advertisement

---

Editors Selection IGR 20-4

Risk factors: Risk calculator

Comment by Franz Grehn on:

13052 Validation of a predictive model to estimate the risk of conversion from ocular hypertension to glaucoma, Medeiros FA; Weinreb RN; Sample PA et al., Archives of Ophthalmology, 2005; 123: 1351-1360

See also comment(s) by Luca Rossetti •

Find related abstracts

---

Medeiros et al. (953) developed a predictive model for the five year risk of conversion to glaucoma derived from the results of the ocular hypertension treatment study (OHTS). The model was assessed in an independent population of 126 cases with ocular hypertension, being studied longitudinally at the authors' institution. Although the results of the OHTS are well known among the glaucoma specialist, the relative weight of risk factors has not yet been put together into a predictive model. A handy risk calculator has been recently derived from that study which may help the general ophthalmologist to assess individual risk of the patient with ocular hypertension. This may both pick up cases with high risk who need treatment as well as avoid unnecessary treatment in patients with very low risk. Case finding in high risk group is obviously more effective than general screening and may be preferable from a cost budget point of view. The authors used a full and a reduced model to describe hazard ratios in both OHTS and DIGS studies, respectively. There was good correlation for age, IOP and central corneal thickness between both studies. The protective effect of Diabetes mellitus which was found in the OHTS, could not be confirmed in DIGS. In the reduced model, pattern standard deviation and vertical cup disk ratio were excluded. There was no statistically significant difference in discrimination between the reduced and the full models. The predicted probability for glaucoma conversion within five years was calculated by the model and applied to the DIGS population.

There is no statistically significant difference in discrimination between the reduced and the full models with Pattern Standard Deviation and vertical Cup Disc Ratio

This resulted in a similar probability rate (14,3 % and 12,6% respectively) as compared to the OHTS Kaplan-Meyer curve (11,6%). There was good agreement between predicted and observed probabilities. With a point score system six risk factors are summed up. From the data given in the paper, the relative weight of central corneal thickness as a risk factor is four times higher than the IOP risk in the given range, and nearly three times higher for vertical cup disk ratio and twice as high for pattern standard deviation as compared to IOP risk in the given range. Diabetes mellitus, when present, would protect glaucoma damage of the same quantity as IOP increases glaucoma risk. These factors although correctly derived from the present studies have to be discussed clinically. The protective factor of Diabetes mellitus is not found in other studies, especially not in the comparative cohort of the DIGS population.

Age, IOP and CCT are factors with high reproducibility that can be used in a risk calculator

The two other factors that were excluded in the reduced model, i.e., vertical cup disc ratio and pattern standard deviation, are subject to high variation in patients with reduced compliance and therefore might artificially influence the outcome, in particular as they have high relative weight in the calculation model. Vertical cup disc ratio has to be exactly defined before being incorporated in such a model. In addition, the vertical cup disc ratio depends largely upon the size of the optic nerve head (macro disc versus micro disc). Hence, the factors with high reproducibility will be age, IOP and CCT. Other studies, such as the European Glaucoma Prevention Study, have confirmed many risk factors of the OHTS with very similar hazard ratios.

See also comments by Herndon on CCT

---

• ← Previous
• Next →

---