Medical Residents Flunk TB Quiz

BALTIMORE, Aug. 2 -- Internal medicine residents scored poorly on a 20-question quiz testing their ability to diagnose tuberculosis and choose appropriate treatment, researchers found.

Asked to answer basic questions about diagnosis and early management, the residents in 2005 got almost half the answers wrong, Petros C. Karakousis, M.D., of Johns Hopkins here, and colleagues, reported online in BMC Infectious Diseases.

Recently, TB diagnosis became the subject of international interest when airline passenger Andrew Speaker was incorrectly believed to be infected with extensively drug-resistant TB, although he had the still-serious multi-drug-resistant form of TB.

The TB knowledge survey was taken by 131 internal medicine residents during a routinely scheduled teaching conference at four urban medical centers in Baltimore and Philadelphia -- Johns Hopkins, Johns Hopkins Bayview Medical Center, the University of Maryland, and the University of Pennsylvania.

Earlier the survey was pilot-tested among 17 infectious disease attending at the participating institutions, eight of whom described themselves as TB experts. The median percent of questions correctly answered was 90%.

The median percent of survey questions answered correctly for all the residents was 55%, the researcher reported. Knowledge did not improve with increasing post-graduate training or a greater number of TB patients managed within the previous year.

Dr. Karakousis pointed out that the survey results were not all bad, with most medical residents understanding the main facts about how M. tuberculosis is transmitted. The median score (95.0%) was highest for questions pertaining to transmission, with nearly half of all residents responding correctly to all three questions in the category.

For example, 81% (n=66) recognized that no specific effective measures exist to prevent acquisition of the bacterium in areas of high prevalence. Seventy-two percent (n=36) correctly recognized that patients with sputum samples negative by acid-fast staining may still transmit the disease.

The correct management and treatment of latent TB infection proved to be a special stumbling block, with the residents getting three-fifths of the answers wrong (median score 40.7%). Interpretation of tuberculin skin tests proved particularly difficult.

• Faced with a foreign-born BCG-vaccinated person with a positive tuberculin test, only 47% of the residents said they would treat with the appropriate regimen of isoniazid for nine months.
• Only 41% of the residents identified the combination of rifampin and pyrazinamide as an unacceptable regimen for patients with latent infection. Only 15% were aware of the proper use of pyridoxine supplementation.
• Nearly half of the residents believed incorrectly that the lifetime risk of developing active disease in non-HIV-infected persons with latent TB was below 2%. Fifty-five percent correctly identified the use of infliximab as the greatest risk factor for progression from latent infection to active disease.

The results for active disease were somewhat better with a median correct score for 57%.

• The great majority (97%) correctly indicated that the diagnosis of pulmonary TB cannot be reliably excluded on the basis of a negative tuberculin skin test.
• However, only 28% of the residents recognized that the diagnosis of pulmonary TB cannot be reliably excluded after three sequential sputum samples are negative with acid-fast staining. Similarly only 40% recognized that the diagnosis cannot be reliably excluded after a single negative bronchoscopic lavage sample by acid-fast staining.

On two questions about the toxicity of current drug therapy for active TB and the link between TB and HIV infection, physicians gave the right answer with a median score of 63.3% for both.

• Many residents (39%) incorrectly thought that the annual risk of developing active disease in HIV-infected patients with latent infection was less than or equal to 1%.
• The majority of residents (84%) correctly identified rifampin as the first-line agent with the greatest potential for drug-drug interactions in patients receiving highly active antiretroviral therapy.

In reviewing the study's limitations, the researchers said that because the survey was conducted at training programs in cities with moderate prevalence of tuberculosis, it is possible that medical resident knowledge would be better in cities with higher rates, such as San Francisco, New York, Houston, or Los Angeles.

Similarly, knowledge may have been overestimated in cities with lower prevalence or in rural areas. However, they said, the data suggest that knowledge does not improve with the number of cases seen.

Finally, Dr. Karakousis wrote, "we cannot be sure that lack of tuberculosis knowledge leads to poor management of patients. Although trainees may score poorly, they may be quick to consult experts in infection control, infectious diseases, or pulmonary medicine to assist in diagnosis, isolation, and treatment."

Because clinical practice paradigms are often ingrained in physicians during their residency, education about guidelines for detection and early management of the disease may be important for future improvement in national tuberculosis control strategies, the researchers concluded.