Katie Reid, MPH, is a senior research associate at America’s Essential Hospitals, and works on multiple projects currently underway at the Essential Hospitals Institute. In the following interview, she explains data’s role in research and offers tips for how to keep analyses on track.
Why is data integral to the research process?
When someone is conducting research, there is always a question that needs to be answered. Why are our member hospitals essential to their communities? Is the hand hygiene initiative a member hospital is implementing effective at reducing nosocomial, or hospital-acquired, infections? Is the medical home model of care an effective way of reducing admissions and lowering the cost of care? To answer these questions, data is absolutely critical, as it allows you to make comparisons and extract the answers to the questions you are looking for.
What are your strategies for completing analysis in a timely and efficient way?
Timeliness and efficiency in data analysis are often tied to organization – you must have a plan in place that will keep you organized and on track while you look at the data. It’s very important to double and triple check your work when analyzing data, as it can be very easy to make mistakes and numbers tend to look alike after a while. So it is also important to have others double check your work and make sure your results match. Using these data checks are essential to keeping an analysis on track – it affords you time to make mid-course corrections if you identify any mistakes, rather than completing the analysis and then uncovering mistakes that need to be addressed.
How often are your findings different than your original hypotheses?
I’d say about half the time, which is why it is very important to always have an open mind when analyzing data. If you go in “knowing” what will happen, you can suffer from data tunnel vision – only seeing the data that supports your hypothesis and not the data that rejects it. Personally, I find the analyses much more exciting when the outcome is different than my original hypotheses.
What are your recommendations for hospital staff who may be struggling with data-related issues?
The most common data-related issue our members ask about is related to lack of resources to allocate to data collection and analysis. Often, hospitals will want to implement a change, but have a hard time collecting data because of the time it may require, or they don’t have anyone with statistical expertise to “know what it all means.” The good news is that we can help. Anyone at our member hospitals can feel free contact me, and we’ll do what we can to help!