Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO 28621:2016 — Medical Devices — Dosage Measurement for Medicinal Products
Technical requirements, accuracy classes, and test methodologies for dosage measurement systems in medical drug delivery devices.
1. Scope and Regulatory Context
ISO 28621:2016 establishes requirements for dosage measurement systems integrated into medical devices used for administering medicinal products. This standard addresses the critical interface between drug delivery devices and patient safety, covering infusion pumps, syringes, nebulizers, and other devices where precise dosage measurement directly impacts therapeutic outcomes. The standard was developed in response to the growing recognition that dosage measurement errors in automated drug delivery systems represent a significant and preventable source of patient harm in healthcare settings worldwide.
The standard specifies accuracy classes, measurement uncertainty limits, and testing methodologies for dosage measurement subsystems. It distinguishes between devices intended for continuous administration (infusion pumps) versus bolus delivery (syringe injectors), with different tolerance bands reflecting their distinct clinical applications and risk profiles. For example, a volumetric infusion pump delivering intravenous fluids in an intensive care unit requires tighter accuracy control than a large-volume pump used for routine hydration, because ICU patients are more likely to experience adverse effects from fluid imbalance.
The regulatory landscape surrounding dosage measurement is complex and varies by jurisdiction. ISO 28621 provides a globally harmonized framework that complements regional regulations such as the EU Medical Device Regulation (MDR 2017/745) and the US FDA’s guidance on infusion pump cybersecurity and accuracy. Manufacturers seeking to market devices in multiple regulatory jurisdictions can use ISO 28621 compliance as a baseline, with additional jurisdiction-specific requirements applied as overlays. The standard also aligns with IEC 60601-1 (medical electrical equipment safety) and IEC 60601-2-24 (particular requirements for infusion pumps and controllers).
Dosage errors are among the most common preventable adverse events in healthcare, affecting an estimated 1.2 million patients annually in the United States alone according to the Institute of Medicine. ISO 28621 requires dual-redundancy measurement paths for devices delivering high-risk medications (e.g., vasopressors, insulin, anticoagulants). This means that two independent measurement channels must agree within a specified tolerance before the device delivers each dose increment, providing a hardware-based safeguard against single-point measurement failures.
| Delivery Mode | Accuracy Class | Maximum Permissible Error | Typical Devices | Clinical Application |
|---|---|---|---|---|
| Continuous infusion (low rate) | Class A | ±2% of set rate | Volumetric infusion pumps | ICU, neonatal, critical care |
| Continuous infusion (high rate) | Class B | ±5% of set rate | Large-volume infusion pumps | General ward, hydration |
| Bolus injection | Class A | ±1% of nominal volume | Syringe pumps, auto-injectors | Chemotherapy, emergency |
| Nebulized delivery | Class C | ±10% of nominal dose | Mesh nebulizers, jet nebulizers | Respiratory therapy |
| Patient-controlled analgesia | Class A | ±2% of demand dose | PCA pumps | Post-operative pain |
2. Technical Requirements and Test Methodologies
ISO 28621 defines specific test protocols for dosage accuracy verification, including flow rate stability testing over the device’s entire operating range, back-pressure sensitivity characterization, and start-up transient analysis. For infusion devices, the standard requires measurement of occlusion alarm response time — the interval between an occlusion event and alarm activation must not exceed specified limits based on the infusion rate. The test protocol specifies that occlusion tests be conducted at three different flow rates spanning the device’s operating range, with the occlusion introduced at a standardized distance of 1 m from the device outlet to simulate clinically relevant conditions.
The standard also addresses environmental factors that can affect dosage accuracy. Temperature cycling tests require that the device maintain accuracy within its specified class over the range of 15°C to 35°C, covering typical clinical environments from cold operating rooms to warm neonatal units. Humidity exposure testing (20% to 90% relative humidity, non-condensing) ensures that electronic measurement components are not affected by moisture-related drift. Electromagnetic compatibility is referenced through IEC 60601-1-2, recognizing that infusion pumps are frequently used in electromagnetically noisy environments alongside MRI machines, electrosurgical units, and mobile communication devices.
Flow start-up transients are a frequently overlooked source of dosing error. ISO 28621 mandates that devices achieve 90% of set flow rate within 60 seconds of start for low-rate infusions (≤ 5 mL/h). Designers should minimize mechanical compliance in the fluid path to reduce start-up delay. In practice, this means using low-compliance tubing (rated < 1 μL/mmHg per meter), rigid fluid path components, and pre-conditioning the pumping mechanism before clinical use. Measurement data from commercial infusion pumps shows that start-up transients can introduce errors of 15-30% in the first 5 minutes of delivery at low flow rates if not properly managed.
2.1 Measurement Uncertainty and Calibration
The standard adopts the Guide to the Expression of Uncertainty in Measurement (GUM) framework for quantifying dosage measurement uncertainty. Calibration intervals are specified based on device type: Class A devices require calibration every 6 months, Class B annually, and Class C every 2 years. On-device self-test capabilities are strongly recommended, with automated drift detection that alerts users when measurement accuracy approaches the tolerance boundary. The standard specifies that the calibration reference standard must have an uncertainty at least 4 times better than the device under test (a 4:1 test uncertainty ratio). For Class A devices with MPE of ±2%, this requires a calibration reference with uncertainty no greater than ±0.5%.
