At several points in the recently published Code of Practice 2000,1 dealing with the control and prevention of tuberculosis in the UK, a number of important actions and decisions turn on the results of sputum microscopy for acid fast bacilli (the “smear”). This simple, cheap, and rapid test is used to assess sputum infectivity and prompts decisions on isolation, initiation of treatment, and the need for and extent of contact tracing. It is also the major criterion used to assess specimen priority for advanced testing, including polymerase chain reaction (PCR) and automated liquid culture.2

However, this procedure is unstandardised and smear positivity relative to culture varies from 60% to over 80%. Ziehl-Neelsen staining continues as a primary screen despite good evidence that auramine-based methods are more sensitive and quicker.3 Processing of sputum before staining may or may not involve digestion and/or concentration by sedimentation or centrifugation, despite advice and evidence that both improve the results.4 Quality control schemes assess the ability to stain and microscopically examine suspensions of mycobacteria, but not the critical issue of specimen processing before staining.

Less sensitive smear techniques may cause delays in the recognition and management of the index case, including the use of isolation facilities, and an unjustified view of low infectivity which will persist even after the culture is positive. Casual, particularly occupational, contacts of such patients will be at a significant disadvantage. Suboptimal smear techniques will also mean that some specimens meriting examination by enhanced methods will not be sent for such examination. The potential benefits of theMycobacterium tuberculosis PCR will be unnecessarily compromised.

It is now nearly 120 years since Ehrlich first described the acid fastness of some organisms, including mycobacteria. We think it is long overdue that all mycobacteriologists accurately, optimally and, above all, consistently exploit his discovery.