top of page
  • Peter

Book Review: Data Science and Predictive Analytics

Updated: May 29, 2020

I remember several years ago before I put together the content to teach my EMR methods course I casually perused books described as being about ‘healthcare analytics’…the pickings were pretty slim. Even though the data science bubble was at peak amplitude, it was still early in the healthcare hype phase.


Things have changed in the intervening years. As preparation for a lecture in a business school I’m giving this semester, I wanted to go out and re-assess the current market of ‘health analytics’ books that a non-technical audience might find looking on Amazon. As opposed to the 2-3 I remember, I now find about 3 dozen books that include health analytics in the title description in some capacity. I thought it would be interesting to take a look at each of these books and try to figure out if any of these books are helpful, or just trying to cash in on the healthcare data and data science gold rush. I can definitely say, just from a cursory glance, many of these books appear to focus on the ‘analytics’ (most of the analytics content being essentially the same book to book and equivalent to a short bootcamp on either R or Python) and pretty light on the healthcare.


First off, semi-randomly chosen, I’m looking at ‘Data Science and Predictive Analytics: Biomedical and Health Applications using R’. The version I’m looking at is the 1st edition, published in 2018 by Ivo Dinov who is a quantitative researcher in the University of Michigan’s nursing school. This book is published by Springer, who I trust for academic and technical texts, at least in the quantitative public health space (epi, health econ, etc). However, right off the bat the title seems to be the unfortunate work of a marketing dept. trying to cram as many buzz words as possible onto the book cover. I’m surprised they didn’t shoe-horn ‘python’ or ‘conda’ in there, too.


In terms of the content – remember my reference about light on the healthcare? This book is ~22 chapters on bread and butter data science using R, it looks like with some discussion of a healthcare case-study thrown in. This may be a fine primer on data science using R (at >800 pages, it does appear to give the topics some real discussion), but I take odds with it’s self-described focus on healthcare or biomedical applications. I’d contend that if you don’t already know a good deal about those topics, the limited case-studies aren’t going to be terribly informative. From what I can tell, the case studies also appear to be from the author’s (or peers) biomedical research, which is fine and justifiable, but I’d argue also pretty different from the health applications suggested in the title.


Assessment: might be a good book for data science using R with some medical vignettes, but it doesn't seem like you'll get too competent at healthcare data science just from this book. if you're already a competent health care researcher this book might get you up-to-speed with modern data science in R, though.








118 views0 comments

Recent Posts

See All
bottom of page