74: Applying Evidence-Based Medicine to Older Patients

Ideally, the care of older patients should be grounded in the best available evidence about the benefits and harms of treatments. Unfortunately, high-quality evidence rarely exists. Good studies of a clinical condition often exclude older persons. Even when older persons are included, enrollment tends to be limited to healthy robust older persons who may little resemble the patient in front of you.

As a result, one can rarely practice true evidence-based medicine in older patients. Instead, one needs to examine available evidence and then critically assess the extent to which the evidence might apply to the patient being treated. To best make optimal clinical decisions, clinicians need to understand the limitations of applying the clinical research literature to their older patients, and thoughtfully apply existing evidence to the individual patient they are treating.

Clinicians are generally trained to consider outcomes from clinical trials as the gold standard for how to apply evidence-based medicine to clinical care. Ideally, clinical trials would include any patient who is a logical candidate for the therapy being examined. Unfortunately, most clinical trials include only idealized patients, excluding patients in whom the therapy has a greater risk of side effects, or patients who are at risk for not completing the trial. However, these nonidealized patients who are commonly excluded from clinical research are just the type of patients seen in geriatric practice, and in whom the therapies will often be used and marketed.

Zulman et al have described a framework that outlines the reasons older persons may be excluded from clinical trials. These reasons include explicit age exclusions, implicit age exclusions, and unintentional age exclusions.

Explicit Age Exclusions

Many studies have age-specific cutoffs in which all subjects above a defined age are denied enrollment. Although these exclusions are common, they can almost never be justified. Most studies with explicit age exclusions present absolutely no rationale to justify the exclusion.

An explicit age exclusion is only justifiable if one clearly would not offer a therapy in clinical practice to persons older than a particular age. In actual practice, most therapies tested in younger patients are eventually offered to older patients. Furthermore explicit age exclusions ignore the vast heterogeneity in health in older persons.

Implicit Age Exclusions

More often, the reasons for excluding older subjects are more subtle. Many studies without an age cutoff have exclusion criteria that differentially restrict the entry of older patients, particularly older patients who are more medically complex. Examples of implicit age exclusions include comorbidity, functional impairments, cognitive impairments, and inability to give informed consent.

Comorbidity

Treatment studies often focus on the effect of a specific treatment on health outcomes in persons with a specific condition. But most older people have multiple conditions. Yet many studies of specific conditions exclude patients with additional health conditions other than the condition being studied. For example, a study comparing the effectiveness of bronchodilator and anticholinergic inhalers for chronic obstructive pulmonary disease (COPD) excluded persons with chronic kidney disease or hospitalizations for congestive heart failure (CHF) in the past year. This is very problematic in older persons, as many older persons with COPD have coexisting chronic kidney disease or CHF. In actual practice, comorbidity is the norm, not the exception, and most treatments are still offered to older patients despite this comorbidity. It is difficult to judge the risk and benefits of treatments studied in idealized patients when applied to real patients with complex comorbidity.

Functional Impairment

Many studies exclude patients with “poor performance status” or poor functional status. Although there are many simple tools available to define functional status, often studies do not include a definition of poor functional status.

Many older patients have functional limitations and, in practice, these are rarely viewed as contraindications for most therapies. Often, the goal of therapy for older adults may be to prevent further loss of function. However, functional problems, such as falls, can markedly alter the risk and benefits of therapies. The failure to account for functional impairments in most studies makes it difficult to gauge how a patient’s functional impairments should affect the decision to offer a treatment.

Cognitive Impairment and Inability to Consent

Studies often exclude subjects who have cognitive impairment or are unable to provide informed consent. Often, the studies fail to describe how cognitive impairment or consent capacity was assessed.

Cognitive impairment is extremely common in older patients and older adults with cognitive impairment are usually offered the same therapies as patients without cognitive impairment. Because most studies fail to even describe the cognitive status of their subjects, it is often impossible to know how cognitive impairment might impact the risks and benefits of a treatment. Additionally, in clinical practice, family members are often asked to consent to treatments when the patient is unable to fully understand the risks and benefits. While methods exist for enrolling these patients in studies using similar approaches for surrogate consent, surrogate consent is often not attempted.

