FIGURE 3.1 Survey responses to questions about routine medical exams, November 2005
There are scores of diagnostic tests—blood tests, x-rays, computed tomography (CT) scans, ultrasounds, magnetic resonance imaging (MRI)—to help the health care practitioner identify the cause of disease. It is important for practitioners to choose tests that not only improve their understanding of the disease but also affect treatment decisions. The decision to order a specific diagnostic test takes into account the test's reliability, validity, sensitivity, and specificity in addition to its risks to the patient and costs in terms of time and dollars.
The Reliability and Validity of Diagnostic Tests
Reliability of diagnostic testing refers to the test's ability to be repeated and to produce equivalent results in comparable circumstances. A reliable test is consistent and measures the same way each time it is used with the same patients in the same circumstances. For example, a well-calibrated balance scale is a reliable instrument for measuring body weight.
Validity is the accuracy of the diagnostic test. It is the degree to which the diagnostic test measures the disease, blood level, or other quality or characteristic it is intended to detect. A valid diagnostic test is one that can distinguish between those who have the disease from those who do not. There are two components of validity—sensitivity and specificity.
THE SENSITIVITY AND SPECIFICITY OF DIAGNOSTIC TESTS
Sensitivity refers to a test's ability to identify people who have the disease. Specificity, on the other hand, refers to a test's ability to identify people who do not have the disease. Ideally, diagnostic tests would be highly sensitive and highly specific, thereby accurately classifying all people tested as either positive or negative. In practice, however, sensitivity and specificity are frequently inversely related—most tests with high levels of sensitivity have low specificity, and the reverse is also true.
The likelihood that a test result will be incorrect can be gauged based on the sensitivity and specificity of the test. For example, if a test's sensitivity is 95%, then when one hundred patients with the disease are tested, ninety-five will test positive and five will test "false negative"—they have the disease but the test has failed to detect it.
However, if a test is 90% specific, when one hundred healthy, disease-free people are tested, ninety will receive negative test results and ten will be given "false-positive" results—they do not have the disease but the test has inaccurately classified them as positive.
The advantages of highly sensitive tests are that they produce few false-negative results, and people who test negative are almost certain to be truly negative. Highly sensitive tests may be useful as preliminary screening measures for diseases where early detection is vitally important, such as the enzyme-linked immunosorbent assay (ELISA) screening test for human immunodeficiency virus (HIV), the virus that produces acquired immune deficiency syndrome (AIDS).
In contrast, highly specific tests produce very few false-positive results and those who test positive are nearly certain to be positive. Highly specific tests are useful when confirming a diagnosis and in cases where the risks of treatment are high, such as the Western blot test to confirm the presence of HIV after it has been detected by the highly sensitive, but less specific, ELISA test.
Laboratory Tests
The editors of Harrison's Principles of Internal Medicine observe that the growing number and availability of laboratory tests has encouraged physicians and other health care practitioners to become increasingly reliant on them as diagnostic tools. Laboratory tests are easy, convenient screening measures because multiple tests may be performed on a single sample of blood and abnormal test results can provide valuable clues for diagnosis.
For screening purposes (to detect disease at its earliest stage, before it produces symptoms), the health care practitioner may order a variety of blood tests, including:
- Fasting blood sugar—This test is a screening and diagnostic test for diabetes; values consistently greater than 126 mg/dl indicate diabetes.
- Calcium—Blood levels of calcium can be elevated as a result of hyperactive parathyroid glands.
- Lipids—Elevated cholesterol, triglycerides, and low-density lipoproteins are associated with increased risk of heart disease.
- Thyroid stimulating hormone (TSH)—High levels of TSH indicate hypothyroidism (underactivity of the thyroid gland), and abnormally low levels indicate hyperthyroidism (overly active thyroid gland).
- VDRL (Venereal Disease Research Laboratory) or RPR (rapid plasma reagin)—These tests screen for syphilis, a sexually transmitted disease.
- HIV—It is important to screen for the presence of the virus that causes AIDS.
- PSA (prostate specific antigen)—This blood test is used to screen for prostate cancer and to monitor treatment of the disease.
- Stool occult blood (also called fecal occult blood test)—This tests for the presence of blood in the stool, which could be an indicator of colon cancer.
