Study of the Efficiency and Workflow of Femtosecond Laser-Assisted Cataract Surgery in a Spanish Public Hospital

Background: To assess the time-efficiency of a designated operation room (OR) workflow in the introduction of Femtosecond laser-assisted cataract surgery (FLACS, LenSx, Alcon®). The study was carried out in a public hospital with high volume of procedures. Methods: We performed this prospective, controlled, surgical intervention study in the Ophthalmology department of a tertiary referral Spanish public hospital. A total of 167 eyes were enrolled, including 62 eyes undergoing conventional phacoemulsification surgery. In phase I, patients were assigned either to FLACS-I (n=63) or conventional phacoemulsification surgery (n=62). One surgeon operated the Femto-second laser, another finished the procedure, whereas another performed a conventional phacoemulsification. In the second phase (FLACS-II), all the surgeries were FLACS (n=42). A surgeon performed the FLACS procedure and two different surgeons completed the surgeries in separated ORs. Surgical and roll-over times of all the patients were recorded. Results: Preparation time was statistically significant lower in FLACS-I and FLACS-II (p<0.001) whereas the time of the cataract procedure per se was higher in FLACS-II compared to conventional phacoemulsification (p= 0.03). Phacoemulsification energy was higher in FLACS-II compared to FLACS-I (p = 0.01) whereas laser-related surgical time was lower (p = 0.001). Surgical complications and total surgical time showed no statistically significant differences between all three groups. Conclusions: This study suggests a time-efficient and suitable workflow model for FLACS considering the specific requirements and restrictions of a saturated public hospital. Even so we have shown that FLACS procedure does not take longer than conventional phacoemulsification following a detailed planning for the OR workflow. Besides, our data reflect an improvement surgical times in FLACS with ongoing experience. Trial registration: NCT03931629 (retrospectively registered)


Background
Cataract surgery is the most common surgical intervention conducted worldwide. 1 It is also one of the most cost-effective ophthalmological procedure, leading to great improvements in patients' quality of life. 2 Technological advances have allowed the introduction of substantial improvements in cataract surgery techniques and outcomes over the past twenty years, including the evolution from intracapsular surgery to phacoemulsification, the development of microincision surgery, the introduction of femtosecond laser technology to assist during the surgical procedure, and the development of sophisticated models of intraocular lenses (IOLs) allowing maximum levels of postoperative spectacle independence. [3][4][5] One of the most recent advances is the introduction and development of femtosecond laser assisted cataract surgery (FLACS). With the precision of currently available platforms for FLACS, some sections of the cataract surgery can be safely performed, with a completion of the surgical procedure with some conventional steps, such as nucleus and mass aspiration, IOL injection and stromal hydration. Several authors have claimed that FLACS outperforms traditional phacoemulsification surgery in terms of safety, accuracy and reproducibility. 6,7 Likewise, FLACS has been shown to be especially useful in certain circumstances, such as in shallow anterior chamber eyes, 5 very hard cataracts, 8 infantile cataracts (easier to perform posterior capsulotomy), 5,9,10 low endothelial cell counts [11][12][13] and unstable zonules. 5,8 However, recent meta-analyses have not shown statistically significant differences between FLACS and conventional phacoemulsification. 1,3 The potential advantages of FLACS over conventional procedures include lower postoperative residual refractive error possibly due to a more optimal creation of incisions 14,15 and the possibility of making additional arcuate incisions to control astigmatism, 5 and less corneal endothelial cell damage due to more reduced effective phaco time. 5,16,17 Likewise, the femtosecond laser technology allows the creation of a precise capsulotomy, facilitating a perfect centration of the IOL into the capsular bag and then avoiding the "lens tilt" phenomenon, which can lead to refractive errors despite correct biometric calculations. [18][19][20][21] Concerning the disadvantages of FLACS, one of the most relevant is the considerable initial investment required to acquire the laser platform, 22 with some concerns about the cost-effectiveness of the technique. 23 Furthermore, although there are advantages of FLACS over conventional phacoemulsification in terms of safety and postoperative refractive error, complications with conventional phacoemulsification surgery are very uncommon, compared to FLACS. 3 In contrast, a learning curve of about 100 cataracts has been described for FLACS, which is significantly shorter in ophthalmologists performing refractive surgery compared to conventional phacoemulsification. 24,25 Most of scientific evidence on postoperative outcomes and safety of cataract surgery using femtosecond laser and subsequent phacoemulsification to this date are based on studies conducted at private centres, 1 but in recent years, growing scientific evidence of the use of FLACS in public hospitals and how its introduction may affect to the efficiency of the public system is being published. [26][27][28] Besides the assessment of safety of FLACS, the aim of our study was also to evaluate the surgical room workflow and its efficiency in a public health center with a large volume of surgeries (21 cataracts per day, on average) in comparison with the conventional procedure. It should be remarked that the same surgeons perform the femtosecond-laser part of the surgery as well as the procedures in the control group following the conventional protocol to minimize the potential bias due to the "surgeon factor". All of the them were consultants with wide experience both in conventional phacoemulsification and refractive surgery.
