Your role is to teach and to guide students toward clinical practice. Why then are you spending so much time and effort creating/synthesizing patients to use in your curriculum? Please don’t get me wrong. Content is extremely valuable to facilitate learning when you are using a medical record (EMR/EHR) system in the classroom.
The effort to add new patients to your EMR/EHR & CPOE environment for education should not be where your focus is when you are updating your program’s curriculum for students. EHR/EMRs for education should make this part of the process simple for educators and administrators.
All curriculum evolves over time, so without a large patient database to draw from, educators are constantly in this patient identification, creation and assignment mode rather than planning curricular enhancements. It is simply not a good use of your time.
We learned this valuable lesson while working with Indiana University School of Medicine (IUSM) during the development of their virtual health system. The EMR/EHR & CPOE system is a central part of their Accelerating Change in Medical Education grant from the AMA that they received in 2013. The Regenstrief Institute’s Center for Biomedical Informatics has been one of their leading technology partners and is responsible for the electronic medical record system (we now call the Teaching EMR) they use today with almost 350 second-year medical students. They are expanding into year-one and year-three classes in 2016. With multiple classes for IUSM planned they will need the right tools for patient identification, selection, and assignment.
How many patients does a program need for education?
I’m not sure I know the exact answer to this question, but to enhance learning and provide educators more options for teaching we have built a database of over 10,000 patients to choose from in the Teaching EMR. As an almost 50-year-old organization rooted in research and HIT, we certainly understand the value of good data and systems. We have a top-notch team of engineers and faculty that have designed a rigorous de-identification and misidentification process to protect patient information and privacy while leaving the records readable.
What is a readable record?
Most of us have probably seen a redacted record or two in our lifetime. When you consider the volume of information that needs to be protected to maintain patient privacy and security of the PHI (Protected Health Information), then you need a well planned and executed de-identification and misidentification process to leave a record readable.
De-identification (think of it as “scrubbed” data), to define for those not as familiar with the concepts, is the process of extracting the private information in a record and replacing it with a character like “X’s.” We have all seen a credit card or social security number online where the numbers are replaced by “XXX-XX-XXXX.” That is done programmatically to a medical record to protect the patients information that is often not germane to the learning point and meets HIPAA Privacy Rules around Safe Harbor and Expert Determination.
Bonus HIPAA Definition: Data becomes de-identified when you remove 18 identifiers or statistical certification that the reidentification is low.
Misidentification (this “scrambled” data you can still read) is the information replacement process of taking the private information from a medical record as describe above and replacing the values with information that maintains the characteristics of the original information while masking the true identity of the patient. Masked demographic and geographic information can be maintained to give the educator and students a record that is readable and consistent with the original record without the concern for risking patient privacy through reidentification. We also misidentify information by a balanced shift of time for the patient history through the rigorous transformation process. As well, we can limit the amount of information from the record that is shown to facilitate different stages of the patient interaction within the health system to facilitate learning objectives.
De-identification and misidentification correctly completed give you a readable record that does not create distraction or confusion from the learning objectives due to missing information. When you consider the volume of information in a medical record, this becomes a rather daunting task without a rigorous way to programmatically accomplish the task. That is why even synthesizing a few patients to make them seem like real patient records takes effort. Having an extensive database to search makes this process simpler.
Where should you be spending your effort?
I hope you can see why it is valuable to use a tool for learning that allows for the speed and flexibility of identifying, selecting, and assigning patients to lessons while using an EMR in the classroom. To learn more or see for yourself, please don’t hesitate to contact a representative at Regenstrief on the Teaching EMR team or myself to discuss your program’s curricular needs with for a Teaching EMR.
Sr. Product Manager – Teaching EMR
Regenstrief Institute – Center for Biomedical Informatics