Journal of Refractive Surgery

Editor's Page Free

Conventional Standards for Reporting Results of Refractive Surgery

George O Waring, III, MD, FACS

One of the major challenges for science in general and for refractive surgery in particular is to verify published reports and to develop cumulative experience that will identify the most useful procedures and techniques. This process was dramatized recently by the publicity surrounding Pons and Fleishmann's observation of "cold fusion in a jar". These investigators chose to release their findings to the public through the popular press rather than through peer reviewed scientific publication. This publicity set off a multi-million dollar flurry of attempts to verify their reports; no convincing evidence emerged that room -temperature fusion can generate commercially useful power. Part of the confusion came from the lack of standards for measuring exactly how the energy was produced.

The lack of conventional standards also makes comparison of the results among reports of refractive surgery difficult. For example, in studies of radial keratotomy the definitions of overcorrection differ; some authors define it as a spherical equivalent refractive error greater than +1.00 D,1 some as more than +2.00 D, and some as "visually symptomatic hyperopia".2 The formats used for scattergrams of the refractive outcome differ: some plot baseline refraction on the X-axis versus change in refraction on the Y-axis,3 others plot baseline refraction on the X-axis versus final refraction on the Y-axis,1 and still others plot baseline refraction on the Y-axis versus final refraction on the X-axis.4 Similarly, a meaningful loss of best corrected visual acuity is one or more Snellen lines in some studies and two or more in others.

Thus, ophthalmologists, researchers, patients, insurance companies, and regulatory agencies who wish to compare different techniques of the same surgical procedure or different surgical procedures for the same refractive error must either make crude estimates,5 recompute the data,1 or engage in meta-analysis6 in order to make rational decisions about the relative effectiveness, safety, predictability, and stability of these procedures.

Conventional standards are commonplace in ophthalmology. For example, most individuals accept an uncorrected Snellen visual acuity of 20/20 as "normal"; most states have selected 20/40 as the visual acuity standard required for an unrestricted drivers license; many agencies acknowledge 20/200 as the visual acuity criterion for legal blindness. Setting standards does not truncate creativity, restrict freedom, force conformity, or interfere with an investigator's rights to report information in any other format. On the contrary, conventional standards allow the world community of refractive and corneal surgeons to communicate more accurately, to speed the development of refractive surgical procedures, to provide more appropriate management of patients, and to improve the quality of scientific reporting.

Therefore, we should create a set of conventional standards on which most refractive surgeons agree, publish these standards in a simple and useful form that serves as a guide or checklist, adopt these conventions formally through societies such as the International Society of Refractive Keratoplasty and the European Refractive Surgical Society, and encourage authors to use these conventions in their reports.


TABLEMinimal Essential Standards for Reporting Studies of Refractive Surgical Procedures


Minimal Essential Standards for Reporting Studies of Refractive Surgical Procedures

In the accompanying Table, I propose a set of minimal essential standards for reporting studies of refractive surgical procedures. Most of this information is available from a patient's clinical record and can be compiled without the use of computers or biostatisticians. Individuals with the computing or statistical resources can supplement this minimal information by more sophisticated methods of analysis such as survival curves, confidence intervals, and multiple regression analysis. These criteria are not a prerequisite for publication of studies in Refractive and Corneal Surgery, but authors who choose to include this conventional information will find more rapid acceptance of their papers and more ready appreciation from their colleagues for clarity of communication.

We seek the comments of refractive and corneal surgeons on this proposal. A set of expanded standards is being developed.7


1. Waring GO III, Lynn MJ, Culbertson W, Laibson PR, Lindstrom RL, McDonald MB, et al. PERK study group: three year results of the prospective evaluation of radial keratotomy (PERK) study. Ophthalmology. 1987; 94:1339-1354.

2. Sawelson H, Marks RG. Two-year results of radial keratotomy. Arch Ophthalmol. 1985; 103:505-510.

3. Arrowsmith PN, Marks RG. Visual, refractive and keratometric results of radial keratotomy: five year follow-up. Arch Ophthalmol. 1989; 107:506-511.

4. Rowsey JJ, Balyeat HD. Radial keratotomy: preliminary report of complications. Am J Ophthalmol. 1982; 93:437-455.

5. Binder PS. Radial keratotomy in the United States. Where are we six years later? Arch Ophthalmol. 1987; 105:37-39.

6. Goodman SN. Meta-analysis and evidence. Controlled Clinical Triah. 1989; 10:188-204.

7. Waring GO III. Standards for reporting results of refractive surgical procedures. In: Waring GO III. Refractive keratotomy for myopia and astigmatism. St Louis, Mo: CV Mosby Co., St. Louis, In press.


Minimal Essential Standards for Reporting Studies of Refractive Surgical Procedures


Sign up to receive

Journal E-contents