Health Screening
Routine laboratory tests are often ordered regardless of findings elicited from the medical history or physical examination.Ordering batteries of tests nonselectively is costly and counterproductive. Ordering tests that are not warranted by history and physical examination may also increase patient risk. Patients can be harmed directly from follow-up of false positive or borderline positive results. They can be harmed indirectly when false positive tests divert a physician’s attention from more important patient care issues or unnecessarily result in postponement of a therapeutic procedure.
Reference ranges are chosen for laboratory tests to include 95% of healthy individuals. Therefore, 5% of test results from patients without disease will be outside the hospital reference range. Patients whose results exceed the upper 2.5% of healthy individuals or fall below the lower 2.5% of healthy individuals are flagged as abnormal. For example, if hemoglobins were performed on100 healthy individuals, 5 would be expected to have abnormal results.
Ordering several tests further increases the chances of a healthy person having at least one abnormal result. If two tests are ordered on a healthy individual, the chances of both being normal are 0.95 X 0.95 = 0.90. The following table illustrates that the chance of one test being abnormal increases with the number of tests ordered.
|
Number of Tests Ordered |
Probability of at Least One Abnormal Test |
|
1 |
5% |
|
2 |
10% |
|
5 |
23% |
|
10 |
40% |
|
15 |
54% |
|
20 |
64% |
The yield of new diagnoses is small in routine screening programs. Serum glucose may be abnormally elevated in 5-8% of adults screened, but the identification of new cases of diabetes mellitus is less than 1%. Likewise, serum cholesterol may be >200 mg/dL in over half of a screened population, but new diagnoses of hypercholesterolemia leading to treatment are reported to be approximately 5%. Occult liver disease is detected in <1% of adults screened. Other chemistry tests have even lower diagnostic and therapeutic yields.
In today’s era of cost efficiency, it is better practice to perform only those tests that provide more benefit for the patient than risk. Suggested screening tests include: hemoglobin, cholesterol, glucose, creatinine, alanine aminotransferase (ALT), and sensitive TSH.
References
1. Cebul RD, Beck JR. Biochemical Profiles: Applications in Ambulatory Screening and Preadmission Testing of Adults. Annals Internal Medicine 1987; 196:403-13.
2. Ruttimann S, et al. Multiple Biochemical Blood Testing as a Case-Finding Tool in Ambulatory Medical Patients. American J Medicine 1993; 94:141-148.
