Secondary Logo

Journal Logo


Propensity and quantification of aerosol and droplet creation during phacoemulsification with high-speed shadowgraphy amid COVID-19 pandemic

Shetty, Naren MD; Kaweri, Luci MD; Khamar, Pooja MD; Balakrishnan, Nikhil MD; Rasheed, Abdur MTech; Kabi, Prasenjit PhD; Basu, Saptarshi PhD; Shetty, Rohit MD, PhD, FRCS; Nuijts, Rudy M.M.A. MD, PhD; Sinha Roy, Abhijit PhD

Author Information
Journal of Cataract & Refractive Surgery: September 2020 - Volume 46 - Issue 9 - p 1297-1301
doi: 10.1097/j.jcrs.0000000000000289
  • Free
  • Associated Video

Severe acute respiratory syndrome coronavirus (SARS-CoV)-2 infection has been declared as a global pandemic by World Health Organization since March 2020.1 Its spread to healthcare workers due to close proximity to patients has been a reason for concern all over the world.2 Person-to-person transmission occurs primarily through direct contact or droplets spread by coughing or sneezing from an infected individual.3 Surgical procedures are also deemed a risk factor after a series of surgeons were reported as COVID-19 positive in China.4 There is a worldwide concern that phacoemulsification, which is the most widely performed type of cataract surgery at present, might release aerosols and droplets. These might transmit the infection within the operating room and its air-handling units. Based on this hypothesis, recommendations were made in the past to stop elective cataract surgeries.5 SARS-CoV-2 virus might be present in the tears and conjunctival secretions of infected COVID-19 patients.6,7 Because the prevalence of the virus in the conjunctiva is very low, the transmission of the same through ocular surface and ocular fluids has been controversial.8

Aerosol- and droplet-generating medical procedures release particles as small (size <20 μm) and large droplets (>20 μm).9 Aerosols and droplets of size 6 μm or greater might get trapped in the upper respiratory tract.9 High-speed shadowgraphy is a widely used imaging technique to study these particles.10 It uses a strobe light source such as a pulsed laser or LED, which is focused toward the camera to create a white image. The dark outline of fast-moving objects can then be captured using a sufficiently fast shutter (short exposure time). In this study, we investigated the propensity of aerosol and droplet generation during phacoemulsification using high-speed shadowgraphy.


This experimental study was approved by the institutional research and ethics committee of Narayana Nethralaya Multispecialty Hospital, Bangalore, India, and conducted in accordance with the tenets of the Declaration of Helsinki. This approval was secured for the part of the study involving the use of human ocular tissue. The study was performed in collaboration with the Indian Institute of Science, Bangalore, India. The study design was as follows: (1) Enucleated eyes of goats were carefully inspected for uniformity and clarity of the ocular surface. The prepared eyes were mounted on a mannequin head and were exposed in a way such that the cornea, limbus, and some part of sclera were clearly visible for surgical maneuvering (Figure 1, A); (2) The human corneoscleral button was loaded onto a Barron artificial anterior (Figure 1, B) chamber (Katena) connected by tubing to a 10 mL syringe filled with a balanced salt solution. The Visalis 100 (Carl Zeiss Meditec AG), a peristaltic pump device with titanium straight tips (21G biconical 2.2 straight 15 degrees and 20 G 2.8 straight 30 degrees), was used for the procedure. The sculpt (80 mm Hg vacuum, 18 mL/min flow rate, and 40 µm of ultrasound) and quadrant removal modes (350 mm Hg vacuum, 34 mL/min flow rate, and 60 μm ultrasound) were used in linear and fixed modes (Figure 1, C).

Figure 1.
Figure 1.:
Experimental set up. A: Enucleated goat eye mounted on a mannequin after ensuring regularity and clarity of ocular surface. B: Cadaveric human corneoscleral rim loaded on Barron artificial anterior chamber connected by tubing to a 10 mL syringe filled with a balanced salt solution. C: Settings for quadrant removal mode of phacoemulsification on Visalis 100. D: Optical setup for high-speed shadowgraphy using the Mini-UX100 high-speed CCD camera coupled with macrolens for imaging.

