Democratic Republic of Congo (DRC) has the second highest burden of smear-positive tuberculosis (TB) in Africa yet many cases go undetected. In the South-Kivu Province, this may be closely related to conflicts and difficult access to health services. Through a TB REACH grant, we addressed low case detection and identification of drug resistance through GeneXpert MTB/Rif technology (GXP). GXP is a molecular test that can be used at lower levels of health services, has a high sensitivity and a turn-around time of 2 hours. In addition, this assay permits to identify Rifampicin resistance. There is little experience using GXP in programmatic settings in DRC. Material and methods Eight GXP machines (consisting of 18 total modules) were placed in 2 referral/academic laboratories, 5 secondary hospitals and one urban primary health care facility. No lab technician had used PCR techniques before training. Device setting and training of lab technicians were performed on-site during 3 consecutive days. In addition, each laboratory was equipped with appropriate electrical power to be able to perform 4 runs of tests per module/day. GXP testing was performed on patients clinically suspected of TB with at least 3 negative Ziehl-Neelsen microscopy tests. Results During the first year evaluation period, 8.614 tests were performed among smear-negative sputum samples. Each module performed on average 19,4 tests per month during the first semester and 45 tests per month during the second semester. The positivity rate of GXP tests performed among smear negative samples was 7,2% (n = 624). Rif resistance was detected among 31 samples, and 20 patients received second line treatment. Two of these patients (10%) died within the 1-year follow-up. Error rates was 9,6% (n = 827). Before one year of use, 7 out of 18 modules (39%) had to be replaced for technical reasons, including unexplained high error rates. Discussion GXP technology can be successfully implemented in settings with very limited resources. In South-Kivu, considering a general positivity rate of 13,6% for smear microscopy, GXP improved bacteriologically confirmed case notification by 50% in areas where this test was available. Detection of drug resistant cases surged and demonstrates that many resistant cases are missed with the current algorithm. Maintenance issues and continuous monitoring of error rates are crucial and should be improved before scale-up of GeneXpert at a country-level.
André, Emmanuel ; Chirambiza, Jean-Paul ; Nyota, Rosette ; Kalumuna, Dieudonné ; Creswell, Jacob ; et. al. Lessons learned from early implementation of GeneXpert MTB/RIF technology in a remote area of the Democratic Republic of Congo.19th Conference of the Union - Africa Region (Kigali, du 20/06/2013 au 22/06/2013).