Secondary Logo

Journal Logo

Laboratory science

Anterior chamber stability during bimanual irrigation and aspirationTable 1. Tested irrigation cannulas: manufacturers' specifications and experimental results.

Theoretical and experimental analysis

Jeng, Bennie H. MDa,1; Huang, David MD, PhD∗,a,1

Author Information
Journal of Cataract & Refractive Surgery: October 2001 - Volume 27 - Issue 10 - p 1670-1678
doi: 10.1016/S0886-3350(01)00860-4
  • Free


Although the advent of small incision cataract surgery has led to safer surgery and more rapid visual recovery for patients, it has brought new surgical challenges. Among these is the task of removing subincisional cortex. This is an awkward maneuver for the standard single-piece irrigation and aspiration (I/A) handpiece. The tip must reach around the incision, iris, and capsulorhexis margin to remove the cortex. Even with modifications, consisting of bends and curves in the aspiration tip,1,2 aspiration of the subincisional cortex remains difficult.

The technique of bimanual I/A was developed to make cortical removal easier. It was described by Brauweiler,3 Buratto (L. Buratto, MD, “Removal of the Sub-incisional Cortex After Phaco: I/A with Two Cannulas Solves the Problem,” Ocular Surgery News, International Edition, June 1995, pages 26–27), and Colvard.4 The technique uses 2 separate I/A cannulas inserted through 2 side-port incisions (Figure 1). Bimanual I/A offers several advantages including improved access to subincisional cortex, bimanual stabilization of the globe, and the ability to retract the iris in eyes with small pupils.4 Bimanual tips allow faster and more complete cortical cleanup and capsule polishing because the tips can easily reach any place in the capsular bag.

Figure 1.
Figure 1.:
(Jeng) Bimanual I/A viewed from an operating microscope. Irrigation cannula (Rhein Huang prototype) is on the left and aspiration cannula on the right.

In our experience, despite its advantages, bimanual IA tends to be more susceptible to surge instability at high aspiration flow rates and high vacuum. This study investigated the factors that affect anterior chamber stability during bimanual I/A and tried to determine which cannula designs were best able to prevent collapse of the anterior chamber at a brisk aspiration setting.

Materials and methods

Irrigation cannulas of various gauges and configurations for use in the bimanual I/A technique were obtained on loan from various manufacturers (Accutome, Duckworth & Kent, Eagle, Katena, Rhein, Storz, and Visitec). The cannulas were tested in an experimental I/A system based on the Legacy® 20000 phacoemulsification system (Alcon Surgical, Inc.). A standard 100 cm water pressure head was established at the beginning of a measurement by adjusting the height of a 500 cc balanced salt solution (BSS®) bottle. A timed-catch method was used to measure flow. Depending on the expected flow rate, a stopwatch was set to 30 seconds (1-piece I/A and phaco tips), 1 minute (21-gauge cannulas), or 2 minutes (23-gauge and smaller cannulas). Fluid catch was read in a 100 cc graduated cylinder. The pressure head at the end of the catch period was determined by measuring the height difference between the fluid level in the BSS bottle and in the graduated cylinder. The pressure head at the beginning and end of the catch measurement was averaged and used in calculating an adjusted flow rate (cc/min) at 100 cm water pressure head.

The dimensions of each cannula (gauge, length, orifices) were taken from the manufacturer's specifications and qualitatively confirmed by visual comparison between cannulas. For comparison, the irrigation flow rate of the standard Alcon 1-piece I/A handpiece and the MicroTip ABS phaco handpiece were also determined.

In addition to measuring the cannulas, the pressure drop across the other components of the irrigation system (tubing and handpiece) was measured. The rate of fluid flow through the tubing and handpiece without any cannula attachment was measured at various bottle heights, and a quadratic equation that relates the pressure drop to the flow rate was determined.

Anterior chamber stability afforded by the various irrigation cannulas was tested using a model anterior chamber system. The test chamber is the silicone test chamber supplied with the Legacy 20000 phacoemulsification handpiece. The open end of the test chamber was plugged, and simulated paracentesis incisions were made with a 20-gauge MVR blade. An irrigation cannula was connected via standard gauge tubing to a bottle of BSS hung at a height of 100 cm. A 23-gauge aspiration cannula with a 0.3 mm orifice was connected to a Legacy 20000 phacoemulsification machine. The I/A cannulas were then inserted into the model anterior chamber. No leak was observed at the incision sites in this experimental closed IA system. The maximum aspiration flow rate and vacuum power that could be sustained for at least 1 minute without collapse of the anterior chamber was measured. This was recorded as the critical aspiration rate for each irrigation cannula.

