EyeWorld is the official news magazine of the American Society of Cataract & Refractive Surgery.
Issue link: https://digital.eyeworld.org/i/78719
92 EW RESIDENTS February 2012 EyeWorld journal club University of Iowa residents' review of "Risk factors for complications in by Shaival Shah and Angela McAllister, University of Iowa residents Thomas A. Oetting, M.D. Director, ophthalmology residency program University of Iowa, Iowa City This month, I asked the University of Iowa residents to review this important study on resident surgical complications from the February JCRS issue. —David F. Chang, M.D., chief medical editor V itreous complications re- main among the most feared for the beginning surgeon. The ability to ac- curately stratify patients pre-operatively and assign them to residents based on skill level is a log- Preston H. Blomquist, M.D., author of the paper, made a surprise visit to Iowa as the residents reviewed his paper during their journal club Source: Thomas A. Oetting, M.D. ical strategy, meant to reduce the likelihood of intraoperative compli- cations and hopefully lead to better outcomes with happier patients, Risk factors for vitreous complications in resident-performed phacoemulsification surgery Preston H. Blomquist, M.D., Marlene E. Morales, M.D., Liyue Tong, M.S., Chul Ahn, Ph.D. J Cataract Refract Surg (February) 2012; 38:208-214 Purpose: To identify risk factors for intraoperative vitreous complications in resident-performed phacoemulsification surgery Setting: Two urban public county hospitals Design: Case series Methods: Phacoemulsification cataract surgeries performed by resi- dents between January 4, 2005, and January 8, 2008, were retrospec- tively reviewed. Clinical characteristics of patients with and without intraoperative vitreous complications were compared and independent factors associated with vitreous complications identified using univari- ate and multivariate analysis. Results: Of 2,434 cases meeting inclusion criteria, there were 92 vitre- ous complications (3.8%). On univariate analysis, significant pre-oper- ative risk factors for vitreous complications included older age (P=.020), poor pre-operative corrected distance visual acuity (CDVA) (P=.007), left eye (P=.043), history of trauma (P=.045), prior pars plana vitrectomy (P=.034), dementia (P=.020), phacodonesis (P=.014), zonule dehiscence (P<.0001), posterior polar cataract (P=.037), white/mature cataract (P=.005), dense nuclear sclerotic cataract (P=.0006), and poor red reflex (P=.002). Factors that remained significant on multivariate analysis were older age (odds ratio [OR], 1.03; 95% confidence interval [CI], 1.01-1.05), poor pre-operative CDVA (OR, 1.52; 95% CI, 1.14- 2.03), left eye (OR, 1.63; 95% CI, 1.05-2.51), prior pars plana vitrec- tomy (OR, 1.88; 95% CI, 1.01-3.51), dementia (OR, 3.65; 95% CI, 1.20-11.17), and zonule dehiscence (OR, 8.55; 95% CI, 3.92-18.63). Conclusion: Elements of the pre-operative history and examination can identify patients at higher risk for intraoperative complications during resident-performed phacoemulsification surgery. Financial disclosure: No author has a financial or proprietary interest in any material or method mentioned. residents, and supervising attending physicians. Preston Blomquist, M.D., and his colleagues addressed this very important issue in their re- cent publication in the Journal of Cataract and Refractive Surgery.1 In their study, they retrospec- tively reviewed 2,434 cases, the largest study to date on this subject. For each case, they collected over 25 pre-op characteristics that could po- tentially have predisposed to vitre- ous loss, ranging from the training level of the resident surgeon to the presence of weak zonules. They ran a univariate analysis to show which factors seemed to be associated with vitreous loss. Most importantly, they ran a multivariate logistic regression analysis, in which they assessed the unique influence of each independ- ent variable (pre-op risk factors) on the dependent variable (vitreous loss) while controlling for all other variables. In this study, odds ratio was used to approximate relative risk for all of the variables studied. From this data, six independent risk fac- tors were found to be associated with vitreous complications during phacoemulsification surgery. Some of their findings were not surprising and confirmed what is al- ready commonly accepted, i.e., zonular dehiscence increases the risk of vitreous loss during surgery. How- ever, several of their variables were interesting and surprising, such as dementia and left eyes increasing the risk of complications. Both of these characteristics can be ex- plained and rationalized but were not generally thought to be signifi- cant risk factors for vitreous loss. There are several variables (such as mature cataracts) that have been shown to increase the tendency for vitreous loss that were not found to be significant in this study. This could be due to a faulty data set (i.e., beginning resident performing pre- op examination and failing to ask or document correctly), or it could mean that increased awareness and improved surgical tools really are improving outcomes in cases such as mature cataracts and small pupils. Most residents in ophthalmol- ogy have written and published articles by the time they get to resi- dency, and virtually all will have done so by the time they complete their training. An equally important skill set is the ability to critically analyze an article and determine whether to believe it and how to in- corporate it into practice patterns. As part of this learning exercise, our program director, Thomas Oetting, M.D., managed to persuade the lead author, Dr. Blomquist, to fly to Iowa City and surprise the residents dur- ing our journal club and join us in a discussion of his article. As a group, we had several concerns about the article, and Dr. Blomquist was very candid in ex- plaining the creative process leading to this study design. Many of the limitations of the study have been noted and acknowledged by the au- thors in their article. These include selection bias, the retrospective na- ture of the study, and the lack of randomization. We found it surprising that only 19 of the patients had a history of tamsulosin and questioned whether this implied a faulty data set. While not mentioned in the article, Dr. Blomquist informed us that it is not the main drug on formulary at the hospitals used in the study. Another concern was the study design and the way the data was gathered. We recognize that a prospective, randomized, controlled trial would not be practical and probably unethical. Nonetheless, retrospective studies rely on accurate chart information, and with so many different variables in the study it is unlikely that every resident in- quired about every one of them for each patient encounter. In dis- cussing this with Dr. Blomquist, it came to light that the study design was largely an IRB issue, and data collection was done by residents at the time of surgery with pre-printed