Chapter S12: The Clinical Laboratory in Modern Health Care

Modern medicine relies extensively on the clinical laboratory as a key component of health care. It is commonly held that at least 60–70% of all critical clinical decisions rely to some extent on a laboratory result. For many diseases, the clinical laboratory provides essential diagnostic information. As an example, histopathologic analysis provides basic information about histologic type and classification of tumors and their degree of invasion into adjacent tissues. Microbiologic testing is required to identify infectious organisms and determine antibiotic susceptibility. For many common diseases, expert groups have produced standard guidelines for diagnosis that rely on defined clinical laboratory values, e.g., blood glucose or hemoglobin A_1C levels form the basis for diagnosis of diabetes mellitus; the presence of specific serum antibodies is required for diagnosis of many rheumatologic diseases; and serum levels of cardiac markers are a mainstay in diagnosis of acute coronary syndromes.

With their ever-increasing number and scope, clinical laboratory tests provide the clinician with a powerful set of tools but pose challenges in terms of appropriate selection and judicious, cost-effective use to deliver effective patient care.

DIAGNOSIS OF DISEASE

One of the most frequent reasons for performing clinical laboratory tests is to support, confirm, or refute a diagnosis of disease that is suspected based on other information sources, such as history, physical examination findings, and imaging studies. The following questions need to be considered: Which clinical laboratory tests may be of value in supporting, confirming, or excluding the clinical impression? What is the most efficient test-ordering strategy? Will a positive test result confirm the clinical impression or even definitively establish the diagnosis? Will a negative result disprove the clinical suspicion, and, if so, what further testing or alternative approach will be needed? What are the known sources of false-positive and false-negative results, and how are these misleading results recognized?

SCREENING FOR DISEASE

Another reason for ordering clinical laboratory tests is to screen for disease in asymptomatic individuals (Chap. 4). Perhaps the most common examples of this application are the newborn screening programs now routinely used in most developed countries. Their purpose is to identify newborns with treatable conditions for which early intervention—even before clinical symptoms develop—is known to be beneficial. In adults, screening tests for diabetes mellitus, renal disease, prostate cancer (measurement of serum prostate-specific antigen [PSA] levels), and colorectal cancer (testing for occult blood in stool), for example, are widely applied to apparently healthy individuals because early diagnosis and intervention lead to improved long-term outcomes.

Differences between Screening Tests and Confirmatory Tests

It is important to distinguish between clinical laboratory tests that can be used to screen for disease and those that offer a confirmatory result. Screening tests are generally less expensive and more widely available than are confirmatory tests, which often require more specialized equipment or personnel. As a general principle, screening tests are designed to identify all individuals who have the disease of interest, even if, in the process, some healthy individuals are misidentified as possibly having the disease. In other words, maximization of the diagnostic sensitivity of screening tests inevitably comes at the expense of reduced diagnostic specificity. Confirmatory testing is intended to correctly separate those individuals who have a disease from those who do not.

These principles are exemplified by screening for hepatitis C virus (HCV) infection. A common approach is to test first for the presence of antibodies to HCV in serum. A positive result generally indicates either a current infection or a previous infection that the patient’s immune system has successfully cleared. In the latter situation, antibodies may persist at detectable levels for life. However, a small proportion of patients have false-positive results in the serologic screening test for HCV. To resolve uncertainty in these instances, a positive serologic screening test should be followed by confirmatory identification of HCV RNA with molecular techniques. This confirmatory testing can provide evidence of current viral infection and can identify patients who are not infected.

RISK ASSESSMENT OF FUTURE DISEASE

Another reason for clinical laboratory testing is to assess a patient’s risk of developing disease in the future. Many diseases are associated with well-established clinical laboratory–defined risk factors that, if present, indicate the need for more frequent monitoring for the disease in question. The need for risk assessment is even clearer if there are useful interventions that decrease the risk of developing disease. For example, hypercholesterolemia is a well-established risk factor for coronary artery disease that may be modified by pharmacologic intervention. Many genetic mutations are known to be associated with increased risk of cancer. For example, hereditary mutations in the BRCA1 and BRCA2 genes predispose to breast and/or ovarian cancer. Individuals who are known to carry these mutations require more vigilant monitoring for early signs of cancer and may even opt for prophylactic surgery in an effort to prevent cancer (Chap. 457). Individuals with factor V Leiden are at increased risk of developing deep-venous thrombosis and may benefit from prophylactic anticoagulation, especially in settings that pose additional risk of venous thromboembolism, for example, in the perioperative period.