The actual aspiration flow rate (cc/min) through the 23-gauge aspiration cannula with a 0.3 mm orifice was then measured at the maximum settings of the Legacy 20000 phacoemulsification machine. This flow rate was calculated by observing the amount of BSS that was aspirated from a graduated cylinder via the aspiration cannula over 1 minute.

To determine the criteria for anterior chamber stability, a theoretical model of the anterior chamber was developed (Figure 2). Chamber stability requires Qi = Qo (inflow = outflow), Pc gt; 0 (anterior chamber pressure greater than atmospheric pressure), and Pc gt; Pp (anterior chamber pressure greater than posterior vitreous pressure). The gauge pressure convention in which the atmospheric pressure is defined as zero was used. The anterior chamber pressure is Pc = Pi − ΔP, where Pi = irrigation pressure head and ΔP = pressure drop across the irrigation system. Since Pi is fixed by the surgeon and Pp is relatively constant, the pressure drop across the irrigation system is the primary dynamic variable that determines chamber stability. The pressure drop ΔP is determined by the flow rate set by the surgeon (depending on the desired aspiration rate) and the flow resistance of the irrigation system. Most of the pressure drop ΔP in the irrigation system occurs in the cannula because of its relatively small lumen and orifice diameter. Thus, the design of the irrigation cannula is critical to the stability of the bimanual I/A system.

Figure 2.
Figure 2.:
(Jeng) Theoretical model of the anterior chamber fluid balance during I/A. Stable anterior chamber volume requires that Q i = Q o and P c gt; P p gt; 0, where Q i = irrigation flow, Q o = aspiration flow, P c = anterior chamber pressure, and P p = posterior pressure. In addition, P c = P i − ΔP, where P i = irrigation pressure and ΔP = pressure drop across the irrigation system. Thus, P iP p + ΔP to prevent anterior chamber collapse. Finally, P i = y i × ρ × g, where y i = bottle height, g = gravitation constant, and ρ = density of fluid.

To calculate the pressure drop across an irrigation cannula, theoretical analysis was performed using the Hagen-Poiseuille and Bernoulli equations. The Hagen-Poiseuille equation describes nonturbulent or “laminar” flow, which occurs at relatively low flow rates over smooth surfaces, such as the lumen of a cannula. For a circular pipe,5 the equation is

where ΔP = pressure drop between pipe outlet and inlet due to flow, Q = flow rate, L = length of pipe, η = viscosity of fluid, and d = bore diameter of pipe.

Laminar flow condition and the Hagen-Poiseuille equation apply when Reynold's number is less than 2000, indicating that turbulence is negligible.5 Reynold's number was computed each time the Hagen-Poseuille equation was applied, which confirmed that the nonturbulence criterion was fulfilled over the range of flow rates and cannula diameters investigated.

The Bernoulli equation describes the pressure drop associated with the acceleration of fluid at a narrow point in the flow path, such as the orifice of a cannula. For a circular orifice,5 the equation is

where ρ = density of fluid.

Using these theoretical and empiric relations, the maximum available irrigation flow under a standard operating condition was calculated. A Pi (bottle height) of 100 cm water and a posterior pressure of zero were assumed. Maximum irrigation flow occurs when the anterior chamber pressure is zero (the limit before collapse) and ΔP = 100 cm of water is produced by laminar resistance in the cannula bore, Bernoulli effect at the orifice, and pressure drop in the remainder of the irrigation system. This theoretical maximum irrigation flow rate was compared with the measured irrigation flow rate at 100 cm water pressure head in the Legacy 20000 phacoemulsification system.


To support anterior chamber stability at all times, the irrigation system must be able to provide flow equal to the aspiration flow. The aspiration flow on the Legacy 20000 phacoemulsification system was measured at various flow settings (Figure 3). Vacuum was set at the maximum 500 mm Hg. A 23-gauge prototype aspiration cannula (Rhein) of 25.0 mm length and with a single 0.3 mm diameter orifice was used. The actual measured flow was lower than the nominal panel setting at all settings, and very little additional flow was obtained above a setting of 20.0 cc/min. At a maximum nominal panel setting of 60.0 cc/min, the actual measured aspiration rate was 20.5 cc/min (Figure 3) with a maximum achieved vacuum of 530.0 mm Hg (panel reading). Based on this finding, the anterior chamber should remain stable at the maximum aspiration setting if the irrigation system can provide up to 20.5 cc/min of flow.

