Syringe Pump Accuracy in Preclinical Infusion Studies

Part 1: The pump’s accuracy specification is only a small part of the story.

Syringe pumps are widely considered to be the most accurate devices for delivering an IV dose to a laboratory animal. Researchers typically adjust flow rates based on body weights to make the dose as precise and consistent as possible, and they rely on the accuracy of the pump to deliver it. But how accurate are these pumps in practice?

Harvard Apparatus’s entry level 11 Elite pump series, commonly used in academic laboratories, has an “accuracy” specification of “±0.50%” and a “reproducibility” specification of “±0.05%.” If you get the more expensive PHD Ultra pumps you get double the accuracy, “±0.25%.” What does this mean?

Baxter's specification for the AS50 syringe pump, a clinical pump designed in 1990s and a model that’s still in use in some preclinical research labs today together with its replacement the SAI 3D™, is more detailed:

• Flow Rate Accuracy: ±3% of full scale plunger travel (not including syringe tolerance)
• Volume Accuracy: ±3% or ±0.007” of travel, whichever is greater (not including syringe tolerance).

Let’s say you need to deliver a 1mL dose over 30 minutes and you’ll use a 3mL syringe.

You might expect the Harvard pump to deliver between 995 and 1005µL (±0.50% accuracy) and if it happens to deliver 1000.0µL the first time, then it should deliver between 999.5 and 1000.5µL each time after that (±0.05% reproducibility).

Likewise, you might expect the Baxter AS50 to deliver between 970 and 1030µL, ±3%. (You’re not sure what to make of “±0.007” of travel.”)

The first sign that these accuracy figures are not the whole story is the statement “not including syringe tolerances.” If you are using plastic syringes, as nearly everyone does, their inconsistency is significantly greater than the accuracy of the pump.

To start, you need to make sure you are using syringes that are designed and certified to be used in pumps. Among other requirements that are important for pump compatibility, ISO standard 7886-2 sets a maximum permissible flow rate error of ±5%, and this is the best you can expect from plastic syringes. BD Plastipak™ syringes are certified to this standard; many others are not. For example, a customer of ours reported a 15% error when using a Nipro 20mL syringe. After scratching our heads for a while we eventually discovered that Nipro sold two different syringes, one with a 20.3mm ID and one with an 18.8mm ID, under the same “JD+20L” reorder number. Monoject brand syringes were subject to an FDA class 1 recall in 2024 for a similar reason.

Ultimately a syringe pump’s job is to move the plunger of the syringe, while the syringe’s inner diameter defines the volume that will be infused as the plunger moves.

This might seem like the end of the story. Syringe pumps should be very good at moving the plunger at an accurate rate: they have a lead screw with a pitch that is machined with high precision and typically a stepper motor that turns it based on a clock that is accurate to a few seconds over years. As long as the pump is rigid and well built, it is reasonable to expect fantastic “pusher travel rate” accuracy. And in fact the lead screw pitch and stepper motor speed will not change over time, so no annual calibration for accuracy is needed. This is probably the type of accuracy Harvard Apparatus is referring to with its “0.50%” specification. The Baxter pump is made from plastic parts that can flex and does not use a stepper motor as it is an old design, so it's reasonable to expect lower plunger rate accuracy. Also, the may have been more conservative since their specifications were subject to FDA review.

Back to our example, if you’re using a B-D 3mL syringe you would expect these errors to add:

• Harvard 11 Elite: ±5.5%, 945-1055µL
• Baxter AS50: ±8%, 920-1080µL.

But it is still not that simple. If you are concerned about the volume of test article actually delivered out the end of a catheter, you need to think about your entire system, not just the pusher travel rate and syringe inner diameter. In fact we have identified over 40 potential sources of error in delivering a dose to a catheterized animal from a syringe pump, listed in the table below. Some are minor, but others can be as much as 100% of your dose. Here are some real world examples:

• You press start on your syringe pump. After several minutes you hear a “click.” The clutch had been riding along the top of the lead screw and only later slipped off and engaged. The motor had been turning but the pusher arm was not moving during that period, and that volume was counted but not delivered. Even worse, you are working with a bank of pumps and you have no idea which one clicked.
• Your pump is located on a shelf above the animal cage. Part way through the dose you notice that the plunger has pulled away from the pusher block on your pump. The syringe has emptied due to siphoning, delivering the dose much faster than you intended.
• You checked patency at the beginning of your study, but part way through the animal changed positions and kinked the catheter. Your pump did not sense the occlusion, pressure built up, and the dose leaked out of a joint in the infusion line and dripped into the bedding without your knowledge.

In future posts we will discuss the ways you can manage and minimize these sources of error. Some can be handled by following best practices, but for others you may need to upgrade your equipment. (Stay tuned: Instech is about to introduce a new syringe pump for laboratory animal research.)

Potential sources of error in delivering a dose from a syringe pump

* Present under all conditions.