Diagnostic Imaging Techniques
Imaging studies are another form of diagnostic testing. In the past all diagnostic imaging studies were obtained using ionizing radiation (x-rays) and recorded on transparent film. Modern imaging studies such as ultrasound and magnetic resonance imaging (MRI) use nonionizing radiation and can be recorded digitally, viewed on computer monitors, sent via electronic mail, and stored on compact discs, digital tape, or transparent film. Most imaging studies are painless and pose little risk to patients apart from minimal exposure to radiation.
X-RAYS AND ULTRASOUND
The images produced by x-rays are the result of varying radiation absorption rates of different body tissues—the calcium in bone has the highest x-ray absorption, soft tissue such as fat absorbs less, and air absorbs the least. Chest x-rays, which offer images of the lungs, ribs, heart, and diaphragm, are among the most frequently ordered imaging studies.
To view tissues normally invisible on x-ray, contrast agents, such as barium and iodine, may be introduced into the body. For example, contrast agents often are used for imaging studies of the GI tract to diagnose digestive disorders.
Another common use of diagnostic x-rays is the measurement of bone density. Bone mass measurement (also called bone mineral density) is performed to evaluate the risk of bone fractures. Bone density usually is measured in the spine, hip, and/or wrist because these are the most common sites of fractures resulting from osteoporosis, a disease in which bones become weak, thin, fragile, and more likely to break.
Mammography also relies on x-ray technology to detect and pinpoint changes or abnormalities in the breast tissue that are too small to be felt by hand. Another imaging technique for breast examination is ultrasound, which can accurately distinguish solid tumors (lumps or masses) from fluid-filled cysts.
Ultrasound images are produced using the heat reflected from body tissues in response to high-frequency sound waves. Whereas x-ray is ideal for examining bone, ultrasound is used to examine soft tissue, such as the ovaries, uterus, breast, and prostate. It is not suitable for looking at bones, because calcium-containing tissues such as bone absorb, rather than reflect, sound waves.
COMPUTED TOMOGRAPHY, MAGNETIC RESONANCE IMAGING, AND POSITRON EMISSION TOMOGRAPHY
For conventional flat x-rays, the patient, x-ray source, and camera remain fixed and immobile. Computed tomography (CT) scans use a mobile x-ray source and generate a series of cross-section pictures, or slices, that are assembled by computer into images. Because CT distinguishes differences in soft tissue more effectively and with higher resolution than conventional x-rays, it often is used to examine internal organs in the abdomen, such as the liver, pancreas, spleen, kidneys, and adrenal gland, and the aorta and vena cava (large blood vessels that pass through the abdomen).
Magnetic resonance imaging (MRI) generates images based on interaction between a large magnet, radio waves, and hydrogen atoms in the body. Stimulated by ordinary radio waves within the powerful magnetic field, these atoms give off weak signals that a computer builds into images. MRI frequently is used to create images of the brain, spinal cord, heart, abdomen, bone marrow, and knee.
CT and MRI scans generate images of the body's structure (anatomy), whereas positron emission tomography (PET) scans offer insight into body function or processes (physiology). To create PET images, positron-emitting atoms are injected into the body, where they travel and strike other electrons, producing gamma rays. The gamma rays then are interpreted into images by a computer. Unlike CT and MRI, PET rarely is used for screening or diagnostic purposes. Instead, it is used to track the progress and treatment of patients with diagnosed diseases such as cancer.
Diagnostic Procedures
Other diagnostic tests commonly performed to screen for the presence of disease include the following:
- Throat culture—This test is used to determine whether streptococcus pyrogenes (commonly called strep) bacteria are the cause of a sore throat. To obtain a sample of the mucus in the throat, the health care practitioner swabs the back of the throat and places the swab in a tube. The swab is transferred into a culture in the laboratory where it is examined for bacterial growth. The results of this test are available in two to three days. A "rapid" strep test that produces results in minutes is also available.
- Urinalysis and urine culture—Chemical and microscopic examination of urine allow identification of infection, diabetes, and the presence of blood in the urine.
- Colonoscopy—Using a long tube fitted with a lens, the health care practitioner is able to look at the entire colon; identify and remove polyps; detect cancer; and diagnose other causes of blood in the stool, abdominal pain, and digestive disorders. To prepare for a colonoscopy, patients must empty their intestines completely before the examination.