The shadowgraphy technique involved the use of high-speed camera. The Mini-UX100 (Photron USA, Inc.) was coupled with a macrolens (ATX 100, 100 mm, f2.8D; Kenko Tokina, Co., Ltd.) for imaging. The resolution of the camera was 1280 × 1024 pixels. The aperture was set to f/32 for maximum depth of field. The illumination using a high-power LED source (Constellation 120, Veritas) was positioned opposite to the camera and used in a continuous mode. The experimental setup was inserted in between the light and the camera for high-speed shadowgraphy. The camera was manually triggered to acquire images during the procedure. The acquisition rate was fixed at 500 frames per seconds (fps) and at a shutter speed of 1/16,000 second (Figure 1, D). In both models, that is, the goat eye on mannequin and cadaveric human corneoscleral rim in the artificial anterior chamber, only main port incisions were made. The tips and sleeves were used in the following combinations:

  1. 2.2 mm incision and 2.2 tip and sleeve;
  2. 2.8 mm incision and 2.2 tip and sleeve;
  3. 3.2 mm incision and 2.2 tip and sleeve;
  4. 2.8 mm incision and 2.8 tip and sleeve;
  5. 3.2 mm incision and 2.8 tip and sleeve.

Nucleotomy was performed in sculpt and quadrant removal mode after embedding the phacoemulsification tip within the nucleus. The same was repeated after exposing the tip of the phacoemulsification probe and keeping it in close proximity to the ocular surface.

A simple 1D calculation was used to calculate the spread distance of the droplets. An airborne droplet of diameter D and with a horizontal velocity ud can be acted on by the ambient convection velocity (uair), which is introduced by the high-efficiency particulate air filtration system present in the operating theater.11 Under such circumstances, the appropriate governing drag equation was as follows11:(1)duddt=92r2μfρdurel(uairud)urel,where urel=(uairud)2+vd2, r = D/2, vd was the settling rate (measure of velocity) of the droplet, μf was the viscosity of air (=18.37 × 10−6 Pa·s), ρf was the density of air (=1.184 kg/m3), ρd was the density of droplet (=997 kg/m3), g was the acceleration due to gravity (=9.81 m/s2), and D was the diameter of the droplet. The value of uair was estimated to be approximately 0.6 m/s, which was as per the certified inspection report of the operating theaters at the Narayana Nethralaya eye hospital. The rated capacity of the air-handling units in our operating theaters (3 in number) ranged from 2364 to 2653 ft3/min, and the number of air changes per hour was not less than 40. The droplet evaporates and settles under gravity simultaneously. The evaporation timescale can be estimated from the D2 law, whereas the appropriate settling timescale can be estimated from the Stokes equation as follows11:(2)vd=(ρdρf)gD218μf

Assuming the point of droplet ejection (location of human cornea) was approximately 1 m from the floor, the timescale of droplet settling on a surface was obtained from equation 2 as follows:(3)ts=1vd=18μf(ρdρf)gD2

By multiplying the droplet velocities calculated from the shadowgraph images with ts, their spread distance was calculated. These distances were plotted as a function of droplet diameter. The shadowgraph image was binarized and analyzed using “Particle Analyzer” plugin in ImageJ (open source JAVA-based image processing software) to quantify the diameters of the aerosols and droplets.


In the goat eyes mounted on a mannequin, phacoemulsification was performed using different incision and tip sleeve combinations. The dynamic optical images were recorded as black shadows against a white background using high-speed shadowgraphy. There was no release of aerosols in a closed chamber surgery where the size of tip sleeve and incisions were similar (Figure 2, A and Video 1, available at The leakage of fluid from the main wound was seen on the shadowgraph but there was no generation of aerosols when the incision size was larger than the sleeve tip size (Figure 2, B and Video 1, available at The atomization of water droplets and generation of aerosols was noted when the tip was exposed and close to the liquid film on the corneal surface (Figure 2, C and Video 1, available at Figure 2, D shows leaking liquid from the main wound (blue arrow) but no optical shadows of aerosols. Figure 2, E shows atomization of a balanced salt solution when the exposed phacoemulsification tip was kept close to the ocular surface. Figure 2, F shows a high-speed shadowgraphy image where optical shadows of large droplets released close to the exposed part of the tip are indicated by the blue arrow.

Figure 2.
Figure 2.:
Clinical and shadowgraphy images of enucleated goat eye mounted on a mannequin. A: Clinical photograph of goat's eye with phacoemulsification probe and 21 G biconical 2.2 straight 15-degree titanium tip inserted through a 2.2 mm corneal main wound. B: High-speed shadowgraphy image showing no optical shadows of aerosols captured during microincision cataract surgery. C: Clinical photograph of phacoemulsification probe and 20 G 2.8 straight 30-degree titanium tip inserted through 3.2 mm corneal main wound. D: High-speed shadowgraphy showing leaking fluid from the main wound (blue arrow) but no optical shadows of aerosols. E: Atomization of a balanced salt solution was seen when exposed phacoemulsification tip was kept close to ocular surface. F: High-speed shadowgraphy image showing optical shadows of large droplets released close to the exposed part of the tip (blue arrows).

