Lessons Learned: Wrong Intraocular Lens

Oliver D. Schein, MD, MPH,1 James T. Banta, MD,2 Teresa C. Chen, MD,3 Scott Pritzker, MBA, RN,1 Andrew P. Schachat, MD4
Objective: To report cases involving the placement of the wrong intraocular lens (IOL) at the time of cataract surgery where human error occurred. Design: Retrospective small case series, convenience sample. Participants: Seven surgical cases. Methods: Institutional review of errors committed and subsequent improvements to clinical protocols. Main Outcome Measures: Lessons learned and changes in procedures adapted. Results: The pathways to a wrong IOL are many but largely reect some combination of poor surgical team communication, transcription error, lack of preoperative clarity in surgical planning or failure to match the patient, and IOL calculation sheet with 2 unique identiers. Conclusions: Safety in surgery involving IOLs is enhanced both by strict procedures, such as an IOLspecic time-out, and the fostering of a surgical team culture in which all members are encouraged to voice questions and concerns. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2012;119:2059 2064 2012 by the American Academy of Ophthalmology.

During the past decade, there has been an increasing national focus on ensuring patient safety, fostering transparency, and stimulating initiatives to minimize errors in healthcare delivery. As a result, a proliferation of protocol, policy, and regulatory/compliance mandates have been instituted publicly and privately to address quality and safety issues. In particular, preventable causes of morbidity have been targeted, and examples of this include surgery involving the wrong patient, the wrong site, and the wrong procedure. In July of 2004, the Joint Commission on Accreditation of Healthcare Organizations promulgated a universal protocol in an effort to prevent such medical errors.1 This protocol provides a consistent approach involving preoperative verication, site marking, and a time-out immediately before incision. Cataract surgery is the most commonly performed surgical procedure, and virtually all involve placement of an intraocular lens (IOL). Wrong lens implants have repeatedly been documented to be the highest frequency medical errors in ophthalmic surgery. In a review of malpractice insurance claims and New York State Department of Health reports that were related to eye surgery between 1982 and 2005, wrong lens implants accounted for 63% of the cases.2 In a multispecialty review of incorrect surgical procedures in Veterans Health Administration Medical centers between 2001 and 2006, ophthalmic procedures had the highest rate of reported adverse events by surgical specialty (1.8 per 10 000 cases), and approximately half of the surgical errors were attributed to wrong implant.3 In a follow-up report on the same population from 2006 to 2009, ophthalmic surgery had the second highest rate (1.06 per 10 000 cases), with the majority of errors due to wrong implants.4 In 2009, Lum and Schachat5 wrote an editorial on the quest to eliminate wrong surgery, drawing attention to the Universal Protocol and its application to ophthalmic surgery. The American
2012 by the American Academy of Ophthalmology Published by Elsevier Inc.

Academy of Ophthalmology (AAO) established a wrong site task force that made its recommendations publicly available.6 This document emphasized standard processes for marking the eye and for the time-out. The importance of using 2 patient identiers to cross-reference the patient and the consent was emphasized. In regard to IOL surgery, the task force emphasized the need to double-check IOL measurements, to ensure clarity and completeness in the ordering of IOLs, and to have some form of written documentation regarding the intended IOL available during the surgery. Examples provided for doing this were writing the patients name, eye, and IOL power on a whiteboard or having a copy of the operative plan taped to the operating microscope. In a recent report from the United Kingdom,7 164 cases of wrong IOL implantation were enumerated via a national incident reporting database between 2003 and 2010. A reason for the error was reported for 102 of these cases. Approximately 30% were attributed to inaccurate biometry. The remainder was due to a variety of human errors, including transcription and ling errors, handwriting misinterpretation, and other patient or eye identication confusions. The authors recommended against transcription and the use of abbreviations and whiteboards, and favored electronic patient records, preoperative checklists, and an adequate stock of IOL ranges. The following cases illustrate how initial errors, even occurring before the day of surgery, can result in a cascade of events culminating in a wrong IOL implant. These cases occurred between 2006 and 2011, over a period of time in which there was considerable evolution nationally in practice related to universal protocols in general and cataract surgery specically. It is well recognized that there are many different reasons for IOL errors, including some that reect limitations in our measurements and equations in subsets of patients with very long or short eyes or after
ISSN 0161-6420/12/$see front matter http://dx.doi.org/10.1016/j.ophtha.2012.04.011

2059

Ophthalmology Volume 119, Number 10, October 2012

refractive surgery.8 The focus of this report, however, relates to IOL implantation that is wrong as the result of human or system error.
was limited to counting ngers. On further evaluation, the patient was found to have approximately 10 D of myopia. The chart was reviewed, and it was discovered that the second IOL calculation that had been requested with a target of 3 D of myopia contained the name of another patient with the same last name and an almost identical rst name. The patient was promptly informed of the error, and 24 hours later an IOL exchange was performed without complication, and the patient did well. An analysis of the case reached what had become an obvious conclusionthe IOL printout must be matched to the patient with 2 identiers, such as name and date of birth or medical record number.