MONITORING DISEASE AND THERAPY

Many clinical laboratory tests offer useful information on the progress of disease and the response to therapy. One example is the measurement of viral load in HIV-1-infected patients who are taking antiretroviral agents. A key goal of treatment is a reduction in the viral load to below the level of laboratory detection, which is typically in the range of 40–50 copies/mL. Other examples of the use of clinical laboratory testing for monitoring disease include measurement of tumor markers such as PSA, especially following surgical removal of tumors. In this situation, the expectation is that successful treatment of a tumor will cause a decrease in the level of the tumor marker. A later increase in the level of the tumor marker suggests a recurrence of the disease. Finally, the clinical laboratory offers direct monitoring of levels of some therapeutic agents, such as drugs. This monitoring is important if a drug has a defined therapeutic concentration range above which it is toxic and below which it is ineffective. Monitoring of drug levels in this situation facilitates optimal dosing and avoidance of toxicity.

It is common practice for clinical laboratories to establish a list of “critical values.” These are values of test results that indicate immediate risk to the health or life of the patient and therefore require urgent communication with the patient’s physician so that appropriate medical intervention may be initiated. Critical values are reported regardless of whether the test was ordered as a “stat” or a routine test. The critical values themselves are generally created by the clinical laboratory medical director in conjunction with the medical staff. A representative list of critical values is shown in Table S12-1.

Tests that are ordered on a “stat” basis receive priority in the clinical laboratory’s testing queue; testing of specimens from other patients may be delayed while a stat specimen is analyzed. Ordering a test stat should be reserved for situations in which a result is needed for urgent medical care—a judgment that must be made by the ordering physician. Stat testing should not be used merely for the convenience of either the patient or the health care provider.

The commonly used metrics of a clinical laboratory test are the diagnostic sensitivity, specificity, and positive and negative predictive values. These concepts are discussed in Chap. 3. In the clinical laboratory, the terms sensitivity and specificity have alternative meanings that are applied to tests, and the different uses of these terms may cause confusion. Analytic sensitivity can refer to the lowest detectable concentration of analyte that can be measured with some defined certainty or to the rate of change of signal intensity as analyte concentration changes. For example, newer generations of laboratory assays frequently exhibit improved sensitivity over that of earlier generations, i.e., they can detect lower concentrations of the analyte—a feature that is often of value in disease diagnosis. Analytic specificity refers to the extent to which other substances in the test system interfere with measurement of the analyte of interest. This concept is frequently applied to immunoassays, in which a detection antibody may also bind with compounds that have a structure similar to the substance sought. For instance, immunoassays for drugs may show cross-reactivity with drug metabolites, and immunoassays for glucocorticoids may show cross-reactivity with other glucocorticoids of similar structure. Certain chemical assays are also subject to nonspecificity. For example, the Jaffé reaction, a chemical method commonly used to measure creatinine, is subject to positive interference from several other compounds, including glucose, certain ketones, and cephalosporin antibiotics. Elevated concentrations of bilirubin, free hemoglobin, or turbidity in plasma or serum specimens may also interfere with some assays. The clinical laboratory should be able to provide advice about the presence or magnitude of these effects in the assays it performs.

Clinical laboratory diagnosis, like all diagnosis, is based on observation of disease-related changes from normality.

1. Tissue injury or necrosis allows leakage of intracellular components into the circulation, with consequent detectable rises in blood levels of these components. Many intracellular molecules are common across tissue types and are therefore not indicative of injury to a specific tissue. Other constituents are selectively expressed in relatively high concentrations—or are even uniquely present—in certain tissues. Therefore, their presence in the blood is evidence of injury to that tissue. This principle forms the rationale for measurement of blood levels of, for example, liver enzymes in evaluating liver disease (Chap. 330), cardiac troponins in acute coronary syndromes (Chap. 269), and myoglobulin in muscle injury. The extent of the rise in blood levels of these markers generally correlates with the extent of tissue damage, although there are exceptions; for example, liver enzyme levels may fall in end-stage liver disease.

2. An increase in blood levels of some analytes indicates failure of normal excretory processes. This principle is illustrated by elevations in conjugated bilirubin that accompany obstruction of the biliary system, by elevations in ammonia in advanced or metabolic liver disease, by rises in creatinine and potassium levels in renal failure, and by increases in PCO₂ in some pulmonary diseases.

