RadioBDC Logo
| Listen Live
< Back to front page Text size +

Wheat from chaff: Making sense of the electronic health record

Posted by Ishani Ganguli  September 10, 2012 07:00 AM

E-mail this article

Invalid E-mail address
Invalid E-mail address

Sending your article

At 5am, Mr. A rolls onto the medicine floor: the fifth and final new patient to be admitted that night. The 70-year-old is well-known to our institution from his near-monthly hospitalizations and his primary care doctor, cardiologist, podiatrist, ophthalmologist, and both of his endocrinologists all work in-house. Unfortunately, for the intern admitting him (and for Mr. A), this translates into a few hours-worth of prior blood test results, MRI reports, visit notes, and discharge summaries to peruse. Where to begin? How to find the key details buried in this hoard of information? 

Electronic health records (EHRs) have brought to health care both a much-needed modernity and an emerging challenge: how do doctors manage the rapidly growing quantities of health records that we are responsible for reviewing and that (theoretically) help us take better care of our patients, so that we can extract critical information while spending more time with patients and less in front of a computer?

There is little question that electronic records trump the tree-killing alternative (eg. that Mark Twain autobiography-sized pile of faxed paper, one line of doctor-scrawl per page, documenting a patient’s stay at an outside facility). But even electronic records can become unwieldy in the form now used in most hospitals, including Mass General: clinical data organized by the date a note was written or a blood test was drawn. 

Sorting through such files as quickly and effectively as possible is a skill that we must pick up early in residency. But what if a computer program could make this easier and more foolproof by pulling out the information we needed when we needed it? The technical term for this is “clinical summarization.” It’s a still rare feature that is gaining traction:

Nearly a decade ago, MGH radiologist Michael Zalis wanted to look up clinical information about the patients whose scans he was reading, so he got the help of computer scientist Mitch Harris to create a search interface for the MGH EHR: QPID. The tool has since grown into a platform for search-based applications throughout the Partners Healthcare EHR: Anesthesiologists use QPID to gather a patient’s drug allergies and prior complications before surgeries. Palliative care doctors use it to find patients who would benefit from their service. Medicine residents use it to search for the stress test results or cholesterol levels of patients coming in with chest pain. “QPID combines a powerful search of the entire EHR data with the ability to prerecord insights that matter to clinician in a particular clinical context,” Zalis tells me. 

Two years ago, as an internal medicine intern at MGH, Gaurav Singal (full disclosure: he’s been a friend of mine since medical school) saw another use for QPID that might improve how he and other medicine doctors take care of patients. Early in that first year of residency, Gaurav started to notice medical missteps that were made because key information was buried in or missing from a hospitalized patient’s health record. He jotted down the examples into a Google document: The patient with severe pneumonia who was initially treated with an antibiotic to which her prior infections had bred resistance. The patient with a history of type I diabetes that was not documented in a transfer note, delaying her from getting the diet and insulin she needed badly. The patient who had four hospitalizations and multiple CT scans in the past month for unexplained abdominal pain and got yet another CT scan before a buried, diagnosis-clinching blood test result was finally noticed. 

What if a search tool like QPID could prevent such errors by auto-populating prior diagnoses, drug allergies, physical exam findings, and diagnostic tests into a user-friendly internal medicine dashboard or summary screen, organized the way an intern on overnight call thinks?

Gaurav worked with the QPID team and a group of residents to create such a dashboard, dubbed QPID MedPortal: For a patient with a new diagnosis like kidney failure, you could see what testing had and hadn’t been done already to avoid repeating costly tests. For a patient with congestive heart failure, you could look at the results of his most recent cardiac ultrasound, his baseline weight, and the medications that had worked for him before: information needed to make middle-of-the-night decisions. QPID MedPortal is in alpha-testing and will launch officially this fall. 

Because tools like QPID work at the so-called output end of the medical record, doctors can still (in theory) write down patients’ stories more or less the way they tell them and not through structured templates that, some argue, endanger clinical reasoning; like Google mail, QPID extracts and organizes the key information afterwards. Of course, the search tool is only as useful as the amount and quality of information included in the database, and we still have a long way to go before health records are ubiquitous and integrated. But clinical summarization tools like QPID and QPID MedPortal will do wonders to help that on-call intern, and all doctors, make sense of EHRs.

This blog is not written or edited by or the Boston Globe.
The author is solely responsible for the content.

E-mail this article

Invalid E-mail address
Invalid E-mail address

Sending your article

About the author

Ishani Ganguli, MD, is a journalist and a second-year resident physician in internal medicine/primary care at Massachusetts General Hospital. She studied biochemistry and Spanish at Harvard College and received her More »

Health search

Find news and information on:

More community voices

[an error occurred while processing this directive]

Child in Mind

Corner Kicks

Dirty Old Boston

Mortal Matters

On Deck

TEDx Beacon Street


Browse this blog

by category