Introduction

Cataract extraction is the most commonly performed surgical procedure worldwide, with an estimated 18 million surgeries performed annually [1]. Phacoemulsification with manual incisions and anterior capsulorhexis has been the standard of care for cataract removal for decades [2]; however, since its inception in 2008, the femtosecond laser has been used to perform multiple steps during cataract surgery, including creation of the main corneal wound, partial-thickness corneal wounds to treat astigmatism, anterior lens capsulotomy, and fragmentation of the lens nucleus [1].

Femtosecond laser-assisted cataract surgery (FLACS) has been established to be safe in most circumstances [3–5], but its overall superiority to manual cataract surgery has not been proven. Many studies have shown that femtosecond fragmentation of the lens leads to lower phacoemulsification power requirements [6] and ultimately less endothelial cell loss [7] and post-operative corneal edema [8, 9]. Supporters of FLACS also point to the possibility of more predictable refractive outcomes due to the precision with which the femtosecond laser creates an anterior capsulotomy [10], leading to less postoperative lens decentration and tilt [11–13].

On the other hand, some data suggest that the potential benefits of FLACS are offset by its costs [14] and longer operative times [15] in comparison to manual cataract surgery. Anterior capsular tears have been reported to occur at a higher rate with FLACS, likely due to incomplete capsulotomies [16], which could potentially lead to further complications and worse outcomes. The learning curve associated with the femtosecond laser is another concern, as complications have been shown to occur more often in the first 200 cases with the femtosecond laser, even when performed by experienced surgeons [17–20]. Given these data, it is easy to understand the uncertainty surrounding FLACS being performed by resident surgeons, whose complication rates with manual cataract surgery have been shown to be higher within their first 80 cases [21].

The purpose of this study is to compare intraoperative factors and post-operative outcomes of femtosecond laser-assisted cataract surgery (FLACS) and manual cataract surgery performed by resident surgeons.

Materials and methods

The internal review boards of Baylor College of Medicine and the Michael E. DeBakey Veterans Affairs Medical Center (MEDVAMC) approved this research. All cases of FLACS performed at the MEDVAMC by third-year ophthalmology residents during the 2013–2014 academic year were retrospectively reviewed and compared to a control group of manual cataract surgery cases performed by the same surgeons on the same day. FLACS cases were performed using the LenSx® Femtosecond Laser Platform (Alcon Laboratories, Dallas, TX, USA) and supervised by experienced faculty, who were certified to use the LenSx® device. Phacoemulsification was performed with the Infiniti® Vision System (Alcon Laboratories, Dallas, TX, USA). Cases were excluded if patients had pre-existing ocular disease that limited visual acuity, including active diabetic macular edema, wet age-related macular degeneration, advanced glaucoma, and prior retinal detachment. Cases were also excluded when the medical record was insufficient or the patient was lost to follow-up.

Pre-operative data collected included patient age, comorbid conditions, grade of nuclear sclerosis, and pre-operative corrected distance visual acuity (CDVA). Intraoperative data noted were pupil size at the start of the case, intraoperative complications, and the cumulative dissipated energy (CDE) used during phacoemulsification. Data collected from postoperative month and year number one included corneal edema, CDVA, and manifest refraction. The percentage of patients achieving a refractive prediction error ±0.25, ±0.50, and ±1.00 D was also calculated.

Statistical and data analysis

All data were collected on an Excel sheet (Microsoft Office 2013), and the Data Analysis Tool Pack was used to calculate statistics. The RPE of each group was calculated by subtracting each patient’s postoperative spherical equivalent from their predicted postoperative refraction (based on the LenStar optical biometry data and the Holladay 1 formula). AMO ZCB00 and Alcon SN60WF intraocular lenses were used along with the manufacturer’s suggested A-constant values. A two-sample F-test was used to determine if there was significant variance between the two groups, and then two-sample T-test for equal or unequal variance was used to assess if the mean arithmetic RPE produced in each group was significantly different from zero. The same method was used to compare the LogMAR visual acuity and CDE values between groups. A p value < 0.05 was considered statistically significant.

Results

Seventy-six patients received femtosecond laser-assisted cataract surgery during this time period, while 101 patients had manual cataract surgery on the same day. Patients in each group were excluded (Table 2), with the most common reasons due to insufficient data and follow-up. Three patients receiving FLACS were excluded due to procedure being aborted due to suction loss or difficulty docking. This left 57 patients in the FLACS group and 68 in the manual group to be analyzed at postoperative month number one. Table 1 shows baseline patient information. By postoperative year number one, data were available for 23 patients in the FLACS group and 31 in the manual group (Table 2).

