Reviewed by Steven Schallhorn, MD
San Francisco—Postoperative keratometry does not appear to affect patients’ night vision or satisfaction with LASIK, according to Steven Schallhorn, MD.
“There are many surgeons who won’t do LASIK if the cornea is made too flat by a myopic procedure,” said Dr. Schallhorn, a clinical professor of ophthalmology at the University of California, San Francisco. “But it has a very small and clinically irrelevant influence.”
Few studies have investigated the influence of a flat postoperative cornea on visual degradation, and these have reached contradictory results.
At the same time, a flatter postoperative keratometry is associated with higher attempted myopic correction, which might confound results.
To investigate the effect of keratometry on patient’s night vision and satisfaction, Dr. Schallhorn and his colleagues evaluated 8,672 myopic eyes of 4,602 patients from the database of Optical Express in Glasgow, United Kingdom.
Before surgery, the mean manifest spherical equivalent was -3.72 D, and the mean keratometry (K) value was 43.64.
The researchers compared these statistics to similar measures 1 month after surgery.
“We wanted to look at the early postoperative time period specifically because symptoms are generally worse in the 1-month time period,” Dr. Schallhorn said.
At 1 month after surgery, 93.7% of eyes were within 0.50 D and 99.1% were within 1 D of emmetropia. Similarly 94.6% of eyes achieved monocular uncorrected-distance visual acuity (UDVA) of 20/20 or better, and 98.3% achieved binocular uncorrected distance visual acuity of 20/20.
At the same time, 48.7% of the eyes had a flat corneal meridian (Kmin) of 40.0 D or less, and 6.0% had a Kmin of 37 D or less.
Overall, 95.6% of the patients said they were satisfied or very satisfied with the procedure, while 2.6% were neither satisfied nor dissatisfied and 1.8% were dissatisfied.
The mean scores for postoperative night-vision phenomena were 1.96 for glare, 1.89 for halo, 1.94 for starburst, and 1.39 for ghosting and double vision.
Postoperative keratometry was a small-but-statistically-significant predictor of patient-reported satisfaction, and the change in halo reports. But it accounted for only 0.3% of the variance in patient satisfaction and 0.1% of the variance in halo complaints.
In addition, it was not a significant predictor of changes in reports of glare, starburst, ghosting, or double vision.
Looking at the subcategory of patients with the highest preoperative manifest SE (-8 D or less), in which the most significant flattening of corneal surface is expected, the researchers found something surprising. The postoperative Kmin was slightly steeper in those patients whose night-vision symptoms increased compared with those whose night-vision symptoms stayed the same or improved.
This difference as not statistically significant, but it was noteworthy because of the expected association of a flatter cornea with night-vision disturbances.
Also in this subcategory, the patients who were satisfied with their visual acuity had a mean postoperative Kmin of 37.25, while the dissatisfied patients in this subcategory had a mean postoperative Kmin of 38.04. This difference was statistically significant (p = 0.02).
The researchers speculated that patients with steeper corneas might have been under-corrected, affecting their postoperative visual acuity and satisfaction.
The researchers also looked at other factors that might predict poor night vision.
“The biggest factor before surgery that predicts who will have quality-of-vision problems (glare, halos, star bursts, etc.) after surgery are people who have pre-existing quality of vision issues,” Dr. Schallhorn said.
But even that factor couldn’t account for much of what makes patients dissatisfied.
“We can explain about 20% of the reasons why patients are dissatisfied,” Dr. Schallhorn said. “However, we cannot explain fully 80% of the variance of why patients are dissatisfied.”
In this study, the researchers did not try to find a cutoff point for postoperative keratometry beyond which patients should not be treated. Because of concerns about excessive flattening, many clinicians have used a borderline of 34 D.
Only 7 patients in this sample had corneas flatter than that. All of them were “very satisfied” or “satisfied” with their visual outcomes and did not report night-vision problems. However, the number was too small for the researchers to draw statistically significant conclusions about this group.
The researchers acknowledged some other limitations in the paper.
Most important, it was retrospective. Retrospective studies are more susceptible to biases, such as selection bias, that are outside the control of the researchers, he noted.
Another weakness is that the researchers used an automated device to measure the keratometry. Also, the number of patients who have severe and significant problems is not high, making them hard to study. And symptoms tend to get better with time.
Editorial: Ten-day workweek for ophthalmologists
“On the other hand, the sample size is enormous and that allows a huge amount of statistical power in being able to analyze the data and subgroups within the population,” Dr. Schallhorn said.
Steven Schallhorn, MD
This article was adapted from Dr. Schallhorn’s presentation at the 2016 meeting of the American Academy of Ophthalmology. Dr. Schallhorn is a consultant for Abbott, AcuFocus, and Carl Zeiss Meditec, and chief medical director of Optical Express.