To our knowledge, this is the first study with the aim of evaluating FLACS workflow and roll over times in a public hospital. The results of this study could provide data that would help to clarify whether investing in femtosecond technology in public health for cataract surgery can be justified.

Methods
Patient recruitment. This single-center, prospective, consecutive, interventional, and nonrandomized study was conducted at the Department of Ophthalmology of the Regional University Hospital of Málaga (Spain) from February to June 2016. Patients were recruited in this clinical setting during routine activity. Inclusion criteria for the study were signed informed consent, patients of more than 18 years old, presence of visually significant cataract (best corrected visual acuity worse than 0.4 logMAR), and possibility of performing cataract surgery under local anaesthesia (topical, sub-tenon or retrobulbar).
Exclusion criteria were patients who refused to give their consent, traumatic, congenital, luxated, subluxated cataracts, or with manifest zonular weakness, vitrectomized patients, small pupils (mydriasis 6 mm), narrow palpebral fissure, advanced optic nerve glaucomatous excavation, irregular astigmatism, significant corneal opacities, severe ocular surface disease, keratoconus, stromal scarring or edema and patients requiring general anaesthesia for the intervention. Patients who complied with the inclusion and exclusion criteria mentioned above signed a specific informed consent for the study and were non-randomly assigned to the FLACS (femtosecond laser-assisted cataract surgery) or MANUAL (manual or conventional phacoemulsification surgery) group. We checked that both groups were comparable in terms of age and gender.The three surgeons involved in the study were highly experienced consultants in performing conventional cataract surgery (from 15 to 20 years of experience, performing from 500 to 800 phacoemulsifications yearly) and refractive surgery (10 to 15 years of experience, performing from 200 to 500 refractive procedures).This study has been carried out following the Declaration of Helsinki and Good Clinical Practices guidelines. Personal data was processed according to guidelines established by the Spanish Law of Data Protection.
Study design.This study was composed of two phases and was carried out by three presumably straightforward, as they were part of the learning curve for the surgeons and were evaluated by consultant ophthalmologists (it must be considered that the maximum best-corrected visual acuity [BCVA] for being eligible for cataract surgery in the Andalusian public System is 0.4 logMAR). Patients with easily accessible sockets and good patient collaboration were included. Eligible cataracts ranged from nuclear opacities from NC1 to NC4, cortical opacities from C1 to C4 and posterior subcapsular from P1 to P4, according to the Lens Opacities Classifications System (LOCSIII); with good pupil dilation (minimum 8 mm, that was measured in the slit lamp as described by Ho et al 29 ) after the standard dilation protocol, i.e., instillation of tropicamide 1% and fenylephrine 10% eyedrops in the conjunctival sac, three times, each time separated by 10 minutes. Fuchs endothelial dystrophy and white cataracts were excluded. To ensure the comparability among groups, cataracts included in the MANUAL group were selected with similar characteristics.In phase II, one of the three surgeons operated the LensX®, while the other two surgeons completed the procedures in two different ORs (FLACS II)( Figure 2).
