Eyeworld

EW RESIDENTS 92 October 2011 during the study period. In addition, in an attempt to ensure the validity of the reported complication rate, the authors investigated the possibil- ity of underreporting, recognizing a limitation to any self-reported out- come study. By capturing a large, representative sample and perform- ing additional analysis to validate the data, the authors attempt to pro- vide a sense of the true complication rate so as to minimize the bias of underreporting. The study confirms previously recognized risk factors for capsule complications, such as poor pre-op corrected visual acuity, diabetic retinopathy, and glaucoma. It also identifies new potential risk factors, such as first eye surgery, and sug- gests that increasing age, which has been associated in some but not all studies with worse outcomes, is in fact a risk factor. Identifying such risks may allow pre- or intraopera- tive adjustments to be made in an attempt to prevent complications or at least permit a more thorough in- formed consent process. However, there are limitations to the study as well. First, in their validation sample, the authors found an underreporting rate of 32%, which was evenly distributed across the study years analyzed (2002-2006). This underreporting rate can be used to adjust the cap- sule complication rate for the popu- lation as a whole, but it is difficult to know how it factors into the evalua- tion of trends and risk factors. Un- derreporting, although shown to be distributed evenly across time, may differ within important variables such as clinic volume, university setting, or other unmeasured con- founders. For example, if a disproportionate amount of underre- porting came from high volume cen- ters, then the association between high volume centers and lower risks of complications may be spurious. We also feel that the validation could have been strengthened by in- cluding the time frame from 2007 to 2009, particularly because the logis- tic regression analysis for the en- tirety of the data includes those years. Disproportionate underreport- ing from those years may signifi- cantly alter the overall trends and conclusions. There are also a number of limi- tations to the specific inferences that can be drawn from the data, which the authors acknowledge. Based on the data, the authors conclude that the capsule complication rate has been decreasing over time. While their data do suggest this trend, there is no way to prove that the de- creased rate is due to any of the hy- pothesized factors, including increased surgical experience for spe- cific surgeons. From the data pro- vided, it is impossible to know whether the findings reflect greater experience on the part of individual surgeons performing more surgeries, improved surgical techniques, tech- nical refinements in equipment, or some combination of these factors. The authors also propose that the complication rate has decreased over time as a result of operating on younger and presumably healthier eyes. Again, however, there are no data to support the assumption that younger patients have healthier eyes, particularly given changing population demographic risk factors over time such as the increasing prevalence of diabetes in younger patients. Furthermore, while the mean age in patients undergoing surgery decreased by 1.29 years over the course of the study, the clinical significance of this finding is un- clear. From our perspective, one of the most intriguing findings is the asso- ciation between high volume centers and lower complication rates (Figure 1). It would have been interesting to include this variable (center volume) in the multivariate analysis, espe- cially as the authors note a trend of more centers performing higher vol- ume surgery over time. Is it therefore possible that the decreased rate of complications over time is due to more surgeries being done at high volume centers? It is well recognized in a number of surgical subspecial- ties that negative outcomes and cen- ter volume are inversely related, due to a variety of reasons, such as more experienced staff and surgeons, bet- ter equipment, etc. We hope that the authors continue to explore this finding in future analyses. Finally, this study raises addi- tional interesting findings regarding the difference in complication rate between university and non-univer- sity clinics, which provokes several questions. Do trainees have more complications? Are trainees less likely to underreport? Are university clinics high volume or low volume? In teaching future generations of surgeons, it would be important to know if in fact this difference is real and if so, whether it is a result of the training aspect of university clinics. Though the authors report that they controlled for ocular comorbidity when comparing university and non-university clinics, none of this analysis was presented, which limits our ability to evaluate the conclu- sions. In summary, this article gives valuable insight into capsule compli- cations in a large population. It re- views well the risk factors for capsular complications during cataract surgery and adds to the body of literature describing previ- ously known risk factors. This study also highlights the NCR as a power- ful tool for answering interesting and important clinical questions. If more such national health registries existed and were able to combine prospective data as Sweden has done for the last decade, we would be able to even more robustly address some of the compelling questions that the authors have raised. EW Contact information Dunn: jpdunn@jhmi.edu Wilmer continued from page 91 89-93 Resident_EW October 2011-DL32_Layout 1 9/29/11 4:35 PM Page 92

• Cover