Abstract
Purpose:
To compare the efficiency, efficacy, and safety, as well as the educational value, of heads-up (three-dimensional visualization system–assisted) and traditional microscopic cataract surgery.
Methods:
This randomized noninferiority trial enrolled 242 eyes of 201 patients who received femtosecond laser-assisted cataract surgery. The questionnaire study enrolled 26 medical interns and 39 medical students. Patients received surgery under either a three-dimensional visualization system (3D group, 117 eyes) or traditional microscope (TM group, 125 eyes) after random allocation. The primary outcome was surgical time. The noninferiority margin of surgical time was 60 seconds. Secondary outcomes included ultrasound power, phacoemulsification time, visual acuity, intraocular pressure, endothelial cell density, central corneal thickness, complications, and observer satisfaction scores for surgical procedures.
Results:
Surgical time was 462.03 ± 80.36 seconds in the 3D group and 452.13 ± 76.63 seconds in the TM group (difference 9.90 seconds; 95% CI, –9.98 to 29.78; P = 0.365). Visual acuity and other perioperative parameters were comparable between the 3D group and the TM group (all P > 0.05). Incidences of both intraoperative and postoperative complications were low and not statistically different between groups (all P > 0.05). Across all observers, 3D surgery was superior to TM surgery for improving the degree of satisfaction (all P < 0.001).
Conclusions:
The surgical efficiency of heads-up cataract surgery is not inferior to traditional microscopic surgery. Both methods achieved similar efficacy and safety outcomes. Moreover, heads-up cataract surgery showed a significant advantage in medical education.
Translational Relevance:
Our findings show that heads-up cataract surgery has comparable efficiency, efficacy, and safety, as well as superior medical educational value, to TM surgery, which lays the foundation for promoting and popularizing this technology.
Cataract extraction with intraocular lens (IOL) implantation is the most established method of cataract management so far. However, conducting surgery through a microscope might result in an uncomfortable body posture with chronic cervical and lumbar fatigue for some surgeons, causing cumulative work-related musculoskeletal disorders that may shorten their career.
1
Unlike traditional microscopic (TM) surgery, heads-up cataract surgery involves two high-definition cameras that capture image signals from different angles of view under the microscope. These are then processed by an image processor and transmitted to a high-resolution three-dimensional (3D) screen. Surgeons wearing passive polarized 3D glasses perform microsurgical procedures by directly viewing screen; they do not have to look down through the microscope eyepieces during the surgery.
Heads-up surgery has the advantage of high-resolution visualization, superior stereoscopic sensation, and wide visual field.
2 Its appropriate ergonomic design enables a more natural body posture and reduces the burden on surgeons’ cervical spines.
2–5 Moreover, 3D imaging allows observers to see exactly what the surgeon sees during the surgery, improving effectiveness in teaching and learning.
4,6,7
Since first developed in 2009,
8 heads-up surgery has gained increasing popularity in ophthalmic surgery, including Descemet membrane endothelial keratoplasty (DMEK),
9,10 cataract surgery,
11–14 and vitreoretinal surgery.
4,15–19 Current reports on heads-up surgery mainly focus on its use in vitreoretinal disease because of its advantage of a lower endoillumination level and higher depth of field.
4,15–19 However, as manipulation space is smaller in the anterior segment compared with the posterior segment, heads-up cataract surgery is performed less frequently than TM surgery.
A retrospective study reported the efficiency and safety of heads-up surgery on aspects of surgical duration and complications.
14 However, to date, no prospective randomized trial with a large sample has comprehensively evaluated this technology.
The present study evaluated the efficiency, efficacy, and safety of heads-up cataract surgery compared with TM surgery, as well as its educational value in intraoperative surgical procedures.
This randomized clinical trial was conducted from February 10, 2020, to June 12, 2020, at the Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, China, after obtaining approval from the institutional review board and in accordance with the Declaration of Helsinki. It was registered with the Chinese Clinical Trial Registry (ChiCTR2000029466) in February 2, 2020. Written informed consent was obtained from all patients after full explanation of the study.
The primary outcome was surgical time (initiated by side-port corneal incision and finished with hydration of primary corneal incision). The secondary outcomes were other perioperative clinical parameters.
All eligible patients were interviewed regarding their medical histories and underwent a comprehensive ophthalmic examination, including uncorrected distance visual acuity (UDVA), slit-lamp biomicroscopy, dilated fundoscopy, intraocular pressure (IOP) measurement (NT-510; Nidek, Gamagori, Japan), and corneal topography by Scheimpflug imaging (Pentacam; Oculus Optikgeräte GmbH, Wetzlar, Germany). The nuclei were graded according to the Emery–Little classification.
The evaluated intraoperative parameters included the surgical time, ultrasound (US) power, absolute phacoemulsification time (APT), and effective phacoemulsification time (EPT). Intraoperative complications—including posterior capsule rupture with or without vitreous loss, intraoperative iris floppy syndrome or iris prolapse, iris injury, lens materials dropped into vitreous and suprachoroidal effusion with or without hemorrhage— were recorded.