Figure 3.
Figure 3.:
(Jeng) Actual aspiration flow rate achieved at various flow rate settings of the Legacy 20000 phacoemulsification system using a 23-gauge aspiration cannula (25.0 mm length, 0.3 mm orifice).

The irrigation flow was then measured with a variety of side-port irrigation cannulas at a pressure head of 100.0 cm water (Table 1). The cannulas were also tested in the artificial anterior chamber to determine the maximum aspiration flow setting at which chamber stability could be sustained. These measurements correlated well when the discrepancy between actual aspiration flow and the panel setting are considered (Figure 3). The irrigation cannulas were primarily differentiated by the lumen or bore gauge. Flow increased with increasing lumen size. None of the irrigation cannulas with lumens smaller than 21 gauge was able to provide stability at maximum aspiration (60.0 cc/min nominal setting and 20.5 cc/min actual flow). Four of the irrigation cannulas with 21-gauge lumens were able to provide the high flow necessary to maintain stability. Among these cannulas, higher flow was found in the shorter cannulas. Two of these cannulas (Duckworth & Kent CE0120 and Rhein Huang prototype) use a special thin-walled cannula that has a lumen size equal to a standard 21-gauge cannula but an outer diameter almost as small as a standard 23-gauge cannula (Figure 4).

Figure 4.
Figure 4.:
(Jeng) Schematic diagrams of cross sections of various-sized cannulas. Note that the inner diameter of the thin-walled 23-gauge cannula is the same as that of a standard 21-gauge cannula. However, the outer diameter of the thin-walled 23-gauge cannula is only 0.01 mm larger than that of a standard 23-gauge cannula.

For comparison with the side-port cannulas, the irrigation routes of the Alcon 501917 1-piece I/A handpiece and 0.9 mm MicroTip ABS phacoemulsification handpiece were measured. These tips are intended to be inserted through the main phacoemulsification wounds, which are larger than the side-port incisions. The experiments confirmed that the larger bore sizes provided greater flow (Table 1). The intuitive sense that larger bore and shorter cannulas provide more flow was confirmed by our theoretical calculations.

The Hagen-Poiseuille equation was used to calculate the pressure drop along the length of the cannula. Figure 5 demonstrates the linear relationship between pressure drop and flow rate in laminar flow. For a 10 mm length of cannula, the larger 21- and 22-gauge (lumen) cannulas were able to provide >20.5 cc/min of flow with less than 100.0 cm water pressure drop. Based on this simple simulation alone, it can be determined that side-port irrigation cannulas need 22-gauge or larger lumen diameters to provide the high flow required for chamber stability.

Figure 5.
Figure 5.:
(Jeng) Pressure drop versus flow rates through various gauge cannulas 10.0 mm in length as determined by the Hagen-Poiseuille equation. The vertical line at 20.5 cc/min represents the flow rate required to match maximum aspiration measured on the Legacy machine.

Pressure drop also occurs because of the Bernoulli effect, in which potential energy in pressure is translated to kinetic energy in the velocity of the irrigation fluid. The kinetic energy is then dissipated in turbulence when the irrigation jet exits the cannula at the orifice. The Bernoulli equation can be applied to any strictures in the pipeline that produce fluid acceleration. Most of the tested cannulas have open ends, which means the bore diameter determines the fluid velocity inside the cannula and in the jet at the exit of the orifice. The size specifications for various gauges are shown in Figure 4. If the orifice at the cannula tip is smaller, the cannula orifice diameter in the Bernoulli equation would be used. Some of the tested cannulas have closed ends and 2 side orifices. In all cases, the sum of the orifice areas was greater than the cross-sectional lumen area of the cannulas. Therefore, the lumen diameter was used in calculating the Bernoulli effect in all cases. The Bernoulli pressure drop increased as the square of the flow rate (Figure 6). Therefore, this effect is more important at higher flow rates. At the previously measured maximum aspiration rate of 20.5 cc/min, 0.32 mm (23 gauge), 0.40 mm (22 gauge), and 0.50 mm (21 gauge) orifices dropped pressure by 92.0 cm, 38.0 cm, and 15.0 cm water, respectively. Based on this simulation, it was determined that single orifices must be 0.32 mm (23 gauge) or larger to support chamber stability using 100 cm water pressure head.

