Femtosecond laser–assisted cataract surgery and corneal transplantation have generated intense interest in recent years. With its ability to create extremely precise incisions in the cornea, lens capsule, and nucleus, the femtosecond laser has the potential to revolutionize anterior segment surgery. We are at a crossroads, on the cusp of widespread adoption of this new technology. As we consider its role going forward, we should be asking: 1) Does it improve outcomes? and 2) At what cost?

In this issue of the Journal , van den Biggelaar and associates answer these questions for endothelial keratoplasty (EK). They compare outcomes and costs of EK with posterior lamellar grafts prepared with the femtosecond laser (femtosecond laser–assisted Descemet stripping endothelial keratoplasty; FS-DSEK) to those prepared using standard microkeratome techniques (Descemet stripping automated endothelial keratoplasty; DSAEK) and to standard penetrating keratoplasty (PK). Outcomes were determined based on best spectacle-corrected visual acuity (BSCVA), refractive astigmatism, improvement on the National Eye Institute’s visual function questionnaire (VFQ-25), and the percentage of patients who met a combination of these effectiveness criteria.

At 6 and 12 months, mean BSCVA was significantly better in the DSAEK and PK groups than in the FS-DSEK group, though the mean gain in BSCVA was not significantly different. Mean refractive astigmatism was significantly lower in both EK groups than in the PK group. Improvement in mean VFQ-25 score was similar among the groups, and in each group was significantly higher than the preoperative value. Performance on combined effectiveness criteria was greater in the DSAEK group than in the PK or FS-DSEK groups. DSAEK was marginally more costly than PK, but costs for femtosecond-prepared tissue were $7050 (€5372) greater than for microkeratome-prepared tissue.

The authors conclude that FS-DSEK is not cost-effective compared to DSAEK or PK, even though it is clinically effective. However, outcomes might improve and costs could decrease as femtosecond laser technology matures. DSAEK, while slightly more costly than PK, is also more effective than either PK or FS-DSEK, and therefore remains cost-effective. The authors acknowledge that their study does not capture any differences between the groups in costs of follow-up beyond 1 year. Nor does it account for the economic effects of patient productivity. One might expect that faster visual rehabilitation after EK would result in greater economic productivity than after PK.

Although cost-effectiveness analyses such as this may not be as generalizable as cost-utility analyses for comparisons between specialties, they are certainly useful for comparing different techniques for treating the same disorder. Similar comparisons of femtosecond laser–assisted to standard penetrating and lamellar keratoplasty and cataract surgery techniques would be welcome.

Early reports of femtosecond laser–assisted PK and anterior lamellar keratoplasty (ALK) promised reduced astigmatism and increased wound integrity. Although improvement in acuity is more rapid and sutures are typically removed earlier (or may even be unnecessary in the case of ALK) after femtosecond laser–assisted keratoplasty, no significant long-term differences in acuity or astigmatism have been demonstrated. There have yet been no clinically meaningful controlled studies of long-term wound integrity.

There is even less data with respect to femtosecond laser–assisted cataract surgery. The capsulotomy can be done with extreme precision. Whether this results in more predictable effective lens position and more accurate intraocular lens calculations remains to be seen. A precise capsulotomy may be important for proper centration and function of some multifocal and accommodating intraocular lenses, but there have been no prospective clinical comparisons between laser and manual capsulotomies. The lens nucleus can be pretreated with the laser, but claims that this reduces complications remain speculative, with comparisons available only to historical controls. In a report of the first 200 cases of femtosecond laser–assisted cataract surgery at one institution, capsular tears attributed to the capsulotomy occurred in 4% of cases, with posterior extension in 3.5% and posterior dislocation of nuclear material in 2%. These are similar to or higher than historically reported rates for capsular tears and higher than rates for nuclear dislocation with standard phacoemulsification. A learning curve was observed, with fewer capsular tears after the first 100 cases. There have been no prospective randomized clinical trials comparing femtosecond technology to the highly successful current state of the art.

Exceptionally precise arcuate corneal incisions for correction of astigmatism can be created with the femtosecond laser. Comparisons of toric intraocular lenses (IOLs) to manual arcuate incisions suggest better outcomes with the toric IOL. There are no controlled, prospective studies comparing the predictability and stability of astigmatism reduction or costs of femtosecond arcuate incisions to either manual keratotomy or toric lens implantation.

Inability of a new technology to immediately demonstrate superior cost-effectiveness or cost-utility should not be discouraging, whether it is attributable to lack of improvement in outcomes, new or unexpected complications, or increases in costs. History shows that outcomes and efficiencies improve and complications decrease as successful new technologies mature. Equipment costs often come down with improved manufacturing methods and efficiencies of scale associated with widespread adoption of new equipment. Society and payers, however, will demand a demonstration of improved outcomes before they are willing to shoulder the burden of increased costs. As physicians and patient advocates, we must be rigorous in our analysis of outcomes. We must avoid the temptation to adopt new technologies before they are proven, based on a promise of improved outcomes or driven by financial considerations. Once pioneering clinical investigators have blazed the trails to better outcomes and greater efficiencies under the protection of well-controlled and properly consented clinical trials, we can all reap the benefits promised by femtosecond laser technology. It should not be the responsibility of our patients, directly or indirectly through their payers, to finance these studies. This is the function of industry and investors. It is their role to take the financial risks, and to realize the financial rewards when they succeed.