Future Research Needs

Publication Details

The available evidence regarding glaucoma treatments demonstrates definitively that intraocular pressure can be lowered by medications, laser treatments, and surgery. High-quality randomized controlled trials have also shown that reduction of intraocular pressure slows the development and progression of damage to the optic nerve and slows visual field loss. Although logical to presume that slowing glaucoma damage would lead to preservation of vision-related quality of life and reduction in visual impairment, this link has not been demonstrated in the research literature. Establishing this link is perhaps the most important next step in fleshing out the analytic framework presented above, thereby establishing (or not) the impact of treatment on the more meaningful final outcomes.

As part of this systematic review, it also became clear that most of the available literature on glaucoma treatments was deficient in one or more ways and therefore could not be used to answer the questions incorporated in the analytic framework. First of all, many studies had to be excluded because they were non-comparative or too small. In general, the lack of a control or comparison group precludes the kind of conclusions one would like to draw from a study, specifically whether a particular treatment is superior to another. Similarly, studies with inadequate sample sizes are also less informative than desired. The glaucoma treatment literature also contains many non-randomized studies. This seems to be a general deficiency in surgical fields where it is possible to implement new procedures or devices without the same level of oversight or approval found with medications. Regardless of the underlying reasons, the fact that subjects were not drawn from the same population and randomized to one of a number of treatments means that it is again difficult to draw any meaningful conclusions. All of these problems could be overcome with more rigorous study design, which should include at least two groups, preferably randomized, and some kind of a priori sample size calculation to increase the likelihood of being able to answer the question motivating the study in the first place.

It is not true that only prospective randomized studies should be conducted but other study designs need to be purposefully designed to answer some question. For example, large observational studies or clinical data registries could be used to assess harms of treatment that might not be detected in smaller, randomized studies. Similarly, these non-randomized designs could also provide information about the incidence or prevalence of visual impairment or disability among various groups.

Another category of deficiency in much glaucoma literature is with regard to the reporting of outcomes and harms for study populations. This may be due to the facts that there is no universally accepted measure of glaucoma or its progression (outcomes) and that there has also been no consensus regarding which harms should be reported and how. As such, the ability to combine results from multiple studies would be enhanced by more commonality with regard to study design. Fortunately, such a consensus is now available and the World Glaucoma Association publication “Guidelines on Design and Reporting of Glaucoma Surgical Trials” should serve as a basis for all trials of new and existing treatments.15

One area of glaucoma that has only recently received attention is that of formal risk estimation for patients. In the case of patients who are similar to the subjects of the Ocular Hypertension Treatment Study, it is possible to use a calculator to estimate their risk of conversion to glaucoma within 5 years. Production of similar calculators for other stages of disease (early, moderate, severe) and for other groups of glaucoma suspects (African Americans, Latinos, etc.) may help identify those patients most at risk of incident or progressive disease, thereby allowing treatment efforts to be better focused. Based on the results of the studies described above, there is not yet compelling evidence that these groups would benefit from any particular approach to treatment.

Below, we utilize the Population, Intervention, Comparison, and Outcome framework to outline areas of research that might help resolve the deficiencies in prior work identified as part of this review.

Lack of Association Between Treatment and Visual Impairment

Population:

  • Patients with moderate visual loss from glaucoma, that is, those at highest risk for visual impairment

Interventions:

  • Studies evaluating all interventions are needed: medical therapy, laser trabeculoplasty, incisional surgery

Outcomes:

  • Visual impairment as measured by standard definitions (e.g., International Classification of Diseases)
  • Functional measures of impairment: reading, driving, other activities of daily living
  • Long-term trials or patient registries (i.e., greater than 10 years) are needed to determine the relative impact of treatments on visual impairment
  • All studies of glaucoma treatments should routinely include generally accepted measures of visual impairment

Lack of Association Between Treatment and Patient-Reported Outcomes

Population:

  • Open-angle glaucoma patients in need of treatment

Interventions:

  • Studies evaluating all interventions are needed: medical therapy, laser trabeculoplasty, incisional surgery

Outcomes:

  • Assessment of patient-reported outcomes prior to the start of therapy to provide appropriate basis for assessing these outcomes after therapy
  • Potential outcomes for consideration in future research include satisfaction with therapy, self-assessment of visual function, and concerns about future vision loss.

Assessment of the Relative Risks and Benefits of Treatment

Population:

  • Glaucoma patients in need of treatment
  • Provide subanalysis or complete stratification by risk

Interventions:

  • Studies evaluating all interventions are needed: medical therapy, laser trabeculoplasty, incisional surgery

Outcomes:

  • All studies of glaucoma treatments should be designed to provide information on the comparative effectiveness of one treatment versus the most appropriate “standard.”