Causal Inference in Clinical Practice


Big Data.  It’s all the rage.  But how does it factor into day-to-day clinical decision making?  Most practices have assistants who collect data prior to the doctor seeing the patient.  The doctor conducts her or his own tests and measures, and factors those results together with the data collected by the assistants in order to formulate an assessment and treatment plan.  Ironically the area of optometry that was well ahead of its time in terms of algorithms to handle “Big Data” was functional vision.  Those algorithms were  graphical analysis or checking, chaining and typing, and they smoothed out voluminous data that most practices don’t bother to collect anymore.  Why have these systems been discarded?  It’s easy to say that most practices don’t have the luxury of time to collect the data, or that technology has largely replaced many of the data points collect, but it’s deeper than that.  And I believe Dr. Samantha Kleinberg’s handy paperback may hold some of the answers.


Samantha Kleinberg, Ph.D.  is an Assistant Professor of Computer Science at Stevens Institute of Technology located in Frank Sinatra’s hometown, Hoboken, New Jersey.  In Why: A Guide to Finding and Using Causes, Dr. Kleinberg delves into the psychology behind how we use data.  One of the better known elements is confirmation bias, leading to the fact that we tend to find what we look for, and overlook that which we don’t look for, even when it is readily apparent.  To quote Kleinberg: “Confirmation bias leads to seeking evidence that supports a particular belief … In addition to bias in the types of evidence people seek and use, there can also be bias in now they interpret the evidence.”

Kleinberg emphasizes the potential disconnect between cause and effect.    The reality is that the complexities of clinical practice often dictate that even when we’re confident of a relationship between cause and effect, it is frequently non-linear and often multifactorial.  Where Kleinberg most shines in in Chapter 9, simply titled Action.  There can be a significant difference between efficacy and effectiveness.  Efficacy is defined as the ability to produce an intended or desired result.  It’s been known for a long time that optometric vision therapy has a scientific basis and is efficacious.  In reality, though, effectiveness is going to be determined by context, which are the conditions that are needed to enable the cause to produce an effect.

A delightful read here that helps explain, from a statistical perspective, how we literally and figuratively arrange the conditions for patients to succeed.



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s