Citation: Farjas P, Regard A, Grevin G, “Influence of the Human Factor on Nasal Drug Delivery Device Evaluation”. ONdrugDelivery Magazine, Issue 85 (Apr 2018), pp 4-8.

Pascale Farjas, Alain Regard and Guillaume Grevin, discuss what is required of a modern nasal spray pump and how introducing the human factor to a study of five nasal spray pumps, including Nemera’s own Advancia device, showed a significant impact on the variability of delivered volume.


Intranasal delivery is a common route of administration for treating various indications, from allergic rhinitis to breakthrough cancer pain. On the one hand, nasal delivery is an attractive option for locally acting medications (e.g. saline solutions, decongestants, corticosteroids or antihistamines) which treat allergic rhinitis and nasal congestion. On the other, the nose is also the entry point for systemic delivery of numerous drugs and therapies for a variety of diseases. In these cases, common drugs used include calcitonin (osteoporosis), fentanyl, triptans (pain management), estradiol (hormone replacement therapy), nicotine (smoking cessation), desmopressin (enuresis), and metoclopramide (motion sickness).1

“Ever more drugs targeting other therapeutic fields and diseases may join the increasing ranks of marketed products for systemic delivery using the nasal route, such as drugs that act upon the central nervous system to treat disorders like Alzheimer’s disease and obesity…”

Ever more drugs targeting other therapeutic fields and diseases may join the increasing ranks of marketed products for systemic delivery using the nasal route, such as drugs that act upon the central nervous system to treat disorders like Alzheimer’s disease and obesity.2 The main reason for this trend in achieving systemic delivery via the nasal route is the advantage of delivering treatments directly from the olfactory region into the brain, allowing the drug to circumvent the blood-brain barrier.

Furthermore, nasal vaccination is an attractive alternative to injection that causes little discomfort to patients. Mucosal vaccines not only promote good local immune protection, but also a systemic response similar to that of injection.3 FluMist® (MedImmune (AstraZeneca), Gaithersburg, MD, US), is a nasal influenza vaccine currently on the market and is a popular alternative to the traditional influenza vaccine injection, particularly for children.

Also, preservatives, such as benzalkonium chloride, are commonly used in nasal drug formulations. However, preservatives can irritate the mucosa, deteriorate ciliary clearance and cause unpleasant adverse effects, such as itching, which may negatively impact compliance. For instance, long-term use of intranasal corticosteroids with benzalkonium chloride can lead to high-grade dysplasia in the nasal mucosa.4 The development of preservative-free nasal medications is especially important for chronic treatments which, by their nature, require daily use over several months (e.g. for allergic rhinitis therapies) and require appropriate multi-dose delivery systems that prevent contamination.

Nemera has developed a new preservative-free nasal pump, designed to increase compliance and deliver a precision dose independent of user actuation profile, ensuring the full dose is administered every time.


Nasal spray pump performance can be evaluated in accordance with different regulations, such as those set out by the EMA, US FDA or one of the pharmacopeias. Physical characteristics of the spray are then measured with precise methods.

Dose delivery, or shot weight, consists of single actuation weighing and gives information about the consistency of dose delivered to the patient.

  • Droplet size distribution is measured by laser diffraction.
  • Particle size is measured by cascade impactor.
  • Spray geometry consists of the plume geometry and spray pattern measurements.

These characteristics are also important for evaluating the performance of the pump in order to predict nasal deposition. In particular, taken together the droplet size distribution and particle size measurements inform about the overall particle size distribution of the spray, enabling prediction of how the spray deposits in the airways.

Although these physical parameters are relevant to describe the spray produced by nasal pump, in vitro spray characteristics evaluated in accordance with different standards/guidelines are mainly assessed using automatic spray actuation controlling the velocity, timespan and strength of actuation. This is restrictive as it does not mimic the variation in actuation profile demonstrated by humans. Previous studies have demonstrated the influence of the actuation parameters on spray characteristics,5 questioning the robustness of standard in vitro methods for predicting the spray performances when, in practice, the device is manually actuated by patients.


Nemera proposed a study in which the evaluation of intranasal delivery devices took into account the human factor. First, Nemera evaluated the dose delivered under real use conditions with volunteers. In vivo measurements were taken in terms of delivered volume via manual actuation for evaluating the influence of the human factor on the variability of the nasal spray pump’s performances. The same devices were then evaluated using an automatic actuator system to measure the delivered volume as per standard methodology. In vivo and in vitro results were then compared to determine in vitro/in vivo correlation.

