Resources

Viscosity Is the Design Constraint Nobody Budgeted For: Engineering Subcutaneous Delivery of High-Concentration Biologics

Author

DMDivya Menon
Human Factors and Device Engineering Lead

Expertise

Product Design and Development

Service

Medical Device Design & Development
New Product Introduction

Sector

Drug Delivery

Topic

Biologics & Advanced Therapies
Human Factors & Usability Engineering

Published

13 July 2026
8 min read

TL;DR

• Subcutaneous delivery of biologics is volume-limited, which forces concentration up, which forces viscosity up non-linearly. Protein formulations above 200 mg/mL frequently exceed 20 cP, and 300 mg/mL formulations have been reported as high as 128 cP.
• Flow through a needle follows the Hagen–Poiseuille relationship, in which flow rate scales with the fourth power of the internal radius. Needle internal diameter, not gauge label, dominates injection force.
• The upper bound on injection force is not a mechanical specification. It is the thumb strength of the intended patient population — which for elderly patients and those with rheumatoid arthritis is substantially reduced.
• Silicone oil, the lubricant that makes the syringe usable, is also a protein aggregation risk and a source of subvisible particles. Optimising for glide force and optimising for drug stability pull in opposite directions.
• Temperature is an uncontrolled variable in home use. Viscosity rises steeply as a refrigerated product approaches injection, and a patient who does not wait for equilibration receives a different injection experience than the one that was validated.

The device engineering problem in modern drug delivery is almost entirely a consequence of a biological constraint: the subcutaneous space will only accept so much volume. Practical single-injection volumes have historically been capped around 1 to 2 mL before pain, backpressure and leakage become limiting, with larger volumes requiring either infusion over time or permeation-enhancing excipients.

That volume ceiling, colliding with the high milligram doses required for monoclonal antibodies, Fc-fusion proteins, and bispecifics, produces a single design consequence: concentration must rise. And viscosity does not rise linearly with concentration — it rises steeply, and often exponentially, once intermolecular interactions and reversible self-association begin to dominate.

The Numbers Are Worse Than Most Teams Assume

For reference, water at 20 °C has a viscosity of about 1 cP. Formulations above roughly 15 cP are conventionally treated as "high viscosity" in the injection device world.

Published data on immune globulin and antibody formulations places protein concentrations above 200 mg/mL frequently above 20 cP, with 300 mg/mL formulations reaching viscosities as high as 128 cP.

That is a two-order-of-magnitude range across a formulation space that a development programme may traverse between early clinical supply and commercial presentation. A device architecture that works comfortably at 10 cP may be entirely non-viable at 70 cP, and the formulation team frequently does not know the final concentration when the device platform is selected.

The market context explains why this problem is now unavoidable rather than niche: the prefilled syringe market was valued at approximately USD 22.79 billion in 2024, projected to grow at around 10.3% CAGR through 2030, driven substantially by the shift toward self-administered, high-concentration biologics for chronic conditions.

Hagen–Poiseuille Explains Why Needle Selection Dominates Everything

For laminar flow of a Newtonian fluid through a cylindrical tube:

where Q is volumetric flow rate, ΔP is pressure drop, r is internal radius, η is dynamic viscosity, and L is length.

Two consequences follow directly, and they govern the entire design space:

Flow scales with the fourth power of internal radius. Halving the internal diameter requires a sixteen-fold increase in pressure to maintain the same flow rate. This is why the difference between a 27G thin-wall and a 27G ultra-thin-wall needle — same nominal gauge, different internal diameter — produces a large and often decisive difference in injection time.

Force scales linearly with viscosity. A formulation that doubles in viscosity doubles the required force at constant geometry. There is no clever mechanism that avoids this; it can only be traded against needle geometry, injection time, or drive energy.

The gauge trade-off is genuinely uncomfortable. Larger-bore needles reduce force and time but increase perceived pain and needle anxiety, which depresses adherence. Finer needles improve the patient experience but require force that patients may not be able to generate, or drive springs that make the autoinjector larger, louder, and more expensive.

Vendor evaluation data illustrates the sensitivity: platform studies pairing a 2.25 mL autoinjector with four different needle configurations — 29G extra-thin-wall 8 mm, 27G special-thin-wall 12.7 mm, 27G special-thin-wall 8 mm, and 27G ultra-thin-wall 8 mm — across viscosity levels of 11, 23, 36, 53 and 70 cP found materially different mean injection times and, importantly, different variability in injection time. Consistency, not just mean performance, is a design output.

The Force Limit Is a Human Variable, Not an Engineering Specification

For manual syringes, the ceiling on injection force is set by the thumb push strength of the intended user, and that population is often precisely the one with reduced strength.

Hand and finger strength declines with age. Patients with rheumatoid arthritis — a major indication for subcutaneous biologics — frequently have reduced grip and pinch strength, joint pain, and limited dexterity as a direct feature of the disease being treated. Designing an injection force limit against a healthy-adult population, and then deploying to an RA population, is a use-error generator.

This is why the injection force specification is a human factors output, not a mechanical one. It should be derived from measured strength data for the intended population at the relevant percentile — typically a low percentile, since the design must accommodate the weakest intended user, not the average one.

The same logic applies to every other manual interaction step. Autoinjector cap removal has been repeatedly associated with use errors and difficulties in the human factors literature, and it is a step that requires grip and torque from users who may have neither. Needle shield removal, device orientation, and the requirement to hold the device against the skin for a specified dwell time after activation are all steps where population capability, not device capability, determines success.

