John Devin Peipert’s research while affiliated with University of Birmingham and other places

We aimed to develop an Explanation & Elaboration document to provide explanations, rationales, and good examples of reporting for each item in the COSMIN Reporting Guideline 2.0. The development of the Explanation & Elaboration document was conducted in four phases. In phase 1, experts on measurement properties were invited as writers to draft explanations and examples for each item. The original COSMIN Reporting Guideline items into 22 clusters, each containing two to six items, which were self-assigned by writers based on expertise. In phase 2, a draft of the Explanation & Elaboration document was created. For each item, an ‘Explanation’, ‘Essential elements’, and ‘Item examples’ were written. The draft was reviewed by a team of reviewers for grammar, conciseness, and consistency with other COSMIN tools. In phase 3, the Explanation & Elaboration document was extensively revised considering the newly created COSMIN reporting guidelines 2.0. In phase 4, the final version incorporated feedback from all writers and final approval. Eighteen writers drafted the item explanations and provided examples of good reporting. The Explanation & Elaboration document contains explanations of the General and Specific items. We explain the reasons why reporting of the item is recommended and how to report the items with examples of good reporting. The Explanation & Elaboration document of the COSMIN Reporting Guideline 2.0 facilitates comprehensive reporting of studies on measurement properties of PROMs and should be used in conjunction with the COSMIN Reporting Guideline 2.0.

Background Aromatic L-amino acid decarboxylase deficiency (AADCd) is a rare genetic disorder characterized by movement disorders, motor and autonomic dysfunction, and developmental delays. The gene therapy eladocagene exuparvovec has become available in some regions; pooled clinical trial results demonstrate continuous long-term improvement in motor development and cognitive function. We sought to characterize clinically meaningful change in motor function, as measured by Total Peabody Developmental Motor Scales-Second Edition (PDMS-2) score, and assess correlations with cognition and language domains of the Bayley-III and motor milestone (MM) achievement. Methods Data from N = 30 patients from three single-arm clinical studies of eladocagene exuparvovec were analyzed. Anchor-based estimation of the meaningful score difference (MSD) of Total PDMS-2 score was conducted using mean-difference and receiver operating characteristic curve (ROC) approaches. MM achievement served as the anchor defining meaningful change. Results An MSD of 40 points for Total PDMS-2 score was selected for analysis as it yielded specificity > 0.95 using the ROC approach, and generally aligned with the mean-difference approach. Cumulative incidence analysis reflected that 50% of patients treated with eladocagene exuparvovec may achieve the MSD of 40-point change in Total PDMS-2 score at 6 months, and 86% at 18 months. Correlations between measures were of large magnitude and improved over time (Month 6: r = 0.599 [ p = 0.0032]; Month 18: r = 0.796 [ p = 0.0002]; Month 60: r = 0.861 [ p = 0.0007]). Conclusions The MSD of 40 points for Total PDMS-2 score enables the interpretation of changes observed in patients with AADCd, and suggests that treatment with eladocagene exuparvovec leads to significant improvements in motor and cognitive function.

PURPOSE Financial toxicity (FT) has been linked to higher symptom burden and poorer clinical outcomes for patients with cancer. Despite the availability of validated tools to measure FT, a simple screen remains an unmet need. We evaluated item 12 (“My illness has been a financial hardship to my family and me”) of the COmprehensive Score for Financial Toxicity (COST) measure as a single-item FT screening measure. METHODS In this secondary analysis, 711 patients with cancer (690 with breast cancer) were recruited via a web-based survey from a philanthropic organization. COST items 1-11 were scored according to Functional Assessment of Chronic Illness Therapy scoring guidelines, with lower scores indicating worse FT. Analyses focused on establishing a correlation, examining item properties, and sensitivity/specificity of item 12 relative to the total COST score. RESULTS Item 12 had a correlation of r = 0.53 with the COST-11 score, and an increase of one point on item 12 is associated with a decrease of approximately three total points on the full scale ( b , 3.35; P < .001; adjusted R ² , 0.28). Item analysis with the graded-response item in response theory modeling showed very good discrimination ( a , 2.096) for item 12, indicating that it can reliably distinguish between low and high FT in patients. Sensitivity ranged between 75.6% and 95.7% on all item 12 thresholds to screen positive for FT using two COST cutoffs as criteria. Maximizing both sensitivity and specificity was to be found for higher item 12 scores. CONCLUSION To our knowledge, this is the first validation of a single-item screening measure for FT. Overall, these results illustrate that item 12 from the COST measure is a good candidate for a single-item screener. Clinicians can choose among item 12 screening thresholds depending on their tolerance for low specificity.