3. Increases in the blood concentration of tissue-specific markers may result from expansion of the total volume of that tissue. This principle forms the basis for the measurement of levels of many tumor markers such as PSA (prostate cancer), CA-125 (ovarian cancer), CEA (colon cancer), and CA-19–9 (pancreatic cancer). In practice, the usefulness of these markers varies with the degree to which they are produced by a tumor and by the tumor’s size. Small colon cancers, for example, may not produce a significant rise in CEA levels, whereas small prostate cancers often produce detectable rises in PSA concentrations.

4. Disease processes often manifest characteristic patterns of coincident changes in levels of several analytes. These patterns of change can be understood by consideration of the underlying pathophysiology. For example, acute intravascular hemolysis is characterized by a fall in levels of hemoglobin and haptoglobin and by a rise in unconjugated bilirubin. In endocrine diseases, there are often changes in concentrations of several hormones because of disturbance of feedback loops. Primary hyperthyroidism, as an example, is characterized by increases in thyroxine and by suppression of thyroid-stimulating hormone. In diabetic ketoacidosis caused by insulin deficiency, there are concomitant elevations of plasma glucose, ketones, and (frequently) potassium. In response to metabolic acidosis, levels of bicarbonate are typically reduced.

5. Genetic changes underlie many diseases, both inherited and acquired. In the era of molecular medicine, there is increasing recognition of the contribution of hereditary factors to many common diseases. Often, the epidemiology of common diseases such as hypertension is characterized by a minority of families that have mutations in recognized genes, whereas the genetic basis of the same disease phenotype in the larger population is unclear. The search for the genetic factors that contribute to many common diseases remains a topic of intense research interest. It is now clear that essentially all tumors have genetic abnormalities. Although there is an inherited predisposition in some families, most of these genetic changes are acquired. Identification of the genetic abnormalities in cancer offers new tools for clinical laboratory diagnosis and classification of tumors in ways that surpass traditional histopathology and also provides insights into cellular processes that may be targets for treatment.

6. Clinical laboratory results should always be interpreted in the context of the patient’s history and physical examination as well as any other relevant information (e.g., imaging studies). The clinician should be circumspect about treating laboratory results rather than the patient.

7. Recommended clinical laboratory tests change with time. As new markers of disease emerge, they may replace older markers. For example, measurement of serum creatine kinase (CK) levels was introduced for diagnosis of acute myocardial infarction in the 1980s. Use of the cardiac-specific isoenzyme CK-MB later became widespread in clinical practice. Today, cardiac troponins are replacing CK (or CK-MB) measurements in recommended guidelines. Many other assays have fallen out of use as better assays have become available. Measurement of urine 17-ketosteroids (arising from androgens) and of urine 17-hydroxycorticosteroids (arising from glucocorticoids) has been supplanted by immunoassays or mass spectroscopy determinations of specific steroid hormones. Today, many steroid hormones are measured by mass spectrometry, often with better analytic specificity than is provided by immunoassays. As new tests are introduced, it is essential that they be evaluated critically before adoption for clinical use. At a minimum, consideration needs to be given to questions of clinical validation, specimen stability, diagnostic sensitivity and specificity, positive and negative predictive values, analytic accuracy and precision, and relative costs.

In the interpretation of clinical laboratory results, comparison is usually made to a reference range (sometimes called a normal range) that defines the values seen in health or considered to be desirable for health. Several common methods are used to describe reference ranges in the clinical laboratory.

1. For many quantitative clinical laboratory tests, the range of observed values in a healthy population shows an approximately Gaussian distribution. The factors that contribute to this range include the inter- and intraindividual variation in the concentration of the analyte and the analytic imprecision. When there is an approximately Gaussian distribution of values in the population, the reference range is commonly defined as being the central 95% of the range of distribution of those values. According to this method, 2.5% of the population will have a measured value that is below the reference range for the analyte, and 2.5% will have a value that is above the reference range. The fact that 5% of healthy individuals will have a test value that is outside the reference range has important implications when multiple tests are ordered. If N independent tests are performed on a specimen, then the probability that at least one result will be outside the reference range is (1–0.95^N). The greater the number of tests ordered (even for a healthy individual), the greater is the likelihood of an abnormal result (Fig. S12-1). If 20 independent tests are performed on a healthy subject, the probability of at least one abnormal result is almost two-thirds.

   In some settings, more stringent criteria are used to define abnormality. For example, current American Heart Association guidelines recommend the use of a serum level of cardiac troponins that is greater than the 99th percentile of values found in a healthy population as evidence of acute myocardial infarction.

