Lamellar procedure outcomes may not surpass penetrating keratoplasty
"Providing high long-term value for the total care delivered to patients is imperative in the changing health care environment," says Sanjay V. Patel, M.D., an ophthalmologist at Mayo Clinic in Rochester, Minnesota. "That value equation requires effective and meaningful definition and measurement of ideal outcomes."
In an article published in Ophthalmology in 2014, Dr. Patel, in conjunction with W. John Armitage, Ph.D., University of Bristol, and Margareta Claesson, M.D., Ph.D., University of Gothenburg, commented on results of the Australian Corneal Graft Registry that suggested that the outcomes of lamellar keratoplasty procedures — endothelial keratoplasty (EK) and deep anterior lamellar keratoplasty — were worse than those of penetrating keratoplasty (PK).
Australian study outcomes support PK
The Australian Corneal Graft Registry collects outcome data from subjects undergoing corneal transplantation. In a study published in Ophthalmology in 2014, Douglas J. Coster, M.D., Flinders University of South Australia, and others found that lamellar procedures, whether endothelial or deep anterior, were associated with worse graft survival and visual acuity compared with PK for the same indication and over the same period. The authors attribute this finding, in part, to the difference between real-world registry data from multiple surgeons versus data from a few single-center surgeons in optimal conditions.
"The difference is important and should not be ignored," says Dr. Patel. "Although level I evidence is the gold standard, randomized controlled trials may not always be feasible or achievable within a reasonable time frame. As a result, the adoption of new techniques is often driven by single-center studies reporting the outcomes from specialist units.
"Dr. Coster's study shows that the wider picture revealed by Australian registry data does not necessarily reflect the optimism generated by the excellent results from single-center studies. This finding reinforces the need for individual surgeons to determine their own outcomes. They should not assume that their outcomes are the same as those achieved by experienced and high-volume surgeons. It is hoped that the Australian registry will show improving outcomes for lamellar procedures as more surgeons gain experience in these techniques."
Swedish and American registries
Longitudinal, observational data collected on large numbers of patients in national registries help to detail the rate of uptake of new techniques and the outcomes across multiple centers. At the EyeNet Sweden registries, surgeons enter data directly, which reduces time, improves accuracy and enables them to compare individual results with national outcomes.
In 2015, the American Academy of Ophthalmology launched the Intelligent Research in Sight Registry (IRIS), which will gather data directly from electronic medical records and enable benchmarking against national outcomes.
"In addition to contributing substantive, evidence-based data for the evaluation of the efficacy of new techniques, these registries also provide preliminary data for initiating randomized controlled trials," says Dr. Patel. "The seemingly poorer outcomes of the lamellar techniques compared with PK in Dr. Coster's study are not necessarily an indictment of these newer techniques. Excellent results can be achieved in single-center studies. Comparing visual outcomes of different techniques, especially, may require carefully designed randomized controlled trials."
Ideal keratoplasty outcome measures
Graft failure has been the primary outcome in most large keratoplasty studies because of its relative ease of assessment. Although graft failure is associated with patient dissatisfaction, graft survival is not always representative of visual outcomes or amelioration of patient visual disability.
Visual outcomes are becoming more important in the current era of EK, but assessing vision with rigor requires standardized refraction and testing protocols, and a defined minimum level of ability and experience in refracting patients. There also can be incongruence between visual outcomes and patient-reported outcomes.
"Perhaps we should be making a more concerted effort to relate patient-reported outcomes to clinical outcomes by using validated vision-related quality-of-life instruments," says Dr. Patel. "Such tools could be simple to administer and could reveal a point of diminishing return in terms of patient-reported outcome in relation to vision and graft survival."
For more information
Patel SV, et al. Keratoplasty outcomes: Are we making advances? Ophthalmology. 2014;121:977.
Australian Corneal Graft Registry. Flinders University.
Coster DJ, et al. A comparison of lamellar and penetrating keratoplasty outcomes: A registry study. Ophthalmology. 2014;121:979.
EyeNet Sweden. Blekinge Hospital.
IRIS Registry. American Academy of Ophthalmology.