Figure 6.
Figure 6.:
(Jeng) Theoretical pressure drop versus flow rate at various orifice diameters as determined by the Bernoulli equation. The vertical line at 20.5 cc/min represents the flow rate required to match maximum aspiration measured on the Legacy machine.

Although most of the pressure drops in the irrigation system were expected to occur in the cannula, the pressure drop in the remainder of the system was also measured. The flow rates through the cannula-less irrigation system were measured at various pressure heads (bottle heights). The resulting pressure-flow relationship (Figure 7) was well described by a quadratic equation:

Figure 7.
Figure 7.:
(Jeng) Pressure drop versus flow rate in the cannula-less system as determined by the best-fit equation: ΔP = .4141 × Q i + 0.00402 × Q i 2.

At a flow rate of 20.5 cc/min, the predicted pressure drop was 10.1 mm Hg.

A total model of the irrigation system was constructed by adding the pressure drops in the cannula lumen (Hagen-Poiseuille), orifice (Bernoulli), and the remainder of the system. Based on a standard pressure head of 100.0 cm water, the predicted irrigation flow was computed as a function of cannula length for various gauge cannulas (Figure 8). According to this model, 23-gauge and smaller lumen diameters cannot provide sufficient flow to match the 20.5 cc/min of maximum aspiration flow measured in our system. Standard 21-gauge irrigation cannulas up to 29.0 mm in length and 22-gauge irrigation cannulas up to 8.0 mm in length were able to provide the high flow.

Figure 8.
Figure 8.:
(Jeng) Irrigation flow rate versus cannula length calculated by our theoretical model of the total irrigation system. The model assumes the cannula tip is open. The horizontal line at 20.5 cc/min represents the flow rate required to match maximum aspiration measured on the Legacy machine.

The model predictions were compared with the experimentally determined flow rates (Table 1) in the cannulas tested and plotted in Figure 9. There was approximate agreement. However, there was a tendency toward overestimating flow in the higher-flow cannulas and underestimating flow in the lower-flow cannulas. At higher flows, there may be more turbulence in the irrigation system than the model predicted. Alternatively, the nominal size specifications obtained from the cannula manufacturers may have been slightly inaccurate.

Figure 9.
Figure 9.:
(Jeng) Experimental flow rates versus theoretical flow rates of various side-port irrigation cannulas tested at 100 cm water pressure head. The diagonal line represents equality between experimental and theoretical results.


Our investigation revealed that for the typical I/A settings on the Legacy 20000 phacoemulsification system, a cannula with a bore size of 22 gauge or larger was required to provide irrigation when matched with a 23-gauge aspiration cannula. This was due to the difference in the available pressure heads in the 2 systems. The aspiration vacuum can be as high as 530.0 mm Hg, while the irrigation pressure head is only 100.0 cm water (= 73.5 mm Hg). While it is possible to use slightly greater bottle heights in the irrigation system, it is eventually limited by the pressure tolerance of the eye. When the aspiration flow is deactivated or blocked by port occlusion, the intraocular pressure rises to the level of the irrigation pressure. Depending on the arteriolar blood pressure, our setting of 100.0 cm water is already capable of temporarily stopping ocular perfusion. Further increasing the irrigation pressure is not a viable strategy. An alternative strategy may be to reduce the aspiration flow rate setting. We find that with the maximum aspiration setting and the 23-gauge aspiration cannula, the pace of cortical and viscoelastic aspiration is optimal. Significantly lowering the aspiration flow would undesirably prolong the surgical time. In fact, high-volume surgeons who desire to conclude their surgery in less than 10 minutes may find it necessary to use a higher flow aspiration cannula that is shorter or has larger bore and orifice diameters. In that case, an even larger irrigation cannula would be needed to provide stability.

There is some disadvantage in performing I/A using a 23-gauge aspiration cannula and a 21-gauge irrigation cannula. The side-port incisions must be equal so irrigation and aspiration sites can be switched to expedite cortical cleanup. If the side ports are large enough to accommodate the 21-gauge cannula, leakage would occur around the 23-gauge cannula. Smaller side ports would make insertion more difficult and may be stretched by the 21-gauge cannula, making the wound non-self-sealing.

