Is 'routine' laboratory testing a thing of the past?

Current recommendations regarding screening

Gary H. Tabas, MD; Marian S. Vanek, MBA

VOL 105 / NO 3 / MARCH 1999 / POSTGRADUATE MEDICINE

CME learning objectives

• To understand why biochemical profiles do not contribute to improved health in asymptomatic patients
• To become familiar with recent recommendations for use of individual screening tests
• To learn the rationale behind current recommendations for preoperative testing

This page is best viewed with a browser that supports tables

Preview: A 55-year-old man comes to you for a routine physical examination. He is a nonsmoker who takes no medications and has no signs of acute or chronic disease, and he has not seen a doctor in years. What blood work should you order for this patient? The authors of this article help you answer this question in light of recent advances in technology, restrictions in healthcare reimbursement, and increased sophistication in cost-benefit analysis for laboratory testing.

Laboratory screening has become a hot topic in the last few years because of reimbursement restrictions imposed by managed care as well as a growing understanding of risk-benefit analysis. When the technology became available, biochemical profiles became popular among physicians, hospitals, and patients alike because of their perceived benefit in detecting disease. They became widely used without the benefit of cost-effectiveness or outcomes analysis. However, with the publication of guidelines by the American College of Physicians (1) and the US Preventive Services Task Force (2), awareness has increased regarding the "burden of suffering" for specific diseases and the effectiveness of various screening tests. This new knowledge has resulted in greater sophistication regarding screening.

This article summarizes recent recommendations regarding the use of biochemical profiles ("chem screens") in asymptomatic adults and discusses laboratory testing as part of routine physical examinations, preoperative testing, and recent recommendations regarding use of specific tests. Guidelines for screening in pregnant women and patients with existing conditions (eg, testing for microalbuminuria in diabetic patients) are not addressed.

Principles of screening

Screening is an attempt to detect disease in asymptomatic persons. For screening to be effective, consideration must be given to the characteristics of the population being screened, the disease targeted for detection, and the specific test used (table 1). The population must be accessible (eg, patients presenting to physicians' offices, nursing home residents). The disease must have a high enough prevalence or a great enough impact to be of concern and a more favorable outcome when treated early, in the asymptomatic phase, rather than after symptoms have appeared (eg, as in breast cancer but not pancreatic cancer). Finally, the screening test must be sensitive enough to detect the target disease but also specific enough to limit false-positive results, which would lead to unnecessary follow-up testing or treatment. In addition, the test should not be so expensive nor the technique so cumbersome that its use would be limited.

Table 1. Criteria for effective laboratory screening

Characteristics of the population being screened Sufficiently high disease prevalence Accessible (eg, patients presenting to physicians' offices, nursing home residents) Likely to comply with subsequent diagnostic tests and necessary therapy Characteristics of the disease Significant effect on the quality or length of life Prevalence or burden of suffering sufficiently high to justify cost of screening Acceptable methods of treatment available Asymptomatic period during which detection and treatment significantly reduce morbidity, mortality, or both Treatment in the asymptomatic phase yields a better outcome than treatment delayed until symptoms appear Characteristics of the test Sufficiently sensitive to detect disease during asymptomatic period Sufficiently specific to provide acceptable predictive value Acceptable to patients

The availability of a new test does not, in itself, mean that the test should be used. Rather, before widespread use, the test should be scrutinized using the aforementioned criteria to determine its usefulness. The same criteria should be used to determine the appropriateness of any test or panel of tests being considered for an asymptomatic patient.

Biochemical profiles

Other than hypercholesterolemia, the diseases targeted by biochemical panels have very low prevalence (0.001% to 5%) (1). One study (3) found that with a panel of 23 tests, new diagnoses were made in less than 2% of patients tested and changes in management occurred in only up to 0.6%. The exception was cholesterol testing, which led to new diagnoses in 7.6% of patients tested and changes in management in 4.4%.

Nevertheless, biochemical panels have been widely used because with older technologies, it was less expensive to run a whole panel of tests than to perform individual tests. With the current generation of analyzers, however, it is often less expensive to order individual tests. For instance, at our institution, the cost of a panel that includes tests for glucose, serum urea nitrogen, creatinine, and electrolytes is $5.44, whereas the cost of a serum urea nitrogen test alone is $0.30. An estimated 13% of the $3.45 billion spent annually on laboratory testing in the United States pays for biochemical profiles, half of which contain 19 or more separate components (4).

