A hard truth to swallow: critically evaluating the MD Anderson Dysphagia Inventory (MDADI) as an endpoint in human papillomavirus-associated oropharyngeal cancer trials

The MD Anderson Dysphagia Inventory (MDADI), a measure of swallowing-related quality of life (swQol), has become the preferred patient-reported outcome measure (PROM) in contemporary clinical trials evaluating the experience of human papillomavirus-associated oropharyngeal squamous cell carcinoma (HPVOPSCC) survivors. With many potentially practice-changing studies using the MDADI composite score as either a primary or co-primary endpoint, or a key secondary endpoint, it is important to understand its psychometric properties as judged by contemporary PROM standards with a particular focus on its application to contemporary HPVOPSCC populations. In this critical review, we evaluate contemporary HPVOPSCC studies reporting MDADI outcomes, followed by a detailed evaluation of the psychometric properties of the MDADI. While the focus of this review was the MDADI, the issues discussed are not unique to the MDADI and have broader applicability to the evaluation and assessment of other PROMs currently in use. First, it may be possible to improve administration of the instrument, as related to missing items, scoring, and the number of items required. Second, while in many instances the MDADI has been intended as a direct or surrogate measure of swallowing physiology, the MDADI composite score captures a broader health-related quality of life construct affected by both swallowing and eating, the latter of which may be affected by a range of nonswallowing treatment-related toxicities. Finally, a clinically meaningful change (CMC) of 10 in the MDADI composite score, widely accepted and applied to the clinical trial setting, represents an undoubtably clinically relevant difference in unselected HNC survivors. However, the smallest difference that might be clinically important to a highly functional HPVOPSCC cohort remains uncertain. Understanding the purpose and properties of the MDADI instrument and furthering the sophistication with which we apply it in this population would improve its interpretation in clinical trials.