Glaucoma Risk Calculators: Adding Up An Accurate Diagnosis
A look at the pros and cons of using this tool for ocular hypertensives.
By James D. Brandt, MD
Risk calculators derived from the Ocular Hypertension Treatment Study (OHTS) and the European Glaucoma Prevention Study (EGPS) now provide clinicians with important tools to help decide on a treatment plan for ocular hypertensives. While elevated intraocular pressure has long been considered the leading indicator of a patient at risk of glaucoma, the OHTS has brought recognition of the many other factors that must be taken into consideration.
The OHTS examined whether reducing IOP in patients with ocular hypertension would result in a lowered risk of developing primary open-angle glaucoma. While the study showed that at five years only 4.4% of treated patients developed POAG — in comparison to 9.5% of control patients — with longer follow-up, the study has also demonstrated that most ocular hypertensive patients never develop glaucoma, regardless of treatment. In conjunction with data from the EGPS, investigators at Washington University in St. Louis created an online tool for estimating an individual's risk of developing POAG over the next five years. While the calculator allows clinicians to think beyond elevated IOP by incorporating other data, there are still factors that the tool does not include, but are still vital to the decision to initiate or withdraw treatment.
An iPhone app of the calculator is available via iTunes.
In this article, I will briefly review the factors taken into account when the calculator was developed along with precautions to consider when using it in treatment decisions.
Risk Calculator Factors
The OHTS identified the following five predictive factors important in assessing the risk of POAG. First and foremost, when using the risk model it is important to ask if the patient sitting in front of you in your office is similar to subjects in the trial — the further your patient strays from the study's entry qualifications, the less likely the calculator is going to be relevant. If your patient isn't similar to the OHTS subjects (eg, much younger or older, history of exfoliation syndrome, glaucoma already diagnosed in one eye, etc.), you should use the calculator cautiously, if at all.
● Age. This is the only factor that can be considered in your exam room the exact same way it was considered in the study. OHTS enrolled patients ranging in age from 40 to 80.
● Untreated IOP. The calculator is based on three measurements of IOP by Goldmann tonometry per eye. OHTS subjects were washed out of any treatment they had been on and to be enrolled had IOPs ranging from 21 to 32 mm Hg. IOPs outside this range weren't considered in the OHTS, so using the calculator for someone with an untreated IOP of 15 (or 45) isn't valid. Remember that baseline IOP was untreated. If you are evaluating a previously-treated ocular hypertensive, consider a drug holiday to re-evaluate risk.
● Central Corneal Thickness. CCT turned out to be one of the most significant predictive factors — a thinner cornea raises the likelihood of developing glaucoma independently from IOP and other factors. The OHTS took five measurements by ultrasound. Measurements from patients with corneal disease or who have had corneal refractive surgery or penetrating keratoplasty shouldn't be used to determine risk.
● Vertical Cup-to-Disc Ratio. In the OHTS, this was determined by a masked reader viewing stereoscopic photos, not by the individual clinician. Your ratio estimates should be based on contour rather than color, viewed stereoscopically.
● Pattern Standard Deviation. In the study, PSD was derived from at least three visual field tests using Humphrey perimeters performed at baseline. In the clinic, you usually don't have the luxury of three baseline visual fields. All of the OHTS subjects were also highly reliable perimetry subjects. What do you do when you have limited (or unreliable) perimetry data? The OHTS statisticians are looking at how data should be used when this information is excluded, but we don't yet know how the risk model performs without PSD.