Large panels (>12 tests per panel) have been eliminated by Medicare and have been replaced by smaller, disease- or organspecific panels (table 2). As of April 1, 1998, Medicare no longer pays for most previously common chemistry profiles. At the direction of the Health Care Financing Administration (HCFA), laboratories now must either bill Medicare separately for each test or use one of the new HCFA-approved panels. These panels are not appropriate for screening in asymptomatic patients but may be warranted in those with specific diseases for which multiple tests are indicated.

Table 2. New biochemical panels approved by the Health Care Financing Administration

Comprehensive metabolic panel (80054) Albumin Alkaline phosphatase Aspartate aminotransferase Bilirubin, total Calcium Chloride Creatinine Glucose Potassium Protein, total Serum urea nitrogen Sodium Basic metabolic panel (80049) Carbon dioxide Chloride Creatinine Glucose Potassium Serum urea nitrogen Sodium Hepatic function panel (80058) Alanine aminotransferase Albumin Alkaline phosphatase Aspartate aminotransferase Bilirubin, total and direct Electrolyte panel (80051) Carbon dioxide Chloride Potassium Sodium

With the elimination of many test panels (previously represented by CPT codes 80002 through 80019) and the introduction of the four new panels, billable outpatient tests have decreased by 15% to 25%. If fully implemented, these new Medicare requirements would save $60 million annually (4). The ordering physician is responsible for knowing which tests are included in the panels and avoiding subsequent duplicate testing. Medical necessity requirements must be followed when ordering these panels, but documentation is not required for every test in the panel. The HCFA considers a panel medically necessary if there is reasonable justification for most of the tests in the panel (5).

The concept behind the HCFA's medical necessity requirement is to control and ultimately reduce test ordering and subsequent reimbursement. Although the HCFA introduced a mandate for medical necessity in 1987 with a statement in the Social Security Act (section 8162[a][1] states that "the [Medicare] program will only pay for disease treatment and diagnosis"), serious enforcement did not occur till the mid-1990s (5). Documentation of medical necessity by the ordering physician may be in the form of a narrative diagnosis, but the testing laboratory must then convert this to an ICD-9 code (6). Many facilities do not allow their staff to assign CPT codes because Medicare may consider it an illegal kickback for a laboratory (which cannot determine medical necessity) to bill for a test on a sample that does not have a diagnosis code (5). Recent HCFA rulings state that the Department of Health and Human Services is authorized to "impose an exclusion or a monetary penalty" if the "practitioner or other person violated the statutory obligations to render medically necessary and appropriate care or failed to provide evidence of medical necessity and quality" (5).

Individual tests

Cost-benefit analysis and new healthcare financing regulations also necessitate careful scrutiny of the use of individual tests for screening purposes. The following discussion focuses on current recommendations for use of common laboratory tests.

Cholesterol screening
Routine measurement of serum cholesterol levels is recommended for men 35 to 65 years of age and women 45 to 65. For patients over age 65, decisions should be made on a case-by-case basis, but unfortunately, Medicare does not cover lipid tests ordered for screening purposes. Multiple studies have shown the benefit of cholesterol reduction therapy on cardiac end points (a 2% to 3% reduction in coronary artery disease end points for each 1% decrease in low-density lipoprotein [LDL] cholesterol level or 1% increase in high-density lipoprotein [HDL] cholesterol level) and, more recently, on overall mortality. The National Cholesterol Education Panel recommends screening with random total cholesterol and HDL cholesterol tests. Many physicians prefer to screen with a fasting lipid profile, but HCFA has eliminated the lipid panel, CPT code 80061. Such tests must now be ordered individually and require medical necessity documentation.

Cholesterol tests that require the patient to fast necessitate a separate office visit (although many such patients also need a fasting serum glucose test) but allow for accurate determination of LDL cholesterol and triglyceride levels. The benefit of screening for hypertriglyceridemia to improve outcomes in coronary ar-tery disease is unproved, but a high triglyceride level may be a marker for previously undiagnosed diabetes. Elevated levels of plasma homocysteine have been found to be an independent risk factor for coronary artery disease, but no randomized trials have assessed the benefit of mass screening and subsequent treatment on morbidity and mortality (7).