- Flexible sigmoidoscopy—This test is similar to the colonoscopy in its use of a tube fitted with a camera to examine the colon. However, because the instrument is shorter than a colonoscope, it does not enable views of the entire colon. Through the flexible sigmoidoscope, the practitioner can examine only the sigmoid (lower portion) of the colon to detect polyps and cancers.
- Electrocardiogram—This test assesses the electrical function of the heart, detects abnormal heart rhythms, and aids in the diagnosis of myocardial infarction (heart attack) and other heart diseases.
Prenatal Diagnostic Testing
Ultrasound is used routinely to monitor the progress of pregnancy; evaluate the size, health, and position of the fetus; and detect some birth defects. Fetal ultrasound assists in the prediction of multiple births (more than one baby) and sometimes provides information about the gender of the unborn child.
Chorionic villus sampling (CVS) enables obstetricians and perinatologists (physicians specializing in evaluation and care of high-risk expectant mothers and babies) to assess the progress of pregnancy during the first trimester (the first three months). A physician passes a small, flexible tube called a catheter through the cervix to extract chorionic villi tissue—cells that will become the placenta and are genetically identical to the baby's cells. The cells are examined in the laboratory for indications of genetic disorders such as cystic fibrosis, Down syndrome, Tay-Sachs disease, and thalassemia. The results of the testing are available within seven to fourteen days. CVS provides the same diagnostic information as amniocentesis; however, the risks (miscarriage, infection, vaginal bleeding, birth defects) associated with CVS are slightly higher.
Amniocentesis involves analyzing a sample of fluid that surrounds the fetus in the uterus. The fluid is obtained when a physician inserts a hollow needle through the abdominal wall and the uterine wall. Like CVS, amniocentesis samples and analyzes cells derived
from the baby to enable parents to learn of chromosomal abnormalities and the gender of the unborn child. Results usually are available about two weeks after the test is performed.
Blood tests are also available to help diagnose fetal abnormalities. The enhanced alpha-fetoprotein test (also called a triple screen) measures levels of protein and hormones produced by the fetus and can identify some birth defects, such as Down syndrome and neural tube defects. Two of the most common neural tube defects are anencephaly (absence of the majority of the brain) and spina bifida (incomplete development of the back and spine). Test results are available within two to three days. Women with abnormal results often are advised to undergo additional diagnostic testing, such as CVS or amniocentesis.
DIAGNOSING MENTAL ILLNESS
Unlike physical health problems and medical conditions, there are no laboratory tests such as blood and urine analyses or x-rays to assist practitioners to definitively diagnose mental illnesses. Instead, practitioners generally rely on listening carefully to patients' complaints and observing their behavior to assess their moods, motivations, and thinking. Sometimes mental health disorders may accompany physical complaints or medical conditions. The presence of more than one disease or disorder is termed comorbidity.
Although there are varying opinions about the personality traits and characteristics that taken together constitute optimal mental health, historically it has been somewhat easier to define and identify mental illness—deviations from, or the absence of, mental health. Within the broad diagnosis of mental illness, there is more consensus about the origins, nature, and symptoms of mental disorders—serious, and often long-term conditions in which changes in cognition (thinking), behavior, or mood impair functioning—than exists about mental health problems—shorter term, less intense conditions that often resolve spontaneously, without treatment.
Because many mental health disorders are identified by primary-care physicians (general practitioners, family practitioners, internists, and pediatricians), the World Health Organization (WHO) developed educational materials and guidelines to assist practitioners in general medical settings—as opposed to psychiatric or other mental health settings—to assess and treat the mental health problems and disorders of patients in their care. The guidelines describe an assessment interview as a series of screening questions for which predominantly positive answers suggest the patient has an "identified mental disorder," or a "subthreshold disorder"—the patient responds positively to many questions but not enough to fulfill the diagnostic criteria for a disorder as defined by the WHO's tenth revision of the International Classification of Diseases (ICD-10), the European guide for diagnosis of mental disorders. In North America the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; Washington, DC: American Psychiatric Association, 2000) is used for the same purpose as the ICD-10. Practitioners are encouraged to ask open-ended questions that encourage patients to freely express their emotions, assure confidentiality, to acknowledge patients' responses, and to closely observe their body language and tones of voice.