For bimanual I/A, we make 2 side-port paracentesis incisions with a 20-gauge MVR blade (1.4 mm width), which works optimally with 23-gauge side-port cannulas. Matched with the 23-gauge aspiration cannula, we had trouble with anterior chamber stability with all the standard 23-gauge irrigation cannulas we have used. Therefore, we set out to find a special design that would allow a 23-gauge irrigation cannula to provide as much flow as a standard 21-gauge cannula. The solution was to use a special thin-walled 23-gauge cannula for irrigation. The thin-walled cannula has an inner bore diameter equal to that of a standard 21-gauge cannula while maintaining an outer diameter that is almost as small as a standard 23-gauge cannula (Figure 4). Two of the tested cannulas use this design (Duckworth & Kent CE0120, Rhein Huang prototype). The Huang prototype (Figure 10) further increases flow by using a 2-stage design that shortens the 23-gauge thin-walled section to the distal 9.0 mm. The proximal portion is 16 gauge; because of its larger diameter, it produces negligible resistance to flow. Part of the 9.0 mm 23-gauge section is hidden inside the 16-gauge sleeve. We have used the Huang prototype cannula for I/A during actual cataract procedures and found that it does provide sufficient flow to stabilize the anterior chamber through the whole range of aspiration settings on the Legacy 20000 phacoemulsification system. The design, of course, can be scaled up to 20- or 19-gauge internal lumens if a surgeon wishes to have higher speed I/A.

Figure 10.
Figure 10.:
(Jeng) Huang prototype irrigation cannula manufactured by Rhein. The short 9.0 mm segment (3.0 mm exposed and 6.0 mm hidden) of a thin-walled 23-gauge cannula (21-gauge lumen and 23-gauge external diameter) at the end of a 16-gauge sleeve cannula allowed higher flow than other tested 23- and 21-gauge cannulas while maintaining the ability to be inserted through a small incision.

Based on our experimental and theoretical results, we recommend that cataract surgeons carefully choose their combination of I/A systems to maintain the stability of the eye. For bimanual I/A, the irrigation cannula should have a larger lumen diameter than the aspiration cannula (ie, 21/23-gauge combination). Special thin-wall/short-length design for the irrigation cannula can reduce the side-port incision size requirement. Our results can be generalized to any closed-loop fluid circuit during surgery. For most ocular surgery, the desired aspiration vacuum is generally higher than the available safe irrigation pressure head. Therefore, the irrigation system must have lower flow resistance to maintain stability. We note that Alcon flared phacoemulsification tips use this principle by purposely constricting the waist of the aspiration tip, which also increases the cross-sectional area of the annular irrigation flow inside the sleeve. This type of design increases the stability of the anterior chamber and the safety of surgery.

This principle was recognized by pioneers in the use of the anterior chamber maintainer (ACM) during I/A. The ACM is similar to the irrigation cannula or handpiece except that it is meant to be secured in place rather than manipulated by hand. Blumenthal and coauthors6–8 and Chawla and Adams9 describe using a 20- or 21-gauge ACM in combination with a small-port aspiration handpiece to maintain chamber stability.


1. Hagan JC III. A new cannula for removal of 12 o'clock cortex through a sideport corneal incision. Ophthalmic Surg 1992; 23:62-63
2. Hagan JC III. Irrigation/aspiration handpiece with changeable tips for cortex removal in small incision phacoemulsification. J Cataract Refract Surg 1992; 18:318-320
3. Brauweiler P. Bimanual irrigation/aspiration. J Cataract Refract Surg 1996; 22:1013-1016
4. Colvard DM. Bimanual technique to manage subincisional cortical material. J Cataract Refract Surg 1997; 23:707-709
5. Sabersky RH, Hauptmann EG, Acosta AJ. Fluid Flow: a First Course in Fluid Mechanics, 4th ed. Upper Saddle River, NJ, Prentice Hall, 1999
6. Blumenthal M, Moisseiev J, Hashomer T. Anterior chamber maintainer for extracapsular cataract extraction and intraocular lens implantation. J Cataract Refract Surg 1987; 13:204-206
7. Blumenthal M, Assia EI, Chen V, Avni I. Using an anterior chamber maintainer to control intraocular pressure during phacoemulsification. J Cataract Refract Surg 1994; 20:93-96
8. Blumenthal M, Grinbaum A, Assia EI. Preventing expulsive hemorrhage using an anterior chamber maintainer to eliminate hypotony. J Cataract Refract Surg 1997; 23:476-479
9. Chawla HB, Adams AD. Use of the anterior chamber maintainer in anterior segment surgery. J Cataract Refract Surg 1996; 22:172-177
© 2001 by Lippincott Williams & Wilkins, Inc.