Other Factors to Consider
The risk calculator is not an all-encompassing decision maker, but a tool for having a two-way discussion with your patient. Though it gives us more information than we used to have, the decision to escalate or withdraw treatment in a patient with ocular hypertension should not come only from the risk calculator, but a discussion of the results and what they mean. Do not blindly use risk models; before arriving at a treatment plan, other factors should also be considered:
● Family History. This is probably the risk factor that trumps everything else. Numerous studies show that risk greatly increases if there is a family history of glaucomatous vision loss or a first-degree relative who needed to have a trabeculectomy. Many have criticized the OHTS for not finding a risk relationship with family history. In retrospect, that is almost certainly related to the fact that we did not do a very good job of assessing family history. At the beginning of the study patients were simply asked if a relative had glaucoma — unreliable since they didn't know a lot about glaucoma at the time and didn't really know whether or not their family members had glaucoma, ocular hypertension or even cataract. You, however, can ask for more detail from your patient.
● Life Expectancy. A healthy 49-year-old and an 80-year-old on renal dialysis have very different time horizons to consider when deciding on how aggressive they want their glaucoma prevention treatment to be. If they have a normal optic nerve, the decision to treat becomes a clinical one. The 49-year-old may have a low risk of glaucoma by the calculator but has memories of a grandmother blind from the disease and want to be treated. The 80-year-old may choose to be treated or choose watchful waiting instead — ultimately it is their choice, and this is where the art of medicine comes in through a thoughtful discussion with your patient.
● Cost to Patient. Patients may decline medications because they don't have the money or insurance.
● Optic Disc Hemorrhage. This was not included in the risk calculator because patients with disc hemorrhage prior to entry were excluded from the OHTS. The OHTS showed that disc hemorrhages that occurred during the study resulted in a much higher risk for developing glaucoma. Don Budenz, MD, professor of ophthalmology at the Bascom Palmer Eye Institute, studied the increased risk for glaucoma associated with intercurrent optic disc hemorrhages. While he found that occurrence of hemorrhage increased risk for POAG fourfold, 86.7% of disc hemorrhage cases did not ultimately develop glaucoma. The study was only half completed then, with patients only observed for a median of 30.7 months after hemorrhage. Additional follow-up available now may reveal more patients developing glaucoma. “One could study optic disc hemorrhages and other potential intercurrent risk factors and develop a unique risk calculator that would help in the course of following a patient with OHT longitudinally, which is what we do in clinical practice,” Dr. Budenz says.
Conclusion
When using the OHTS/EGPS risk calculator, it is imperative that you understand how the data were derived, make sure your data are acquired in a similar manner and avoid applying the calculations to patients for whom it wasn't designed. With that said, you can still use the measurements to get a rough estimate to determine high, medium or low risk.
Clinically, I find the calculator most useful when I'm counseling an ocular hypertensive patient and I'm deciding on whether to institute or withdraw treatment — it is particularly useful in someone who has been treated in the past and risk calculations show that they are at very low risk (eg, a patient with thick cornea, healthy nerves and IOPs in the low 20s).
Clearly, high pressures and thin corneas are bad to have if you're worried about glaucoma, but there's no such thing as the calculator being “right” or “wrong” in a given patient. The calculator can only say what percentage of a group of patients with similar characteristics will develop glaucoma over the next five years. In comparison to cardiovascular risk calculators (which predict the likelihood of cardiovascular mortality based on things like blood pressure, cholesterol, etc.), the OHTS/EGPS calculator compares quite favorably.
The calculator is most accurate for people who closely resemble the patients in the study. Are these same risk factors as useful for established glaucoma patients? We don't know, but the answer is probably yes. We don't yet have a risk model for glaucoma progression, but data from the Early Manifest Glaucoma Trial suggests that the risks identified by the OHTS likely apply to a patient with established glaucoma. However, the OHTS risk model was not designed for patients with established disease and the calculator should be applied in such patients with a big grain of salt. Ultimately risk calculators are not a replacement for good clinical judgment, but rather an additional tool to guide that judgment. OM
The OHTS Glaucoma Risk Calculator is available at http://ohts.wustl.edu/risk.
James D. Brandt, MD, is professor of ophthalmology & vision science and director of the glaucoma service at the University of California, Davis. He has served as principal investigator of numerous clinical trials, including the Ocular Hyper tension Treatment Study. He may be reached at jdbrandt@ucdavis.edu. |