2. An alternative approach to using population means and standard deviations is to define a range of analyte values that is judged to be consistent with health on the basis of expert consensus opinion. These ranges are often referred to as decision limits. Examples of reference ranges established in this way include those for total, high-density, and low-density cholesterol (Table S12-2). Such ranges may deviate considerably from those that would be established if the analyte concentrations of the population (mean ± 2 standard deviations) were used as a basis for establishing the reference range. For example, the desirable total cholesterol value as defined by the National Cholesterol Education Program is <200 mg/dL. This value is very close to the mean concentration found in U.S. adults, and so less than half of U.S. adults have a desirable cholesterol level, and 31 million have cholesterol levels >240 mg/dL, which is considered to be undesirable. If the central 95% of cholesterol concentrations in the population were taken as the reference range, the upper end of that range would be well beyond what is considered desirable.

   Reference ranges may vary with age, gender, ethnic background, and physiologic state (e.g., pregnancy, high-altitude adaptation). The existence of different reference ranges poses challenges for interpretation of results. In particular, creatinine stands out as an analyte for which conventional reference ranges are not always easy to apply in clinical practice. Plasma levels of creatinine vary with age, gender, and ethnic group. This fact makes it difficult in practice to use a simple reference range for this analyte when attempting to gauge a patient’s renal function. A large decrease in glomerular filtration rate (GFR) is associated with slight increases in the plasma creatinine concentration within the typical reference range provided by many laboratories (Fig. S12-2). A 60-year-old white woman with a serum creatinine level of 1.00 mg/dL, which is well within the typical reference range, has an estimated GFR of only 57 mL/min per 1.73 m², whereas the same creatinine concentration in a 20-year-old African-American male is consistent with normal renal function. To better estimate the GFR, which is widely considered to be the most useful index of overall renal function, it has become customary to use equations that incorporate plasma creatinine with other parameters. The most widely used of these equations in current practice are the 4-parameter Modification of Diet in Renal Disease (MDRD) equation and the CKD-EPI equation. Both equations incorporate plasma creatinine, age, gender, and ethnic group (African American or not African American). Recommended clinical laboratory practice is to report the estimated GFR with all creatinine measurements in adults.

Errors can arise at all stages of the testing process, from specimen collection to result interpretation. An error arising at any stage may adversely affect patient care. In clinical laboratory practice, it is customary to divide the testing process into three phases: preanalytic, analytic, and postanalytic. Examples of each type of error are shown in Table S12-3. The most common error in the testing process is specimen mislabeling, in which a specimen from one patient is placed in a container labeled with another patient’s name or identifiers. Specimen mislabeling errors may have very serious consequences for a patient. For example, if erroneous typing of a patient’s blood group results from specimen mislabeling and is followed by transfusion of a mismatched unit of blood, the outcome may be fatal. A mislabeled biopsy specimen can lead either to an erroneous diagnosis and inappropriate therapy or to a failure to make a diagnosis and institute appropriate therapy.

In addition to errors, many preanalytic factors can influence clinical laboratory results. Posture (i.e., recumbent versus upright), exercise, diet, recently ingested food, and use of prescribed or recreational drugs (including tobacco, alcohol, caffeine, and herbal supplements) can influence a variety of analyte concentrations. After blood has been collected, certain analytes undergo changes in their concentration during storage or transportation. Glucose levels fall as a result of red cell metabolism. Ammonia levels rise as a result of protein breakdown. Increasing permeability and breakdown of red cell membranes leads to increases in plasma potassium and free hemoglobin levels. Bacterial contamination can lead to overgrowth of specimens. To minimize these precollection alterations, specimens should be processed or transported to the clinical laboratory as soon as possible after collection. The list of known preanalytic variables and their effects is extensive, and the reader is referred to Young’s compendium on this subject (see Further Reading).

The great majority of tests continue to be performed in dedicated clinical laboratory facilities, but for several decades there has been a trend toward point-of-care testing. This change has been made possible by the development of portable analytic devices, including single-purpose instruments such as glucometers and oxygen saturation monitors, and multifunction instruments that can perform a wider variety of analyses, particularly in chemistry and hematology but also in some areas of microbiology. The use of these devices is driven largely by the convenience of faster result availability. In some settings (e.g., in rural areas and developing countries), there may be no easily accessible clinical laboratory and a point-of-care device may be the best or only option for testing. However, the per-specimen cost of point-of-care testing, in terms both of reagents and supplies and of personnel, is often greater than that of centralized testing. Other concerns relate to the adequacy of personnel training for point-of-care testing, the quality of the results, and the incorporation of results into the medical record.