Surgeons rotated in subsequent days. The FLACS II group included patients with harder cataracts (nuclear opacities from NC4 to NC6, cortical opacities from C3 to C5 and posterior subcapsular from P3 to P5, according to the LOCSIII), smaller orbits, suboptimal mydriasis (from 6 to 8 mm, measured in the slit lamp as mentioned before), Fuchs endothelial dystrophy and white cataracts. (Images 1-3). Therefore, this group included the whole range of cataract surgeries that are commonly performed at a public hospital.
Procedures.Prior to surgery, the routine dilation protocol (instillation of tropicamide 1% and fenylephrine 10% eye-drops in the conjunctival sac, three times, each time separated Finally, intraoperative complications were also recorded, including loss of suction, loss of fixation/follow-up and incomplete capsulorrhexis for the femtosecond laser-assisted stage, and incomplete rexis, capsular rupture, IOL dislocation and incisions without coaptation for the phacoemulsification stage. Statistical analysis.A descriptive analysis of the variables studied including mean, standard deviation, maximum and minimum. To analyse the differences between FLACS and MANUAL groups in terms of the continuous quantitative variables, the unpaired Student t test was applied whenever the normality condition was satisfied (verified with the Shapiro-Wilk test). Otherwise, the corresponding non-parametric test was used (Mann-Whitney test). Comparisons of categorical variables were performed using the Chi-square test. A homogeneity analysis was performed between groups in terms of age and gender to ensure the comparability of groups. All these statistical analyses were performed using  (Table 1). Tables 2 to 4 display the main outcomes obtained in the two phases of the current study.
Regarding the comparison of FLACS I vs MANUAL (Table 2), both groups were comparable in terms of patient age (p=0.85). Likewise, no statistically significant differences in CDE were found between groups, although less energy was used in the FLACS I subgroup Regarding the comparison of FLACS II vs MANUAL (Table 3), both groups were also comparable in terms of patient age (p=0. 20). No statistically significant differences between groups were found in CDE (p=0.18). Preparation time (time elapsed from the arrival of the patient to the surgical premises until the entrance to the laser or operating room); was statistically significantly lower (p<0.001) in FLACS II subgroup compared to MANUAL group. In contrast, Total time did not differ significantly between groups (p=0.99). Surgical time (defined as time from the insertion of blepharostat to its removal after ending the surgical procedure) was significantly lower in the MANUAL group compared to FLACS II subgroup (p=0.005).
Finally, the comparison of FLACS I vs FLACS II (Table 4) revealed that both groups were also comparable in terms of patient age (p=0.39). The CDE used in FLACS II was significantly higher than in FLACS I subgroup (p=0.01), whereas femtosecond laser time procedure was significantly lower in FLACS II subgroup (p<0.001). Preparation time (p=0.07) and preparation plus laser time (p=0.56) did not differ significantly between subgroups. In contrast, circulation time between the laser room and the operation room where the phacoemulsification was performed was significantly shorter in FLACS II compared to FLACS I subgroup (p=0.008). Similarly, whereas femtosecond laser time procedure (the time elapsed from the moment the patient entered the OR room where the femtosecond laser procedure was conducted to the time the patient left having completed the procedure) was significantly lower in FLACS II subgroup (p<0.001). Total time between both groups did not reach significant difference (p=0.71). In contrast, surgical time (defined as time from the insertion of blepharostat to its removal after ending the surgical procedure) was shorter in FLACS I subgroup compared to FLACS II (p=0.04). No statistically significant differences were detected in intraoperative phacoemulsification time (total time elapsed since the patient enters the OR in which the phacoemulsification was performed and exits to the recovery room) between both subgroups (p= 0.99).
