Methods for Estimating Analgesic Efficacy in Acute Pain Clinical Trials Christopher Joseph Miller

Format:

Book

Thesis/Dissertation

Author/Creator:

Miller, Christopher Joseph, author.

Contributor:

University of Pennsylvania. Epidemiology and Biostatistics., degree granting institution.

Language:

English

Subjects (All):

0308.

0564.

0622.

0766.

Local Subjects:

0308.

0564.

0622.

0766.

Physical Description:

1 electronic resource (150 pages)

Contained In:

Dissertations Abstracts International 87-07B

Place of Publication:

Ann Arbor : ProQuest Dissertations and Theses, 2025

Language Note:

English

Summary:

The U.S. Food and Drug Administration (FDA) requires at least two adequate and well-controlled trials to support the approval of novel non-opioid analgesics. However, the interpretation of efficacy from randomized controlled trials in acute pain is challenged by the lack of validated thresholds for determining whether differences in outcomes between treatment groups are clinically meaningful. Additionally, estimating the efficacy of an analgesic requires accounting for the effects of rescue medication on pain intensity scores. The current approaches for handling rescue medication involve single-imputation procedures, which are known to introduce bias and underestimate variability when their assumptions are not scientifically justified. This dissertation develops new statistical methods to address these limitations and evaluates them using individual participant data from trials previously submitted to the FDA. First, we applied a novel statistical method for the double-stopwatch technique to estimate the minimum clinically important difference (MCID) across different surgical models of acute pain. A 50% reduction from baseline was found to be a consistent, clinically meaningful threshold across surgeries. Second, we developed the Responder Outcome Over Time (ROOT), a more clinically interpretable, patient-centric endpoint that quantifies the proportion of time an individual experiences meaningful improvement, which excludes periods affected by rescue medication and study withdrawal. Compared to the conventional efficacy endpoint, the Sum of Pain Intensity Differences (SPID), ROOT provided a similar ability to detect statistically significant treatment effects while avoiding the strong assumptions required for imputation. Third, we developed Multiple Imputation for Rescue Adjustment (MIRA), a reference-based multiple imputation procedure that uses historical information on rescue drug efficacy to adjust pain scores after rescue use. Through extensive simulation studies, we showed that the single-imputation methods currently used to adjust for rescue bias both group-level estimates and treatment differences. In contrast, MIRA produced unbiased estimates when rescue effects were correctly specified and demonstrated robustness to misspecification. Together, these advances provide more principled approaches for evaluating analgesic efficacy in acute pain trials that could enhance both regulatory decision-making and future trial design

Notes:

Advisors: Farrar, John T. Committee members: Harhay, Michael O.; Connor, Jason T.; Long, Qi; Mitra, Nandita; Neuman, Mark D.

Source: Dissertations Abstracts International, Volume: 87-07, Section: B.

Ph.D. University of Pennsylvania 2025

Vendor supplied data

Local Notes:

School code: 0175

ISBN:

9798276005195

Access Restriction:

Restricted for use by site license