Glucose measurement
The American Diabetes Association (8) recommends screening with a fasting serum glucose test for patients with the following risk factors for diabetes:

• Family history of diabetes
• Obesity
• African American, Hispanic American, Native American, Asian American, or Pacific Islander race
• Age over 44 years
• Previous impaired glucose tolerance
• Hypertension
• HDL cholesterol level less than 36 mg/dL
• Triglyceride level greater than 249 mg/dL
• Physical inactivity
• Gestational diabetes or delivery of an infant with a birth weight over 4 kg (9 lb)

A fasting serum glucose level of higher than 126 mg/dL is considered abnormal (8).

In the past, the US Preventive Services Task Force and the American College of Physicians have stated that there is insufficient evidence to recommend for or against routine glucose screening in nonpregnant adults but that physicians may decide to screen those at high risk because of the "potential, but unproved, benefits of early intervention." (2) However, evidence to support such screening is growing. A recent decision model used data from the Diabetes Control and Complications Trial and other trials that estimated rates of progression of microvascular disease in patients with type 2 diabetes (9). The investigators concluded that substantially greater benefit would be realized from treating relatively younger patients with high hemoglobin A1c levels than patients with late-onset diabetes. More recently, the UK Prospective Diabetes Study Group (10) found that intensive blood glucose control decreased the risk of microvascular complications, particularly retinopathy.

Creatinine testing
Some authorities recommend measurement of serum creatinine levels to screen for renal dysfunction because of increasing evidence that dietary protein restriction and use of angiotensin-converting enzyme (ACE) inhibitors can retard progression once renal insufficiency develops. A meta-analysis of trials evaluating the effect of a low-protein diet on progression of renal disease (11) found a relative risk of 0.67 (95% confidence interval [CI], 0.50 to 0.89) for renal failure or death in nondiabetic patients with elevated creatinine levels. Similar results were seen in trials involving patients with type 1 diabetes, although the criteria in these trials were less stringent than those used for nondiabetic patients. However, the sensitivity of serum creatinine testing is not high, and a low-protein diet is difficult for many patients to follow. More studies are needed before a strong recommendation can be made for mass screening.

Another meta-analysis (12) assessed the effect of ACE inhibitors on the progression of nondiabetic renal disease. The investigators found a relative risk of 0.7 (95% CI, 0.51 to 0.97) for end-stage renal disease and no significant effect on mortality. They concluded that ACE inhibitors are more effective than other xantihypertensive agents in reducing the risk of end-stage renal disease but that the effect may be due to the greater decline in blood pressure achieved rather than other specific attributes of ACE inhibitors.

Calcium screening
Serum calcium screening tests are usually not recommended because of the low prevalence of target diseases. For example, the annual incidence of primary hyperparathyroidism, which has been decreasing for reasons that are unclear, is estimated to be 4 cases per 100,000 persons (13). Moreover, the benefit of early therapy for this disease is unproved.

Liver enzyme tests
Routine testing for liver injury is not indicated. Although screening with an alkaline phosphatase test would lead to detection of more cases of asymptomatic Paget's disease, treatment is usually reserved for symptomatic patients (2).

Thyroid hormone tests
Routine screening with thyroid hormone tests is usually not recommended in asymptomatic patients. However, thyrotropin testing should be considered in patients with symptoms suggestive of thyroid disease and in those in whom symptoms may be overlooked (eg, the elderly, patients with Down syndrome) (2).

Complete blood cell count
Routine ordering of complete blood cell counts is not indicated in asymptomatic adults, but individual blood tests are warranted in some patients. For example, a hemoglobin or hematocrit is recommended in pregnant women, high-risk infants and, perhaps, recent immigrants and the institutionalized elderly. However, these tests are not useful in asymptomatic patients because of the low prevalence of target conditions, limited sensitivity and specificity for iron deficiency, lack of proven benefit from early detection, adverse effects of some treatments, and cost. Similarly, a leukocyte count is not useful for screening because the test is unreliable and physiologic variation is common. Abnormal leukocyte counts are nonspecific, the prevalence of target conditions is very low, and early therapy does not improve outcomes.

A recent study (14) recommended screening with a transferrin saturation test and, possibly, serum ferritin measurement in middle-aged patients to detect hemochromatosis in early stages. The prevalence of gross iron overload was shown to be 0.2% to 0.7% in the various groups studied, and rates were estimated to be as high as 1.8% on the basis of elevated transferrin saturation levels. (Standard textbooks have reported prevalence of 0.16% to 0.5% for the homozygous state.) The study population was limited to western Germany, and some of the subjects were not truly asymptomatic. Nevertheless, cirrhosis resulting from undiagnosed hemochromatosis is a serious complication, and treatment with phlebotomy in the early stages is effective and relatively uncomplicated.

