Discussing Benefits, Risks, Probabilities, and Expected Outcomes With Patients

Physicians and their clinical team members need to educate patients regarding their disease, including potential preventive measures, treatments, and outcomes.

Clinical research defines risk as the conditional probability of an event, usually an adverse event. Risk difference is the absolute difference in the risk between 2 groups of individuals. Relative risk is the ratio of 2 risk measures. The risk difference depends on the unit of measure, whereas relative risk is dimensionless because it involves division of 2 risk measurements. In the Ocular Hypertension Treatment Study (OHTS), the risk difference of glaucoma development for subjects who were not treated compared with those who were treated was 5% (9.5%–4.5%) across 5 years. The relative risk of not being treated compared with receiving treatment was 211% (9.5%/4.5%). Both measures are consistent with the data, but clinicians or patients may interpret them very differently. Numerically, a 5% increased risk of glaucoma (if ocular hypertension is not treated) might seem small to a patient, while a 211% increased risk might seem large. A key piece of information that may help in the interpretation of such results is the baseline probability of the outcome. For example, the baseline probability of developing glaucoma with untreated ocular hypertension is 9.5%. In most cases, baseline probabilities or expected outcomes can help patients understand risk and make an informed decision about a procedure.

The number needed to treat (NNT) can also be helpful when describing how likely a treatment or medication will improve an outcome for an individual patient. In the above example, the study shows a 5% absolute risk reduction (ARR) in the proportion of patients without glaucoma if they use ocular hypotensive medications. The ARR is 5%, and the number needed to treat is 20 (100/5). In other words, a clinician could tell the patient that in order to prevent 1 patient from developing glaucoma over 5 years, 20 patients would need to be treated with ocular hypotensive medications. Armed with this information, the patient can decide whether to use an ocular hypotensive medication.

For those providers with capitated payments, it may be useful to consider the impact of the new procedure or treatment on societal cost. This cost can be determined by calculating the added cost using the NNT. For example, if the retail cost of a generic glaucoma medication for 1 patient is $70/month, or $4200 over 5 years, the excess cost of using ocular hypotensive medications calculated with the NNT would be $84,000 ($4200 × 20) for 1 patient to see the benefits from using these ocular hypotensive medications.

The advantage of the NNT is that it only uses absolute risk reduction and is less likely to exaggerate the impact of a new procedure or medication. However, the NNT does not take other important factors into consideration, such as the negative effects on quality of life related to eyedrop use, nor other possible factors such as systemic adverse effects resulting from the use of ocular medications. The NNT does provide a straightforward attempt to describe the likelihood that a patient will be helped, harmed, or unaffected by a treatment.

In observational studies, investigators may also present their results as odds ratios. In an odds ratio, the odds of a subject with the disease (case) having an exposure (eg, smoking) are compared with the odds of a subject without the disease (control) having the exposure. When the disease is rare, the odds ratio approximates the relative risk of the exposure, because the denominators for both the odds under comparison are close to 1. For example, in the meta-analyses on potential risk factors for late AMD, which occurs relatively infrequently, the odds ratio for smoking is 2.35, meaning smokers have a 235% risk for development of late AMD compared with nonsmokers. As previously stated, the baseline risk is key; if it is low (ie, less than 1%), then an odds ratio of 2.0 still results in a low risk for an individual patient.

Exposure to specific factors may or may not be clinically significant, and it may not be causal, or one of the root causes of the disease. Because it is difficult to distinguish causal risk factors from noncausal risk factors in observational studies, researchers often use causal criteria to identify which risk factors are causal and which are not.

Medical providers use risk calculators in a number of ways, for example, to predict an individual patient’s risk of cardiovascular disease, risk of having a child with Down syndrome, likelihood of survival from an intensive care unit, and likelihood of experiencing other medical conditions. Ophthalmologists have used risk calculators to help simplify complex study results and apply these results to individual patients. For example, the OHTS regression equation used 5 variables (cup–disc ratio, central corneal thickness, untreated IOP, pattern standard deviation from the visual field, and age) to predict the risk of a patient developing glaucoma as a result of ocular hypertension. The OHTS risk calculator is available for free online (https://ohts.wustl.edu/risk/). Other available risk calculators include those for macular degeneration, keratoconus, and glaucoma progression. Many risk calculators can be downloaded onto a mobile device. Overall, the advantage of risk calculators is that they simplify complex results to provide an estimate of the mean baseline probability of disease development or surgery complications in individual patients. Risk calculators can also identify those patients at high risk of developing disease and select them for a lower NNT and lower societal costs.

• De Moraes CG, Sehi M, Greenfield DS, Chung YS, Ritch R, Liebmann JM. A validated risk calculator to assess risk and rate of visual field progression in treated glaucoma patients. Invest Ophthalmol Vis Sci. 2012;53(6):2702–2707.

  Kass MA, Heuer DK, Higginbotham EJ, et al. The Ocular Hypertension Treatment Study: a randomized trial determines that topical ocular hypotensive medication delays or prevents the onset of primary open-angle glaucoma. Arch Ophthalmol. 2002;120(6):701–713.

  Mansberger SL. A risk calculator to determine the probability of glaucoma. J Glaucoma. 2004;13(4):345–347.

Excerpted from BCSC 2020-2021 series: Section 1 - Update on General Medicine. For more information and to purchase the entire series, please visit https://www.aao.org/bcsc.