The study below was published recently. It is very powerful study because of its high number of patients and as it is registry data it is very powerful at tracking true outcomes. Many studies may indicate excellent outcomes but often patients may be lost to follow up leading bias in the interpretation of the results. Registry evidence like that below tracks those patients no mater who does their revision surgery giving it excellent follow-up which can be a problem in many studies as mentioned

In this study looking at over 14,000 ACL surgeries it shows as you can see in the blue box at the bottom that Patellar Tendon ACLs have a revision (repeat surgery for failure)  rate of 0.7% at 2 years. The remaining categories are all Hamstring grafts with various methods of fixation. Some have a revision rate as high as 5.5% or greater than 7 times the patellar tendon failure rate at 2 years. You can see this represented in the graph where the black line is the failure rate over time of patellar tendon grafts and the colour lines are different hamstring graft combinations.

This is the reason Shane prefers to use BTB (Patellar tendon) grafts for his ACL reconstructions.

For those of you interested in statistics because often they can be interpreted different ways, the P value in the bottom left of the diagram indicates how significant the finding is. Any value under 0.05 is said to be scientifically significant and as a doctor you should consider the result of this evidence. In this case the P values are mostly below 0.001 meaning they are extremely significant.
ACL Evidence