Issue link: https://www.e-digitaleditions.com/i/1496325
28 April 2023 Inhalation pressure fluctuations at sea level in the nominal sam- pled volume of two liters is predicted to be less than -1% of nominal (Figure 2). Likewise, the equivalent bias at the two higher altitude examples where DPI performance testing likely takes place is predicted to be <1.5%. Similarly, deviations from the nominal 4-kPa pressure drop at the inlet to the DUSA are pre- dicted to be only marginally sensitive to local changes in atmospheric pressure, being in the range -2% to +0.5% at the highest altitude (worst case condition), decreasing slightly to vary from -1.5% to +0.5% at sea level (Figure 3). We conclude that for perfor- mance testing at a fixed altitude, corrections of the sampled air volume and inlet pressure drop for ambi- ent pressure variations are unnecessary. Further, there appears to be no need for different set-up procedures at laboratories located significantly above sea level. However, we take this opportunity to remind users of this method that for the purposes of best practice, fil- ter resistance in the DUSA should be investigated if a change is made to a different filter product, especially if nominal porosity is different. Further, the volumet- ric flow rate should be set frequently when a series of measurements is being made, ideally before every replicate, but at least daily. Finally, it is important to ensure a protective filter is always present during the flow rate measurement; if it is present at some sta- tions and not others, or changes from test to test, the gas density inside the total dose tube during measure- ment can be different among the multiple tests and test set-ups in a major manufacturing facility. We have not considered the effect of a leak at the mouthpiece adapter because our purpose has been to examine the method robustness when followed under ideal conditions. However, such an investiga- tion would be of interest in a future extension of the model, most likely based on well-defined and realis- tic leak path geometries. Perhaps of more immedi- ate importance, however, would be to continue this modeling in the context of the compendial methods for passive DPI-generated aerosol APSD determina- tions, given the added potential for changes to the dynamics of powder deagglomeration, subsequent aerosolization and transport into the multi-stage cas- cade impactor associated with the flow rate-rise time profile following actuation [9]. Achieving this goal will therefore be our next task. Acknowledgement e authors wish to acknowledge Mr. Mark Copley of Copley Scientific, Ltd. (Nottingham, UK) for providing the accurate value of 34.85-mm diame- ter for the open area of filters held in ordinary DPI DUSA tubes. postulate the same (fractional) range of ambient pressure variation, but starting at two representa- tive locations that are significantly above sea level, namely Albuquerque, NM, US (5,310 feet) and Mexico City, Mexico (7,350 feet) [11]. e calculation results presented in Figure 2 show that the sampled volume can be up to 1% lower than its expected value if the ambient pressure at the time of measurement drops by a few percentage points. is result derives from the pressure drop of the flow meter in step 2 being smaller than the device pres- sure drop in steps 1 or 3. In addition, as the atmos- pheric pressure changes at the time of testing, the flow meter pressure drop offset is a different fraction of the atmospheric air pressure. ere is, therefore, an inherent failure to deliver the target sampled vol- ume of two liters even when the atmospheric pressure does not change, simply because the pressure drop of the flow meter is smaller than 4 kPa. Figure 3 shows that the device pressure drop can also fall by a percentage point or more if or when the ambient pressure at the time of testing decreases by a few percentage points from the atmospheric pres- sure at the time of set-up. e curves are somewhat steeper as the altitude increases, reflecting the slightly non-linear dependence of the filter flow resistance to the volumetric flow rate approaching the filter face. A change in the device pressure drop can intrinsically affect the powder dispersion [9], and we anticipate that this type of change is more important than an offset in the sampled volume. Discussion and conclusions e findings from this exploratory analysis of the underlying physical processes involved with the deter- mination of total dose content from a DPI following methodology advocated in the pharmacopeial com- pendia [1, 2] support the robustness of these meth- ods, in that bias arising from normal, local, ambient Figure 3 Influence on device pressure drop of ambient test pressure Baseline conditions Albuquerque, NM: alt. 5,310 ft Mexico City: alt. 7,350 ft 4.04 4.02 4.00 3.98 3.96 3.94 3.92 3.90 70,000 75,000 80,000 85,000 90,000 95,000 100,000 105,000 Ambient Pressure (Pa) Device Pressure Drop (kPa) Datum Pressure Drop: 4 kPa