The choice of graft for anterior cruciate ligament reconstruction remains controversial. A systematic review was performed to compare bone-patella tendon-bone and 4-strand hamstring grafts. Medline (1966 onwards), EMBASE (1980 onwards) and the Cochrane database were searched retrieving 6312 possible articles but only 6 studies fulfilled all the inclusion criteria. To be included, the study had to be prospective, randomised or quasi-randomised, comparing 4SHS and central third BPTB autografts, inserted using an arthroscopically assisted technique and have a minimum 2-year follow up for all patients. These studies recruited 526 patients and 475 were followed for at least 2 years with 235 patients receiving a bone-patella tendon-bone graft and 240 receiving a 4-strand hamstring graft. Overall, there was a greater chance of extension loss (p=0.007) and a trend towards increased patellofemoral joint pain (p=0.09) with a patella tendon graft. With a 4-strand hamstring graft there is a greater loss of hamstring power (p=0.008) and a trend towards an increased chance of a pivot shift > 1 (p=0.12). There was no difference between the 2 groups in terms of lachman testing, chance of returning to the same level of sport, clinical knee scores, graft ruptures or other complications.

The choice of graft for anterior cruciate ligament reconstruction remains controversial. A systematic review was performed to compare bone-patella tendon-bone and 4-strand hamstring grafts. Medline (1966 onwards), EMBASE (1980 onwards) and the Cochrane database were searched retrieving 6312 possible articles but only 6 studies fulfilled all the inclusion criteria. To be included, the study had to be prospective, randomised or quasirandomised, comparing 4SHS and central third BPTB autografts, inserted using an arthroscopically assisted technique and have a minimum 2-year follow up for all patients. These studies recruited 526 patients and 475 were followed for at least 2 years with 235 patients receiving a bone-patella tendon-bone graft and 240 receiving a 4-strand hamstring graft. Overall, there was a greater chance of extension loss (p=0.007) and a trend towards increased patellofemoral joint pain (p=0.09) with a patella tendon graft. With a 4-strand hamstring graft there is a greater loss of hamstring power (p=0.008) and a trend towards an increased chance of a pivot shift > 1 (p=0.12). There was no difference between the 2 groups in terms of lachman testing, chance of returning to the same level of sport, clinical knee scores, graft ruptures or other complications.