The delivery system is a critical element for nasal spray performance, in particular it needs to deliver a uniform dose upon each actuation. Hence, the device must be user-friendly and convenient for “on the go” use so that the patient can rely on the nasal spray at any time during the treatment, especially during a migraine or allergy related symptoms, which often occur outside the convenience of the patient’s home.

Figure 1: In vivo study protocol.

Furthermore, adherence is a key parameter which can influence the efficacy of the treatment. Indeed, not only should the therapeutic efficacy and molecule safety be taken into account for a treatment, but ease of use and comfort of the dispensing system for the patients as well. The nasal spray should help the patient to accept the treatment and therefore improve patient compliance. To achieve patient acceptance and improved compliance, ergonomics should be applied to the nasal device design to ensure overall attractiveness and user-friendly features, such as intuitive handling, good grip, uniform delivery accuracy regardless of actuation profile, etc. These attributes should allow the patient to use the device properly and receive their daily dose of medication required, by improving overall patient compliance.

As with all self-administered drugs, the most critical parameter affecting device performance is the patient themself. A patient, most of the time untrained, relies on their personal appreciation and the instructions for use to operate the device properly. This perspective constitutes the fundamentals of Human Factors Engineering (HFE),6,7 an inclusive design process that aims to identify and mitigate all user-induced risks. The HFE process is based on user studies to identify risks and user misunderstandings, then improve the device design accordingly. HFE also highlights user competence and satisfaction as equally important in ensuring patients’ adherence to their treatment. Verifying both a safe and user-friendly device eventually relies on a combination of very different factors; ranging from functional to more perceptive ones, such as overall ergonomics or daily-use adaptability.

In the second step of this study, Nemera evaluated the performance of the devices in terms of perception and feeling because even the best drug, in the best container, with the best delivery device, can end up being useless if patients don’t, or can’t, use it properly.


Figure 2: In vivo volume delivery test with Advancia nasal pump.

Thirteen healthy adult volunteers (seven male, six female) were included in this single-centre study (Figure 1). The study was performed in 2017 at Nemera (Innovation Center, La Verpillière, France). Five different nasal spray pumps from different companies were filled with water and were tested in a randomised order, including Nemera’s own next-generation pump, Advancia (Figure 2). A first interview was conducted with the volunteers before the test. Each product was observed and manually manipulated. A second interview was conducted for product comparison. The device was manually primed and then given to the healthy volunteer. The device was actuated in ambient air and the weight difference was measured to calculate the delivered volume by the volunteers. Three actuations were performed per device and per user, resulting in 39 results per nasal pump. After the five pumps were tested manually by the volunteers, a last interview was conducted about the five nasal pumps. Participants were asked to rank the different nasal pumps in three boxes: the green box (for those they would like to use for a long-term daily treatment), the red box (for those they wouldn’t like to use for this treatment) or the yellow box (for those in between). Results to these questions gave a ranking from the most appreciated pump to the least appreciated one.


Figure 3: In vitro volume delivery test with Advancia nasal pump.

The same devices used in the in vivo study (provided from the same batches) were tested in a randomised order in an in vitro study (Figure 3). They were filled with water and manually primed. Devices were placed in an automatic actuator (Proveris NSx, actuation speed of 80 mm/s, acceleration of 7000 mm/s²). Devices were actuated in ambient air and the weight difference was measured to calculate the delivery volume. 39 actuations were performed per device to match the number performed in the in vivo study.


Figure 4 clearly shows how the volunteers ranked the different pumps. Advancia proved to be the pump that volunteers felt most positively about, with no appearances in the red box and almost 70% of the time being placed in the green box (nine in the green box). Contrarily, Devices A and B were the two pumps that the volunteers felt the least affinity for (one and two in the green box respectively).

Figure 5, and the associated data in Table 1, shows the mean average of the delivered volume across the devices for the in vitro and in vivo studies. All products have a relatively low in vitro variability, with a maximum of 5% for device B and a minimum of 0.6% for device D. However, manual actuation by the 13 volunteers shows larger variability in terms of delivered volume, with the exception of Advancia (device E). There was a good correlation between the mean of in vitro delivered volume (x) and the mean of the in vivo delivered volume (y) (y=0.93x, R2=0.97). Nevertheless, Figure 5 shows a trend towards lower volumes delivered by manual actuation, most keenly observed in device D, with a dose decrease of 12% between in vitro and in vivo actuation.

