Background Latvia has consistently ranked among the countries with the highest rates of MDR-TB in the world. In the first Global Tuberculosis Drug Resistance Survey (1996), 14.4%, or 1 out of 7, of all newly diagnosed sputum smear-positive tuberculosis cases in Latvia were diagnosed as MDR-TB. (1) Data also show that the proportion of cases with additional resistance to second-line anti-tuberculosis drugs is high. In the meantime, HIV seroprevalence is increasing among TB patients. Latvia, with an estimated population of 2.35 million, joined the European Union on 1 May 2004. Latvian economy had been severely affected by the collapse of the Soviet Union, with gross domestic product (GDP) per capita falling by nearly 35% in real terms in 1992. GDP per capita in 1999 was US$ 4200, increasing to US$11 500 in 2004. Latvia concurrently experienced dramatic increases in TB morbidity and mortality peaking in 1998, together with the appearance of drug-resistant and MDR-TB. (2,3) Latvia adopted WHO's recommended DOTS strategy for TB control in 1996 and subsequently introduced MDR-TB management (4) in 1997. This relies on MDR-TB treatment with individualized regimens under the consilium or expert consultation process. The treatment is provided at four inpatient treatment centres (including a prison TB ward) followed by outpatient directly observed therapy. All funding for TB and MDR-TB control comes from the government. In 2000 Latvia's National Tuberculosis Program (NTP) sought MDR-TB management support from the Green Light Committee (GLC) and got approval to treat 350 more MDR-TB patients. The GLC enabled Latvia to treat all patients diagnosed with MDR-TB. Epidemiology In 1991, the incidence of TB was 29 cases per 100 000 population, (5) increasing to 74/100 000 in 1998 and then declining to 53.5/100 000 in 2005. Case finding shows 49% case detection by smear microscopy. Drug-resistant TB case detection strategy in Latvia is based on drug sensitivity tests (DST) on solid media. For high-risk MDR-TB cases, the BACTEC/MIGT system is used, as well as the INNO LiPA test to detect rifampicin resistance in 2-4 days. Extensive resistance to first- and second-line drugs among MDR-TB patients is well known in Latvia. One of the reasons is the country's long and extensive use of second-line drugs before implementing the DOTS strategy. Extensive resistance affects the MDR-TB treatment regimen and outcomes. For cohorts registered from 2000 to 2005, resistance to kanamycin was 49%; capreomycin, 39%; ofloxacine, 9%; protheonamide, 30%; para-aminosalycilic acid, 31%; and thiacethasone, 23%. In the first worldwide survey, published in May 2006, estimates for years 2000-2004 showed that 19% of MDR-TB patients have resistance to first-line drugs defined as MDR-TB plus resistance to three drugs of six classes of second-line drugs. (6) Using the new revised extensive drug resistance (XDR-TB) definition of resistance to at least rifampicin and isoniazid, additional resistance to any fluoroquinolone and to any of three second-line injectable drugs (capreomycin, kanamycin or amikacin), such extensively resistant TB was found in 39 cases, or 5.2% of all MDR-TB cases registered during the past six years. Two-thirds, or 67%, of MDR-TB patients out of 820 treated in the years 2000-2003 were cured; 6% were dead; 14% defaulted; and treatment failed in 13% of cases. (7,8) The treatment success rate for XDR-TB patients is low: out of all 48 patients treated from 2000 to 2005 (including MDR-TB retreatment cases with XDR-TB), only 18 (38%) were cured, while treatment failed for 22 (46%). Among all MDR-TB cases in the cohorts, 3% were co-infected with HIV; this proportion increased to 12% among XDR-TB cases. Treatment success for TB/HIV co-infected new patients, at 74%, is similar to overall treatment success for new TB patients, but the HIV-associated MDR-TB success rate is 56%. …