One of the largest markets for point-of-care testing is home testing by patients, which has long been an important element in the management of persons with diabetes who monitor their own blood glucose levels. Over-the-counter kits for home pregnancy testing have been available for decades. More recently, kits have become available for home testing of the international normalized ratio or prothrombin time by patients taking oral anticoagulants. Kits are also available for cholesterol monitoring, fecal occult-blood detection, and hemoglobin measurement. In these areas, there is often little information on the quality of test performance, the accuracy of the results, or the correctness of result interpretation and resulting action. Genetic tests (e.g., 23andMe) are also being marketed to provide information to individuals about their genetic backgrounds, genetic traits, and potential risk for a limited number of diseases, or to assess telomere length (e.g., Teloyears). At this stage, such testing has little clinical implication but patients are likely to come to their clinicians with questions about how to interpret the results.

The principles of genetic medicine in clinical practice are discussed in Chaps. 456, 457. Here we will concentrate on issues related to clinical laboratory testing for genetic disease.

The distinction between genetic testing for inherited disorders and that for acquired disorders affects the type of tissue that should be obtained for analysis. In inherited disorders, all nucleated cells are expected to carry the inherited mutation; thus white blood cells or buccal cells (obtained by scraping the inside of the cheek) are convenient sources of DNA for clinical laboratory testing. For prenatal testing of the fetus, chorionic villi or amniocytes are commonly used. Fetal DNA is also present in maternal blood, which can provide a convenient source of material for genetic testing of the fetus. In tests for acquired genetic disorders (e.g., cancer), the tissue of interest that contains a suspected mutation is sampled. It is often useful to compare tumor DNA with the patient’s normal DNA in order to identify acquired mutations (e.g., testing for microsatellite instability in colorectal cancer; Chap. 67).

Although it is assumed that all clinical laboratory testing is performed with the consent of the patient (or, in the case of minors, their parents), regulations may require formal written consent for genetic testing. Such regulations vary among jurisdictions, and the practicing clinician should be aware of local regulations. In some jurisdictions, there are regulations on the storage and use of genetic information and on the maximal period for which genetic specimens may be stored.

For some late-onset genetic diseases, such as Huntington’s disease (Chap. 427), genetic testing allows a prediction about the future development of the disease. The degree of certainty that is possible on the basis of this testing surpasses that associated with identification of more traditional disease risk factors (e.g., hyperlipidemia as a risk factor for future myocardial infarction). When deciding to undertake predictive genetic testing, it is important for the patient to consider the broad implications of a positive or negative test result, to be made aware of any support and counseling that is available, and to understand the implications of a result for other family members. In dealing with these issues, genetic counselors play an important role (Chap. 457). Their expertise includes the ability to explain genetic disorders at an understandable level to patients and their families, to arrange for support services, and to provide genetic risk assessments to members of families with genetic disorders.

When testing for genetic disorders, the clinical laboratory will use different analytic approaches according to the disease of interest. Some disorders, such as sickle cell anemia, are caused by single-point mutations. Testing for these disorders involves mere assessment for one or a few mutations in a single gene. Other disorders (e.g., hyperphenylalaninemias) may be caused by numerous mutations in a single gene. Still others (e.g., hereditary breast cancer) may be caused by mutations in many genes. The number of possible mutations and genes that underlie a clinical phenotype affects the cost for clinical laboratory testing as well as the likelihood of finding a disease-causing mutation.

If a disease phenotype can be caused by many mutations, a clinical laboratory result that is negative should be interpreted with care. For example, it is common to screen healthy pregnant women (and their partners) for mutations in the CFTR gene, which is mutated in patients with cystic fibrosis (CF). The goal of this screening is to identify women who are carriers of a CFTR mutation and therefore are at increased risk of having a baby with CF. Because CF is an autosomal recessive disorder, a fetus has a 1:4 chance of being affected if both parents are carriers of disease-causing CFTR mutations. The screening test approach that is commonly used to identify mutations in carriers detects 80–85% of all known disease-causing CFTR mutations in Caucasians and up to 97% of mutations among Ashkenazi Jews. A negative screening result therefore does not completely eliminate the possibility that a woman (or her partner) actually has a mutation. What can be inferred from a negative test result is that the risk of having a CF-affected baby has decreased significantly to an extent that depends on the woman’s ethnic group and the mutations that were examined. The clinical laboratory may calculate and report the woman’s new risk of being a carrier if the screening result is negative.

