Description

van Rheenen developed an algorithm for making the diagnosis of tuberculosis (TB) in children living in an HIV-endemic area of Africa. The author is from The Netherlands with the study done in St. Theresa's Mission Hospital in Zambia.


 

Patient selection: Children <= 12 years of age with chronic cough for > 3 weeks and/or low body weight. (I also included weight loss and night sweats in the implementation.)

 

Diagnostic findings:

(1) positive smear for acid fast bacilli (Ziehl-Neelsen)

(2) positive mycobacterial culture

(3) positive PCR for M. tuberculosis

(4) positive skin test

(5) chest X-ray suggestive of pulmonary tuberculosis

(6) caseating granulomas on biopsy

(7) history of contact with a patient with tuberculosis

 

If (1), (2) and/or (3) were positive, then the person was considered to have confirmed TB.

 

If (4), (5), and/or (6) were positive, then the person was started on antituberculosis drugs.

(1) If the patient improved, then the patient was considered to have probable TB.

(2) If the patient did not improve, then the patient was considered to have possible TB.

 

If only (7) was positive, then the patient was started on antibiotics for 2 weeks.

(1) If the patient improved, then a diagnosis other than TB was made.

(2) If the patient did not improve, then antituberculosis drugs were started.

(3) If the patient did not improve on antituberculosis drugs, then a diagnosis other than TB was made. If the patient improved on antituberculosis drugs, then the diagnosis was possible TB

 

If none of the findings were present, then other diagnoses were considered.

 


To read more or access our algorithms and calculators, please log in or register.