(M1230-05-30) Application of Quantitative Models to Systematically Evaluate Factors Impacting Swallowability of Tablets

Purpose: Swallowability may be defined as the patient being able to take the drug without gagging or choking per the U.S. Food and Drug Administration (FDA)’s Guidance for Industry [1]. The usability of drug products is greatly impacted by its swallowability, since swallowing difficulties may lead to many issues such as the misuse risk from inappropriate manipulation of opioid tablets to make swallowing easier [2-4]. The U.S. FDA has also published a guidance on tablet size to help reduce swallowing difficulties [4]. Many research studies including some clinical studies have been conducted on assessing factors that impact swallowability. However, most of the reported results were based on a single clinical study for a certain group of subjects and used conventional statistical methods, like a chi-test for analysis [5-8]. In this abstract, we reviewed study designs used for clinical swallowing evaluation that were submitted to FDA, collected data from eight completed studies and conducted a systematic evaluation on factors affecting swallowability based on machine learning models.

Methods: This work first established a dataset by collecting individual subject data from eight clinical studies with swallowability assessment submitted to the FDA. Data were standardized/integrated from these eight clinical studies retaining individual level swallowability data. In the standardized database, 1028 records from 667 subjects were collected in total (with 352 of these subjects having two or three records of swallowing one drug product with different formulations). Each record had an outcome variable indicating the tablet swallowability from the subject, five subject-specific factors (i.e., “age”, “sex”, “height”, “weight”, and “healthy status”), and four product-specific factors (i.e., “drugCoating” indicating whether the product is coated, “drugMaxDimension” indicating the product maximum physical dimension, “drugColor” indicating the product physical color, and “drugShape” indicating the product physical shape). With the constructed dataset, exploratory analysis (e.g., correlation analysis) and modeling approaches (e.g., logistic regression and machine learning model - Conditional Decision Tree) were conducted to assess the impact of these variables on swallowability of drug products.

Results: when the size of tablets is larger than 15mm (tree splitting 1à5à9àNode 10, comparing with tree splitting 1à5à6). Lastly, subjects older than 11 years of age in general have no swallowability issue (tree splitting 1à5à9à11).

Conclusion: This research work complements the existing swallowability literature by applying a quantitative model across multiple clinical studies. The findings identified age and the maximum dimension as primary factors affecting swallowability which is consistent with previous literature. Considering that the Conditional Decision Tree is a machine learning model for pattern recognition, this work offers a great potential for future research in predicting swallowability of specific tablets and its target subject groups.

References: 1. U.S. Food and Drug Administration (FDA) Draft Guidance for Industry: Use of Liquids and/or Soft Foods as Vehicles for Drug Administration: General Considerations for Selection and In Vitro Methods for Product Quality Assessments.
2. Schiele, J.T., et al., Difficulties swallowing solid oral dosage forms in a general practice population: prevalence, causes, and relationship to dosage forms. Eur J Clin Pharmacol, 2013. 69(4): p. 937-48.
3. McCarberg, B.H., et al., An abuse-deterrent, microsphere-in-capsule formulation of extended-release oxycodone: alternative modes of administration to facilitate pain management in patients with dysphagia. Curr Med Res Opin, 2016. 32(12): p. 1975-1982.
4. U.S. Food and Drug Administration (FDA) Guidance for Industry: Size, Shape, and Other Physical Attributes of Generic Tablets and Capsules (2022).
5. Bracken L, McDonough E, Ashleigh S, et al. Can children swallow tablets? Outcome data from a feasibility study to assess the acceptability of different-sized placebo tablets in children (creating acceptable tablets (CAT)). BMJ Open 2020;10:e036508. doi:10.1136/bmjopen-2019-036508.
6. Overgaard, A. B. A., et al. "Patients' evaluation of shape, size and colour of solid dosage forms." Pharmacy World and Science 23 (2001): 185-188.
7. Yamamoto, Shinya, et al. How do tablet properties influence swallowing behaviours? Journal of Pharmacy and Pharmacology, 66.1 (2014): 32-39.
8. Kelly, Jennifer, Gibson D’Cruz, and David Wright. Patients with dysphagia: experiences of taking medication. Journal of Advanced Nursing, 66.1 (2010): 82-91.

Acknowledgements: This abstract reflects the views of the authors and should not be construed to represent FDA’s views or policies.

Figure 1.