Development, validation and application of a device to measure e-cigarette users’ puffing topography

This study was designed to develop and validate a puffing topography device adapted for use with e-cigarettes and to measure users’ topography for two e-cigarette products. Changes to the topography device head were made to address the known limitations of puffing topography devices when used with e-cigarettes—condensation of aerosol, ability to measure only a limited number of puffs and reliable measurement at low flow rates. Our findings demonstrate the robustness of the modified topography head to measure puff volumes following repeated puffing (up to 150 puffs) on the e-cigarette devices. This development enables collection of reliable puffing topography data to be added to the scientific literature and will help in the development of standardised laboratory testing protocols for e-cigarettes that better reflect actual consumer behaviour.

Puffing topography devices have been extensively used to study smokers’ puffing behaviours, and several studies have reported little effect of the measurement process on puffing topography parameters and the smoking sensory experience. Blank et al.⁹ compared the effect of using direct observation measurements against those of portable and desktop puffing topography devices, and found relatively small differences between methods. Ross et al.¹¹ reported no systematic differences in how cigarettes are smoked with topography devices relative to natural smoking. Lee et al¹⁴ investigated whether smokers changed their puffing behaviour over time when smoking through a puffing topography device and found no significant effect, although day to day variability led to 7% variance in topography measures. Those results demonstrate that topography measurements can yield exposure data that are representative of actual use.

The study by Blank et al.⁹ comparing video recording and puffing topography devices showed longer puff durations with video recordings than with topography devices. The increased puff durations with direct observation may have been caused by participants holding the cigarettes in their mouth before puffing, or keeping them there during the mouth-hold phase. Initial studies of the puffing behaviour of vapers involved analysis of video footage to measure puff durations. Hua et al.¹⁵ used a stopwatch to record puff durations, and found that those for e-cigarettes were significantly longer than those measured for smokers. Hua used the time the e-cigarette LED light was on to determine puff duration, which could limit the effect of keeping the device in the mouth. Farsalinos et al.¹⁶ used video-processing software to measure puff duration on a frame-by-frame basis. Results from this study agreed with those reported by Hua et al.

Some button-activated e-cigarette devices are available that offer the opportunity to approximate puffing duration by recording, within the e-cigarette, the length of time the user activates the heating element^12,13. However, many users pre-heat their heating coils before puffing and, therefore, this approach can overestimate actual puff durations.

A small number of studies have used the commercially available CReSS puffing topography device (Borgwaldt) to obtain e-cigarette users’ puffing topography data, but researchers have reported reliability issues and data capture limitations. Norton et al.¹⁷, reported that device failure led to a loss of nine participants’ data. Behar et al.¹⁸ noted that although the CReSS devices were supplied with e-cigarette adaptors, no specific user instructions were available for use of the device with e-cigarettes, and 4–5 months of method development was necessary to be able to use the device reliably. In addition, the authors reported limitations of the device when collecting more than 43 puffs, leading to inaccurate puff number determinations in 26% of the user sessions. Other researchers have reported fewer issues with their own non-commercial devices^19,20.

Spindle et al.¹⁹ reported two potential challenges to measuring e-cigarette users’ topography: first, condensation and build-up of aerosol within the topography device, and second, the ability of the device to measure low flow rates generated by e-cigarette users. The modifications we made to our topography head limited the susceptibility of the device to report inaccurate flow rates and puff volumes. Whilst initial measurements of puff volumes were successful using the unmodified device in a cleaned condition, the accuracy of the measurements decreased following repeated puffing on a disposable e-cigarette through the device (see Supplementary Table S3). Visible droplets of deposited aerosol were observed within the tubing of the topography device on prolonged usage with the e-cigarette and led to the development of the modified device to limit the effect of aerosol deposition. For the modified topography device, continual puffing on three different e-cigarette types to exhaustion of the device battery or a maximum of 150 puffs, under three puffing regimes, resulted in puff volumes within the range of pre-set tolerances. Whilst it is unlikely that an e-cigarette would be used from battery recharge to exhaustion or for 150 puffs during a single vaping session, the accuracy of the measured puff volumes provides evidence of the modified topography device’s suitability to be used for extended periods of time. The ability of the topography device to accurately measure low flow rates was tested during validation by inclusion of 33 calibration flow rates down to 2 mL/s. Ensuring this capability results in more of each individual puff being captured and, therefore, leads to increased accuracy of puff volume and duration measurements. Coupled with the device’s ability to record data at a sampling rate of every 40 ms, which results in a more precise measurement of the transient features of each individual puff, puffing topography datasets can be accurately duplicated in the laboratory to provide a measure of users’ exposure to aerosol emissions²².

