DMC reports in the 21st century: towards better tools for decision-making

Abstract

Data Monitoring Committees (DMCs) have the important task to protect the safety of current and future patients during the conduct of a clinical study. Unfortunately, their work is often made difficult by voluminous DMC reports that are poorly structured and difficult to digest. In this article, we suggest improved solutions. Starting from a principled approach and building upon previous proposals, we offer concrete and easily understood displays, including related computer code. While leveraging modern tools, the most important is that these displays support the DMC's workflow in answering the relevant questions of interest. We hope that the adoption of these proposals can ease the task of DMCs, and importantly, lead to better decision-making for the benefit of patients.

Keywords: DMC report; Good graphical principles; Data Monitoring Committee; Data Safety Monitoring Board; Statistical graphics.

Fig. 1

Demographics from the Xanomeline study (alternative). Top: Box-whisker plot with a red dot representing the mean of the data, the vertical line is located at the median of the data. The box covers the second and third quartile, while the whiskers extend to the last data point within 1.5 times the interquartile range. Bottom: bar plot of the percentage of female subjects in the respective treatment groups

Fig. 2

Adverse event overview for the Xanomeline study

Fig. 3

Adverse event “volcano plot” for the Xanomeline study (Xanomeline High dose vs. Placebo)

Fig. 4

Laboratory parameter distribution over time from the Xanomeline study (here: cholesterol). Box-whisker plot with horizontal line representing the median of the data. The box covers the second and third quartile, while the whiskers extend to the last data point within 1.5 times the interquartile range. n_P, number of patients on Placebo; n_X, number of patients on Xanomeline High Dose

Fig. 5

Abnormal laboratory parameter values over time from the Xanomeline study (here: cholesterol). n_P, number of patients on Placebo; n_X, number of patients on Xanomeline High Dose

Fig. 6

Spaghetti plot of creatinine kinase over time. Trajectories that exceed 3 times the upper limit of normal are colored, with annotated subject numbers. Horizontal reference lines can be toggled on for limits of normal if they are the same for every subject (which is not the case for this parameter)

Fig. 7

A graphical alternative to a laboratory shift table. Black lines are limits of normal

Fig. 8

Patient profile. Description of the most relevant clinical information at a glance