Biostats for clinical research

[zoombini01]

Hey everyone,

I'm in the middle of a summer research project and am realizing how little I actually know about the clinical research process and ESPECIALLY the stats involved. My school doesn't have a formal biostats course and the little we have learned is very very basic. I'm looking for sort of a "Clinical Research 101" book/resource that covers some fundamentals, targeted to future MDs who want to eventually do some clinical research. I realize that a lot of this stuff is probably learned "on the job" (I'm learning some stuff just by being part of this research project this summer), but I'm the type of person who likes learning by reading and would love a good supplemental reference text on the ins-and-outs of how to conduct clinical research well and some of the important statistics. Anyone come across anything good?

Thanks!

---

Members don't see this ad.

 

[aegistitan]

I'm interested in this as well

 

[trino]

Me 3

 

[spicysoup55]

Stanford is offering an online course called "Statistics in Medicine" starting on 6/24. Judging from the brief description and intro video, it seems like an introductory statistics course with medical applications. Here's the link:

http://online.stanford.edu/Statistics_Medicine_CME_Summer_15

 

[Freezer]

So there are a ton of stats books out there, many have way too much detail than what you may need. For someone new to stats that wants a little more beef than when a med school review text will give you, I like Gould's Biostats Basics: A Student Handbook. It focuses primarily on your bread and butter univariate analysis (T tests, Anova, Chi Square, ect) with some discussion on non parametric analysis. What I love most about this book is that it has this handly flow diagram to help you choose exactly what test you need (I would buy this book just for this flow chart). It's functional. It is by no means beefy enough to go into multivariate analysis or Bayesian stats, but if you need that computing power, you should be working with a statistician. Which I must point out, gone are the days of doing your own stats. The first step of designing any project should be a discussion with a statistician to determine how your data should be collected to best answer your question. In med school, your goal is to be able to look at stats in a publication and formulate an educated opinion as to the quality of the work. If you are looking at clinical research, get a statistician involved early in your project and keep them in your project discussion through the entire project. Most hospitals have an in-house statistician to meet this exact need.

 

[typhoonegator]

Couple of points:

1. In most situations, statisticians don't work for free and those that are on staff (meaning they're basically free to you) tend to have a large backlog.
2. You absolutely can do your own stats. Too many people become slaves to their data because they don't feel comfortable interpreting it themselves. So take some courses, read some books, and wade in there. It's not magic -- it's just math, logic, and assumptions. I find very often that building the models, writing the R scripts, and interpreting the results provides much more insight into the questions you are asking than looking at the STATA output someone else generated for you. As a scientist you want to live as close to your data as possible, so get help where you need it, but roll up your sleeves and get dirty. It's not like you can't publish in JAMA or NEJM or whatever unless a statistician has performed your analyses for you.

 

[ewax]

If you can find them, the 2010 version of Kaplan has 1-2 videos by Dr. Steven Daugherty on interpreting statistical data. Great resource.

 

[supremus]

Stanford is offering an online course called "Statistics in Medicine" starting on 6/24. Judging from the brief description and intro video, it seems like an introductory statistics course with medical applications. Here's the link:

http://online.stanford.edu/Statistics_Medicine_CME_Summer_15

Signed up! This looks great.

 

[wholeheartedly]

Epi grad student here with lots and lots of biostats credits.

The Stanford course is probably good and the Johns Hopkins school of public health has some great free online coursework to help you shore things up. I've got a few biostats textbooks, but to be honest, I've actually found biostatistics for dummies to be a very thorough yet concise overview of almost everything I've covered in grad school so far. I bought it as a review book for my masters project and the public health certification exam and have been pretty impressed with the content. It'd be a good intro/overview before diving into some of the more detailed stuff.

 

[Carmiche]

Here are some papers my mentor gave me at the beginning of my summer research program on statistics and how they are applied to medical research.

I would suggest using your school's library to access them; don't know if there are free versions available or not.

Greenfield ML, Kuhn JE, Wojtys EM. A statistics primer. Am J Sports Med. May-Jun 1996;24(3):393-395.

2. Wojtys EM, Greenfield ML, Kuhn JE. A statistics primer. Statistical terminology--Part 2. Am J Sports Med. Jul-Aug 1996;24(4):564-565.

3. Greenfield ML, Kuhn JE, Wojtys EM. A statistics primer. P values: probability and clinical significance. Am J Sports Med. Nov-Dec 1996;24(6):863-865.

4. Greenfield ML, Kuhn JE, Wojtys EM. A statistics primer. Power analysis and sample size determination. Am J Sports Med. Jan-Feb 1997;25(1):138-140.

5. Kuhn JE, Greenfield ML, Wojtys EM. A statistics primer. Types of studies in the medical literature. Am J Sports Med. Mar-Apr 1997;25(2):272-274.

6. Kuhn JE, Greenfield ML, Wojtys EM. A statistics primer. Prevalence, incidence, relative risks, and odds ratios: some epidemiologic concepts in the sports medicine literature. Am J Sports Med. May-Jun 1997;25(3):414-416.

7. Kuhn JE, Greenfield ML, Wojtys EM. A statistics primer. Statistical tests for discrete data. Am J Sports Med. Jul-Aug 1997;25(4):585-586.

8. Greenfield ML, Kuhn JE, Wojtys EM. Current concepts. A statistic primer. Descriptive measures for continuous data. Am J Sports Med. Sep-Oct 1997;25(5):720-723.

9. Greenfield ML, Wojtys EM, Kuhn JE. A statistics primer. Tests for continuous data. Am J Sports Med. Nov-Dec 1997;25(6):882-884.

10. Greenfield ML, Kuhn JE, Wojtys EM. A statistics primer. Confidence intervals. Am J Sports Med. Jan-Feb 1998;26(1):145-149.

11. Greenfield ML, Kuhn JE, Wojtys EM. A statistics primer. Correlation and regression analysis. Am J Sports Med. Mar-Apr 1998;26(2):338-343.

12. Greenfield ML, Kuhn JE, Wojtys EM. A statistics primer. Validity and reliability. Am J Sports Med. May-Jun 1998;26(3):483-485.