Barrett / SRK-T Sim

While the original SRK regression formula was P = A - 2.5L - 0.9K (where L = Axial Length and K = Average Keratometry), the SRK/T advanced this by incorporating a theoretical vergence foundation. It integrates five major algorithmic corrections, optimizing postoperative Anterior Chamber Depth (ACD) prediction, correcting for retinal thickness based on axial length, and refining the corneal refractive index.

There are two critical errata for the 1990 SRK/T publication: one in 1990 correcting the axial length LCOR thresholds (JCRS 1990;16(4):528), and another in 1993 fixing the geometric loop for extremely short eyes where computed corneal height H became mathematically negative (JCRS 1993;19(5):444).

The SRK family of formulas was born from the discontent of Dr. John Retzlaff, Dr. Donald Sanders, and Dr. Manus Kraff regarding post-cataract refractive surprises. They leveraged the new power of mainframe and high school computers to conduct linear regression on mass data sets. Starting with coin-flip naming (SRK), the basic regression P = A - 2.5L - 0.9K became the global standard. Prompted by the emergence of purely theoretical formulas (like Holladay 1), the doctors eventually merged their vast empirical regression data with a theoretical eye model to forge the SRK/T in 1990.