comparison of roll stitch technique and core suture technique for extensor tendon repair at the metacarpophalangeal joint level

conclusions roll stitch technique had superior outcome compared to the modified kessler technique, when performed in the mcpj area. level of evidence: therapeutic (level iii) background proper suturing technique is needed to ensure good outcome in extensor tendon surgery. different techniques have been reported for the repair of extensor tendon injuries at the level of the metacarpophalangeal joint (mcpj). these reports were in vitro studies on cadaver models. repair techniques must be clinically tested, to determine results. objectives the purpose of this in vivo study was to compare results of extensor tendon repair, using roll stitch and core suture techniques. patients and methods forty two fingers, in 38 patients (aged 15- 45 years), with simple complete extensor tendon injuries in the mcpj area, were identified and operated by a single surgeon. the patients were divided into two groups, according to the technique used for tendon repair. the first group consisted of 21 digits, in 19 patients, who were repaired with roll stitch technique, while the second group consisted of 21 digits, in 19 patients, who were repaired with core suture technique. the same splint and rehabilitation regimen (early passive range of motion) were given to all patients. the splints were removed at 6 weeks after surgery and range of motion of the operated fingers was measured and compared to uninjured hands, after 12 weeks. results five patients were lost to follow up or excluded from the study. there was no rupture of the repaired tendons in the groups. there was no statistically significant difference in mean mcpj flexion, proximal interphalangeal joint (pipj) flexion, distal interphalangeal joint (dipj) flexion and total range of motion of the fingers, between the two groups. however, extension lag was significantly more common in the second group (11 of 19 digits) compared the first group (four of 17 digits).