Reply to "reducing endophthalmitis in India: an example of the importance of critical appraisal".

Dear Editor,

We thank Dr. Thomas, for his interest[1] in our article published in Journal of Cataract and Refractive Surgery.[2] We regret that by choosing the Indian Journal of Ophthalmology to comment about our paper published in a different peer reviewed journal, those reading his letter may not have had access to our original paper. This makes some of the puzzling misrepresentations of our conclusions in his letter all the more troubling.

Dr. Thomas poses a number of criticisms regarding our data reporting, which we will respond to. However, one basic implication of his letter seriously misrepresents our intent as being to recommend or establish new standards for global surgical (operating room) and sterilization protocols. In fact, our discussion does not suggest this at all. Our paper reviews the unique mission and patient population of the Aravind Eye Care System, in which the majority of patients undergoing cataract surgery are unable to pay for care. Our mission of providing high quality care in the most cost effective manner to as many underprivileged patients as possible, motivates us to examine potentially unnecessary operating room practices, such as requiring the long autoclave sterilization cycle for every case. This retrospective study illustrates our determination to continuously monitor whether modifying these practices has any deleterious effect on outcomes, such as postoperative endophthalmitis (POE) rates. At no point in this article do we recommend that all international eye surgical facilities adopt practices that we ourselves use.

Dr. Thomas begins his letter by implying that the national endophthalmitis rate in India is 0.6%, and with respect to our reported endophthalmitis rate of 0.09% misstates that the paper “attributes this impressive reduction to their unconventional surgical protocols.” We are puzzled by this misstatement because at no point in our paper do we claim to have a superior surgical protocol; nor do we claim to have superior outcomes. Instead, we stand by the conclusion of our abstract, which states that “the rate of endophthalmitis in this generally underserved patient population with multiple risk factors for infection was comparable to that reported in other modern settings.”

To support using 0.6% as a representative endophthalmitis rate in India, Dr. Thomas references a national survey on visual outcomes after cataract surgery covering 15 randomly selected districts during the period 1998–2002.[3] In that report, 69.5% of the surgeries were performed by intracapsular cataract extraction (ICCE) and 26.5% of these surgeries were done in makeshift camps. Among the causes of poor vision (best corrected vision < 3/60) reported in this survey, 19.9% were due to postoperative complications, 8.9% had phthisis, and 5.4% apparently had enucleation.[3] We question why Dr. Thomas would choose this survey to provide a contemporary estimate of endophthalmitis prevalence in our country.

Dr. Thomas goes on to question the general validity of drawing any conclusions from retrospective studies. To our knowledge, there are only two other papers on the incidence of post-cataract endophthalmitis in similar sized patient populations in India. Jambulingam reported the incidence to be 0.053% over an 8 year period. This prospective study only reports culture-positive cases and may therefore underestimate the true incidence of endophthalmitis.[4] Lalitha reported a 0.05% POE incidence among 36,072 private cataract patients.[5] A meta-analysis looking at the incidence of cataract POE worldwide notes that most studies are retrospective and only case-controlled studies looking at different intervention options were prospective in nature.[6] In this context, we believe that our retrospective study of more than 42,000 consecutive cataract surgeries from a single eye hospital contributes useful information regarding POE rates to the ophthalmic literature, particularly because standardized instrument cleaning and sterilization protocols were used for all patients. Interestingly, prior to 2009, there were no studies in the ophthalmic literature reporting on POE rates when using short cycle steam sterilization. In its discussions with American regulatory agencies, our paper was cited by the American Society of Cataract and Refractive Surgery as evidence that when used appropriately, a shorter steam autoclave cycle is an acceptably safe method of ophthalmic instrument sterilization.

Dr. Thomas questions the accuracy of endophthalmitis diagnosis at our institution because it was not clearly defined and may have been based on the clinical impression of physicians in training. Indeed, culture negative infectious endophthalmitis is purely a clinical diagnosis based on signs and symptoms that are well known to every clinician. As we suspect would be true at all teaching institutions, a diagnosis as dire as infectious endophthalmitis was always made or confirmed by a senior full time staff (one of the three authors, RDR, BT, MRV) rather than solely by a trainee.