The standard introduces the concept of “total measurement chain” validation, requiring that the entire dosage measurement path — from sensor through signal processing to digital display or delivery mechanism — be validated as a complete system. This approach recognizes that component-level accuracy does not guarantee system-level accuracy, as interface effects, signal conditioning artifacts, and firmware rounding errors can introduce cumulative measurement errors that exceed the sum of individual component tolerances.
3. Engineering Design Insights
The most technically demanding aspect of ISO 28621 compliance is managing the interaction between mechanical actuation tolerances and fluidic behavior at low flow rates. At infusion rates below 1 mL/h, the combination of syringe plunger stick-slip friction, tubing compliance, and back-pressure variations from venous access can introduce errors exceeding 15% — far beyond the Class A limit of ±2%. Advanced designs incorporate active pressure sensing and closed-loop flow control to compensate for these disturbances. The closed-loop approach uses real-time flow measurement (typically using thermal or ultrasonic flow sensors) to adjust the drive mechanism in milli-second timeframes, effectively canceling out mechanical nonlinearities.
Material selection is critical: drug-contact materials must satisfy ISO 10993 biocompatibility requirements while maintaining dimensional stability and low friction over the device’s service life. Perfluoroelastomer seals and ceramic plunger coatings have emerged as preferred materials for high-precision applications due to their low coefficient of friction (COF < 0.15) and chemical inertness. The plunger-syringe barrel interface is particularly challenging because it must simultaneously provide a leak-free seal at high infusion pressures (up to 100 kPa for occlusion detection tests) and low-friction movement at low flow rates. Surface finish specifications for the syringe barrel interior are typically in the range of Ra 0.1-0.2 μm, requiring precision manufacturing processes such as diamond honing or injection molding with polished tooling.
Software-related considerations are equally important. The standard requires that dosage calculation algorithms be validated using boundary value analysis across the entire operating range, including edge cases such as minimum flow rate combined with maximum back pressure. Alarm management follows the principles of IEC 60601-1-8, with prioritized alarms for occlusion, air-in-line, battery depletion, and system faults. Alarm thresholds must be configurable by clinical staff but protected against unauthorized modification through administrative access controls.
During design validation, perform a sensitivity analysis identifying the top three contributors to dosage uncertainty. Target a combined uncertainty budget ≤ 60% of the allowable MPE to provide sufficient margin for production variability and long-term drift. For a Class A device with ±2% MPE, this means the design target should be ±1.2% or better. Common dominant contributors include: (1) plunger position measurement resolution and accuracy, (2) tubing compliance compensation algorithm, and (3) temperature-induced viscosity changes in the fluid being delivered.
4. Frequently Asked Questions
Q: Does ISO 28621 apply to all medical devices that deliver medication?
A: It applies to devices where dosage measurement is performed by the device itself. Devices that rely on external measurement (e.g., manually administered syringes where the clinician reads the volume from the barrel markings) are not within scope. However, smart syringes with integrated measurement electronics would fall under the standard’s requirements.
A: It applies to devices where dosage measurement is performed by the device itself. Devices that rely on external measurement (e.g., manually administered syringes where the clinician reads the volume from the barrel markings) are not within scope. However, smart syringes with integrated measurement electronics would fall under the standard’s requirements.
Q: How does this standard relate to IEC 62304?
A: ISO 28621 is complementary. While IEC 62304 covers medical device software lifecycle processes, ISO 28621 specifies the metrological requirements for the dosage measurement function specifically. Both standards apply concurrently to programmable infusion devices, and software validation must address both the general requirements of IEC 62304 and the measurement-specific requirements of ISO 28621.
A: ISO 28621 is complementary. While IEC 62304 covers medical device software lifecycle processes, ISO 28621 specifies the metrological requirements for the dosage measurement function specifically. Both standards apply concurrently to programmable infusion devices, and software validation must address both the general requirements of IEC 62304 and the measurement-specific requirements of ISO 28621.
Q: What is the maximum permissible occlusion alarm delay?
A: For low-rate infusions (≤ 5 mL/h), the occlusion alarm must activate within 30 minutes. For high-rate infusions (> 5 mL/h), within 10 minutes. These time limits reflect the clinical reality that low-rate infusions deliver small volumes and therefore pose less immediate risk from undetected occlusion, but the delayed alarm still requires careful clinical risk management.
A: For low-rate infusions (≤ 5 mL/h), the occlusion alarm must activate within 30 minutes. For high-rate infusions (> 5 mL/h), within 10 minutes. These time limits reflect the clinical reality that low-rate infusions deliver small volumes and therefore pose less immediate risk from undetected occlusion, but the delayed alarm still requires careful clinical risk management.
Q: Are there special requirements for pediatric applications?
A: Yes, pediatric devices must meet Class A accuracy across a wider flow rate range (0.1-100 mL/h) and include additional safety features such as dose limits, air-in-line detection, and pressure history logging. Neonatal applications have even more stringent requirements, with flow rate accuracy of ±1% required for rates as low as 0.1 mL/h, demanding advanced drive mechanisms and compensation algorithms.
A: Yes, pediatric devices must meet Class A accuracy across a wider flow rate range (0.1-100 mL/h) and include additional safety features such as dose limits, air-in-line detection, and pressure history logging. Neonatal applications have even more stringent requirements, with flow rate accuracy of ±1% required for rates as low as 0.1 mL/h, demanding advanced drive mechanisms and compensation algorithms.