Materials and Methods

An informal consortium of faculty responsible for quality and safety programs at their respective institutions was formed to discuss areas of common interest and concern. These faculty identied 7 cases related to the implantation of a wrong IOL that resulted in a formal review or root cause analysis (RCA). Although the exact format of the RCA varied by institution, all shared the use of a multidisciplinary team approach to document the event, explore multiple possible contributing causes, and outline specic plans to reduce the likelihood of recurrence. These cases, their review, and resulting changes in clinical policy were summarized.

Case 3
A patient with cataract and corneal astigmatism was scheduled for implantation of a toric IOL. An Alcon SN6AT6, 6.5 D lens (Alcon Inc, Hnenberg, Switzerland) was ordered. The packing slip conrmed the order was for an SN6AT6, 6.5 D lens. In the OR, during the time-out preceding implantation, the circulator read aloud the label on the IOL box, indicating an SN60AT, 6.5 D lens. The surgeon, after looking at the chart, indicated correct, and that IOL was implanted. Postoperatively, although the patient did well and was pleased with his vision, it was discovered that the patient had not received the intended toric IOL, but instead, the spherical SN60AT IOL. The patient was informed, and the additional charge for that IOL was refunded. Despite the satisfactory clinical outcome, a wrong implant had been used. On review of the case, several factors were recognized as contributory to this error. First, the similarity in lens model names facilitated the confusion. Second, it was recognized that it is preferable for the surgeon to repeat back to the circulator the desired IOL specications, not simply to respond yes, correct. Third, an alert on the OR schedule, indicating the plan for a nonstandard IOL (e.g., toric, multifocal) would be helpful.

Results
Case 1
A surgeon consulted the printed IOL calculation sheet in the chart just before patient transport into the operating room (OR). The surgeon reviewed the calculations, circled the desired IOL, and brought that IOL into the OR. Uncomplicated surgery was performed, the circulating nurse conrmed that the IOL selected matched the one circled on the printout (correct brand, model, and power), and the patient was transported to the recovery room. While completing the postoperative paperwork, the surgeon noted that the patients chart had 2 separate IOL calculation printouts, and the one from which the patients IOL was chosen was the incorrect one. It belonged to another patient and had been misled. The surgeon immediately informed the patient of the error, brought the patient back to the OR, and safely exchanged the IOL for the correct power. The patient did well postoperatively. An RCA was undertaken that concluded the error had largely occurred because the IOL printout had not been matched to the patient by name. A policy to do so was then instituted.

Case 4
A patient underwent uncomplicated cataract surgery and was found to have blurry vision in the operated eye postoperatively. Manifest refraction revealed approximately 2.5 D of hyperopia. Investigation revealed that the surgeon had mistakenly chosen the power of the lens from the column designated for the anterior chamber IOL (ACIOL) when a posterior chamber IOL (PCIOL) had been intended. Further investigation revealed that the position of the PCIOL and ACIOL lens power options on the IOL measurement printout had been switched from their traditional positions on the IOL printout, although both columns were clearly marked. The patient underwent an uneventful IOL exchange and did well. An RCA was performed that led to the institution of a new IOL-specic time-out before insertion of the IOL during surgery. The IOL calculation sheet is now matched to the patient by name and date of birth by the circulating nurse. The surgeon is then required to review the IOL calculation sheet and conrm the IOL style and power to be implanted before the implant is opened for the scrub technician to load.