Prostate-specific antigen test
Before undergoing screening with a prostate-specific antigen test, patients should be informed of the possible benefits as well as the risks (eg, patient anxiety over false-positives or inconsequential positives, adverse effects of subsequent testing and treatment). The benefits of this test are still unproved, and further study is needed.

Rubella screening
Women of child-bearing age should undergo screening for rubella unless there is documentation of previous vaccination or routine vaccination is planned (2).

Tests for osteoporosis
Various markers in serum (eg, osteocalcin) and urine (eg, N-telopeptide of type I collagen) appear to be linked to increased bone resorption in postmenopausal women (15,16). However, there have been no large, randomized trials to determine whether treatment given on the basis of abnormal marker levels results in decreased morbidity.

Preadmission testing

Biochemical profiles are not recommended for elective hospital admissions, because of the low yield of new diagnoses (4%) and new management (0.5%) (1). A complete blood cell count is not useful unless infection or a hematologic disorder is suspected. In addition, determination of prothrombin time and partial thromboplastin time is not indicated because of the very low prevalence of target diseases.

Preoperative testing

A complete blood cell count is not necessary before minor surgical procedures in which minimal blood loss is expected. Some investigators recommend obtaining a hemoglobin measurement or hematocrit in all presurgical patients; our hospital's guidelines include testing for women over age 40 and men over age 50.

Prothrombin time and partial thromboplastin time have poor predictive value, and measurement is not indicated in patients who have no clinical indicators of bleeding risk (eg, personal or family history of prolonged bleeding, liver disease, malnutrition, or malabsorption). In low-risk patients, postoperative bleeding is no more likely in those with abnormal test results than in those with normal results (1). Platelet counts below 50,000/mm (3) are associated with postoperative bleeding, but thrombocytopenia to this degree is rare and should be detectable from the patient history and physical examination. Biochemical profiles are not indicated for the reasons discussed previously.

Preoperative liver enzyme tests are indicated only in cases in which liver enzyme levels may be expected to be abnormal postoperatively. Routine glucose testing is not recommended. However, diabetes has been shown to be a risk with vascular or coronary artery bypass surgery, and screening should be considered before these procedures (17). Creatinine testing is not needed routinely, but because renal impairment is an operative risk, screening should be considered for patients receiving drugs excreted by the kidneys, in whom postoperative renal failure would be expected, and for patients over age 50.

Malpractice concerns and physician responsibilities

Physicians have cited concerns about malpractice as a reason for ordering some laboratory tests. However, when standards of care that incorporate rational testing guidelines are set by physicians and hospitals, individual patients are at limited risk (18). Furthermore, the risk of malpractice suits may, in fact, increase with unnecessary routine testing: Studies have shown that many of these test results are not seen, recorded, or acted on, particularly in the preoperative setting (19).

To minimize malpractice risk regarding laboratory testing, physicians should ensure that all tests that are ordered are indeed performed. It is also important to follow up on all abnormal laboratory test results. A list of abnormal results, or "panic values," can be made available to the office staff so that they can immediately contact the physician about results that fall outside predetermined normal ranges. Also, staff members should document all phone calls to patients as well as management decisions based on test results.

Conclusion

A summary of current laboratory screening recommendations is given in table 3. In response to technological advances, changes in healthcare reimbursement, and application of cost-benefit analysis, physicians should re-examine the practice of routine testing in asymptomatic patients. Although some tests, most notably serum cholesterol measurement, are important for detection of diseases in which early treatment decreases morbidity and mortality, the vast majority of routine testing fails to improve health, often leads to increased costs and adverse effects, and is a misuse of time for both patients and physicians.

Table 3. Summary of current laboratory screening recommendations

Office screening in asymptomatic patients Random serum cholesterol and high-density lipoprotein cholesterol, or fasting lipid profile Fasting serum glucose in patients at risk for diabetes* Rubella test in women of child-bearing age, if no previous vaccination documented and no routine vaccination planned ?Serum creatinine ?Thyrotropin (TSH) (in patients with possible symptoms of thyroid disease, patients with Down syndrome) ?Hemoglobin or hematocrit (in pregnant women, high-risk infants, recent immigrants, the elderly) ?Transferrin saturation Screening on hospital admission No routine screening; test as appropriate for admission diagnoses and comorbid diseases only Preoperative screening Biochemical profiles not indicated Complete blood cell count, or hemoglobin and hematocrit if substantial blood loss expected during surgical procedure; consider for patients over age 40 and those with hepatic or renal disease or suspected anemia White blood cell count if leukocytosis or leukopenia suspected Prothrombin time, partial thromboplastin time, and platelet count if bleeding disorder suspected from patient history and physical examination Liver enzymes in patients undergoing surgical procedures likely to affect hepatic function Serum creatinine in patients undergoing surgical procedures associated with possible postoperative renal failure, those receiving drugs excreted by kidneys, and patients over age 50 Serum glucose in patients at risk for diabetes and those undergoing vascular or coronary artery bypass surgery *According to American Diabetes Association guidelines.

