The dataset does not include indicators of the quality of care and does not allow to assess whether changes associated with PBF are resource-driven or due to the incentive mechanism itself. The results are largely consistent with other impact evaluations conducted in Burundi and Rwanda. The fact that PBF is mostly associated with positive changes in the use of services that became free suggests an important interaction effect between the two strategies. A possible explanation is that the removal of user fees increases accessibility to health care and acts on the demand side while PBF gives medical staffs incentives for improving the provision of services. More empirical research is needed to understand the sustainability of (the incentive mechanism of) PBF and the interaction between PBF and other health policies.

Abstractobjective Community participation is often described as a key for primary health care in lowincome countries. Recent performance-based financing (PBF) initiatives have renewed the interest in this strategy by questioning the accountability of those in charge at the health centre (HC) level. We analyse the place of two downward accountability mechanisms in a PBF scheme: health committees elected among the communities and community-based organizations (CBOs) contracted as verifiers of health facilities' performance.method We evaluated 100 health committees and 79 CBOs using original data collected in six Burundi provinces (2009)(2010) and a framework based on the literature on community participation in health and New Institutional Economics.

Several unresolved questions remain, concerning the integration of vertical programmes and the sustainability of the system given the considerable costs, since funding is not yet fully ensured by the State and its partners.

A 56-year-old patient presented with neurological disorders resulting from an extrinsic medullary compression extending from T3 to T5 associated with a T4 corporeal invasion by a high-grade non-Hodgkin's malignant lymphoma. Treatment consisted in a laminectomy without spinal stabilization followed immediately by chemotherapy (VCAD) and locoregional radiotherapy (20 cycles of 50 Gy between T3 and T7 followed by nine sequences of 18 Gy). The evolution revealed unfavourable with reappearance of a compressive syndrome thus requiring surgical decompression combined with spinal stabilization. The immediate postoperative period was simple but a secondary wound dehiscence was observed surrounded by an inflammatory area of 15 cm on 12 cm. The authors describe the reconstruction by means of a trapezius musculocutaneous island flap. The results were satisfactory and corroborated those previously reported by various authors after use of this flap in similar situations.

No abstract