The procedure was repeated on cadaveric human corneoscleral rim mounted on an artificial anterior chamber model. The results were similar to those in the goat eye model. The aerosols were seen only when the tip was exposed and kept close to the corneal surface (Figures 3, A and B and Video 1, available at The aerosol formation happened only when the exposed part of the tip came in direct contact (Figure 4, A) with a balanced salt solution outside the ocular surface. The nominal droplet size distribution was extracted from the shadowgraph and illustrated in Figures 4, B and C. Figure 4, D shows that the nominal diameter of the aerosolized droplet was approximately 50 μm. As shown in Figure 4, E, the spread distance of the droplet increased for diameters ranging from 30 to 50 μm and then decreased monotonically for larger droplets. From the high-speed shadowgraphs, the value of ud used for these calculations was estimated to be approximately 1 m/s.

Figure 3.
Figure 3.:
Clinical and shadowgraphy images of cadaveric human corneoscleral rim mounted on Barron artificial anterior chamber. A: Clinical photograph showing aerosol generation at the tip of exposed phacoemulsification tip kept close to the main wound. B: High-speed shadowgraphy image showing shadows of large droplets released close to the tip.
Figure 4.
Figure 4.:
Quantitative analyses of droplets created when probe was close to the ocular surface. A: Shadowgraph of the cadaveric human corneoscleral rim. B: Magnified view of the same. C: Image is binarized and analyzed using “Particle Analyzer” plugin in ImageJ. Red circles indicate the detected droplets. D: Droplet size distribution. E: Plot of droplet spread distance vs droplet diameter. (Inset) shows the condition of calculations where the height of the cornea on the operating table was assumed to be ∼1 m, the velocity of the ejected droplet was u d, the settling velocity of the droplet was v d, and the room convection was u air.


Aerosol generation during phacoemulsification became a major concern among ophthalmic surgeons after dissemination of videos of aerosol production in fluorescein-stained saline water by Wong et al.12 Darcy and Liyanage demonstrated profuse aerosolization using corneoscleral rim mounted on an artificial anterior chamber model.13 In another video using a model eye, Wong et al. highlighted that there was less chance of aerosol visible outside the eye and even lesser with forward movement of phacoemulsification probe.14 The integrity of the model eye was dissimilar to the normal cornea. Therefore, the earlier results cannot be extrapolated to the human cornea.12–14 We used an enucleated goat eye in this study because its role as a training model for phacoemulsification has already been established.15 The artificial anterior chambers are used in corneal graft surgeries. We used human cadaveric corneoscleral rim mounted on an artificial anterior chamber to simulate the experimental setting used by Darcy and Liyanage.13 Published literature is not available to validate the efficacy of this model for phacoemulsification.

The shadowgraph technique is based on the shadow cast on the recording plane (in this case, the camera sensor) because of refractive deflection of the incident light rays caused by the density difference between air and object (aerosols and droplets in this case).16 Although an imaging technique similar to the Schiliren17 might possess better resolution and lower detection limit, shadowgraph was simpler to implement and adequately efficient when performed at high frame rate (500 fps) and using a fast shutter speed (1/16,000 second). Because the particle size in aerosolized pathogen transmission ranged between 0.05 μm and 500 μm, the resolving power of the imaging system in addition to the acquisition rate was also critical in determining the lower limit of the droplet detection.18 Based on the camera settings and illumination light source power spectrum, a resolution of approximately 20 µm per pixel was possible (Figures 3 and 4).

Phacoemulsification works on the principle of piezoelectric effect.19 There was a significant amount of heat generated at the incision site during the procedure, which was usually dissipated from the wound by the irrigating liquid.20 In our setup, we used the longitudinal or axial mode of phacoemulsification because it caused more heat build-up at the tip in comparison with the torsional mode.21 Because there was a lot of heat generated during the procedure, the formation of aerosols was expected. In contrast to the earlier videos, we did not see any aerosol generation in either of our models.12–14 This was true for coaxial microincision cataract surgery and standard phacoemulsification performed in a closed chamber. There was minimal liquid leak seen at the main wound.22 This finding was independent of the energy used in sculpt or quadrant mode. In cases where the incision was large, there was a slow leak taking place at the main port. There were no visible aerosols (small droplets) generated even in this setting. This is because the exposed part of the tip was near the center of the eye and in direct contact with the nucleus. The part in contact with the wound was cushioned by the sleeve, and the continuous flow of liquid on its side dissipated all the heat that was generated.

