128.

78.

Ahalt

Walter

Yourman

et al . “Knowing is better”: preferences of diverse older adults for discussing prognosis. J Gen Intern Med 2012; 27: 568–575.

79.

Schoenborn

Lee

Pollack

et al . Older adults’ preferences for when and how to discuss life expectancy in primary care. J Am Board Fam Med 2017; 30: 813–815.

80.

Lee

Nishimura

Ngu

et al . Predictors of completeness of patients’ self-reported personal medication lists and discrepancies with clinic medication lists. Ann Pharmacother 2014; 48: 168–177.

81.

Sarzynski

Luz

Zhou

et al . Medication reconciliation in an outpatient geriatrics clinic: does accuracy improve if patients “brown bag” their medications for appointments? J Am Geriatr Soc 2014; 62: 567–569.

82.

Weiss

Brega

LeBlanc

et al . Improving the effectiveness of medication review: guidance from the health literacy universal precautions toolkit. J Am Board Fam Med 2016; 29: 18–23.

83.

Fried

Tinetti

Towle

et al . Effects of benefits and harms on older persons’ willingness to take medication for primary cardiovascular prevention. Arch Intern Med 2011; 171: 923–928.

84.

Lorgunpai

Grammas

Lee

DSH

et al . Potential therapeutic competition in community-living older adults in the U.S.: use of medications that may adversely affect a coexisting condition. PloS One 2014; 9: e89447.

85.

Tinetti

McAvay

Chang

et al . Contribution of multiple chronic conditions to universal health outcomes. J Am Geriatr Soc 2011; 59: 1686–1691.

86.

Guiding Principles for the Care of Older Adults with Multimorbidity: An Approach for Clinicians. Guiding principles for the care of older adults with multimorbidity: an approach for clinicians: American Geriatrics Society Expert Panel on the Care of Older Adults with Multimorbidity. J Am Geriatr Soc 2012; 60: E1–E25.

87.

AGS Expert Panel on the Care of Older Adults with Multimorbidity. Multiple chronic conditions action steps. (under review).

88.

Kutner

Blatchford

Taylor

Jr et al . Safety and benefit of discontinuing statin therapy in the setting of advanced, life-limiting illness: a randomized clinical trial. JAMA Intern Med 2015; 175: 691–700.

89.

Garfinkel

Poly-de-prescribing to treat polypharmacy: efficacy and safety. Ther Adv Drug Safety 2018; 9: 25–43.

90.

Bayliss

Edwards

Steiner

et al . Processes of care desired by elderly patients with multimorbidities. Fam Pract 2008; 25: 287–293.