Abstract: Objective To evaluate the accuracy of Barrett True-K formula, Shammas formula and Haigis formula in the intraocular lens (IOL) of cataract patients after radial keratotomy (RK). Design Retrospective case series. Participants 28 patients (49 eyes) diagnosed with cataracts after RK at Beijing Tongren Hospital from January 2017 to February 2024. Methods IOL Master was used to obtain the parameters of the anterior segment of the patient before surgery. SRK/T formula, Shammas formula or Barrett True-K formula were used to calculate IOL diopter before surgery. The refractive status of the eye was obviously checked by optometry one month after surgery. The actual diopter was recorded by equivalent spherical (SE). Log on to the official website of American Refractive Association (www. ascrs.org) and apply Barrett True-K formula, Shammas formula and Haigis formula to calculate the predicted IOL diopter by replacing the actual postoperative diopter and the measured data of IOL Master. Then the actual implanted IOL diopter was subtracted from each formula to predict IOL diopter, denoted as IOL diopter error, and its absolute value was denoted as absolute diopter error. Main Outcome Measures Refractive error and absolute refractive error. Results For the 28 patients (49 eyes) included in the study, the refractive error of Barrett True-K formula method was 0.13 (-0.75, 0.89) D at one month after surgery, which was significantly lower than that of Shammas formula method -1.20 (-1.94, -0.54) D, Haigis formula method -1.59 (-2.54，-0.49) D (S=39.837, P<0.001). The absolute refractive error of Barrett True-K formula method was 0.79 (0.39, 1.58) D, which was lower than that of Shammas formula method 1.21 (0.57, 1.94) D and Haigis formula method 1.59 (0.49, 2.54) D (S=9.959, P=0.007). In all 28 patients (49 eyes), Barrett True-K formula method accounted for 39%, 67% and 96% of refractive errors in the range of ±0.5 D, ±1.0 D and ±2.0 D, respectively.  Shammas formula method accounted for 29%, 51%, 88%, and Haigis formula method accounted for 33%, 43%, 80%, respectively. The difference of refractive error between the three formulas in the range of ±1.0 D (χ2=6.130, P=0.047) and ±2.0 D (χ2=6.078, P=0.048) was statistically significant. The absolute refractive error of Barrett True-K formula was negatively correlated with the axial length of the eye (r=-0.397, P=0.008). For patients (19 cases, 33 eyes) who applied Barrett True-K formula method to calculate IOL diopter before surgery, the diopter error of Barrett True-K formula method was 0.30 (-0.78, 1.56) D, which was significantly lower than that of Shammas formula method-1.18 (-1.77, -0.59)D, -1.31 (-2.46, -0.34)D (S=28.42, P<0.001) for Haigis formula method. There was no statistical significance in the absolute refractive error of Barrett True-K formula compared with Shammas formula and Haigis formula (S=3.697, P=0.157). In the patients with IOL diopter calculated by Barrett True-K formula before surgery, Barrett True-K formula method accounted for 24%, 61% and 97% of refractive errors in the range of ±0.5 D, ±1.0 D and ±2.0 D, respectively. Shammas formula method accounted for 15%, 42% and 87%. Haigis formula method accounted for 30%, 39% and 64%, respectively. The proportion of refractive error of Barret True-K formula method within ±2.0 D was higher than that of Shammas formula method and Haigis formula method (χ2=13.778, P=0.001). Conclusion For cataract patients after RK, there are some differences in the accuracy of Barrett True-K formula method, Shammas formula method and Haigis formula method. The absolute refractive error of Barrett True-K formula method is negatively correlated with the length of the axis of the eye, that is, the longer the axial length of the eye, the lower the absolute refractive error and the higher the accuracy. (Ophthalmol CHN, 2024, 33: 94-98)

Key words: radial keratotomy, intraocular lens diopter calculation