UDVA, IOP, endothelial cell density (ECD), and central corneal thickness (CCT) were measured at 1 day and 1 week postoperation. UDVA measured with the Snellen visual acuity chart was converted to the logarithm of the minimum angle of resolution equivalent for statistical analysis. ECD was measured with a noncontact autofocus EM-3000 specular microscope (Tomey, Nagoya, Japan). Endothelial cell loss (ECL) was calculated as a percentage of the preoperative ECD. CCT was measured using Scheimpflug imaging.
All measurements were conducted by the same technician, who was masked to the patients, and the conditions were kept consistent for all operated eyes. Postoperative complications were recorded, including IOP spike (defined as IOP ≥25 mm Hg or an elevation of IOP ≥8 mm Hg from the baseline IOP
22), corneal edema, toxic anterior segment syndrome, intraocular lens decentration and dislocation, retained lens materials, wound leak or rupture, hyphema, and endophthalmitis.
The trial was framed as a noninferiority design to demonstrate that surgical time of heads-up cataract surgery is not inferior to that of traditional microscopic cataract surgery. We set the noninferiority margin of surgical time for the primary outcome at 60 seconds based on published data
14 and expert opinion. Interpretation of the trial results is based on the 95% confidence interval (CI) for the difference that lies wholly to the left of the noninferiority margin. If there is no difference between groups in surgical time, then 58 patients (29 per group) would provide 90% power to assess the noninferiority at a one-sided 2.5% significance level,
23 assuming a common standard deviation (SD) of 70.2 seconds.
14 For secondary outcomes, we explored evidence of a difference, rather than noninferiority.
Statistical analyses were performed using SPSS software (version 24.0; SPSS, Inc.). The Kolmogorov–Smirnov test was used to evaluate data distribution normality. Continuous data were expressed as mean ± SD, and between-group statistical comparisons were made by independent sample t-test or the Mann–Whitney U test, depending on normal distribution. Categorical data were described in numbers and percentages, and a Pearson χ2 analysis or a Fisher exact test was performed appropriately. A paired-sample t-test or a Wilcoxon signed-rank test was carried out to compare the questionnaire score data.
To account for the intereye correlation of some patients, we performed a sensitivity analysis on the perioperative parameters by analyzing data with marginal linear regression models using generalized estimating equations (GEEs).
24,25 An exchangeable correlation structure was used in the models.
Two-sided 95% CIs of the difference between groups were reported for each outcome measure. Two-sided P values less than 0.05 were considered statistically significant.
To our knowledge, this is the first study to report the efficacy of heads-up cataract surgery. Regardless of anterior or posterior segment surgery, the present study enrolled more than 240 eyes, making it the largest prospective randomized trial to comprehensively compare the clinical outcomes of heads-up surgery and TM surgery to date. We found similar intraoperative and postoperative parameters as well as complication incidence between the two groups. Moreover, the questionnaire results revealed that heads-up surgery was more valuable in medical education.
An increasing number of studies, including three small prospective trials with approximately 50 eyes in total,
4,16,18 have revealed 3D surgery to be as safe and effective as TM surgery for vitreoretinal diseases.
15,17,19 For the anterior segment, only a few reports focused on the application of 3D systems in DMEK
9,10 and cataract surgery.
11–14 Weinstock et al.
14 discussed their initial experience of using the 3D visualization system for cataract surgeries in retrospect and reported that the surgical time and complications were comparable between the two groups. Another trial comprising 20 eyes also evaluated the feasibility of 3D cataract surgery.
13 However, the small sample size may not have been able to provide sufficient statistical power, which might limit the generalizability of the findings. Our result showed that operation time was not significantly different between the two types of surgery performed by a single surgeon while controlling for other factors, indicating the feasibility of new surgical manipulation.
There is well-established evidence that ultrasound energy consumption is positively related to ECL.
26–28 In the present study, similar intraoperative US power, APT, and EPT were observed in the two groups, which was consistent with the results of postoperative ECD and ECL. As expected, elevations in postoperative IOP and CCT were found in both the 3D and TM groups but were not of statistical significance. These results indicate that 3D surgery may not increase the inflammatory response and corneal edema over what is induced by TM surgery. Moreover, postoperative visual outcomes were improved and statistically equal between the two groups at both 1-day and 1-week follow-up.
All results mentioned above indicated that 3D surgery was not time-consuming and had comparable visual improvement and recovery efficacy to TM surgery. In addition, in line with a previous study,
14 both groups displayed a similar and low incidence of complications, further demonstrating the safety of heads-up cataract surgery.
Heads-up technology may improve the teaching and learning of intraoperative procedures. In some cases, only surgeons could understand surgical details, but they could not communicate it to medical students by words or two-dimensional image. In 3D surgery, observers can see exactly the same high-quality stereoscopic surgical experience as a surgeon by wearing 3D glasses, helping them to observe more details and improve their understanding and knowledge retention.
6 Therefore, in our results, medical interns and medical students’ satisfaction scores for 3D surgery were significantly better than for TM surgery, especially in terms of depth of field and educational value. These results were consistent with those of recent studies, which found that 3D technology was more comfortable for both beginner surgeons and observers regarding vitreoretinal diseases.
