discovering dry eye
The Ups and Downs of Clinical Grading
BY JASON J. NICHOLS, OD, MPH, PHD

Clinical grading of contact lens complications can prove difficult at times — especially when more than one practitioner is involved — because of the subjective nature of most clinical grading. Further complicating the issue is the tremendous variability in the presentation of lens-related complications and tear film abnormalities, added to the fact that practitioners vary in training and expertise relative to these outcomes. Finally, we see many different facets or presentations of contact lens complications and tear film disorders that we need to assess as they all relate to the severity, treatment and prognosis of the condition. For example, we can grade corneal fluorescein staining in terms of its extent as well as of its depth, type and location.

Statistically Speaking

It's important to think about subjective clinical grading in somewhat of a statistical sense to make improvements. For example, you can sum up clinical grading or measurements as a "true value" plus an "error" term. You can further break down this error term into systematic error (bias) and non-systematic error (random noise). It's important that we limit error associated with the true value, which ultimately makes our estimate more precise (closer to the true value).

For an individual grader who is consistent, most of the error associated with his grading results from random noise, or human variability. But, when we compare two or more graders (between-examiner reliability), more error occurs because of noise and bias between examiners.

Improving Reliability

We can improve the reliability of clinical grading in several ways. We could attempt to objectify clinical measurements. For instance, many studies and techniques have objectified traditionally subjective measurements such as bulbar hyperemia and tear breakup time. We could also train examiners, which often occurs during a practitioner's clinical training or at the outset of a new study. Lastly, we could use grading scales, which provide some standardization to the clinical grades, especially between examiners. Table 1 shows some of the grading scales available.

Note that some grading scales use photographic images, while others use artistic renditions.  It's important that practices employ only one grading scale for clinical practice, as consistency in this regard is extremely important.

Dr. Nichols is assistant professor of optometry and vision science at The Ohio State University College of Optometry.