How many of your patients have multidrug-resistant TB (MDR-TB)?

In our DOTS projects in the Former Soviet Countries, we realised a while back that between 10% and 20% of our patients didn't respond to treatment, so we conducted drug sensitivity surveys. Our results showed varying rates of resistance: in North Karabagh, none of the new cases and 10% of the previously treated cases had MDR-TB. In the Kemerovo region prisons, the rates were much worse: 21% among new cases and 38% among previously treated cases.

How do you treat patients with MDR-TB?

Treating MDR-TB takes two years using expensive second line drugs. It's essential to have adequate treatment with first-line drugs already in place, otherwise resistance to second-line drugs can emerge.

Duration of treatment and side-effects increase the risk of treatment interruption, so we have to use inclusion criteria to select patients - for instance the patient's history of adherence to treatment and presence of co-disease. In the programmes where we don't have second-line drugs, we monitor the MDR-TB patients and show their families how to protect themselves. It's very hard for all concerned, because many patients don't get better and die of their TB.

What kind of laboratory facilities do you have for diagnosis and drug susceptibility testing?

In Abkhazia, we send samples to a supranational reference lab in Rome. In Nagorno Karabagh, they are sent to a lab in Antwerp and in the Aral Sea Area, to a German lab.

It sounds like you always send samples to laboratories abroad - are there no lab facilities in the countries you work in?

Our TB project in the Koromovo prison in Siberia is the only one that has a bacteriological lab which can conduct drug sensitivity testing on site. Ideally, we really should have access to a lab nearby. But in Abkhazia and Karabagh, our programmes are small and run in unstable areas where it's impossible to set up a lab. So we continue to send samples to a supranational lab.

How many supranational labs are there in the world? What difficulties do you encounter using them?

I know of 23 laboratories in the supranational reference network, mostly in rich countries. There is still no supranational laboratory in Russia, despite its very high TB burden. It took MSF a whole year to find a lab that would accept samples for a resistance surveillance study. They asked the New Delhi lab first, because it was closest, but they refused. Then Bangkok refused, and the English lab asked for an expensive fee for each sample. Finally the German lab accepted to take them.

What big challenges will your projects face in the years to come?

Our biggest problem is improving adherence to treatment - the shortest treatment courses are still six months, which is enormously long. We have to find ways of making observed treatment easier - we're trying to offer patients more support and involve the community.

Another challenge is to get easier and cheaper access to the supranational labs. A lab in Moscow should be receiving the supranational title soon and another one in Almaty should become a referential lab for Central Asia - so things are moving in the right direction. But this won't be sufficient - we need new drugs for TB.

Reducing the duration of treatment to a few weeks would really help us make progress against this disease. At the very least we should develop drugs that are easier to use, such as blister packs for example. There is also an urgent need to develop drug sensitivity tests that are cheaper and faster. The rapid method that exists today is too expensive and difficult to implement in field conditions.