Complications.In FLACS group, the following complications were described: incomplete capsulorrhexis (1 eye in FLACS I and 1 eye in FLACS II), not completely opened incisions (4 eyes in FLACS I and 2 eyes in FLACS II), capsular rupture (2 eyes in FLACS I), one reconversion to conventional phacoemulsification in FLACS I and one suspended surgery also in FLACS I. In MANUAL group the following complications were reported: poor mydriasis being needed the use iris hooks (1 eye) and zonular disinsertion (1 eye). All complications were successfully managed and no permanent damage occurred in these eyes. No statistically significant differences between FLACS and MANUAL groups were found in the complication rate (p=0.56).

Discussion
In the last years, the safety, accuracy and reproducibility of FLACS has been widely reported in a great variety of studies. [5][6][7] Most of the results published have been obtained in private centers, but in recent years, growing scientific evidence of the use of FLACS in public hospitals and how its introduction may affect to the efficiency of the public system is being published. [26][27][28] Besides the evaluation of the safety of FLACS and its outcomes, the aim of the current research project was to evaluate the impact of the use of FLACS on the efficiency of a public health center with a large volume of surgeries (21 cataracts per day, on average) according to the design of intraoperative room patient circulation.
Unlike other reports 5,30,31 , in our sample FLACS procedure did not take longer than conventional phacoemulsification. In our study, "total time", defined as the time from the moment the patient enters the surgical area to the patient is discharged once the intervention is completed, was not significantly different between our groups of study, cataract surgery and they also measured the "total time" of performing a FLACS surgery, including the transportation and roll-over times. However, these authors stated that although there was statistically significant difference between FLACS and manual phacoemulsification, the results may not be indicative of practical differences. 27 Vasquez-Perez and colleagues 27 reported in another study conducted in public clinical setting that a mobile FLACS system of a different technology than that used in our study and housed in the same operating room increased the surgical duration by 5.2 minutes.
Our center is public funded, so the regular seven cataracts per OR per surgical turn had to be maintained. Therefore, an efficient workflow had to be designed and as a consequence, FLACS intervention was not more time consuming than conventional phacoemulsification in our study. We believe that our data can be mainly explained by the careful control of patients' circulation in the OR. Specifically, patients started LenSx® procedure half an hour earlier, being placed immediately afterwards at entrance of the OR, where the cataract surgery was going to be completed. It should be mentioned that the contribution of a highly motivated auxiliary staff was also very important for achieving the optimal circulation of the patient in the surgical area. Another factor that may have influenced our outcomes was the learning curve effect, with more experience and more optimization of the surgical technique as the surgeon performed more FLACS procedures. The maintenance of the regular number of surgeries after the FLACS learning curve was also reported by Ibrahim et al. 26 In any case, the development of an optimized design of the workflow is crucial for increasing the efficiency of the FLACS surgical procedures in public clinical setting. Van Vliet and colleagues 32 stated that the environmental context and operational focus primarily influenced process design of the cataract pathways. These authors concluded that multi-skilled team and pull planning were two crucial factors to consider when optimizing processes to increase efficiency in high-volume cataract pathways. 32 Regge et al 33 also confirmed that the use of specialized scrub nurses and the eye clinic design benefit flow efficiency of the cataract surgery process in hospitals. More research is still needed to determine the workflow designs that are the most adequate to obtain the maximum level of optimization of cataract surgery in public clinical settings.
Phacoemulsification procedure time was not significantly different between FLACS I subgroup and MANUAL group. Possibly, this was due to the learning curve of surgeons in FLACS I subgroup as the learning curve effect has been demonstrated to increase the surgical time. 34 Likewise, the time needed to perform the phacoemulsification procedure was significantly lower in MANUAL group and FLACS I subgroup compared to FLACS II subgroup. This is consistent with previous studies showing laser cataract surgery took longer to complete compared to conventional phacoemulsification without accounting for the time taken to complete the laser procedure itself, being this due in part to irrigationaspiration requiring more instrument maneuvers and taking longer to complete. 29,32 This fact may also be explained by the fact that more complex cataracts were included in FLACS II subgroup. Previous scientific evidence on the surgical time required with FLACS vs. conventional phacoemulsification have shown contradictory outcomes, with some authors reporting comparable surgical times, [35][36][37] whereas others concluding that FLACS was more time-consuming. 24,26,30 Some factors may account for such differences among studies, such as differences in terms of cataract hardness, surgical procedures, surgeon experience or technology used.
