Premium
How should we quantify the performance of KPro's? VATI
Author(s) -
GRABNER G,
STOIBER J,
HITZL W
Publication year - 2009
Publication title -
acta ophthalmologica
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.534
H-Index - 87
eISSN - 1755-3768
pISSN - 1755-375X
DOI - 10.1111/j.1755-3768.2009.4221.x
Subject(s) - visual acuity , computer science , usable , medicine , artificial intelligence , ophthalmology , world wide web
Purpose To report a method of standardized data collection and reporting and statistical assessment that can be used for all KPro´s available on the market. The database (will be presented) should be • Usable for different types of KPro´s • Easily adaptable to changes in technique • Allow for complete entry of relevant data Methods Visual Acuity data should be reported in different international surgical centres in a standardized manner: Best spectacle corrected VA, unless BCVA only possible with CL (> useful time of wear). A complete entry of all relevant data is possible in this database. The statistical analysis should be agreed upon by all centres. For Survival Time = Retention of KPro > the Kaplan‐Meier method For Visual Acuity over the Course of Time > the Monte‐Carlo method Results A database will be demonstrated that can be used free of charge by all KPro centres interested. The VISUAL ACUITY BY TIME‐ INDEX (VAT‐ Index) will also be presented, whose theoretical basis published in: Journal of Theoretical Medicine, 2002 / 4, 183‐190, W. Hitzl and G. Grabner [bdquo]Application of the Monte Carlo Method for the Assessment of Long‐term Success in Keratoprosthesis Surgery“. Example of its use will be give, based in data, courtesy Barraquer Eye Clinic, Barcelona. Conclusion With the Kaplan‐Meier method: + analysis is done quickly, uses all data available, hypotheses tests are available for comparisons and mean and median survival time can be computed ‐ no information about relation between time and best corrected visual acuity and the definition of terminal event is arbitrary to a certain extent. Monte‐Carlo method (VAT‐index): + Method is based on a so‐called non‐parametric longitudinal model + Reliabel estimation of relation between time and best corrected visual acuity at any given time point (patient as well as surgeon is basically interested in this relation). + statistically valid analysis and better comparison of different KPro techniques + easy comparison of defined postoperative periods + comparison of different initial clinical findings and diseases possible + long‐time follow‐up of BCVA ‐ shorter follow‐up time as compared with Kaplan‐Meier method (e.g. with strict [bdquo]80%data complete” criteria)