Case 2
A patient with a common last name presented to the clinic for consideration of cataract surgery in his right eye. He was a high myope who had previously undergone cataract surgery in the left eye elsewhere and had purposely been left with approximately 4 diopters (D) of myopia in that eye. The surgeon received the IOL calculations back from the ophthalmic technician soon after the patient had left the clinic. These calculations had been centered on a target of plano. The surgeon recalled that the intended target for the right eye was approximately 3 D of myopia and asked another technician to repeat the calculations with a target of 3 D. The technician went directly to the IOL Master (Carl Zeiss, Inc, Oberkochen, Germany), pulled the name off the pick list provided by that instruments software, performed the calculation, and placed the printed output in the chart. Several weeks later, the patient underwent uncomplicated cataract surgery, and the IOL power was veried by a clinical fellow and a circulating nurse, both of whom compared the name on the patients consent form with that on the IOL calculation sheet. In addition, just before insertion of the IOL, the circulating nurse read aloud the patients name and the IOL selected from the IOL printout, and the attending surgeon afrmed his agreement. On the rst postoperative day, the patients eye was found to be anatomically intact, but the vision

Case 5
A patient was scheduled for cataract surgery. The surgeon requested a 14.5 D PCIOL on an IOL order form. The order form was faxed to the ambulatory surgical center, and the nurse who pulled the lens was unsure whether the note read 14.5 or 17.5 D. She guessed it was 17.5 D but never mentioned her uncertainty to the surgeon or any other staff member until questioned later. During the time-out before lens implantation, the nurse read the lens order aloud, calling for a 17.5 D lens, and the surgeon

2060

Schein et al Lessons Learned: Wrong IOL

conrmed by looking at the IOL packaging (not at the original order form) that this was correct. On dictating the case, the surgeon noted that the IOL power used was signicantly different from that previously implanted in the fellow eye. On review of the chart, he saw that a 17.5 D lens had been implanted instead of the intended 14.5 D lens. The patient was notied immediately, and the decision was made to wait and see if she tolerated the unintended monovision. In fact, the anisometropia was minimal, and the patient was happy with her vision postoperatively, declining further intervention. An RCA was performed and led to a change in the IOLspecic time-out before insertion of the IOL during surgery. Subsequent to this case, the IOL policy was amended to require that the IOL calculation sheet be matched to the patient by name and date of birth by the circulating nurse. The surgeon is then required to review the IOL calculation sheet and conrm the IOL style and power to be implanted before the implant is opened for the scrub technician to load.

Discussion
In the United States, more than 3 million cataract surgeries are performed annually, and almost all involve the selection and placement of an IOL. With such an enormous denominator, it is not surprising that IOL cataract surgery is associated with the greatest number of surgical confusions in ophthalmology2 or that ophthalmology has been disproportionately represented in raw counts of adverse events occurring in ORs.3 There are more opportunities to choose a wrong IOL than there are to make analogous errors for any other implanted device. This is a simple function of the volume of cases performed and the large number of potential IOL choices that reect the range of dioptric powers and lens designs available. The advent of multifocal and toric IOLs has both increased the options we can offer patients and our opportunity for error. Other commonly implanted devices (e.g., orthopedic, vascular stents) are produced in far fewer versions. The true rate of wrong IOL implantation is unknown. Because of differences in case denition, ascertainment, and time surveyed, we do not have an accurate pooled rate of wrong IOL implantation across our institutions. However, if the rate of wrong IOL due to error in selection were on the order of 1 per 10 000 (99.99% accuracy), as implied by Neily et al,3 this would yield approximately 200 such cases annually in the United States. The institutions represented by the authors of this article have made signicant efforts during the past 5 years to minimize such errors. Although none of the IOL errors reported in this article resulted in permanent patient harm, that potential exists. Furthermore, IOL errors undermine patient condence in the surgeon or institution and may be a source of considerable distress for both patient and surgeon, not to mention a potential source of litigation. The regulatory requirements for reporting IOL errors vary signicantly by state and may also vary on the basis of the practice setting in which the IOL surgery occurred (e.g., whether the surgery was performed in a hospital OR or a physician ofce setting). In some states, the implantation of a wrong IOL is considered wrong site surgery, no different than if the wrong eye, limb, or lung had been operated. Such states require reporting of IOL errors to various regulatory agencies, which may result in a range of consequences to the reporting institution or surgeon, even including nes and community service. Just as for the reporting requirements, the potential consequences of reporting IOL errors are state-specic and vary tremendously. The errors, and often information about steps taken to reduce the chance of recurrence, were disclosed by the surgeon to each patient included in this report, in keeping with the disclosure and apology movement in risk management. Cases 1 and 2 both involved having the incorrect IOL printout sheet present in the OR. Both cases would likely have been prevented with time-out procedures that match the IOL printout to the consent form with 2 patient identiers, such as name and date of birth. The AAO task force stipulates the use of 2 patient identiers to prevent wrong site or wrong patient surgery but does not specically indicate that the IOL printout itself be reconciled to the patient