Nursing Home Patients

Virtually all treatments offered in the community are offered to persons in nursing homes. It is extraordinarily rare to find a study of any therapy that even considers nursing home patients as candidates for enrollment. As a result, there is an extraordinary lack of evidence to guide most therapeutic decisions in nursing home patients.

Unintentional Exclusions

Even when studies have few exclusion criteria, study processes may unintentionally exclude older patients who are potential targets for therapy. For example, many studies have complex procedures for follow-up. Often follow-up procedures require patients to report to a study center for examinations and blood draws. But many older patients no longer drive, making participation difficult. This type of subtle exclusion is important because the same factors that make study follow-up difficult may make the type of monitoring and followup that is needed difficult in the less mobile older patient being considered for treatment. Also, while sensory problems, such as vision and hearing impairment, may not be an explicit study exclusion, sensory impairment may make study enrollment and follow-up difficult for older patients. For example, older adults with hearing impairment may be excluded from studies that screen enrollees over the phone or require answers to phone-based health status questionnaires. Many studies do not budget the extra cost needed for in-person interviews for older participants who are hard of hearing.

Applying evidence from clinical research to an individual patient can be difficult. However, addressing a series of questions can help a clinician make the best possible inference about how existing evidence might inform the best clinical decision in their patient. In addition to addressing whether patients in existing studies are similar to your patient, one should also think about what outcomes matter to your patient and whether limited life expectancy might affect the risks and benefits of the treatment in question.

How Do the Characteristics of My Patient Differ from Those Patients in the Research Studies?

It is particularly useful to consider the types of aging patients who are excluded in clinical studies. For a relatively healthy older patient, there may be no reason to believe that a treatment that is indicated in a younger patient will not also be useful in the older patient. However, this conclusion might change as one considers the comorbid burden, degree of functional impairment, and cognitive problems in a frailer older patient. In a patient with many comorbid conditions, it is possible that side effects of a treatment may exacerbate one of the comorbid conditions. In a patient with functional impairment, falls may be a particular concern. A medicine that causes slight problems with balance may not be a big concern in a younger patient, but could precipitate a complicated fall in an older patient.

Similarly, if a medicine such as warfarin requires close titration and monitoring to minimize possible complications, the presence of cognitive impairment can markedly impact the risks/benefit titration.

What Outcomes Matter to My Patient?

Many clinical trials focus on outcomes such as mortality, or disease-specific outcomes such as cardiovascular hospitalizations. Even though these outcomes are particularly relevant to older patients, other outcomes, such as preserving mobility, preventing falls, or quality of life, may matter more to many older patients. Because these outcomes are often not included in many studies, clinicians must often make their best inference about how a treatment impacts these outcomes.

Because quality of life is best judged by the patient, a patient’s own sense of how a treatment is impacting him or her can be a crucial criterion in assessing the risks and benefits of therapy. This can be especially important for a patient taking multiple medicines. Some patients may not be bothered by taking many medicines, while for others, the addition of another medicine may be sufficiently burdensome and distressing to outweigh a small benefit seen in clinical studies.

How Does My Patient’s Life Expectancy Impact the Risks and Benefits of Treatment?

Some treatments confer immediate risks to patients, whereas the benefits accrue over time. Older patients with limited life expectancy may be subject to all of the risks of treatment with few benefits. This concern becomes increasingly important with increasing age, as well as in patients with limited life expectancy as a result of major comorbidity, or severe functional or cognitive impairment. These concerns are addressed in greater depth in Chapter 8, “Prevention & Health Promotion.”

Overall, the exclusion of older participants in clinical research can affect the ease with which a clinician applies the evidence base to the clinician’s patient. Knowing the important ways that one’s patient is similar or different from participants in clinical research, as well as the quality of life factors that are most important to one’s patient, should guide all evidence-based medical decisions in geriatric care.