Clinical decision making, including diagnosis, treatment, and desired outcomes, differs between younger and older adult patients. The primary goal of medical care in younger adult patients usually is diagnosis of the disease causing the presenting symptoms, signs, and/or laboratory abnormalities. Treatment is targeted toward the pathophysiologic mechanisms deemed responsible for the disease. Relevant clinical outcomes are determined by the specific diseases and include cure if the disease is acute, and control or symptom modification if the disease is chronic.

The conventional disease-specific approach is not optimal in older patients for several reasons. First, age-related physiologic changes in most organ systems affect diagnostic test interpretation and response to treatments and may be difficult to differentiate from disease. In addition to age-related physiologic changes, the average 75-year-old suffers from 3.5 chronic diseases. With multiple coexisting chronic diseases, there is a less consistent relationship between pathology and disease or between disease and clinical manifestations. One disease may obscure or change the pathology, manifestations, or accuracy of laboratory evaluation of coexisting diseases. Treatment of one disease may increase the severity of another. With multiple coexisting diseases, it becomes difficult, and often impossible, to assess the severity or manifestations of individual diseases and to ascribe health and/or functional status to specific disease processes.

Second, many distressing symptoms or impairments among older persons, such as pain, dizziness, fatigue, sleep problems, sensory impairments, and gait disorders cannot be ascribed to a single disease; instead, they result from the accumulated effect of physical, psychological, social, environmental, and other factors. A clinical focus solely on diagnosing and treating discrete diseases may lead to expensive diagnostic testing with inconclusive results, to unnecessary, or even harmful, interventions, or, conversely, to ignoring potentially remediable symptoms. While clinicians may be reluctant to treat symptoms in younger and middle-aged patients without a specific diagnosis, treatment focused on improving symptoms in multiply ill and impaired older patients is often appropriate, because comfort and function are primary goals of health care in this population.

Third, diagnostic test characteristics may be altered by age and comorbidity, making selection and interpretation of tests more complicated than for younger patients. Furthermore, both the benefits and harms of treatment regimens may differ in the face of age-related physiologic changes and coexisting health conditions.

Fourth, older patients vary in the importance they place on potential health outcomes. When asked, older persons are able to prioritize among the often competing goals of increased survival, comfort, cognitive function, and physical function. Optimal clinical decision making in the care of older patients includes the articulation of patient preferences or goals of care; the identification of the diseases, impairments, and nondisease-specific factors affecting the attainment of these preferences and goals; and the selection of treatment options based on the modifiable impediments to individual patient goals. The multiplicity of impairments and diseases; the contribution of psychological, social, and environmental factors to health conditions; the enhanced likelihood of harm as well as benefit from many interventions; and the interindividual variability in patient preference all combine to make clinical decision making in the care of older persons very complex.

Fifth, clinical decision making is further complicated in older patients because other persons, including the spouse, adult children, other relatives, and significant others, are often actively involved, particularly when cognitive impairment is present. Involvement of family and friends is helpful and often crucial, since they may provide additional sources of information, facilitate adherence to treatment recommendations, and offer both emotional and instrumental support. Conflicts may arise, however, when goals of the patient and family differ. Striking a balance between patient confidentiality and family involvement, between independence and support, and between patient and family goals is a constant challenge. When based on an understanding of these factors, however, the clinical care of older persons is both effective and immensely gratifying.

At least three factors affect clinical presentation in older persons: underreporting of symptoms and impairments, changes in the patterns of presentation of individual illnesses, and an altered spectrum of health conditions. Contrary to a popular perception of older persons as complainers, they tend to underreport significant symptoms. One reason for underreporting is that both older persons and their clinicians often dismiss treatable symptoms and impairments as age-related changes for which nothing can be done. Denial, resulting from fear of economic, social, or functional consequences, is suggested as another reason for underreporting health conditions. Cognitive impairment and depressive symptoms may further limit the ability or desire of some older persons to report symptoms and health conditions. This tendency to underreport means that clinicians must actively inquire about symptoms and concerns.

Altered presentation is a second characteristic of illness in older persons. While both acute and chronic illnesses may, and often do, present with “classic” signs and symptoms, age-related changes and coexisting conditions may combine to obscure these classic presentations in older persons. Symptoms or signs of one condition may exacerbate or mask those of another condition, complicating clinical evaluation. For example, arthritis, if it limits physical activity, may mask the presence of severe cardiovascular disease. Manifestations of clinically important disease may be attenuated in older persons, particularly those who are frail. Chest pain may be absent in older persons presenting with myocardial infarction, as may shortness of breath in persons with congestive heart failure. Another common phenomenon is that symptoms in one organ system may reflect disease in another system. Pneumonia may present as confusion or anorexia; a urinary tract infection may present with behavioral or functional changes. A corollary of these altered presentations is that signs and symptoms are often nonspecific. That is, while suggesting that the older persons is experiencing an acute illness or an exacerbation of a chronic condition, the signs and symptoms may offer limited help in determining what the illness or condition might be. These altered presentations mean that the clinician must be particularly diligent in ascertaining all symptoms and signs. She must rely on combinations of findings from the history, physical examination, and ancillary testing to determine the diagnosis, or, as is often the case, to identify the treatable contributors to the illness or health condition.

Third, the spectrum of health conditions in older persons differs from younger patients. Important clinical entities include not only acute and chronic diseases but also geriatric syndromes as well as cognitive and physical disabilities. Geriatric syndromes are health conditions common in older persons that result from the accumulated effect of multiple predisposing factors and that may be precipitated by an acute insult. Examples described in other chapters include delirium, falls, and incontinence. Geriatric syndromes and disabilities are relevant because they may be the presenting manifestation of another underlying illness and because they are treatable causes of morbidity in their own right.

