Sign In

Dave P. Miller

Principal at DPM Biostatistics LLC

Dave P. Miller is a distinguished professional with over 30 years of experience in designing and analyzing clinical trials, observational studies, and health outcomes data.

With expertise in randomized trials, sample size estimation, prognostic models, disease registries, observational studies, HEOR, and various statistical techniques, he has co-authored over 100 papers in medical journals.

He holds a Master's degree in Statistics from Carnegie Mellon University and has pursued further studies in Biostatistics at the University of Michigan, along with a Bachelor's degree in Statistics from Carnegie Mellon University.

Throughout his career, Dave has held key roles in prominent organizations such as DPM Biostatistics LLC, Unlearn.AI, Verily Life Sciences, Genomic Health, ICON plc, R2 Technology, Quintiles, Ischemia Research and Education Foundation, Cleveland Clinic, and the University of Michigan.

His professional journey includes positions like Principal, Chief Science Officer, Head of Biostatistics, Lead Program Biostatistician, Senior Director, Director, Vice President, and Research Associate, showcasing his diverse skill set and leadership capabilities.

Dave P. Miller has provided valuable insights through presentations to biostatisticians, clinicians, regulators, and peers in both academic and industry settings, making significant contributions to the field of biostatistics and healthcare research.

His areas of expertise cover a wide range of fields including PAH, asthma, oncology, cardiovascular diseases, and medical imaging, demonstrating his versatility and in-depth knowledge across various medical domains.

Highlights

Aug 1 · linkedin

Today is the one year anniversary of my consulting business, DPM Biostatistics, and I want to thank all of my former managers, employees, and co-workers who have made this first year so successful with your referrals. Those referrals are really my entire business, and it's kept me very busy and fulfilled. Thanks also to my clients who are doing such important and innovative work bringing sophisticated new medical devices and diagnostics into the healthcare system.

Feb 14 · linkedin

In a previous post, I shared that the 2 specialty areas for DPM Biostatistics LLC are:

  1. Diagnostic Device Studies
  2. Observational studies, RWD and RWE

Of the two, the diagnostic device studies reflect the more highly specialized niche. One of the many aspects of this work that I find interesting is the wide variation from one company to another in the style and level of detail of protocols and analysis plans. It’s rare for device/diagnostic companies to have large clinical operations or biostatistics departments, so the role of each individual in developing documents from scratch is much greater.

For the documents I write, the most important guiding principle regarding style, formality, and detail is “Who is the audience”? If the document is intended for a PreSub to FDA, specifically CDRH in the case of device studies, it’s critical to be explicit about any aspect of study design or the planned hypothesis testing that could affect the acceptability of the study for getting the device cleared. There’s a temptation with PreSub documents to leave certain things vague to “keep your options open”. I see this as a lost opportunity to get good advice from the same people who will later determine the regulatory fate of your device.

Of course, the FDA is not the primary audience for every document, and, hard as it may be for some, FDA should not be top of mind for every word you write about a clinical study. Sometimes the key audience is medical advisors while at other times the key audience may be a CRO or a statistical programmer. Is the goal of the document to get high-level feedback on whether the objectives and measures are clinically meaningful or is the goal of the document to provide instructions to somebody who has to implement what you’ve written?

Lastly, it’s important to acknowledge that most documents have multiple audiences. It can be a tough balance to give a reviewable level of detail for highly technical considerations without making a document unreadable by non-statisticians. I won’t claim I always get that right, but it’s fun and rewarding when I do, which is why I keep at it.

Dave P. Miller
Add to my network

Location

Albany, California, United States