The study was performed at Moorfield’s Eye Hospital, a large tertiary referral center and teaching hospital, and its satellite sites across London, UK. It was registered and approved under the local Clinical Effectiveness Unit of Moorfields Eye Hospital. The hospital data was retrospectively accessed without interacting with patients. Therefore, informed written consent was waived. The personal identity was delinked from other data before analysis. The tenets of the Helsinki Declaration were strictly abided at different stages of the research.
This was a two-armed retrospective cohort study. All adult patients undergoing ISBCS at Moorfields Eye Hospital between March 2020 and March 2023 were included. Any patients having a combined non-cataract procedure were excluded. For our patient search and data collection, an electronic database was used (OpenEyesTM, UK). No paper records were used since all information required is recorded in the electronic database. This produced a data set of 624,397 patients (1,245,794 eyes). Data collected included the patient’s age, operating surgeon, level of the operating surgeon, preoperative visual acuity and refraction (autorefraction), IOL type and power used, predicted post-operative spherical equivalent from biometry, intraoperative complications, postoperative vision and refraction, and postoperative complications for both eyes in the three months after surgery.
Many consultant ophthalmologists, specialist trainees, and fellows performed the surgeries. The specialist trainees were subdivided into Lower House (LH) (specialty trainee year 4 (ST4) or below) and Upper House (UH) (Specialty trainee year 5 (ST5) or above). For analysis, eyes operated by trainee surgeons were grouped into Gr1, and eyes operated by consultant surgeons were grouped into Gr2. The consultants operated cataracts under topical anesthesia. The trainee ophthalmologists operated on cataracts under the supervision of a consultant under sub-tenon local anesthesia if needed.
We assumed that senior residents had a success rate of achieving a spherical equivalent (SE) within 0.5D of the predicted post-operative refraction in 77% following ISBCS. [ref 11] The same outcome was 85% when operated by consultants. [ref 12]
To achieve a 95% confidence interval and 90% power to a cohort study with a 1:1 ratio in a cohort study, we need at least 553 ISBCS in each arm of the cohort. Since Gr1 had 553 eyes and Gr2 had more eyes during the study period, we included all the cases. We used open epi software to calculate the sample size for a cohort study. [ref 13]
Four trainees’ ophthalmologists and two consultants were the study investigators.
All patients listed for surgery were assessed in one-stop cataract clinics. They all had anterior segment and dilated fundal examinations, OCT macula (SPECTRALIS OCT, Heidelberg Engineering, Germany), and biometry performed using the IOL-Master 700 (Carl Zeiss Meditec AG, Germany). All patients were offered a mono-focal lens, and this was selected based on the SRK/T formula or Hoffer Q formula, depending on whether the axial length was above or below 22mm respectively. All patients were offered the choice of DSBCS and ISBCS.
All patients received a standard regime of pre-operative drops in each eye, consisting of Mydriasert (Thea Pharmaceuticals Ltd, France) or a combination of topical eyedrops e.g. tropicamide 1% and phenylephrine 2.5%. In theatre, ISBCS eyes were treated as two separate procedures with consumables from different baches as a mandatory requirement. Before the start of each eye, a separate WHO checklist and biometry check were done. [ref 14] The case selection, surgical procedures, and safety measures for undertaking ISBCS in the present study are described elsewhere. [ref 15] If both eyes were at similar risk, the consultant randomly selected the right or left as the first eye for surgery, and the trainee ophthalmologist operated on the fellow eye. The eye with advanced ocular comorbidity was operated on by the consultant.
All patients were given a standard post-operative drop regime, which consisted of Pred-forte 1% drops four times daily for two weeks followed by twice daily for two weeks and chloramphenicol 0.5% drops four times daily for two weeks. This may have been altered if there was an intraoperative complication. If a patient was at risk of post-operative cystoid macula edema, they were given acular 0.5% (Allergan, USA) drops four times daily for four weeks.
The post-operative review was face-to-face and generally at four weeks unless the procedure was complicated when the post-operative review was expedited. The pre and postoperative autorefraction enabled us to document the spherical and cylindrical values of the refractive status of each eye. The spherical equivalent (SE) of the eye was calculated using the formula spherical refraction (D) + (cylindrical refraction (D)/2). If SE 6 to 8 weeks after surgery was within ± 0.5D, we considered it an excellent outcome. If it was within ± 1.0D and UCVA after surgery was better than 6/18, we thought it an acceptable refractive outcome. If the post-operative deviation from predicted refraction was > ± 1.0 D with symptoms, it was considered a refraction surprise. [ref 16]
Posterior capsular rupture (PCR) included was with or without vitreous prolapse. [ref 17] The PCR risk is the predicted probability of PCR with or without vitreous loss vs composite Odds Ratio’ after adjusting for the patient's and surgeon's risk for PCR. The predicted refraction is defined as SE in the diopter of residual predicted refraction derived from the biometry sheet while selecting the IOL in the individual eye. [ref 18, 19]
We entered the data of the spreadsheet of the statistical package for Social Studies (SPSS 25) (IBM, NY, USA). We performed univariate analysis using the parametric method. The qualitative variables were presented as frequency and percentage. To compare the two variables in two groups, we used the Student’s t-test to calculate the two-sided P value. The continuous variables were plotted to study their distribution. Since they were skewed, we presented the median and interquartile range (IQR). To compare the outcomes in two groups, we used the Mann-Whitney U test to estimate the coefficient and two-sided P value. P value < 0.05 was considered statistically significant.