Case 6
In preparation for surgery, a surgeon ordered 4 separate IOLs. Before the patient was transported into the OR, the surgeon selected a monofocal spherical IOL and placed that IOL on a separated stand in the OR. The other 3 IOLs that had been ordered, one of which was a toric IOL, were placed on a shelf in the OR. Uncomplicated surgery was performed, and the patient was transported to the recovery room. When examining the patient on postoperative day 1, the surgeon noticed that a toric IOL had been implanted instead of the intended spherical IOL. The surgeon immediately informed the patient of the error, and the patient underwent uncomplicated surgery to rotate the toric lens into the optimal position. The patient did well postoperatively. An RCA was undertaken that concluded the error had occurred because there was no verication between the nurse and the surgeon regarding which was the intended IOL. The circulating nurse had taken the toric IOL from the shelf, unaware that the surgeon had already placed the intended IOL on a separate stand. A new process was then instituted requiring the surgeon and nurse to check IOL printouts together and mutually conrm the intended IOL before the surgery. That process also requires, once the preferred IOL has been chosen, that all other IOLs are removed from the OR.

Case 7
Just before a patients cataract surgery, the circulating nurse reviewed the IOL that had been delivered to the OR for that patient. She noticed that a 5 D lens had been delivered to the OR, whereas the surgeon had ordered a 5 D lens. When the nurse veried with the surgeon that the 5 D lens was indeed the intended lens, the correct 5 D was brought to the OR. Uncomplicated surgery was performed, and the correct IOL was implanted. An RCA, undertaken to investigate this near miss, concluded that the wrong IOL had been delivered to the OR because the staff member selecting the IOL had misread the IOL boxs label. The minus sign in front of the 5 D power was in a smaller font than used to specify the IOL power and had been overlooked. A letter was then written to the company that produces and packages the IOL chosen, suggesting a modication to the label that would highlight the presence of the minus sign by changing the font or color. The company responded afrmatively and subsequently improved the packaging in that regard.