Comparison reports on Efficient Phaco time (EPT) in FLACS compared to conventional phacoemulsification are controversial: in some authors' work did not differ significantly 13,38,39 , whereas others reported statistically significant differences in favor of FLACS 3,16,29,37 . In our study, FLACS subgroups and MANUAL group did not show statistically significant differences. This was possibly due to the learning curve effect in the FLACS group. However, EPT was significantly lower in FLACS I subgroup compared to FLACS II subgroup, which may be explained by the inclusion of more complex cataracts FLACS II subgroup (harder cataracts, smaller sockets and mydriasis up to 8 mm). Although compared with conventional phacoemulsification, FLACS for hard nuclear cataracts has shown to conserve phacoemulsification power, provide a significant reduction in corneal endothelial damage, and lead to faster visual rehabilitation, 40 the level of EPT required even with FLACS is higher in hard than in softer cataracts.
Finally, it should be remarked that no statistically significant differences were found between MANUAL and FLACS subgroups in terms of intraoperative complications, confirming the safety of both procedures. The minimal number of complications occurring in each group were satisfactorily resolved without affecting patients' final visual outcome.
As any clinical study, our study has some limitations that should be acknowledged. First, it is a non-randomized study and the sample size is small, especially in FLACS I subgroup, where from the initial 63 patient that underwent the surgery, only 27 eyes could be included for the statistical analysis: the surgical time recorded in 32 patients was disregarded due to a protocol deviation due to the use of Verion® system (Alcon® Fort Worth, Texas, USA) between the laser procedure and the end of the procedure in the OR.
Also, in seven patients (5 in MANUAL, 2 in FLACS I), data was not properly recorded. In FLACS II six patients did not show up, probably because this part of the study was conducted in July and patients tend to be on holidays in this part of the year (Table 1).
Second, although the learning curve for FLACS is significantly shorter in ophthalmologists performing both conventional phacoemulsification and refractive surgery compared to only conventional phacoemulsification, 24,25 which is the case of the three surgeons involved in the study and provided that each of them performed prior to the study an initiation trial with 25 eyes each, their learning may have influenced the outcomes. Third, it was somewhat difficult to find "equivalent" patients for all groups given the characteristics of our center, which is a public hospital with an elderly population, multiple eye diseases and an indication for cataract surgery with BCVA of 0.4 logMAR. However, we tried to minimize the impact of this effect in the second part of the study (FLACS II) by including patients with white or harder cataracts (nuclear opacities from NC4 to NC6, cortical opacities from C3 to C5 and posterior subcapsular from P3 to P5, according to the LOCSIII), smaller orbits, suboptimal mydriasis (from 6 to 8 mm), white cataracts and/or Fuchs endothelial dystrophy CONCLUSIONS In conclusion, our study suggests that a time-efficient and suitable workflow model for FLACS can be reached considering the specific requirements and restrictions of a saturated public hospital. In our setting, a high volume of cataract procedures is managed daily, and both our target population and their cataracts are frequently nonstraightforward. Even so, FLACS procedure did not take longer than conventional phacoemulsification following a detailed planning for the OR workflow. Finally, our data also suggests an improvement of surgical times in FLACS with ongoing experience, although more challenging cases were included in our FLACS II subgroup. This study has been carried out following the Declaration of Helsinki and Good Clinical Practices guidelines. Personal data was processed according to guidelines established by the Spanish Law of Data Protection (LOPD).

Consent for publication
In the informed consent signed by the patients there was a specific mention of the possible use of the data in publications.

Availability of data and material
Data of this study are available on request. Tables   GROUP  INITIAL NUMBER  FINALLY INCLUDED  EXCLUDED  AND CAUSE  MANUAL  62

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.