Aerosol formation happened only when the exposed part of the tip came in direct contact (Figure 4, A) with a balanced salt solution because of the atomization of salt solution secondary to the heat generated. With a nominal droplet size of approximately 50 μm, the spread of infection is highly dependent on the distance traveled by the ejected droplet. The final distance traveled by the droplet was dependent on both the evaporation time scales and was determined from the smaller of the 2 quantities. As shown in Figure 4, E, the spread distance could be as high 1.3 m or 4.3 feet. This distance decreased with larger droplet sizes because the droplet motion was governed by the settling timescale for larger droplets and by the evaporation and settling timescales for smaller droplets. If actual ophthalmic viscosurgical device materials were used instead of BSS (Figures 3, A, B and 4, A), then the generation and spread of aerosols and droplets was intuitively expected to be less because the viscosity of ophthalmic viscosurgical materials is greater than a balanced salt solution. Thus, this would make ophthalmic viscosurgical materials more resistant to aerosol and droplet formation. But, these findings were not relevant to the actual scenario of phacoemulsification because phacoemulsification is always recommended to be performed inside the anterior chamber away from the corneal lip.

To the authors' knowledge, there are no studies in literature that demonstrated visible or invisible (to the naked eye) aerosol generation during phacoemulsification. We did not test longitudinal and torsional modes separately. We used a straight tip only. Flared tips have been known to produce more heat at the exposed part.23,24 Therefore, the same results might not be applicable to other tips. This study was performed with the aim of assessing the risk of transmission of SARS-CoV-2 virus from infected patients to the operating surgeon, nurses, and technicians during the procedure. However, these results cannot shed light on the viral load present in invisible aerosol. The SARS-CoV-2 might be transmitted through asymptomatic carriers and might occur in the conjunctiva of 2% of symptomatic patients.6,7,25 Furthermore, the aerosols generated were from the balanced salt solution moving within the eye and not from the aqueous humor because aqueous is washed off after making an entry and injecting ophthalmic viscoelastic gel in the eye.

Virus present on the ocular surface could pose a threat to cataract surgeons.6 The treatment with topical povidone–iodine for 2 minutes, prior to cataract surgery, reduced the virus count to below detectable levels.26 To conclude, phacoemulsification can be considered as a safe surgery with minimal or no risk of aerosol generation. To the authors’ knowledge, there is no known transmission of COVID-19 through phacoemulsification, and this study supports it. We recommend covering the face with a mask during the procedure to minimize the risk from invisible or smaller droplets. Cleaning of the ocular surface with Betadine before the procedure, use of similar-sized tip, sleeve, and incision, and ultrasound power usage only after the tip is completely inside the anterior chamber would reduce or eliminate the risk of transmission. Because of the fear of posterior capsular rupture, there is a common tendency (not advisable although) of emulsifying the last quadrant of the nucleus in the anterior chamber near the main port.27 If some part of the tip was exposed in this process, there might be plumes of white smoke and droplets, particularly in harder cataracts requiring higher ultrasound energy. A limitation was that ultrasound-induced aerosolization might be proportional to the fluid flow amount, height of bottle, fluid pressure inside the anterior chamber, and the amount of ultrasound power. These need to be investigated in future studies.


  • The severe acute respiratory syndrome coronavirus-2 can spread through aerosols and droplets.
  • In ophthalmology, procedures exist that might have the propensity to create aerosols and droplets.


  • There were no aerosols or droplets generated during phacoemulsification in experimental models using animal and human tissue.
  • Profound droplets where created only when the phacoemulsification tip was in close proximity to the exterior ocular surface, which was not a true representation of phacoemulsification.
  • Phacoemulsification was a safe procedure to perform for cataract surgeons while maintaining precautions against other forms of transmission.

Funded by the Defence Research and Development Organization, India, Chair Professorship.