4,7 Eckardt et al.
29 reported that heads-up surgery was particularly suitable for surgical training in a situation where a teacher used cellular phones to instruct a trainee to perform surgeries. In that case, the teacher led the trainee through each individual step of the surgery, commenting frequently and in great detail. The trainee perceived this method as being more effective than receiving short, direct comments. An additional potential advantage of 3D systems with heads-up display in the teaching field is that the surgical assistant and students are able to view the steps of surgery while maintaining adequate interpersonal distance, which has become important since the coronavirus disease 2019 pandemic began.
30
Heads-up surgery has several advantages over TM surgery. In addition to the educational value mentioned above, the most beneficial advantage is its ergonomic design, which enables a more physiologically comfortable and stable body posture to relieve fatigue and musculoskeletal stress for surgeons, thereby extending their careers.
2–5 This view is in accordance with some surgeon-oriented questionnaire studies.
4,5,7,11,31
With the development of 5G data transmission and virtual reality technology, ophthalmologists are expected to experience remote cataract surgery broadcasting easily in the near future, which will benefit learning and communication using surgical technology. Heads-up surgery may overcome the visualization limitations of standard microsurgery with increased magnification, extended depth of field, and improved depth resolution, enabling surgeons to distinguish intraocular tissue structure better.
2 In addition, 3D visualization system can decrease illumination to reduce the risk of phototoxicity.
5,32,33 With the screen being approximately 2.2 m away from the surgeon, as shown in
Figure 2, a wider field of view is available to the surgeon compared with looking down through the eyepiece. The surgeon can magnify the image to a high magnification without experiencing any discomfort or eyestrain under the same microscope magnification, thereby relieving eye fatigue.
2 In addition, surgeons can more conveniently receive instruments from technicians by observing from the corners of their eyes, which improves operational efficiency.
The current technology still has room for improvement. The learning curve is the primary issue with implementing a new technology. The setup of the 3D screen, best surgical posture, and magnification vary by surgeon and require gradual adjustment and adaption. However, the learning curve seems to be short, since it has been reported that just a few surgical practices are enough to make the surgeon feel familiar and comfortable with 3D surgery.
2,7,31
Some 3D visualization systems only retain the heads-up path and cover the microscopic eyepiece. This design may cause issues for beginning heads-up surgeons if they encounter complex or unexpected situations. For example, eye socket hydrops occurs more often in patients with narrow palpebral fissure, resulting in increased reflected light. In this circumstance, microsurgical procedures conducted with a viewing screen can be difficult and unsafe. Therefore, both a heads-up path and microscope eyepiece should be available and switched between as needed in surgery.
In terms of the visualization system, system delay between the steps of surgery and the video projected on the screen is difficult to avoid and can lead to deviations in sophisticated operation. The lag may be more evident in anterior segment surgeries due to the higher instrumental speed during surgical manipulations. However, in the present study, we did not find the duration to be extended significantly in 3D surgery performed by an experienced surgeon. The surgeon also did not report any delay between the procedure and the screen display throughout the cataract surgery.
The present study had several limitations. First, the follow-up period was only 1 week. For patients who received uncomplicated small-incision surgery, they usually recovered well a short term after surgery. We set the follow-up period as 1 week because the efficacy and safety outcomes were supposed to generally get stabilized at 1 week.
20 Second, to maintain better surgical homogeneity, only one experienced surgeon used both approaches to perform surgery. In the future, multicenter randomized controlled trials with multiple surgeons and a larger patient sample size should be conducted to confirm our observations and evaluations. Third, in our questionnaire study, the observers rated their satisfaction with knowledge retention and educational value by completing the 10-point scale questionnaire rather than the posttest assessment. However, none of the observers had the qualification to perform a cataract surgery. As such, the aim of our questionnaire study is to preliminarily evaluate the perception of the observers regarding the learning-teaching methodology. Follow-up studies are warranted to further investigate the teaching value of the heads-up cataract surgery by posttest assessment. Fourth, even with a sample size of more than 100 eyes per group, we cannot comment on the incidence of the rare but severe complications that do occur with cataract surgery. Fifth, the statistical comparisons in our study were not adjusted for multiple comparisons. Sixth, the learning curve of surgeons is worthy of further investigation.
In conclusion, the current study prospectively demonstrates that heads-up cataract surgery has comparable efficiency, efficacy, and safety to TM surgery in femtosecond laser-assisted cataract surgery when an experienced surgeon is performing both procedures. Moreover, heads-up surgery shows a significant advantage in terms of teaching and learning intraoperative surgical procedures. It is hoped that with continuous refinement, the 3D visualization system can shorten system delay and improve resolution to serve patients, surgeons, and the development of ophthalmic surgery better.
Supported by the National Key R&D Program of China (2018YFC1106104) and Key Research and Development Project of Zhejiang Province (2020C03035).
Disclosure: K. Wang, None; F. Song, None; L. Zhang, None; J. Xu, None; Y. Zhong, None; B. Lu, None; K. Yao, None