2061

Ophthalmology Volume 119, Number 10, October 2012

or patient consent form using the same 2 identier standard. We highly recommend this practice. In addition, it would be helpful if 2 patient identiers were required to access a patients IOL data from the IOL calculation database. This would reduce the possibility of printing out the incorrect patients data, especially in cases where there are similar or even identical names involved. Case 3 would likely have been prevented if the surgeon had taken a greater role in conrming the IOL before initiation of the surgery. If the surgeon were consciously planning for a toric IOL, the axis of astigmatism would need to have been marked and the IOL axis aligned appropriately. The AAO checklist recommending that the written surgical plan be visible to the surgeon intraoperatively would have been helpful. Case 4 highlights the importance of having at least 2 people, typically the surgeon and circulating nurse, review the IOL printout form and the intended IOL together. The power of the ACIOLs and PCIOLs was confused because this step was not taken, and the likelihood of surgeon error was compounded because of the inadvertent change in the routine format for the IOL printout at that institution. Case 5 illustrates the potential consequences of a transcription error. In this case, a 4 was mistaken for a 7, and the potential for error was compounded by the reluctance of the nurse to disclose her uncertainty. A goal of all surgical teams should be to create an environment where all members of the team are comfortable in voicing uncertainty and contributing to a culture of safety. In regard to transcription errors, these can be best minimized by avoiding transcription whenever possible. Transcription occurs most frequently in settings where surgeons or their surrogates transcribe onto order forms a desired IOL. An alternative is to circle and initial the intended IOL on a copy of the IOL printout. Transcription occurs by necessity when the axial length is measured using an ultrasound technique and the value then manually entered into an IOL calculator. Likewise, transcription of keratometric values is usually necessary to perform the calculations needed to select a toric IOL. In such settings where transcription is practiced, the surgeon and staff should make efforts to conrm that the transcription was accurately performed. Case 6 is another example of failure to involve a direct communication between the surgeon and the circulating nurse regarding a conrmation of the intended IOL. It also highlights the enhanced potential for error when multiple IOLs are in the OR simultaneously. Finally, case 7 was a near miss. A 10 D error in IOL choice was averted by an alert nurse, again demonstrating the notion that a surgeon and nurse team trump either alone. This case also illustrates, as does case 3, that packaging and labeling are important. This has long been recognized as a potential problem in other settings. For example, many ophthalmologists are aware of patients who have confused their articial tears or contact lens lubricants with products designed for ear care or even with adhesives such as superglue. Industry can play a greater role in promoting safety by making greater and more logical distinctions between the model numbers of its IOL designs. The cases reported occurred at 4 large ophthalmic training institutions over a 5-year period. Each institution faces a common challenge that is also shared by many others: how to institutionalize a culture of safety that will apply equally to part- and full-time faculty and to residents and fellows in training. That challenge is in turn magnied by the relative rarity of errors and the multiplicity of paths that may lead to the error. Although not an exhaustive list, it is critical to recognize the multitude of potential causes for IOL selection errors, such as the following: 1. An IOL calculation sheet for a different patient with a similar or same name is in the medical record. 2. The previous patients IOL is inserted. 3. The IOL power for the wrong eye is inserted. 4. The wrong IOL A-constant or formula is used in IOL calculations. 5. The surgeon misreads intended IOL power (e.g., 28.0 instead of 23.0). 6. The power for an ACIOL is selected instead of the intended PCIOL power. 7. The wrong IOL model is picked from IOL calculation sheet. 8. The axial length is confused with the IOL power on the printout. 9. The wrong IOL is chosen when multiple IOL types are present in the OR. 10. A minus is confused with a plus in choosing the target refraction or IOL power. 11. A transcription error is made when transferring data for keratometry or axial length data into IOL calculation software. 12. The patient specically requests myopia or monovision, but the surgeon targets emmetropia. 13. The patient requested (and paid for) a different type of premium IOL than implanted. 14. The patient did not want a toric or presbyopiacorrecting IOL, but one was implanted, or vice versa. 15. The requested special-order IOL was not available in the OR after lens extraction. Although it is acknowledged that the most critical moments in preventing IOL error occur in the OR, it has become equally apparent that the path to IOL error often begins earlier. Errors, once committed, may be propagated downstream and may be more difcult to detect than to prevent in the rst instance. Quality-control efforts must begin at the time of initial measurement and decision for surgery. Because of some variability in practice and patient ow in clinics, preoperative holding areas, and stafng, there is not a single, rigid plan that is optimal for every setting. However, we were able to identify a set of common elements that we believe will minimize IOL errors. These may be summarized as follows:

A surgical plan regarding the type of IOL (e.g., spherical, toric, presbyopia correcting) and general refractive target (e.g., better for distance or near) should be documented in the medical record. The intended IOL, in particular any special-order IOL, should be veried to be present in the OR before the patient is taken to the OR.

2062

Schein et al Lessons Learned: Wrong IOL

A patient label that contains name, medical record number, and date of birth is present on every IOL calculation printout. Each technician performing IOL calculations should use 2 patient identiers (name and either date of birth or medical history number) to match the name on the IOL calculation printout to that of the specic patient. If additional calculations on a patient are subsequently requested, 2 patient identiers are used to conrm that the correct patient has been accessed from the IOL database. If the difference in axial length between the 2 eyes is 0.3 mm, we recommend that this difference be reconciled clinically by the surgeon or the measurement repeated. If the axial length is measured by ultrasonography, or corneal power measured manually and then transcribed into an IOL calculation software, the data transcribed should be subsequently conrmed by a technician or the surgeon. The use of IOL order forms that require manual transcription from the IOL printout should be minimized. If the IOL calculations are missing on the day of surgery, they should be transmitted to the OR properly labeled with name, date of birth, and medical record number before the patient enters the OR. IOL verication on the day of surgery: The circulating nurse reviews the patient chart with the surgeon and conrms that the IOL calculation sheet matches the patient by name and date of birth or medical record number and that the IOL model and diopter power circled on the printout sheet and signed or initialed by the surgeon has been brought into the OR. Only the IOL for that one patient is brought into the OR. The time-out in the OR that conrms the correct patient, procedure, and site occurs before the rst incision. For patients who will be receiving an IOL, all 4 institutions participating in this review concur that an IOL-specic time-out is a necessity. However, there is variability across the institutions as to when this should take place. At one of our institutions, the IOL verication occurs as an integral component of the initial time-out before the rst incision, and the intended IOL in its sterile package is typically placed on the Mayo stand in real time. That institutions logic for its timing is that if a concern regarding the IOL is to be discovered, it would rather see that conict resolved before the incision, rather than have to do so in the middle of surgery. At the other institutions, a separate IOL time-out is held just before the insertion, with the logic that the closer in time the IOL time-out is to the actual insertion, the better. For all 4 institutions, the components of the verication include a matching (visual and auditory) of the patients name and date of birth (or medical record number) as recorded on the consent form with that on the label of the IOL printout. If a change in IOL is requested intraoperatively, the surgeon and circulating nurse will repeat the IOL verication procedure.

