My previous two posts lamented the incredibly bad the documentation I get from a local hospital is. In truth, the documentation I get from everyone is terrible. Seldom does it tell me what I actually want to know, and if there is useful information it is buried in an avalanche of yada yada. The main reason for this is that documentation is driven by our ridiculous payment system, which requires us to follow arcane rules to generate notes that justify the obscure codes we submit for money from the payors. This is the reason for much of the gibberish. These rules, combined with computers\’ ability to quickly and efficiently generate drivel (see also The Entire Internet) are the hot house and Miracle Gro for meaningless words. Sprinkle the rules on a computerized medical record and stand back! Useless words and codes will spew out at you like milk from the mouth of an overfed baby.
Sorry. Bad image. But it is actually quite accurate.
When I started my new practice, one which is outside of those rules, I was excited about the possibility of actually documenting based on patient care. I asked myself, \”what would a patient record look like if the only reason for it was for patient care?\” It was a compelling question; one which I undertook in earnest to answer over the past 18+ months. Much of the reason for the downturn in my blogging volume is that I\’ve spent an incredible number of hours working on a charting system that did exactly that. It has sucked out a lot of my creative juices, leaving me only the mental energy at the end of the day to play solitaire on my iPad.
Undertaking this endeavor had taught me an important fact: documentation in itself, without the stupid rules, is still really painful. As many bells and whistles I put on my computer system (I recently put a \”coffee break\” button on we can press when we want to take a break), it still is difficult to generate good documentation.
Why? Imagine having to go through each day and keep track of every conversation you have with people. Imagine trying to not only have meaningful interactions with people and to make good decisions about important things, but to make sure all of the facts were accurately documented in a form that will be useful in the future. Imagine too that you had to do this with all of your emails, phone calls, and text messages you have with people. This is a real pain in the gluteus maximus!
Add to that the joy of patient confidentiality (and our dear friend HIPAA) along with the gobbledygook we get from other practices and hospital documents, and things get really tough. Not only do I have to find a place for each of the pieces of \”documentation\” I get from hospitals, consultants, and old records, but I also need to go through them and pull out the bits of useful information and put them in as \”structured data.\”
What, you may ask, is structured data?
Structured data is information that is sorted so the important stuff can be gotten to easily. Some structured data are numbers (like a blood pressure or blood glucose), some are words (like mammogram reports, heart exam findings, and pathology reports), some are dates (date of last colonoscopy, flu shot, or office visit for diabetes), and some are images/documents (like a picture of that rash you had in July, or the video of accident you had with the golf cart that got posted to YouTube). Not all data should be structured (it gets way more confusing that way), just the stuff you might need later on.
You see? Gluteus Maximus acquires trauma and inflammation. It is really hard to document things well, and human nature means that by the end of the day you\’ve spent your time doing things and not documenting it all. It\’s really hard to be diligent about this stuff and not require a double espresso Adderal latte with a valium mojito chaser.
Still the process I\’ve been trying to build has several strategies to make this better:
- Focus on office workflows to make sure important things get done. When I order a lab test, I need to make sure that it gets done, we get the report, the report is sent to the patient along with a plan, and then a follow-up interval is determined. This needs to be baked into any record system a doctor uses. There are lots and lots of these kinds of workflows that need to be automated.
- Make the record a collaborative record. There is one person with far more at stake than the folks in the medical office: the patient about whom the record is about. I have yet to implement, but my plan is to give people the ability to see and edit their own records in a way that maximizes accuracy. When a person goes to a consultant, why not have them enter it into their record? When they get results back from another doctor or hospital, why not let them at least notify me in the record that this happened? Accurate records are far more important to patients than anyone else, so why to we keep them out of them (and why the hell do we charge people for their own records??)?
- Embed communication tools into the record itself. Why have an email or voice message transcribed into a record system? Why not just have the email message go right to their record? CRM (customer relations management) programs already do that, keeping track of all communications with each customer. Why doesn\’t an EMR do that?
- Use other common tools, like timelines, task management, and tag clouds, to make data easier to understand and compile. There are lots of things done in other information systems that are painfully absent from medical records. Why, for example, don\’t we have a medication timeline for each patient that has start and stop times overlapping with things like symptoms, lab tests, etc.
I am not saying that documentation will become any more fun than cleaning a cat\’s litter box, but if we gain enough from the tool, we will use it better. Right now medical records are caught up in the vortex of codes, ACO\’s and meaningful use. Getting out of that vortex is only step 1 in the process. I\’ve been busting my butt (more gluteal inflammation, sadly) to get something that does even a small part of this. I\’m getting closer, but of course I have to build that new record while also seeing patients each day and properly documenting each encounter.
I\’ll go into more detail as things develop, but I think I need to stand up for a bit.
4 thoughts on “The Pain and Inflammation of Documentation”
Like your post. It sounds like you had similar ideas about an EHR as I was in a solo private primary care practice when I built mine. Perhaps we can talk sometime and share ideas as it sounds like what I’ve already done with my open source EHR attempts to address the issues you have brought up (ie tagging, automatic insertion of email, automated and easy-to-understand workflows for outpatient practices). I’m currently working on using my EHR as a pilot project for a patient-centric EHR to answer the collaboration issue and upend the way things are currently going (dismally) regarding interoperability. I think if we work collaboratively, we can truly make some disruptive change in EHRs.
Michael Chen, MD
firstname.lastname@example.org, NOSH ChartingSystem
Agree. We do need to do some show and tell. Will send email.
Amen, preach it! You are in an enviable position given your practice situation of forging new ground that vendors can’t since they have to support the documentation-as-billing-support model for the foreseeable future. Oh, and "Why, for example, don’t we have a medication timeline for each patient that has start and stop times overlapping with things like symptoms, lab tests, etc" – actually this exists: https://itunes.apple.com/us/app/allscripts-wand/id458026987?mt=8
That’s pretty cool, actually. Too bad the Allscripts program is required. I actually used their stand-alone e-Rx app and it was near impossible to use. We really need to put this in the hands of the patients, not just prescribers.