Dr. Thomas suspects that because our microbial spectrum was different from that of other international hospitals (Nocardia was the most common isolate), therefore, “commoner organisms and early cases were likely missed.” We diagnosed 17 of the 35 (48.5%) POE patients among the charity group in the first postoperative week, when many of the patients were still hospitalized (Table 3 of the article).[2] It is well known that the spectrum of infectious ocular pathogens can vary by geographic region and therefore may not be accurately represented by surveys conducted elsewhere, such as the endophthalmitis vitrectomy study (EVS).[7] We believe that three large Indian studies from Chennai, Hyderabad, and Chandigarh are more relevant to our own geographic location.[489] Compared to other published international surveys, the Hyderabad and Chennai studies showed a reduced percentage of gram-positive bacteria (42-47%), and increased percentages of gram-negative bacteria (26-42%), fungi (17-22%), and polymicrobial infections (13-17%).[48] Gupta reported an even higher incidence of fungal infection (58%) in their study, which had a 13% culture positive incidence of both gram-positive and gram-negative bacteria.[9] Along with a prior published report on endophthalmitis pathogens in our private patients,[5] these collective studies show that Nocardia, for whatever reason, is uniquely endemic to our region. As correctly pointed out, this is a soil organism and its preponderance probably reflects the poor hygiene and general health of underprivileged patients, rather than implicating poor surgical care.

We agree that a weakness of our study was the fact that 6% of private and 16% of charity patients did not have long-term follow-up. However, we are bewildered as to why Dr. Thomas believes it to be a useful analytical exercise to assume that all of these patients developed so rare and visually catastrophic a complication such as infectious endophthalmitis. Both private and charity patients had direct access to and were instructed as to how to seek postoperative care in the event of visual loss or other problems. Because infectious endophthalmitis is accompanied by obvious and alarming symptoms, it is very unlikely that a significant number of infectious cases were missed.

Dr. Thomas asserts that charity status could be a confounding factor in our study. We have reported higher proportion of cases of endophthalmitis in charity patients. However, in large study populations with many varying characteristics, it is necessary to do a subset analysis to assess which factors are most relevant to endophthalmitis risk. In doing so, the incidence in the manual SICS group + ECCE group was statistically significantly higher than the incidence in the phacoemulsification group. However, the endophthalmitis rate between charity and private patients was not statistically different either for phacoemulsification (P = 0.329) or for manual small incision cataract surgery (MSICS)/ECCE (P = 0.367). Our results also show that compared to private patients, charity patients underwent MSICS more frequently (87% compared to 23% among private) and we believe that this explains the higher incidence of endophthalmitis among charity patients rather than any differences in sterilization protocols.

Dr. Thomas makes assumptions about the surgical care and management of charity patients from eye camps in our study, such as the use of different operating rooms. At the Pondicherry Aravind Eye Hospital, private and charity patients undergo cataract surgery in common operating rooms, using the same equipment and the same pre and post operative protocols. Charity patients from outreach camps are typically hospitalized for three days postoperatively. Those with worrisome inflammation or potential complications are kept for longer observation prior to discharge. A 1 month postoperative follow-up eye camp is always conducted in the same rural location where the initial screening camp occurred or at the regional eye hospital if feasible and as a routine we closely monitor postoperative follow up rates.

Finally, although we acknowledge that our efforts to reduce unnecessary costs may raise controversy, we take exception to the disdainful tone of this letter, such as in the use of terms such as “short cuts” and the insinuation that we are suggesting that “lens implants be treated with less rigor than a bone marrow transplant or hip replacement.” In fact, there are major differences among these procedures and imposing universal rigid standards for every procedure across all of medicine probably does result in unnecessary and wasteful practices. For example, bone marrow transplantation is an inpatient procedure that would routinely include a preoperative chest X-ray and laboratory blood work. At one time, routine preoperative laboratory testing and inpatient hospitalization were the standard for cataract surgery in the United States. As we all know, studies have since shown that routine preoperative laboratory testing prior to cataract surgery is indeed of no significant value to the patient.[10] Were such questions never asked, we can hardly imagine the ongoing cost of misallocating precious healthcare resources for such a high volume procedure as cataract surgery.

The global backlog of cataract blindness is increasing at a discouraging rate. With cataract surgery already being the highest volume procedure for an increasingly older global population, there is no health care system in the world that will not be challenged by the burden of providing this care. We are not suggesting that our own institutional protocol should become the new international standard. However, we stand by the concluding sentiments of our article and hope that other institutions will conduct similar studies: “In many countries, sterilization and aseptic protocols for ophthalmic surgery have been arbitrarily devised by regulatory agencies. Many of these measures originated from work performed outside ophthalmology and may not be necessary for our specialty. Because of the increasing need for cataract surgery to be both safe and cost effective, it is important to periodically re-evaluate potentially unnecessary practices based on carefully monitored studies of surgical outcomes.”[2]