Changing Criteria for Mental Illness
There are many controversies in mental health diagnosis, beginning with the definitions and classification of mental illnesses. Which criteria distinguish conditions as mental illness rather than normal variations in thinking and behavior? Should conditions such as attention deficit hyperactivity disorder (ADHD) be classified as learning problems or mental disorders? Should practitioners distinguish between neurological conditions that cause brain dysfunction and cognitive impairment such as Alzheimer's disease and mental illness involving brain dysfunction such as depression that may result from an imbalance of chemicals in the brain?
DSM-IV, as mentioned above, is the authoritative encyclopedia of diagnostic criteria for mental disorders. This definitive guide, which expands upon the ICD-10, is the most widely used psychiatric reference in the world and catalogs more than three hundred mental disorders. Table 3.1 lists the major classifications of mental disorders contained in the DSM-IV.
An examination of past versions of the DSM reveals that the definitions of mental illnesses have changed
TABLE 3.1
Major diagnostic classes of mental disorders (DSM-IV)
Disorders usually first diagnosed in infancy, childhood, or adolescence
Delerium, dementia, and amnestic and other cognitive disorders
Mental disorders due to a general medical condition
Substance-related disorders
Schizophrenia and other psychotic disorders
Mood disorders
Anxiety disorders
Somatoform disorders
Factitious disorders
Dissociative disorders
Sexual and gender identity disorders
Eating disorders
Sleep disorders
Impulse-control disorders
Adjustment disorders
Personality disorders
SOURCE: "Table 2-5. Major Diagnostic Classes of Mental Disorders (DSM-IV)," in Mental Health: A Report of the Surgeon General, U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration, with NIH, 1999, http://www.surgeongeneral.gov/library/mentalhealth/chapter2/sec2.html (accessed January 18, 2006)
dramatically from one edition to another. Persons diagnosed with a specific mental disorder based on diagnostic criteria in one edition might no longer be considered mentally ill according to the next edition. Critics of the DSM, which has expanded more than tenfold since its inception, claim that diseases are added arbitrarily by the American Psychiatric Association (APA) and that while some entries represent changing ideas about mental health and illness, others are politically motivated. For example, homosexuality was once considered a mental illness, but today, largely in response to changing societal attitudes, it is no longer termed an illness.
Skeptics also question the sharp increase in the number of diagnoses and the number of Americans receiving these diagnoses. Does the increasing number of diagnoses reflect rapid advances in mental health diagnostic techniques? Have mental health professionals simply improved their diagnostic skills? Are the stresses of twenty-first-century life precipitating an epidemic of mental illness in the United States? Or are mental health professionals—psychiatrists, psychologists, clinical social workers, marriage and family therapists, and other mental health practitioners—simply labeling more behaviors and aspects of everyday life as pathological (diseased)?
Further, there is dissent even within the mental health field about diagnosis that is rooted in the ongoing debate about the origins of mental illness. After taking into account all of the relevant medical research, Mental Health: A Report of the Surgeon General (1999, http://www.surgeongeneral.gov/library/mentalhealth/home.html) concluded that for most mental illnesses there is no demonstrable physiological cause. This means there is no laboratory test, imaging study (x-ray, magnetic resonance imaging, or positron emission tomography), or abnormality in brain tissue that has been definitively identified as causing mental illness. The majority of persons suffering from mental illness apparently have normal brains, and those with abnormal brain structure or function are diagnosed with neurological disorders rather than mental illnesses.
Finally, there are those who view mental illness as a social condition rather than one requiring medical diagnosis. They observe that even the Surgeon General's report, which favors biological explanations of the origin, diagnosis, and treatment of mental illness, concedes that mental health is poorly understood and defined differently across cultures. If mental health and illness are rooted in cultural mores and values, then they are likely socioeconomic and political in origin. The proponents of societal causes of mental illness contend that if mental illness is in part defined as functional impairment, and during the course of their lives perhaps half the U.S. population will be impaired (according to the National Institute of Mental Health [NIMH], 2006, http://www.nimh.nih.gov/publicat/numbers.cfm), then perhaps it is not the individual who is ailing, but the society. This theory is supported by the fact that the WHO, in their World Health Report (2001, http://www.who.int/whr/2001/en/), estimates that 25% of individuals will be diagnosed with mental illness in developed and developing countries, half of the proportion of the American population estimated to be at risk.
