Clinical Decision Data
When you’re responsible for guiding someone out of a pain that derails life day in and day out… you NEED to have good data to stand on.
Does data relate to being a doctor?
My clinical life would be unconsciously tied to data. There are business KPIs to monitor the financial health of a clinic. there are outcome measures indicative of patient prognosis. All this time I’ve been drifting through analytics and data modeling among the countless hours reading medical research. Data is behind every intervention that is supported as a standard of care.
Often, patients would come to me with a goal and set of symptoms that had little to do with that goal. A good clinician, acts much like a data analyst, by asking questions about the observations provided in order to extract statistically significant data to then categorize and organize a comprehensive clinical profile. This clinic version of ETL is paramount to being a doctor.
As the saying goes, ‘bad data in = bad data out.
I’ve discovered that patient success parallels thorough data modeling in various ways:
Time series data presenting as personal health histories; database schema existing as patient records, multivariate analysis of symptoms correlated with a working diagnosis; or even regression analysis of laboratory biomarkers. Having good clinical decision making enabled my patient’s success. Having good data analysis enabled my patient’s success.
The similarities do not end there.
This data also needs to be presented and clearly communicated to relevant stakeholders who exist as patients, clinic directors, case managers, insurance claim representatives, personal injury attorneys, family members, or even other’s within the healthcare team. Each one requires a different method in conveying understanding of that information, often as ad-hoc reporting in passing between appointments or formally as medical reports submitted under scrutiny of litigation.
I’ve come to the conclusion that data analysis is crucial to being a doctor.
Clinical decision making is so very similar to business intelligence data analysis - but the context and tools are different.