Comparison of femtosecond laser and mechanical microkeratome for flap thickness accuracy

In their article comparing flap thickness results with 2 microkeratomes and a femtosecond laser,1 Rosa et al. focused discussion of the results on (1) how far the flap measurements differed from the targeted flap thickness and (2) the effect on flap thickness measurement of waiting 20 minutes after femtosecond flap creation to measure the stromal bed. The results showed a mean deviation from target that was greatest for the femtosecond laser when measured immediately after flap creation but least for the femtosecond laser when measured 20 minutes after flap creation. They theorized that temporary stromal dehydration after femtosecond flap creation may account for the difference when the readings were 20 minutes apart. Based on this study, waiting 20 minutes to measure a femtosecond-created flap thickness appears to add to the accuracy of the measurement itself, but it does not address the superior accuracy of the femtosecond laser, whether measured immediately or 20 minutes later, compared with that of either microkeratome documented in the study and in other articles. The authors focus much of their attention on the mean flap thickness and how close to the “target” the mean result was. However, it is not the mean result that is most clinically relevant but the standard deviation (SD) of that mean and the range. If a surgeon knows the microkeratome or femtosecond laser creates flaps that have a certain mean result off target, he or she simply takes that into account when planning the desired residual postoperative stromal bed thickness. In the case of a femtosecond laser, adjustments can be made by technical staff to bring the mean result closer to the target. But the key point is that the surgeon must be able to rely on a consistent mean to plan a safe outcome. In this regard, it is very important for the surgeon to be aware of the SD of the mean result and the range of that result. In the study, the authors present a mean SD based on immediate readings of 24.9 μm for the Hansatome, 23.8 μm for the Zyoptix, and 18.4 μm for the femtosecond laser with an immediate measurement and 12.5 μm with a measurement 20 minutes later. This certainly shows a tendency for less variability for the femtosecond laser even when measured immediately. These results are similar to those reported in 3 studies cited by the authors. Kezirian and Stonecipher2 showed the SD to be 14 μm for flaps created with a femtosecond laser, 26 μm for those created with the Carriazo-Barraquer microkeratome, and 29 μm for those created with the Hansatome. Patel et al.3 reported an SD of 16 μm for femtosecond laser–created flaps and 22 μm for Hansatome-created flaps. Javaloy et al.4 reported an SD of 1.70 μm for the femtosecond laser and 14.03 μm for the Moria M2. However, all these results were for one SD only; ie, 66% of cases. We can learn more from the data presented for the range of flap thickness. Of all the measurements presented, the range of the results is the most clinically relevant. Risk increases at the extremes of thickness, whether that be on the thin (buttonholes, partial flaps) or thick (ectasia) end of the spectrum. The range for the Hansatome was 96 μm, for the Zyoptix 93 μm, for the femtosecond laser 57 μm (immediate measurement) and 58 μm (measurement 20 minutes later). Although waiting 20 minutes caused the measurement of the flap to be closer to target, it did not change the range of flap thicknesses. The femtosecond-created flaps had a significantly smaller range than either microkeratome. This is corroborated by the results of the 2 other studies: Kezirian and Stonecipher2 measured a range of 77 μm for the femtosecond laser, 151 μm for the Carriazo-Barraquer microkeratome, and 225 μm for the Hansatome, and Patel et al.3 measured a range of 62 μm for the femtosecond laser and 85 μm for the Hansatome. Thus, Rosa et al. present valuable data in regards to the accuracy of the various flap-making technologies discussed. I would encourage others, when interpreting such data, to recognize the greater clinical value of placing emphasis on flap variability and not on the proximity of the mean result to the target.