Despite the challenges of diagnosing metal illnesses, there is consensus that early diagnosis is vital because untreated psychiatric disorders can produce more frequent and more severe episodes, are more likely to become resistant to treatment, and may lead to the development of co-occurring mental illnesses. NIMH cites research revealing that while about 80% of all people in the United States with a mental disorder eventually do seek treatment, there are personal and public health consequences that result from long delays in seeking and receiving treatment. Untreated mental disorders are associated with school failure, teenage childbearing, unstable employment, marital instability, and violence ("Mental Illness Exacts Heavy Toll, Beginning in Youth," NIMH, June 6, 2005, http://www.nimh.nih.gov/press/mentalhealthstats.cfm).
Projective Testing Techniques
Most diagnoses of mental illness are made on the basis of symptoms reported by the patient, the practitioner's observations, and the use of designated guidelines or criteria for distinguishing between disorders and establishing diagnoses. There are, however, additional diagnostic tests practitioners may perform to confirm diagnoses. Projective tests are thought to provide insight into clients' unconscious minds and have been used to characterize and describe symptoms, as well as to detect physical abuse, sexual abuse, and child abuse. They include the Rorschach test, Thematic Apperception Test (TAT), human figure drawings, and the Washington University Sentence Completion Test. There is widespread agreement that projective tests should be just one component of a comprehensive diagnostic study and that results from the tests should be integrated with history and interview information, because test results should be weighted only when they are consistent with other data.
There has been harsh criticism of projective tests during recent years, with detractors detailing their shortcomings, concluding that they lack a scientific underpinning and produce exaggerated estimates of pathology. In an effort to address this controversy, Howard Garb and his colleagues analyzed the efficacy of a variety of projective tests and reported their findings in "Effective Use of Projective Techniques in Clinical Practice: Let the Data Help with Selection and Interpretation" (Professional Psychology: Research and Practice, vol. 33, no. 5, October 2002).
The investigators concluded that for making diagnoses, psychologists should rely primarily on interview and history information, but that results from psychological tests, including self-report personality inventories and projective techniques, may be helpful. They advised psychologists that they were likely to be "on safer ground when they use projective techniques as an aid for exploration in psychotherapy rather than as an assessment device."
Further, the investigators exhorted psychologists to rely heavily on history and interview data to predict behavior. They recommended that to evaluate psychiatric symptoms and personality traits, practitioners should depend on interview and history information, self-report personality inventories, and, in selected instances, projective tests. Although evaluation of symptoms and personality traits are ostensibly the ideal task for projective techniques, the investigators found that findings derived from Rorschach, TAT, and human figure drawings have not been independently and consistently replicated.
Automated and Online Diagnostic Testing
There are other diagnostic tests and case-finding instruments—tools practitioners may employ to screen for and identify persons suffering from mental illness. Jonathan Shedler and his colleagues from Harvard Medical School and the Clinical Research Unit of Kaiser Permanente in Colorado evaluated the utility and validity of Quick PsychoDiagnostics (QPD) Panel, an automated mental health test. They reported their findings in "Practical Mental Health Assessment in Primary Care: Validity and Utility of the Quick PsychoDiagnostics Panel" (Journal of Family Practice, vol. 49, no. 7, July 2000).
QPD was designed to meet the needs of primary-care physicians who do not have sufficient time to administer even brief diagnostic tests. The test combines features of an inventory and a structured interview and screens for nine frequently occurring psychiatric disorders and requires no physician time to administer or score. Patients respond to a core set of fifty-nine questions and, when responses suggest a possible psychiatric disorder, the test offers pointed questions that, like a structured interview, probe in depth. Although the test contains more than two hundred diagnostic questions, patients see and respond to a customized, relevant subset of them. Scoring is performed electronically.
The investigators evaluated validity by correlating QPD Panel scores to the Structured Clinical Interview for DSM-IV (SCID) and established mental health measures. They assessed utility, in terms of acceptability to physicians and patients, by administering satisfaction surveys to both groups. The researchers concluded that the QPD Panel is a valid mental health assessment tool with the capacity to diagnose a range of common psychiatric disorders. They deemed it practical for routine use in busy primary-care practices and observed that "routine screening would benefit the many patients who currently go undiagnosed and untreated."