Silicone Oil Is Two Problems Wearing One Coat

Prefilled glass syringes are siliconised to reduce break-loose and glide force. Without lubrication, the plunger stiction is unacceptable.

Silicone oil is also a well-documented risk factor for protein aggregation and a source of subvisible particulate matter, both of which are quality attributes regulators examine closely for parenteral biologics. Silicone migration into the formulation over shelf life can shift both.

This creates a genuine and irreducible tension:

• More silicone → lower, more consistent glide force → better injectability and device reliability
• Less silicone → lower aggregation and particulate risk → better drug product stability

Mitigations exist — baked-on siliconisation for a more stable, thinner layer; cross-linked silicone; silicone-oil-free systems using alternative barrier coatings or cyclic olefin polymer barrels — and each introduces its own trade-offs in cost, break-loose force consistency, and container closure integrity.

The important process point is that this trade cannot be resolved by the device team or the formulation team alone. It requires a joint specification, set early, with both parties agreeing to what glide force window and what particulate limit they are designing to.

Temperature Is the Uncontrolled Variable in Home Use

Most biologics are refrigerated at 2–8 °C. Viscosity is strongly temperature-dependent, rising as temperature falls.

The instruction "allow the device to reach room temperature for 30 minutes before injecting" appears in most IFUs, and adherence to it in practice is poor. A patient who injects a cold product is injecting a substantially more viscous fluid than the one the device performance was validated against, which means:

• A manual syringe requires more force than the patient can comfortably apply
• An autoinjector takes longer to complete, increasing the chance the patient lifts the device before delivery is finished, causing partial dosing
• The cold injectate itself increases injection site pain

Design responses include specifying and validating device performance across the full realistic temperature range rather than at room temperature only, using end-of-dose feedback that is unambiguous enough that a patient will not lift early, and — where feasible — extending the hold-time indication until delivery is actually complete rather than after a fixed interval.

Where Drug Delivery Programmes Break Down

Device platform selected before final formulation viscosity is known. This is the most common and most expensive sequencing error. A platform chosen at 15 cP that must later deliver 55 cP may require a complete redesign of the drive system.

Human factors validation performed with healthy volunteers. If the indication is rheumatoid arthritis, the summative evaluation population must include participants with the hand function limitations characteristic of the disease. FDA human factors expectations are explicit on the representativeness of user groups.

Injection time treated as a mean rather than a distribution. A device with a 10-second mean injection time and a 6-second standard deviation will produce a meaningful number of injections long enough that patients lift early. Variability is a specification.

Testing with a surrogate that is Newtonian when the drug product is not. Glycerol and PEG solutions are common viscosity surrogates and are Newtonian. Many protein formulations are shear-thinning, which means their apparent viscosity at the high shear rates inside a needle differs from their bulk measured viscosity. Surrogate testing is useful but must eventually be confirmed with drug product.

Container closure and device interaction validated separately. The combination product is the system. Stopper dimensional tolerance, barrel internal diameter tolerance, siliconisation uniformity and autoinjector spring force interact, and tolerance stack across the combination determines real-world performance.

Regulatory pathway ambiguity resolved late. Combination products carry dual obligations — device requirements under 21 CFR Part 4 and the applicable device regulations, drug requirements under the relevant marketing application. The ISO 11608 series governs needle-based injection systems and should be the design reference from the start.

The Framing That Helps

A subcutaneous injection device is a system in which a formulation scientist's decision about concentration, a mechanical engineer's decision about needle geometry, and a human factors engineer's finding about population hand strength all determine the same output: whether the patient reliably receives the full dose.

Those three decisions are usually made by three teams on three schedules. Programmes that force them into one shared specification, early, tend to converge. Programmes that do not tend to discover the conflict at summative usability testing, which is the most expensive possible place to find it.

RhythmRx develops drug delivery and combination device systems, with formulation viscosity, needle geometry, drive mechanism and human factors specification treated as one coupled design problem rather than sequential handoffs.

DMDivya Menon is Human Factors and Device Engineering Lead at RhythmRx, working on injection device architecture, combination product design and use-related risk analysis.

Sources

1. 1.Optimizing High-Dose Delivery, Ypsomed — protein concentrations above 200 mg/mL frequently exceeding 20 cP; 300 mg/mL formulations reaching 128 cP; YpsoMate 2.25 / BD Neopak needle configuration study across 11, 23, 36, 53 and 70 cP.
2. 2.Improving prefilled syringe injectability: influences of critical component attributes and human factors, European Journal of Pharmaceutics and Biopharmaceutics, 2025.
3. 3.Advancing injection force modeling and viscosity-dependent injectability evaluation for prefilled syringes, European Journal of Pharmaceutics and Biopharmaceutics.
4. 4.User-Centric Approach to Specifying Technical Attributes of Drug Delivery Devices: Empirical Study of Autoinjector-Cap Removal Forces, PMC.
5. 5.The Medicine Maker — prefilled syringe market valued at USD 22.79 billion in 2024, projected 10.3% CAGR 2025–2030.
6. 6.Drug Formulation Impact on Prefilled Syringe Functionality and Autoinjector Performance — silicone oil, glide force and interfacial effects.
7. 7.ISO 11608 series (needle-based injection systems); ISO 11040 (prefilled syringes); IEC 62366-1:2015+A1:2020; 21 CFR Part 4 (combination products).