BACKGROUNDMicroendoscopic discectomy (MED) is a minimally invasive operation that allows rapid recovery from surgery for lumbar disc herniation, but has replaced traditional open surgery in few hospitals because most surgeons avoid its long learning curve. We evaluated the effectiveness and safety of lumbar MED at stages of spinal surgeons' learning curve.

METHODSFifty patients receiving MED from June 2002 to February 2003 were divided into chronological groups of ten each: A - E. The control group F was ten MED patients treated later by the same medical team (September - October 2006). All operations were performed by the same team of spinal surgeons with no MED experience before June 2002. We compared groups by operation time, blood loss, complications and need for open surgery after MED failure.

RESULTSOperation times by group were: A, (107 +/- 14) minutes; B, (85 +/- 13) minutes; C, (55 +/- 19) minutes; D, (52 +/- 12) minutes; E, (51 +/- 13) minutes; and F, (49+/- 15) minutes. Blood loss were: A, (131 +/- 73) ml; B, (75 +/- 20) ml; C, (48 +/- 16) ml; D, (44 +/- 17) ml; E, (45 +/- 18) ml; and F, (45 +/- 16) ml. Both operation time and blood loss in groups C, D, E and F were smaller and more stable compared with groups A and B. Japanese Orthopedic Association assessment (JOA) score of each group in improvement rate immediately and one year after operation were as follows (in percentage): A, (79.8 +/- 8.8)/(89.8 +/- 7.7); B, (78.6 +/- 8.5)/(88.5 +/- 7.8); C, (80.8 +/- 11.3)/(90.8 +/- 6.7); D, (77.7 +/- 11.4)/(88.9 +/- 9.3); E, (84.0 +/- 8.7)/(89.6 +/- 9.0); and F, (77.8 +/- 11.6)/(86.9 +/- 8.4). Groups showed no statistical difference in improvement rates. Complications developed in three patients in group A, two in group B, and none in the other groups.

CONCLUSIONSSpinal surgeons performing MED become proficient after 10 - 20 operations, when their skill becomes fairly sophisticated. Patients' improvement rate is the same regardless of surgeons' phase of learning curve.