Even if computerized diagnostic capabilities are imperfect and unproved, University of Pittsburgh Department of Psychiatry researcher Howard Garb suggested that computer programs will become more prominent in mental health practice in "Computers Will Become Increasingly Important for Psychological Assessment: Not That There's Anything Wrong with That!" (Psychological Assessment, vol. 12, no. 1, March 2000). Garb contended that computers would be widely used for psychological assessment because mental health professionals are not good at some judgment tasks and that the use of computers to make judgments might prevent problems associated with clinicians' judgments. He concluded, "Using computers to make judgments and decisions in personality assessment can lead to dramatically improved reliability, a decrease in the occurrence of biases, and an overall increase in validity and utility."
Tom Buchanan from the University of Westminster Department of Psychology described the strengths and weaknesses of Internet-mediated, or online, psychological assessment in "Online Assessment: Desirable or Dangerous?" (Professional Psychology: Research and Practice, vol. 33, no. 2, April 2002). Buchanan cited the strengths of online personality testing as allowing more people to complete them than would otherwise be possible. It enables persons who were previously unable to do so, because of distance or time constraints, to access mental health services. (Those who favor online therapy or counseling make the same arguments.) There also is the possibility that people may be more candid when completing tests online and willing to disclose more information about themselves to computers than to other people.
Weaknesses of Web-based assessment include computer anxiety that may affect participants' responses and the observation that online respondents tend to report higher levels of negative affect than those who complete conventional paper questionnaires. Mental health researchers wonder whether online respondents are simply more inclined to self-disclosure or whether they are actually a more depressed group. There are also professional and ethical concerns such as the likelihood of well-meaning but untrained individuals who offer tests and opinions about a range of psychological conditions. One concern is the consequences of providing already troubled individuals with information that might be potentially distressing. Without appropriate follow-up or counseling, delivering such sensitive and potentially emotionally charged information is ethically and professionally unacceptable.
Other potential problems center on the technology itself. Interruptions in connectivity may interfere with the assessment process. This is especially true when assessment is performed live during a videoconferencing session. Further, without the use of a secure server and encrypted communications, electronic communications may be intercepted by a third party. Buchanan concluded that "online clinical tests are both desirable and dangerous. There is clearly great potential, but a lot of work must be done before this potential is realized. Only time and extensive research can tell us whether these instruments will become a useful tool in behavioral telehealth contexts."
SECOND OPINIONS
Although many insurance plans will pay for second opinions, a 2004–05 Gallup Poll found that nearly half of Americans said they would never seek a second opinion when their physician "diagnoses a condition, prescribes a treatment, drug, or operation." Forty-one percent said they sometimes sought second opinions, and a scant 3% said they would always seek one. (See Figure 3.2.)
Women were more likely than men to assume responsibility for where their family members obtain medical care, and Gallup data as well as other research confirms that women visit physicians more often than do men. Rick Blizzard, Gallup's health and health care editor, opines that the frequency of physician visits affords women more opportunities than men to cultivate trusting relationships with their physicians. Still, women (47%) were more likely than men (42%) to say they always or sometimes seek a second opinion, and while 46% of women said they never seek a second opinion, more than half of men (52%) said they would never seek a second opinion. (See Figure 3.3.)
Educational attainment also is related to the propensity to seek a second opinion. Fifty percent of the survey respondents who were college graduates and 52% of respondents with postgraduate education said they
FIGURE 3.2 Public opinion on how often you get a second opinion from another doctor, 2004–05
Blizzard also observed that while it seems intuitively correct to assume that persons in poor health would be more likely to seek second opinions, Gallup Poll data revealed that respondents who rated their health as excellent or good (44%) were almost just as likely to seek second opinions as those who said their health was fair or poor (45%). (See Figure 3.4.)
FIGURE 3.3 Public opinion on how often you get a second opinion, by gender, age, and educational attainment, 2004–05
FIGURE 3.4 Public opinion on how often you get a second opinion, by health status, 2004–05
User Comments Add a comment…