Purpose Physical functioning and symptom severity are important factors in the experience of people with living with non-small cell lung cancer (NSCLC). This study evaluates the reliability, validity, and meaningful within-person change (MWPC) thresholds of 2 patient-reported outcome (PRO) measures in NSCLC: the Patient-Reported Outcomes Measurement Information System (PROMIS®) Physical Function (PF) short form (SF) 8c and the Non-Small Cell Lung Cancer Symptom Assessment Questionnaire (NSCLC-SAQ). Methods Data came from 2 Phase 3 clinical trials among people living with NSCLC. PROMIS PF-SF analyses included data from 300 participants in the PAPILLON trial, and NSCLC-SAQ analyses included 615 participants in the MARIPOSA 2 trial. Prespecified expected relationships between target PRO measures and relevant study variables were used to evaluate validity evidence. Additionally, MWPC thresholds were estimated using anchor- and distribution-based analyses. Results Both PRO measures exhibited adequate internal consistency for clinical trial use. All examined correlations with reference variables and score differences between clinically meaningful groups conformed to expectations for both measures. Estimated thresholds for meaningful worsening were a decrease of 6–7 points on the PROMIS PF-SF and an increase of 2–3 points on NSCLC-SAQ. Conclusions This study is the first to examine longitudinal measurement properties of PROMIS PF-SF and investigates thresholds for meaningful change on the PROMIS PF-SF and NSCLC-SAQ measures. Results support the validity of these measures in NSCLC and aid the interpretation of clinically meaningful change in scores over time.

Regulatory agencies, professional societies, and patient groups have encouraged the inclusion of patient-reported outcomes (PROs) in clinical trials. There is a growing understanding of and encouragement for the use of PROs to capture tolerability. Traditionally, tolerability has been defined as the absence of serious adverse events. More recent and patient-centric definitions of tolerability have been put forward, focusing on how patients feel and function while on treatment. This chapter will provide a comprehensive overview of how to measure, analyse, and visualise data from patients about how they feel and function while on therapy. We will describe the potential use of overall summary measures and multi-item scales for assessing symptomatic side effects. General statistical and practical considerations for tolerability data, as well as key study design and statistical considerations for early-phase and late-phase trials, will be discussed. Issues that will be covered include dose optimisation in early-phase trials and the specification of tolerability endpoints and hypotheses in late-phase trials. We will also discuss key considerations when measuring tolerability in paediatric trials, including paediatric-specific measures and considerations of when children are able to self-report tolerability. Finally, considerations for communicating tolerability data will be discussed.

351 Background: Financial toxicity (FT) refers to the financial burden experienced by patients due to direct and indirect costs associated with their medical care. It has been linked to higher symptom burden, poorer clinical outcomes, quality of life, and psychological well-being and has been shown to extend beyond diagnosis and treatment. Despite the availability of validated tools to measure FT, no single-item screening measure has been validated yet. In the context where time constraints are the most cited reason for patients declining FT surveys, it is crucial to develop efficient tools that allow us to routinely screen patients for FT at different stages of their care. We, therefore, sought to establish item 12 (“My illness has been a financial hardship to my family and me”) of the Comprehensive Score for Financial Toxicity (COST) measure as a well-functioning screening item for FT. Methods: In this secondary analysis of 579 patients who completed a survey on financial toxicity screening preferences, items 1-11 were scored (including reversing 7 items per the version 2 scoring guidelines), with lower scores indicating worse FT. Analyses focused on establishing a correlation with the total score as well as examining item properties. Item 12 was also reverse scored, so lower scores mean greater hardship. Results: Item 12 had an item-total correlation of r = 0.63 (adjusted item-total correlation = .531). Similarly, a linear model with item 12 score as a predictor and total score as the outcome showed that increases of 1 point on item 12 is associated with around 3 total points on the full scale (b = 3.35, p < .001, adjusted R ² = 0.28. Finally, item analysis with the graded-response item response theory (IRT) model showed very good discrimination (a = 2.10) for item 12, indicating that it can reliably distinguish between low and high FT in patients. IRT results also showed that the item functions as a nearly binary item with responses of “quite a bit” and “very much” corresponding to a high probability of high FT, while responses of “somewhat,” “a little bit,” and “not at all” corresponding to a high probability of low FT. Conclusions: This is the first validation of a single-item screening measure for RT. Overall, these results illustrate that item 12 is a good candidate for a single-item screener, with “quite a bit” and “very much” responses indicating a high likelihood of high FT.