July 1999
---The
number of millimeters of induration, not erythema, should be measured
using a ruler and a ball point pen.
DANA POINT, Calif. - There are many children who should not receive tuberculin skin tests, and the challenge to physicians performing these tests is determining who should and should not be tested.
"The tuberculosis (TB) skin test may be very useful for high-risk children, but next to useless for low-risk children," said Jeffrey R. Starke, MD, associate professor of pediatrics, Baylor College of Medicine, Houston.
This is a problematic test that in some ways does more harm than good, he said at the Infectious Diseases in Children Symposium West held here. Positive results will be obtained when low-risk children are skin tested for TB, but most will be false positive, said Starke.
It is estimated that sensitivity and specificity of the test are approximately 90%. This means, said Starke, that when children with a high prevalence of infection are tested, most will be true positives.
This is a good test because there are positive predictive values, he said. However, if the same test is applied to a population in which the true prevalence of infection is only 1%, the positive predictive value is less than 10%. "More than 90% of these tests will be false positive."
In most tests in which a screening test yields false positives, there is a definitive test that has a high sensitivity and specificity allowing for differentiation between the true and false positives. Once a patient tests positive he or she is positive.
"The single most important decision is whether or not to place the skin test on a child because once you have placed it, the ball is rolling and you have to interpret the things in a certain way," he said. "There is no TB skin test that discriminates between the true positives and false positives."
This leads to unnecessary costs including clinic visits, physicians visits, drugs, radiographs and family evaluations. Anxiety, which Starke said has been discounted as a "cost," is another problem with TB skin tests. "I have seen pediatricians and families go nuts [when a child has a 13 mm skin test]," said Starke.
Families get anxious over finding the source of the child's TB, said Starke. "The problem of course is that nobody has TB. It is a false positive."
To help certain populations the sensitivity and specificity of the test is being changed. If the test is lowered to 5 mm the sensitivity will be increased, the specificity will be lowered and more false positives will be yielded.
---This
is one method or applying the Mantoux tuberculin skin test. By
anchoring the hand on the side of the child's arm, the tester maintains more
control and can adjust the angle of the needle.
"There are certain groups of people for whom we will tolerate false positives, and these are not the people who necessarily even have a higher risk of infection," he said. These are people who have a higher risk of progressing to disease once they are infected, including HIV-infected and other immunocompromised patients and close contacts.
The 5 mm skin test is reserved for patients who have high risk of progression to TB disease. Some of these, however, will be false positives, said Starke. The 10 mm skin test is reserved for people with other risk factors that place them at risk for infection. The American Academy of Pediatrics, said Starke, added children younger than 4 years of age to this group.
They were not added because they are at higher risk, but because of the perception that they are at higher risk of developing disease than older children and adults if they are infected.
People with no risk factors fall into the 15 mm skin test group. "If they have no risk factors, what do you want to skin test them for?" asked Starke. Even at 15 mm, most positives will be false positives, he said.
The 5, 10 and 15 mm skin tests do present problems and one of those is discrimination. A lot of high-risk children, particularly those born in foreign countries who were not skin tested as part of immigration, are in U.S. schools.
"We would like to get at these kids, without having to get at everyone else," said Starke. The discrimination problem, he said, lies in testing two children from the same class and telling the parents of the foreign-born child that he or she needs to have X-rays, the family needs testing and the child has to be put on medication with side effects.
"This may be great public policy, but it is extraordinarily difficult social and school policy that raises the specter of discrimination." The truth, however, is that TB is a discriminatory disease, and the challenge for public health is in deciding how discriminatory to be in its methodology to control diseases like TB, said Starke.
Another way to look at it, he said, is as providing extra and special services to a vulnerable population. "The language that we use and how we describe this become very important. Anytime you treat different kids differently, it always poses potential problems."
A number of things, however, can be done to screen those at risk, he said.
The willingness and ability of health care workers and family to take a history that reveals the risk factors is important. If the family does not give that history and no other risk factors are present, low risk may be falsely ascribed to a high-risk child.
"There is no substitution for taking a history. A family history going back two or three generations is adequate. If you go back four or five generations, everyone has a family history of TB," he said.
Foreign birth of a child, foreign travel and/or receipt of foreign visitors also contribute to risk. Starke spoke of a recently published paper that showed that children traveling for more than three weeks or receiving foreign visitors from countries with high TB rates correlated with an increase in positive tests.
Contact with HIV-infected or AIDS patients and with incarcerated individuals are also risks. People in prison or jail, said Starke, "are an incredibly at-risk population put under living conditions that are very conducive to the development of TB disease."
The final factor is local epidemiology, which may be migrant workers, homelessness or a particular housing complex. "This is where pediatricians have to work very closely with the health department to find out where TB is occurring in their communities."
For more information:
- Starke JR. Pediatric tuberculosis. Presented at the Infectious Diseases in Children Symposium West. May 22-23. Dana Point, Calif.
![[bar]](/art/gradred.gif){kind=small}
You can express your
views on this article, or other relevant themes, in the
Infectious Diseases in Children
Specialty Forums.