AIMS: To identify and quantify risk factors for posterior capsule rupture or vitreous loss or both (PCR or VL or both) during cataract surgery and provide a method of composite risk assessment for individual operations. METHODS: The Cataract National Dataset was extracted on 55,567 operations from 12 National Health Service (NHS) Trusts using an electronic patient record (EPR) system between November 2001 and July 2006. Risk indicators for variations in the rate of 'PCR or VL or both' were identified by univariate and multivariate analyses. Adjusted odds ratios (ORs) were used to formulate a composite 'bespoke' risk for individual cases. RESULTS: Overall 'PCR or VL or both' rate was 1.92% (95% CI=1.81-2.04%). Risk indicators for this complication were increasing age, male gender, presence of glaucoma, diabetic retinopathy, brunescent/white cataract, no fundal view/vitreous opacities, pseudo-exfoliation/phacodonesis, reducing pupil size, axial length > or = 26.0 mm, the use of the alpha-blocker doxazosin, inability to lie flat and trainee surgeons performing operations. Adjusted ORs for these variables are used to estimate overall composite risk across multiple risk indicators in the form of a predicted probability of PCR or VL or both. Predicted probability for this complication ranged from less than 0.75% to more than 75%, depending on risk profile of individual operations. CONCLUSIONS: Higher-risk cases can be predicted, thus better informing the consent process and allowing surgeons to take appropriate precautions. Case-mix is a major determinant of the probability of an intraoperative complication. A simple composite risk estimation system has been developed.
AIMS: To identify and quantify risk factors for posterior capsule rupture or vitreous loss or both (PCR or VL or both) during cataract surgery and provide a method of composite risk assessment for individual operations. METHODS: The Cataract National Dataset was extracted on 55,567 operations from 12 National Health Service (NHS) Trusts using an electronic patient record (EPR) system between November 2001 and July 2006. Risk indicators for variations in the rate of 'PCR or VL or both' were identified by univariate and multivariate analyses. Adjusted odds ratios (ORs) were used to formulate a composite 'bespoke' risk for individual cases. RESULTS: Overall 'PCR or VL or both' rate was 1.92% (95% CI=1.81-2.04%). Risk indicators for this complication were increasing age, male gender, presence of glaucoma, diabetic retinopathy, brunescent/white cataract, no fundal view/vitreous opacities, pseudo-exfoliation/phacodonesis, reducing pupil size, axial length > or = 26.0 mm, the use of the alpha-blocker doxazosin, inability to lie flat and trainee surgeons performing operations. Adjusted ORs for these variables are used to estimate overall composite risk across multiple risk indicators in the form of a predicted probability of PCR or VL or both. Predicted probability for this complication ranged from less than 0.75% to more than 75%, depending on risk profile of individual operations. CONCLUSIONS: Higher-risk cases can be predicted, thus better informing the consent process and allowing surgeons to take appropriate precautions. Case-mix is a major determinant of the probability of an intraoperative complication. A simple composite risk estimation system has been developed.