Back to Freepaper Session

The quality measurement system in ophthalmology (QMSO): an update of comparative results

Session Details

Session Title: Special Cases

Session Date/Time: Friday 14/02/2014 | 08:30-10:00

Paper Time: 09:04

Venue: Kosovel Hall (Level -2)

First Author: : TimHerbst GERMANY

Co Author(s): :    Joerg Foerster   Dominik Haas   Detlef Holland        

Abstract Details

Purpose:

After describing the measurement system in ophthalmology, the second of three presentations deals with first results of data analyses. On the basis of 16 different outcome variables per case and more than 60.000 cases, possible differences in healthcare outcome among participating surgery centers will be analyzed with the help of appropriate statistical methods. Furthermore, differences in process as well as in structure quality will be analyzed in order to find possible explanations for medical outcome differences.

Setting:

Data analyses depend on the quality measurement system, which transfers each medical finding into a quality index and bases upon the German school grading system. It is built up of three subarea QiBs, which includes measurable and subjectively ascertained findings as well as different aspects of patient satisfaction.

Methods:

Since the first of three presentations puts its focus on the description of the quality measurement system, corresponding results of data evaluations will be presented. Choice of statistical methods depends on level of measurement: For examinations on aggregate levels, we will use ordinary t-test. For comparisons of single finding QiBs, we will come back to Mann-Whitney- and Kruskal-Wallis tests respectively because of ordinal scale levels. Further possible examinations will be realized by appropriate statistical methods. Since total center score bases upon a statistical prefunded weighting of each included medical outcome, underlying statistical method (discriminant analysis) will be explained rudimentary.

Results:

In preparation of analyses, we anticipated to find significant quality differences of medical outcome among different surgery centers and different surgeons. The Reason for this hypothesis seems to be multifactorial. In addition to medical variables, differences in structure quality (medical device, etc.) and process quality (surgical technique, process organization) may also give explanations for possible differences in outcome quality. First results, which have already been presented in Amsterdam give support for our hypotheses: we found significant differences among different surgery centers and surgeons for nearly every single medical outcome as well as for patient satisfaction. Furthermore, we found significant differences for perioperative variables (effective phaco time, type of IOL, IOL centration, etc.). We expect to find similar results in the ongoing data analyses. Since data analysis is still in process, possible explanatory power of structure and process quality for differences in outcome quality has not been analyzed so far.

Conclusions:

In the second of three presentations, an update of results for possible differences in outcome quality, which are captured by a self-developed quality measurement system, will be presented. On the basis of peri- and postoperative outcome variables and with the help of different structure and process variables, possible differences in healthcare outcome among participating surgery centers will be analyzed. Ordinary t-test (interval scale) as well as Mann-Whitney- and Kruskal-Wallis-test (ordinal scale) will be used for statistical analyses. Further examinations will be undertaken with the help of appropriate statistical methods. First results, having already been presented in Amsterdam, show significant quality differences in medical outcome among different surgery centers and different surgeons. Besides different peri- and postoperative variables, which are documented standardly in each case, structure and process quality will be used in order to find explanations for differences in quality outcome. FINANCIAL INTEREST: NONE