Figure 4: Results from interviewing the 13 volunteers.

Figure 5: Volume delivery by automated actuation (in vitro) and by manual actuation (in vivo) for the five nasal pumps (n=39 actuations per device) with the mean value shown in black.

In Vitro In Vivo
Device Mean (µL) SD (µL) Min-Max (µL) Mean (µL) SD (µL) Min-Max (µL)
A 95 1 93-96 92 12 56-101
B 52 3 47-55 50 5 41-59
C 100 1 99-105 96 10 45-105
D 132 1 130-134 116 19 75-131
E (Advancia) 48 1 44-50 48 1 45-50

Table 1: Volume delivery by automated actuation (in vitro) and by manual actuation (in vivo) for the five nasal pumps (n=39 actuations per device) in terms of mean, standard deviation (SD) and minimum-maximum (Min-Max).

Regarding data dispersion, we can observe a high user dependency on volume delivery for devices A, C & D, around 25% in terms of variability, and a lower user dependency for devices B and E (Advancia), around 5% in terms of variability. These results confirm the influence of a manual actuation of the device on delivered volume variability.


A correlation in terms of mean delivered volume was found between in vitro and in vivo tests, showing a lower delivered volume by manual actuation in comparison to automated actuation (-7%). Different manual actuations by users were visually observed, explaining a partial delivered volume from the device when the actuation was incomplete. Furthermore, a high difference in terms of delivered volume variability was observed between in vivo and in vitro results, demonstrating the difficulty of predicting real device variability from automatic actuation only. Nemera’s Advancia device was the only device to show a good correlation between in vitro and in vivo results in terms of delivered volume and variability.

Regarding the user satisfaction for the five nasal pumps, nine out of the thirteen volunteers said they would prefer using the Advancia device for a long-term treatment, more than for any of the other devices in this study. However, whilst Advancia had the best results both in terms of the user satisfaction test and delivered volume variability, interestingly there was no overall correlation found between patient satisfaction and spray volume variability.

The conclusions drawn from this study have limitations, in particular with regard to the low number of participants who tested devices, and therefore complementary tests should be performed in order to confirm these results. However, similar conclusions have been obtained by droplet size distribution measurement,8 showing a higher variability in terms of droplet size by manual actuation compared with automated action.

Nevertheless, an automatic actuator is recommended by multiple regulatory agencies for spray performance evaluation. This may be due to automatic actuation having the advantage of being able to evaluate variation associated with different batches of drug product. Furthermore, controlling the actuation force or the velocity can help to give a better understanding of the inherent device performance by eliminating them as variables. Therefore it can be stated that an automatic actuator is a good option for evaluating device quality but seems poorly suited for predicting in vivo results.


The authors would like to thank Laurent Vecellio for his contribution.


  1. Bitter C, Suter-Zimmermann K, Surber C, “Nasal Drug Delivery in Humans” in “Topical Applications and the Mucosa” (Surber C, Elsner P, Farage MA eds), Basel, Karger, 2011, Vol 40, pp 20–35.
  2. Pardeshi CV, Belgamwar VS, “Direct Nose to Brain Drug, Delivery via Integrated Nerve Pathways Bypassing the Blood-Brain Barrier: An Excellent Platform for Brain Targeting”. Expert Opin Drug Deliv, Jul 2013, Vol 10(7), pp 957–972.
  3. Djupesland PG, “Nasal Delivery of Vaccines”. Pharmaceutical Manufacturing and Packing Sourcer, Winter 2002 Issue.
  4. Tamer et al, “Nasal mucosal dysplasia induced by topical corticosteroids with benzalkonium chloride”. Indian Dermatol Online J, 2018, Vol 9(2), pp 126–127.
  5. Guo C, Doub WH, “The influence of actuation parameters on in vitro testing of nasal spray products”. J Pharm Sci, Sep 2006, Vol 95(9), pp 2029–2040.
  6. “Draft Guidance for Industry and FDA staff: Applying Human Factors and Usability Engineering to Optimize Medical Devices Design”. US FDA, 2011.
  7. “Application of Usability Engineering to Medical Devices”. ISO EN 62366, 2008.
  8. Kippax PG, Krarup H, Suman JD, “Manual versus Automated Actuation of Nasal Sprays”. Outsourcing Resources (suppl to Pharm Tech), 2004, pp 30–39.