You would have to be living under a rock to avoid the data deluge in healthcare. You are either a consumer of data, a provider of data, analyst of data or some combination of the three. We are all trying to get a handle on our new data identity. There are some industries that operate at low-level and just churn along the status quo of misinformation. Throw a few p-values around, claim "significance" and move onto the next project. I am going to proceed with the assumption that 'dear reader' this is not you...
I am not going to pretend that everyone is excited about the prospect of needing to digest the fire-hose of information and the statistical rigor required to create meaningful metrics but what I can offer is a user-friendly guide through the haze of clinical research data (is it meaningful?), EHR data (is it useful or actionable?), medical education outcomes data (am I measuring what I think I am measuring), or just a rosetta stone for asking the right questions.
Here is my evolving definition of an insight analyst. I adopted the term when I observed that quite often I am not involved in the most important phases of biostatistical analysis. Rigorous data collection begins with the design of the study, data collection methodology, analysis, presentation, followed by the interpretation. What can I say about the trends in my data? Because many of the parts of the puzzle are missing I never feel like a true data analyst and because I pull in the scientific context I feel more integrated than a statistician. What I know for sure now that I see a lot of data--there are insights in all steps of data visualization and analyses. Perhaps the level of certitude in any particular outcome is a little less robust but 'all in all' there is a substantive truth to be revealed. In full disclosure I did study statistics throughout my postgrad education but like many of you it was a parallel universe. Too 'mathy' to be functional or accessible outside of a classroom or a specific project.
Years later, as a medical writer, I found the data to be the only trusted source. Well-meaning brand managers, medical education professionals, or other stakeholders are easily biased even if they are unaware of their heuristics. Asking the right questions to be answered by your data team (or you) will provide the infrastructure to help you glean your own insights. And if along the way you need a professional, you can reach out to me. Otherwise look to the posts for useable information applicable to your daily requirements. Feel free to send specific questions or sign up on this site for access to higher-level tools. So far, launching this focused site should help with planning studies (primary question of interest, single or multiple group comparisons, sample size, selection of participants (random or based on pool of interest--take whoever shows up?, and proceed from there.
Now that you have collected the data, how do you summarize, deal with variability, distinguish real patterns from random variation, what can you infer to larger population, what methods to use? What is the best way to visualize the data? What are the main messages? How to communicate uncertainty in estimates?
While it isn't required, you can pick up a copy of a little resource that I pulled together. I might use examples from some of the freely available healthcare data on the internet and you can find them here for less than the price of a latte. If I find a particular useful link for you to dive a little deeper in your analytics I will also provide that for you.
Let's do this...
What I am hoping to do is contextualize the myriad of questions and mistakes in data collection within our specific spheres of focus. Best case scenario? You feel compelled to develop robust data collection methodology within your own organization of 1 or many. Anywhere along the continuum it is a win win. We all need to be able to discern the quality of information in the burgeoning digital society. Whether for personal or professional use--the harm of misinformation is real.
"Statistics are like swimwear - what they reveal is suggestive but what they conceal is vital."
-Ashish Mahajan, Lancet 2007
bonny p mcclain, freelance health economics writer and insight analyst
Sign up for our newsletter!
Browse the archive...
Thank you for making a donution!
In a world of "evidence-based" medicine I am a bigger fan of practice-based evidence.
Remember the quote by Upton Sinclair...
“It is difficult to get a man to understand something, when his salary depends upon his not understanding it!”