First Author: J.Seddon UNITED KINGDOM
Co Author(s): C. Vize
"
Purpose:
Cataract surgery refractive outcomes were reviewed in order to generate departmental benchmarks and assess performance against Royal College of Ophthalmologists 2010 Cataract Surgery Guidelines. To improve performance in the future, we attempted to identify those variables significantly influencing postoperative refraction accuracy.
Setting:
: Hull and East Yorkshire Eye Hospital, UK
Methods:
Data from a consecutive case series of cataract procedures was collected from 2004 to 2010.
Complete preoperative and postoperative data were available for 15428 of 20954 procedures (74%).
The difference between the predicted and achieved spherical equivalent (SE), and the percentage of procedures achieving within ±0.50 diopter (D) and ±1.00 D of planned refraction were calculated. The outcome variables of year, operating surgeon, IOL model and IOL power formula were assessed for their influence upon postoperative refraction using Chi-square and logistic regression analysis.
Results:
Between 2004 and 2010, the mean departmental prediction error (PE) (difference between planned and achieved SE) decreased significantly from -0.139 (SD±0.69) to 0.057 (SD±0.54) (P<.0001). The percentage of eyes falling within 1D of planned refraction increased significantly from 90.5% in 2004 to 95.0% in 2010 (p<0.001). The percentage of eyes falling within 0.5D of planned refraction also increased significantly from 64.7% in 2004 to 71.5% in 2010 (p<0.001). The operating surgeon was a significant influence upon the percentage of procedures falling within ±0.5D and ± 1D of planned refraction (p<0.001). IOL model became a significant influence when other variables under analysis were controlled for with logistic regression (within 1D of target p=0.046, within 0.5D of target p=0.021). The influence of IOL power formula was significant when analysed alone against the percentage of eyes hitting within 0.5D and 1D of target (p<0.001), but the formulas’ predictive abilities did not significantly differ when examined with logistic regression (within 1D of target p=0.351, within 0.5D of target p=0.279).
Conclusions:
Both PE and percentage of procedures falling within 0.5D and 1D of target has followed an improving trend over the 7 years studied and performance for all years surpasses Royal College benchmark standards. We agree with other authors that there is room to raise this benchmark further still. The element introducing the highest variability to postoperative refraction in our study was the operating surgeon; IOL model was also a significant influence. It is important to identify and address the multiple factors responsible for error in order to facilitate a continual improvement in refractive outcomes.
Financial Disclosure:
No"