Table 1 Baseline patient information femtosecond cases with complete data
Full size table
Table 2 Reasons for patient exclusion
Full size table

Mean cumulative dissipated energy (CDE; percent-seconds) for the FLACS group was 14.54 (SD, ±7.45; range, 5.73–39.78), versus 21.61 (SD, ±11.52; range, 2.59–56.35) for the manual group (p < 0.01). Operative complications for each group are listed in Table 3 for all cases reviewed including those lost to follow-up and insufficient data and Table 4 for the cases enrolled in the study that included proper data.

Table 3 Operative complications: all cases
Full size table
Table 4 Operative complication cases with complete data and follow-up
Full size table

At postoperative month number one, CDVA (LogMAR) improved to 0.004 (±0.08) in the FLACS group and 0.024 (±0.11) in the manual group (p = 0.24). Four patients in each group had mild corneal edema at this time. Mean absolute refractive prediction errors were 0.38 (±0.24) D and 0.41 (±0.49) D in the FLACS and manual groups, respectively (p = 0.66). The percentage of patients achieving certain refractive prediction error levels at postoperative month number one is shown in Fig. 1.

Fig. 1
figure 1

Postoperative month number one refractive prediction errors

Full size image

At 1-year postoperatively, CDVA was the same in both groups. (FLACS: 0.013 ± 0.06; manual: 0.032 ± 0.09; p = 0.38). Mean absolute RPE was 0.49 (±0.63) D and 0.34 (±0.26) D in the FLACS and manual groups, respectively, at this time (p = 0.31). No patients had corneal edema at postoperative year number one. The percentage of patients achieving certain RPE levels at postoperative year number one is shown in Fig. 2.

Fig 2
figure 2

Postoperative year number one refractive prediction errors

Full size image

Discussion

To our knowledge, only two papers on FLACS performed by resident surgeons have been published. Cohen et al. were the first to publish an anecdotal report and suggest that FLACS can be safely and effectively integrated into the resident curriculum [22]. Hou et al. reviewed 62 cases of FLACS performed by residents and showed decreased phacoemulsification power and low complication rates when the cases were supervised by experienced faculty [23]. Our results echo those of these prior studies and, to our knowledge, are the first to evaluate visual acuity and refractive outcomes and provide 1-year data comparing FLACS to manual cataract surgery performed by surgeons-in-training.

Intraoperative complication rates were similar; three anterior capsular tears were noted in each group, and the only posterior capsular tears occurred in the manual group in all cases reviewed including those without follow-up and complete data. As expected, CDE was significantly lower in the FLACS group. While endothelial cell density was not measured in these patients, prior studies have shown lower CDE rates to result in less endothelial cell damage and loss. We noted only mild corneal edema in four patients from both groups; this resolved by 1-year postoperatively. It should be noted that additional laser energy in the FLACS group was used as part of the femtosecond laser, but this cannot be compared or added to CDE.

Visual acuity outcomes were excellent in both groups, with no statistically significant difference noted between the two groups at 1 month or 1 year postoperatively. RPE was low in this study, with over 95% of patients in each group within ±1 diopter of their preoperative target refraction at 1-month postoperatively. While a few prior studies have noted a trend in improved refractive prediction errors with the femtosecond laser, this was not seen in this group of patients.

A major limitation of this study is its retrospective nature; many patients were excluded due to insufficient intraoperative documentation of pupil size and CDE, which could have been prevented in a prospective study. Additionally, many patients were not seen at postoperative year number one. This is likely due to the large referral area encompassed by the MEDVAMC in southeast Texas and Louisiana. Patients with no other ophthalmic issues and who are doing well at their 1-month postoperative appointment are routinely discharged back to closer, outlying VA Health Centers for routine care.

While data analysis showed no statistically significant pre-operative differences between the two groups regarding age, grade of nuclear sclerosis, pupil size, and corrected visual acuity, the authors acknowledge that selection bias was present when choosing patients suitable for training residents to use the femtosecond laser. Patients were selected who were less anxious and did not have a small palpebral fissure or prominent brow, as these patients could be difficult to dock to the laser. Although Table 3 shows three posterior capsule tears in the manual group, pupil size was not noted, so it is difficult to assess if that played a role since that may make the case more complex.

Future prospective studies in resident performed cases could measure pre- and postoperative endothelial cell density, monitor intraoperative pupillary size, determine the effectiveness of the femtosecond laser for astigmatic correction, and compare operative times between FLACS and manual cataract surgery.

In summary, this study confirms femtosecond laser-assisted cataract surgery performed by resident surgeons is safe and effective and results in decreased CDE and equivalent complication rates, visual acuity, corneal edema, and refractive outcomes at 1 month and 1 year when compared to manual cataract surgery.