The puffing topography parameters measured in this study fell within the range of values measured with puffing topography devices in the literature; mean puff volumes of 52.3 and 83.0 mL versus 51–133 mL and mean puff durations of 2.0 and 2.2 s versus 1.8-4.16 s^{17,18,19,20,23,24}. Data from several sources have shown that for vaping products puff duration is the main determinant of the amount of aerosol per puff, whereas puffing volume and air flow speeds has little influence⁵. These effects are in sharp contrast to how cigarettes respond, and are thought to comprise the main reason why smokers adapt their puffing behaviour as they learn to vape.

Puff durations have been reported to be longer for e-cigarette users than cigarette smokers^12,15,16, and for duration to increase over time with increasing e-cigarette experience^16,24. However, the time frame involved in adaption appears to be relatively short. Lee et al.²⁴, found significant increases in puff duration after 1 week of e-cigarette use by cigarette smokers, followed by a small decrease in the second week. We found no differences in mean puff duration between users who had been vaping for more than 6 months and those who had vaped for at least 1 month but not more than 6 months. It would thus appear that behaviour is largely stabilised within this first month of use. This interpretation would also be consistent with the change in mean ‘puff duration’, approximated by button activation time that has been recorded over 2 months from initiation of use of certain eGO type open tank products¹². The mean puff duration increased from approximately 3.4 s to 4.1 s within the first 2 weeks of use of the product and then remained stable near that value for the next 6 weeks of study time. In that study users new to using that specific product may not have been new to vaping and, therefore, any adaptation effect would have been less pronounced. On the basis of the discussion above, we assume that the vapers in our study had largely adapted to vaping before participating.

Behar et al.¹⁸ observed significant differences in topography measures (except puff number) between two brands when used by the same participants on the same day, although, in absolute terms (mean puff volumes 56 mL versus 45 mL and durations of 2.75 s versus 2.54 s), the differences were smaller than those previously reported across multiple studies. The aerosol mass emissions from the button-activated product used here, Vype ePen, were significantly higher than those of the “cig-a-like” product, Vype Reload, under standardised machine testing conditions. The significant differences found in mean puff number, volume, interval and peak flow between devices supports the hypothesis that topography is not only determined by user characteristics, but also by product design. This theory is supported by the absence of differences in topography data between study visits for either of the two groups.

Additional data to support the ability of the topography device to measure puffing behaviours are those generated from the use of the unmodified SA7 topography device to measure puffing topography for smokers of a range of cigarette products²⁵. Puff volumes in the range of 42.9 and 54.3 mL reported by Ashley et al.²⁵ fall within the range of 30.8 to 67.5 mL reported by others for cigarette smokers^9,14,17,19. Furthermore, larger puff volumes for e-cigarette users compared with smokers as measured by the same topography device have been reported by Norton et al.¹⁷, which agrees with the larger mean puff volume of 83 mL reported here for the users of Vype ePen compared to those reported for cigarette smokers²⁵.

The wide variation observed across topography studies and devices highlights the challenges in establishing standardised laboratory testing protocols for e-cigarettes before the contributing factors are fully characterised. Contributing factors probably include differences between products (e.g. pressure drop, battery power and nicotine strength), accuracy of topography devices and participants’ use history and demographics. Whether study design parameters alter users’ natural behaviour should be considered when undertaking studies intended to reflect real-life use. For example, participants in the study by Spindle et al.¹⁹ were instructed to take 10 puffs with a fixed interval of 30 s, whereas Behar et al.¹⁸ limited session lengths to 10 min. Both of these designs do not reflect real-world vaping use, and the impact of any of these restrictions on study findings is unclear.

A number of limitations apply to this study. First, the participants were using unfamiliar products, and features such as the nicotine strength and formulation of the e-liquid which varied across the study products might have affected puffing behaviour, but reflects that more powerful e-cigarette devices typically use lower concentrations of nicotine. Future research should focus on participants using their normal product, which could allow a systematic study design to be used thus allowing for the effects of factors such as device type and nicotine strength on puffing topography to be studied. In addition future studies should consider the impact of participants’ characteristics such as dual usage of cigarettes and e-cigarettes on topography data. Second, having to attend a central location for testing and being in the presence of research staff could have shortened session lengths and led to puffs with reduced intervals. Portable topography devices that capture data over a number of usage sessions within a 24 h period or number of days would provide more naturalistic user data^17,20 but at the possible expense of reduced data integrity.