The increasing availability of large-scale (next-generation) sequencing of a patient’s whole genome or exome will greatly affect genetic testing over the next decade, with implications for the number of mutations that can be detected and the increased complexity of result interpretation.

Genetic testing has limitations that are often unique to this field. Results may be inconclusive. For example, a search for mutations in a gene that is suspected of causing a disease may fail to reveal any known disease-causing mutations. A mutation may be discovered that is of unknown clinical significance. In this situation, consideration of the predicted change in the amino acid sequence of the encoded protein may suggest a biologic effect—e.g., replacement of a charged amino acid by one of the opposite charge or by a neutral amino acid; replacement of an amino acid by one of a different size; or replacement of an amino acid that is conserved across multiple species. Further information may be obtained by determining whether the mutation is found in healthy individuals. Even with all of these considerations, it is not uncommon that the biologic significance of an identified mutation remains uncertain, and further research may be needed to assess its significance.

It is also important to understand the limitations of the clinical laboratory approach used to detect mutations. At this time, next-generation sequencing is becoming more widely adopted for both hereditary and acquired genetic mutations. The large amounts of data generated in this approach has the benefit of increasing the chance of finding pathogenic mutations, but also of finding mutations and variants of uncertain significance.

Another unique aspect of genetic testing is the concern that genetic information about individuals may be used to discriminate against them by employers or by insurance companies. In the United States, the Genetic Information Nondiscrimination Act of 2008 (GINA) prohibits the use of genetic information by employers in making decisions related to employment and by health insurance companies in issuing insurance policies or setting premium rates based on knowledge of the applicant’s genetic status. GINA does not cover disability insurance, long-term care insurance, or life insurance policies.

Although public attention has been most closely focused on DNA testing, other clinical laboratory investigations that are not usually thought of as genetic may provide important genetic information about the person being tested. For example, serum protein electrophoresis may reveal α-1 antitrypsin deficiency. Depending on the clinical laboratory technology used, measurement of hemoglobin A_1c, commonly used for monitoring diabetes control, may reveal a hemoglobin variant such as HbS (sickle cell). Measurement of cholesterol and triglyceride levels may reveal any of a number of hereditary disorders. All of these results are types of genetic information.

In the United States, all clinical laboratory testing performed for clinical purposes (but not for research purposes) is regulated by the federal Clinical Laboratory Improvement Amendments Act of 1988 (CLIA). Home monitoring by patients who are testing their own specimens is not covered by CLIA. The statute and the regulations, which are administered by the Centers for Medicare and Medicaid Services, apply to all laboratories, whether located in a physician’s office, a large hospital, or a reference laboratory; and all laboratories are required to hold a valid CLIA certificate that is appropriate for the highest complexity level of the tests they perform. The U.S. Food and Drug Administration is responsible for assigning the complexity level of commercial tests. The lowest category of complexity is the “waived” category, which is followed (in order of increasing complexity) by the categories of “provider-performed microscopy,” “moderate-complexity testing,” and “high-complexity testing.” The category of provider-performed microscopy is used to cover tests such as the use of potassium hydroxide preparations on skin scrapings to examine for fungi, fern tests, and sperm motility tests; it does not encompass histopathology that falls into the high-complexity category. Even if a clinical laboratory is performing only testing in the “waived” category, it must still hold a valid CLIA certificate. Laboratories that hold certificates for nonwaived tests are required to participate in proficiency testing and are regularly inspected to monitor their performance.

The spectrum of diseases, particularly of infectious diseases, varies widely around the world, which influences the type of laboratory testing and personnel skills needed to support local clinical practice. However, clinical laboratory operations, including quality management schemes, share many common characteristics in all developed countries, and are supplied by a limited number of major equipment manufacturers that distribute similar or identical products in many parts of the world. This offers a certain consistency for laboratory testing.

Clinical laboratories are complex entities that depend on an economy that can support the training and hiring of personnel, the availability of equipment and supplies, a reliable infrastructure (e.g., a stable electricity supply), and a robust system to ensure quality of operations. Unfortunately, in developing countries, the resources available may be insufficient, with scarcities of clinical laboratories and the infrastructure needed to support them, trained staff, and pathologists. Such challenges remain among the most important for improvement of global health.