Wednesday, June 11, 2008

The Contents of the Medical Record

On June 17, 2008, I'm facilitating the annual BIDMC IS retreat about the topic of "the transition to the fully electronic inpatient record". To help us prepare for this retreat, I asked several colleagues if they had a scorecard describing the components of the acute care medical record. The hospitals I asked did not have one handy.

Here's the results of my research:

Per the American Health Information Management Association (AHIMA), the legal record is comprised of the component parts detailed in this document.

For BIDMC's 2007 Joint Commission visit, we assembled a roadmap of all our medical records and indicated if they were paper, electronic or a hybrid of the two. That roadmap shows all our medical record systems.

Of note, we scan all our inpatient paper records, so they are available online to our medical record coders and our clinicians. Although scanned paper is not really an electronic record, it does mean that we no longer need to request the paper record for many patients. To organize the record prior to scanning, we divide all aspects of record in the way we believe a clinician will want to intuitively access the record.

To help inform our retreat, I asked our Health Information Management staff to create a Pareto diagram that will guide our priorities toward implementing a fully electronic acute care record. The results are

Graphics/Flowsheets 30%
History/Progress notes 29%
Discharge summaries 15%
Operative Notes 6%
Pre-procedure documentation 5%
Outside records 4%
Orders 4%
Lab Reports 3%
Radiology reports 2%
Diagnostics 3%

This illustrates that the low hanging fruit to creating an electronic inpatient record are graphs/flowsheets, history and physicals, progress notes, discharge summaries, operative notes, and peri-anesthesia testing. Together, these items account for 85% of the paper record. Here's our strawman plan, based on work in progress

Graphics/flowsheets - capture as PDF, HTML forms and scanning
History/Progress notes - capture as structured and unstructured notes using templates, macros, and voice recognition
Discharge summaries - capture as structured documents by assembling all the components from other care processes - medication reconciliation, lab results, team census, progress notes. Note that we already have electronic discharge summaries and we plan on enhancing our applications in 2009 to improve their quality and multidisciplinary content.
Operating Notes - capture via voice recognition. Note that we already have electronic operating notes.
Pre-procedure documentation - capture via scanning (consents) and templates

A few considerations about eliminating paper. We must consider:

1. The ability to render non-repudiable documents over the legal retention period. Except for those documents we send to fixed content storage (Centerra) and stamp, our electronic medical record content is theoretically alterable by a senior system programmer.

2. People must be at least as compliant with electronic signatures as they are with paper document signatures. At present, inpatient document signing (discharge summaries and opnotes) are under good control.

3. The infrastructure underlying a system of record should be disaster proof, or at least as safe as what we do today for paper. Our worst case with paper is a fire in the paper record storage facility. The risk is small as sprinkler protection is in place. We have published service levels for electronic recovery point and time objectives.

4. If we declare an electronic source as the "official" version for that component of the legal record, we should have some policy on when we would produce paper, e.g. only for release to outside entities.

I'll summarize the outcome of our retreat next week which will provide many more details of our plans for FY09 and beyond. Our goal is an 85% electronic inpatient record and transition from a hybrid medical record to an electronic legal record by 2011.


Frank said...

What is interesting to me is the absence of charges from the electronic medical / health record. For what appear to me to be historical reasons - once something is ordered, two data streams are formed - a cost data stream and a content data stream.

Typically the two data streams are reunited and reconciled (specifically, are requisite data supporting the order clearly linked) at the time insurance claims are filed. It would seem prudent to revisit the concept of dissociating cost data from physiologic / procedure / order results. It would seem that the costs associated with redundant management of different elements of the same data stream flowing from each individual patient are real. Are the parallel costs worth it?

John Halamka said...

Our revenue cycle is completely integrated into the electronic health record, but we typically call these "practice management" features rather than clinical features. I'll detail the revenue cycle side of our software in an upcoming blog.

Frank said...

Insights into the practice management element would be very useful - particularly as it relates to patient care. Within the teaching hospital setting, having the house staff overtly aware of the charges associated with each ordered test has been a challenge. If your practice management facilitates clinician awareness of both charge and lab / test results then this is a win win for everyone.