1. World Health Organization. WHO Director-General's opening remarks at the media briefing on COVID-19. Available at media-briefing-on-covid-19---11-march-2020. Accessed March 11, 2020
2. The Lancet. COVID-19: protecting health-care workers. Lancet 2020;395:922
3. Rothan HA, Byrareddy SN. The epidemiology and pathogenesis of coronavirus disease (COVID-19) outbreak. J Autoimmun 2020;109:102433
4. Liu Z, Zhang Y, Wang X, Zhang D, Diao D, Chandramohan K, Booth CM. Recommendations for surgery during the novel coronavirus (COVID-19) Epidemic. Indian J Surg 2020;82:124–128
5. Available at: Accessed April 18, 2020
6. Kumar K, Prakash AA, Gangasagara SB, Rathod SBL, Ravi K, Rangaiah A, Shankar SM, Basawarajappa SG, Bhushan S, Neeraja TG, Khandenahalli S, Swetha M, Gupta P, Sampritha UC, Prasad GNS, Jayanthi CR. Presesnce of viral RNAof SARS-CoV-2 in conjunctival swab specimens of COVID-19 patients. Indian J Ophthalmol 2020;68:1015–1017
7. Xia J, Tong J, Liu M, Shen Y, Guo D. Evaluation of coronavirus in tears and conjunctival secretions of patients with SARS-CoV-2 infection. J Med Virol 2020;92:589–594
8. Seah I, Agrawal R. Can the coronavirus disease 2019 (COVID-19) affect the eyes? A review of coronaviruses and ocular implications in humans and animals. Ocul Immunol Inflamm 2020;28:391–395
9. Tellier R. Review of aerosol transmission of influenza A virus. Emerg Infect Dis 2006;12:1657–1662
10. Castrejón-Pita JR, Castrejón-García R, Hutchings IM. High speed shadowgraphy for the study of liquid drops. In: Klapp J, Medina A, Cros A, Vargas C, eds. Fluid Dynamics in Physics, Engineering and Environmental Applications. Environmental Science and Engineering (Environmental Engineering). Berlin, Heidelberg: Springer Inc.; 2013
11. Sirignano WA. Fluid Dynamics and Transport of Droplet and Sprays. Cambridge, UK: Cambridge University Press; 1999
12. Wong RS, Banerjee P, Kumaran N. Aerosol generated procedure in intraocular surgery. Available at: Accessed May 13, 2020
13. Darcy K, Liyanage S. Aerosol during phaco (cataract surgery). How to make cataract surgery safe during Covid-19. Available at: Accessed May 18, 2020
14. Wong RS, Banerjee P, Kumaran N. Aerosol generation in model eye phacoemulsification. Available at: Accessed May 5, 2020
15. Dada VK, Sindhu N. Cataract in enucleated goat eyes: training model for phaco-emulsification. J Cataract Refract Surg 2000;26:1114–1116
16. Merzkirch W. Density sensitive flow visualization. In: Emrich RJ, ed. Methods of Experimental Physics. Cambridge, MA: Academic Press; 1981
17. Tang JW, Nicolle AD, Pantelic J, Jiang M, Sekhr C, Cheong DKW, Tham KW. Qualitative real-time schlieren and shadowgraph imaging of human exhaled airflows: an aid to aerosol infection control. PLoS One 2011;6:e21392
18. Gralton J, Tovey E, McLaws ML, Rawlinson WD. The role of particle size in aerosolised pathogen transmission: a review. J Infect 2011;62:1–13
19. Parsloe CF, Twomey JM. Safety of phacoemulsification in a patient with an implanted deep brain neurostimulation device. Br J Ophthalmol 2005;89:1370–1371
20. Bradley MJ, Olson RJ. A survey about phacoemulsification incision thermal contraction incidence and causal relationships. Am J Ophthalmol 2006;141:222–224
21. Zacharias J. Thermal characterization of phacoemulsification probes operated in axial and torsional modes. J Cataract Refract Surg 2015;41:208–216
22. Berdahl JP, DeStafeno JJ, Kim T. Corneal wound architecture and integrity after phacoemulsification evaluation of coaxial, microincision coaxial, and microincision bimanual techniques. J Cataract Refract Surg 2007;33:510–515
23. Zacharias J. Laboratory assessment of thermal characteristics of three phacoemulsification tip designs operated using torsional ultrasound. Clin Ophthalmol 2016;10:1095–1101
24. Demircan S, Ataş M, Göktaş E, Başkan B. Comparison of 45-degree Kelman and 45-degree balanced phaco tip designs in torsional microcoaxial phacoemulsification. Int J Ophthalmol 2015;8:1168–1172
25. Bai Y, Yao L, Wei T, Tian F, Jin D, Chen L, Wang M. Presumed asymptomatic carrier transmission of COVID-19. JAMA 2020;323:1406–1407
26. Kariwa H, Fujii N, Takashima I. Inactivation of SARS coronavirus by means of povidone-iodine, physical conditions and chemical reagents. Dermatology 2006;212(suppl 1):119–123
27. Chakrabarti A, Nazm N. Posterior capsular rent: prevention and management. Indian J Ophthalmol 2017;65:1359–1369
Copyright © 2020 Published by Wolters Kluwer on behalf of ASCRS and ESCRS