Only the IOL for the patient currently undergoing surgery is present in the OR; any unused IOL is removed from the OR after each case.

Reducing the rate of IOL errors requires a team approach that involves the efforts and concentration of ophthalmic technicians, surgical schedulers, operative nursing staff, and surgeons. In the future, human error might be further reduced by the introduction of electronic systems, such as bar coding or the use of radio frequency identication,9 which might even be linked to patient specic biometry and electronic medical records. Of course, the use of electronic records may pose other challenges yet to be fully appreciated. For example, an error once in an electronic system can easily be propagated downstream. At present, there is growing evidence that the use of surgical safety checklists10,11 may improve surgical quality and safety. In the setting of cataract surgery, such a checklist should contain detailed and specic information related to the intended IOL. The last moment before IOL implantation in the OR is crucial, but often is not the optimal time to prevent errors whose roots begin elsewhere, such as transcription errors occurring in clinic or at scheduling. Improving processes and their accuracy before the day of surgery will reduce errors. However, the nal verication steps on the day of surgery will always remain paramount as the last, if not the best, way to prevent the implantation of a wrong IOL. In conclusion, the risk of implanting the wrong IOL as the result of human or system error has decreased over recent years. The AAOs 2008 operative checklist was a major milestone in this evolution. We have tried to offer some additional renements to such processes, recognizing and hoping that these too will be supplanted by still more effective mechanisms in the future.

References
1. Joint Commission on Accreditation of Healthcare Organizations. Universal protocol for preventing wrong site, wrong procedure, wrong person surgery [JCAHO News Release]. July 21, 2003. Available at: www.jcaho.org/accredited organizations/patientsafety/universalprotocol/wss_universal protocol.htm. Accessed February 24, 2012. 2. Simon JW, Ngo Y, Khan S, Strogatz D. Surgical confusions in ophthalmology. Arch Ophthalmol 2007;125:151522. 3. Neily J, Mills PD, Eldridge N, et al. Incorrect surgical procedures within and outside of the operating room. Arch Surg 2009;144:1028 34. 4. Neily J, Mills PD, Eldridge N, et al. Incorrect surgical procedures within and outside of the operating room: a follow-up report. Arch Surg 2011;146:12359. 5. Lum F, Schachat AP. The quest to eliminate never events. Ophthalmology 2009;116:10212. 6. American Academy of Ophthalmology. Patient Safety. Recommendations of American Academy of Ophthalmology Wrong-Site Task Force. Available at: http://one.aao.org/CE/ PracticeGuidelines/Patient_Content.aspx?cidd0db838c-28474535-baca-aebab3011217. Accessed February 24, 2012.

2063

Ophthalmology Volume 119, Number 10, October 2012

7. Kelly SP, Jalil A. Wrong intraocular lens implant; learning from reported patient safety incidents. Eye (Lond) 2011;25: 730 4. 8. Jin GJ, Crandall AS, Jones JJ. Intraocular lens exchange due to incorrect lens power. Ophthalmology 2007;114:41724. 9. Fraser SG, Adams W. Wrong site surgery. Br J Ophthalmol 2006;90:814 6. 10. Conley DM, Singer SJ, Edmondson L, et al. Effective surgical safety checklist implementation. J Am Coll Surg 2011;212: 8739. 11. Semel ME, Resch S, Haynes AB, et al. Adopting a surgical safety checklist could save money and improve the quality of care in U.S. hospitals. Health Aff (Millwood) 2010;29: 15939.

Footnotes and Financial Disclosures

Originally received: March 26, 2012. Final revision: April 12, 2012. Accepted: April 13, 2012. Available online: June 14, 2012.
1 3

Massachusetts Eye and Ear Inrmary, Harvard Medical School, Boston, Massachusetts. Cole Eye Institute, Cleveland Clinic, Cleveland, Ohio. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Oliver D. Schein, MD, MPH, 600 N. Wolfe Street, Wilmer 116, Baltimore, MD 21287. E-mail: oschein@jhmi.edu.

Manuscript no. 2012-440.

Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, Maryland. Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Florida.

2064