Article
Prediction of visual results after keratoplasty by hereditary corneal stroma dystrophies
Search Medline for
Authors
Published: | September 22, 2004 |
---|
Outline
Text
Objective
Due to information technologies development and Evidence-based-Medicine principles, the appliance of mathematical analysis methods for prediction of treatment outcome, has become possible. These methods enable to create numerous values of risk factors, to reveal the most sufficient of them and evaluate their reciprocal influence on treatment outcome. The aim was to analyze the risk factors, which decrease visual results after keratoplasty by hereditary corneal stromal dystrophies (HCSD) and evaluate probability of positive visual results achieving after this operation.
Methods
Visual results were retrospectively analyzed after keratoplasty performed since 1.01.1980 till 1.01.04 at 320 eyes in 163 patients with HCSD. Among them were: 135 eyes (42,2%) with granular, 88 (27,5%) - with lattice, 97 (30,3%) with makular HCSD at the age of 41,6±13,0SD. The positive visual results (PVR) was considered increasing postoperative best corrected visual acuity more than 0,2. Excluding criteria were: inflammation in the anterior eye segment and cataract. Twelve possible predictable for PVR risk actors were analyzed. Predictive influence for each risk factor on postoperative visual results was evaluated using the method of logistic regression.
Results
PVR was achieved in 235 eyes (73,4%). The statistically significant differences between PVR and type of keratoplasty (x2=0,76,p=0,35), state of corneal surface (x2=0,72,24,p=0,33), lacrimalsecretion (x2=1,94,p=0,38) and type of dystrophy (x2=5,64,p=0,059) were not found. The strongest significant PVR predictors after keratoplasty were: degree of the corneal stroma opacification (odds ratio 3,3, p=0,01), retransplantation (odds ratio 1,72, p=0,025), age of the first keratoplasty (odds ratio 1,62, p=0,018), depths of the pathological changes in the stroma (odds ratio 1,62, p=0,043), corneal vascularisation (odds ratio 1,57, p=0,085).
Conclusions
Methods of mathematical analysis enabled to design the predictive model of possible evaluation of visual outcome after keratoplasty depending on certain risk factors combination in individual patient with HCSD.