References

1. Sox HC Jr, ed. Common diagnostic tests. 2d ed. Philadelphia: American College of Physicians, 1990
2. Guide to clinical preventive services: report of the US Preventive Services Task Force. 2d ed. Baltimore: Williams & Wilkins, 1996
3. Ruttimann S, Dreifuss M, Clemencon D, et al. Multiple biochemical blood testing as a case-finding tool in ambulatory medical patients. Am J Med 1993;94(2):141-8
4. Profile trak: laboratory medicine review. Lab Industry Rep 1996;5(6)
5. Painter PC. The laboratory, physicians, and medical necessity. Clin Lab News 1997 Sept:10-3
6. Young C. Patient protection from liability: advance beneficiary notices (ABNs), Part One. Vantage Point 1998;2(9):1-6
7. Singh H. Selections from current literature: homocysteine: a modifiable risk factor for cardiovascular disease. Fam Pract 1997;14(4):335-9
8. American Diabetes Association. Screening for diabetes. Diabetes Care 1997;20(Suppl 1):S22-3
9. Vijan S, Hofer TP, Hayward RA. Estimated benefits of glycemic control in microvascular complications in type 2 diabetes. Ann Intern Med 1997;127(9):788-95
10. UKPDS Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998;352:837-53
11. Pedrini MT, Levey AS, Lau J, et al. The effect of dietary protein restriction on the progression of diabetic and nondiabetic renal diseases: a meta-analysis. Ann Intern Med 1996;124(7):627-32
12. Giatras I, Lau J, Levey AS, for the Angiotensin-Converting-Enzyme Inhibition and Progressive Renal Disease Study Group. Effect of angiotensin-converting enzyme inhibitors on the progression of nondiabetic renal disease: a meta-analysis of randomized trials. Ann Intern Med 1997;127(5):337-45
13. Wermers RA, Khosla S, Atkinson EJ, et al. The rise and fall of primary hyperparathyroidism: a population-based study in Rochester, Minnesota, 1965-1992. Ann Intern Med 1997;126(6):433-40
14. Niederau C, Niederau CM, Lange S, et al. Screening for hemochromatosis and iron deficiency in employees and primary care patients in Western Germany. Ann Intern Med 1998;128(5):337-45
15. Garnero P, Delmas PD. New developments in biochemical markers for osteoporosis. Calcif Tissue Int 1996;59(Suppl 1):S2-9
16. Gorai I, Taguchi Y, Chaki O, et al. Specific changes of urinary excretion of cross-linked N-telopeptides of type I collagen in pre- and postmenopausal women: correlation with other markers of bone turnover. Calcif Tissue Int 1997;60(4):317-22
17. Macpherson DS. Preoperative laboratory testing: should any tests be "routine" before surgery? Med Clin North Am 1993;77(2):289-308
18. Lubin MF. Preoperative testing. In: Lubin MF, Walker HK, Smith RB, eds. Medical management of the surgical patient. 3d ed. Philadelphia: Lippincott, 1995:35
19. Carson JL, Eisenberg JM. The preoperative screening examination. In: Goldmann DR, Brown FH, Guarnieri DM, eds. Perioperative medicine: the medical care of the surgical patient. 2d ed. New York: McGraw-Hill, Health Professions Division, 1994:15-23

Dr Tabas is clinical associate professor of medicine, University of Pittsburgh School of Medicine, and director, transitional year residency, UPMC Shadyside, Pittsburgh. Ms Vanek is administrative director, laboratory services, UPMC Shadyside. Correspondence: Gary H. Tabas, MD, UPMC Shadyside, Department of Medicine, 307 School of Nursing, 5230 Centre Ave, Pittsburgh, PA 15232. E-mail: garyt@ssh.edu.

Please send technical questions related to the Web site to Ann Harste