Diagnosis of abdominal tuberculosis * Authors' reply

Abstract

We agree with the argument presented by Mr Rai and Mr Thomas (December 2003 JRSM1) for earlier use of diagnostic laparoscopy in suspected abdominal tuberculosis (TB). However, even in the hands of experienced operators the risk is not negligible (though under 1%). The complications, apart from those of general anaesthesia, include damage to blood vessels, bowel and bladder. In Rai and Thomas's series, seemingly, a polymerase chain reaction (PCR) test was not used on the 11 ascitic fluids when acid-fast bacilli were not found. A positive result might have avoided the need for laparoscopy;2,3 furthermore, PCR can help to distinguish Mycobacterium tuberculosis from M. avium in immunocompromised patients.3 In addition, although in 9 of 28 patients ultrasound showed evidence of TB, it seems that fine needle aspiration was not attempted in those with lymph node abnormalities, or colonoscopy in those with an ileocaecal mass. An ultrasound-guided biopsy4 or colonoscopy5 is indicated in the presence of lymphadenopathy, abscess or a focal lesion in the viscera. Finally, Rai and Thomas refer to low haemoglobin and raised C-reactive protein as a consistent feature but say nothing about Ca-125, which can be high in abdominal tuberculosis.6 In our opinion, diagnostic laparoscopy should be used only when less invasive investigations yield no positive results.