Reason for Performing a Test or Evaluation

The process of establishing a diagnosis is complex for any patient. The goal may be to make a diagnosis in an ill patient, to generate a differential diagnosis for observed signs and symptoms, or to make an early diagnosis of presymptomatic disease among well individuals. Making diagnostic decisions is even more challenging in the older person because many have multiple conditions which can mask, mimic, or increase the symptoms of other diseases. Older persons may present atypically, and with nonspecific signs, symptoms and syndromes such as confusion or falls. Furthermore, contrary to the traditional clinical evaluation that is aimed at identifying the presence of discrete diseases, the aim of the clinical encounter in older patients maybe to identify the impairments, diseases, and other factors impeding the attainment of individual patient preferences and goals, as discussed below.

A clinical evaluation grounded in a clear understanding of these issues can be effective. These factors, however, dictate changes in the conventional clinical encounter in both the content and method of ascertainment for the history, the physical examination, and ancillary testing. It is important for clinicians to understand how to assess, interpret, and use the results of an evaluation.

Before deciding whether to pursue a diagnostic test, clinicians should take into account several issues related to the patient and to the characteristics of the test (Table 10-1). Clinicians should consider what they are going to do with the information obtained from the test and whether its consequences will help patients achieve their goals of care. For example, will the test be used to establish a diagnosis for which there is effective treatment? Is the patient willing to accept the treatment? Are there comorbid conditions or contraindications present that would preclude the patient from receiving the therapy? The diagnostic test is likely unnecessary if the patient is not interested in the therapy or if the therapy would result in significant risk of harm. The diagnostic test could also be done to assist with establishing a prognosis—but does the patient want this information? Will the diagnostic test lead to labeling, which can have disastrous consequences? Diagnostic tests can also provide assistance in monitoring therapy once initiated. Finally, is the test being done for academic curiosity or simply because it is available?

The older population may have multiple target conditions but it may not be necessary or clinically useful to pursue all of these with diagnostic tests. An alternative strategy to consider in the elderly with multiple illnesses and symptoms is to focus on symptom modification.

Issues in Assessing and Interpreting Evaluation Results

Definition of Normal in Older Persons

If a diagnostic test is to be performed, then several issues in assessing and interpreting results need to be considered. Results of interview, examination, and ancillary tests typically are reported as “normal” or “abnormal.” What is normal in the older person? There are at least six definitions of normal (Table 10-2) but this chapter will focus on the fifth definition, which refers to a range of results beyond which the target disorder becomes highly probable. The sixth definition is also worthy of consideration and includes the range of results beyond which treatment does more good than harm. This last definition is particularly important in the older population. Consider, for example, how the definition of normal blood pressure has changed over the past few decades as evidence accumulates that treatment of progressively less pronounced elevations in blood pressure does more good than harm.

Assessing Diagnostic Test Accuracy

Prior to performing a diagnostic test, clinicians should review and understand its accuracy in older patients with a spectrum of comorbidities. When attempting to determine the accuracy of a diagnostic test, three questions must be considered: (1) is the evidence about the accuracy of a diagnostic test valid? (2) Is there evidence that this test can accurately distinguish patients who do and do not have a specific disorder? and (3) How can this valid, accurate diagnostic test be applied to a specific patient? In assessing validity, the three key components to consider are: (1) independent, blind comparison with a reference standard of diagnosis (e.g., auscultation for a heart murmur should occur independent of results of the echocardiogram; pathological interpretation of a biopsy should be done without knowledge of laboratory tests that precipitated the biopsy); (2) performance of the reference standard regardless of the diagnostic result (e.g., either performance of pulmonary angiography or long-term follow-up in the absence of anticoagulation therapy to determine negative test accuracy for suspected pulmonary embolism); and (3) evaluation of the diagnostic test in a spectrum of patients similar to those in whom it will be used in clinical practice. Methods for assessing validity using these components have been described in several excellent reviews. Of particular relevance is whether the diagnostic test was evaluated in older patients with a spectrum of comorbidities similar to those in whom you plan to apply the test.

Ability of Tests to Distinguish Patients WHO Do and Do Not Have the Disease or Condition

Deciding whether a diagnostic test is important requires consideration of its ability to change the probability of disease prior to test completion (called the pretest probability of the target disorder) to a probability of the disease after test completion (called the posttest probability). Diagnostic tests that produce large changes from pretest to posttest probabilities are important and likely to be useful in clinical practice.

Consider, for example, a 76-year-old woman who is admitted to the hospital with community-acquired pneumonia. She responds nicely to appropriate antibiotics but her hemoglobin remains at 100g/L with a mean cell volume of 80. Her peripheral blood smear shows hypochromia. She is otherwise well and is on no incriminating medications. Her family physician found that her hemoglobin was 105 g/L 6 months prior to admission. She has never been investigated for anemia. A ferritin was ordered and the result is 60 mmoL/L. Does this patient have iron-deficiency anemia?

In a systematic review and meta-analysis of the accuracy of ferritin for diagnosing iron-deficiency anemia, displayed in Table 10-3, the prevalence of iron-deficiency anemia was 31%. The posttest probability, also known as the positive predictive value, of iron-deficiency anemia among patients with a serum ferritin <65 mmoL/L was 73%. Conversely, for a serum ferritin ≥65 mmoL/L, the posttest probability of iron-deficiency anemia is 5%, meaning that the probability of not having iron-deficiency anemia after a negative test result (serum ferritin ≥65 mmoL/L) is 95%, which is known as the negative predictive value. Thus, the probability of iron-deficiency anemia has shifted from 31% to either 73% or 5% depending on the test result.