Welcome to HIStalk’s new Monday feature, Curbside Consult. For those of you not familiar with the term, a curbside consult happens when another physician wants to ask your opinion about something, but doesn’t want to write a formal consultation request in the chart or actually send the patient to see you.
Curbsides have a certain appeal for both sides. When busy subspecialists like infectious disease (ID) docs are in short supply, they don’t want their time wasted by what they consider simple questions, like validating an antibiotic selection for a hospitalized patient. If the request is made formally, they have to see the patient and write a note, when often it’s an answer they can just give off the top of their heads (and are happy to do so).
My favorite ID consultant requires that you call his cell phone to curbside him prior to writing a formal consultation request. Eighty percent of his business is in curbsides, because he likes to save the “real” consultations for patients who are truly complex. I think it’s also his way of having some semblance of a home life. If he formally consulted on everyone he was called about, he’d never see the outside of the hospital. (It probably would be a different story if there were ID specialists around every corner, but there aren’t, at least in my market.)
For those of us that ask for the occasional curbside, often the question is one that we know the answer to, but want a colleague to validate. Maybe it’s not something we see every day, or we want to confirm that we’re operating with the latest and greatest information.
The worst kind of curbside is when a non-physician tries to ask for one — also known as the Supermarket Consult. Definitely to be avoided, as no physician ever wants to try to diagnose a rash in the frozen foods aisle.
The worst one of these I ever experienced was when I worked for a small-town hospital that had a float in the town’s Founders’ Day parade. Waving and throwing candy from my Astroturf-covered perch (wearing a white coat and scrubs, no less) a patient stepped out of the crowd in the middle of the parade route and asked me if I had an update on new medications for constipation. (With patients like that, who needs HIPAA, anyway?)
On HIStalk, Curbside Consult is about sharing physician opinions on the hot topics, not only from me (Dr. Jayne), but from the HIStalk Medicine Cabinet. Response from potential Cabinet members has been good. I haven’t gotten back to everyone yet since I worked entirely too much this weekend, so stay tuned.
When Mr. H offered to run my posts on their own night, it was exciting, yet a little bit nerve-wracking. I appreciate the opportunity and think it will be a chance to have more focus on CMIO/CIO/CMO/physician topics independent of the news, rumors, and updates. And so, here we go!
A Study of an Enterprise Health Information System
Mr. HIStalk opened the door on this one in his Monday Morning Update, so I’m going to walk right through.
For those of you who haven’t seen it: Jon Patrick, Professor at the University of Sydney’s Health Information Technologies Research Laboratory, wrote about implementing Cerner FirstNet in emergency departments in New South Wales.
(As an aside, you can take the CMIO out of clinic, but you can’t take the clinic out of Dr. Jayne. The author also had an article about extracting data from narrative pathology reports on melanoma patients, which I downloaded to read on my way to the beach later this month. Using optical character recognition and concept tags to gather data and improve care? Fabulous!)
I agree with Mr. H though. Forget Cerner, as this applies to vendors and applications that are too numerous to count. Being the person who often has to have the less-than-pleasant conversations with vendor CMOs and physicians when there are issues at my hospital, I’d like to share my observations.
Although there are some significant defects out there, many vendors are working hard to remedy these and to implement aggressive protocols to bake quality into their products and design defects out. The best of these organizations are adopting techniques long used by manufacturing to ensure quality. This is still a relatively emerging industry, though — some folks are building Model T Fords, while others are building models that will go for 200,000 miles.
Vendors are under the gun from clients and prospects to make their products pretty, which translates too frequently into something that might look nice but isn’t clinician-friendly, usable, or safe. Vendors are also pressured to make their systems “fully customizable,” which often means allowing users to do things which are, for lack of a better word, stupid. I’d love to see a rebuttal of the article addressing what parts of the system were implemented against Cerner advice or best practices.
Some additional thoughts:
- Too many systems allow clients to “customize” patient safety out of the implementation. When I learned that there was a preference in my system to disable allergy and drug interaction checking, I cringed. (This was years ago when I was but a mere Physician Champion, before some of my colleagues decided I was an “IT sellout” and had crossed from “us” to “them.”) Within a few months, some of my colleagues demanded that we allow users to turn it off. Their reasons (no kidding) included “I know what medications interact with each other” and “I know what the patient is taking, I don’t need the computer” and “I hate it, turn it the [bleep] off.” There’s no class in medical school that teaches you how to deal with peers who act like this. The only thing you can do is make sure your patients and loved ones don’t go see them.
- User errors can be reduced, but they can’t be eliminated. At our hospital, we aggressively track errors in documentation and look at trends between physician offices and across the system. Good staff training, low turnover, and accountability all lead to fewer errors. Surprise!
- Patients give bad information, which no amount of decision support or application design can fix. Until we have fully integrated interoperable platforms that remove human memory or manual data entry from the equation, this is going to be an issue. On the other hand, when everything is fully automated, errors can become magnified and nearly impossible to fix. Talk to anyone who’s been the victim of medical identity theft and you’ll know immediately how horrible this can be.
- The ability to review longitudinal patient data is sorely lacking in many systems. In this world of accountable care and shared risk, the concept of episodic care needs to die a quick death.
- Data integrity has to be paramount. It doesn’t matter how slick your user interface might be or how few clicks it takes to document — if the data gets vaporized, it’s game over. My first EHR had this handy little feature: the templates accommodated unlimited input, but the documents only accommodated X number of characters. Not knowing any better, I trusted my vendor (ha!) and didn’t discover this undocumented functionality until patients came in for follow up visits and many Assessment/Plan sections (traditionally done at the end of the note) were missing. My risk management consultant loved that one, let me tell you. On the other hand, that EHR was extremely easy to use, was highly standards-based, and had a workhorse scanning solution that I still miss. But it ate my notes, so it had to go.
- Several vendors still use design philosophies that revolve around the individual visions of key leaders regardless of user needs. Granted, they’re baking in the Meaningful Use requirements because they have to, but the world has changed and what was cool 20 years ago is no longer what healthcare needs. There are peer-reviewed studies on usability and user interface design, and some of these folks must never have seen the articles. HIMSS did a piece on EMR Usability a couple of years ago and I know the systems that have design elements that hopscotch all over the place or color schemes that induce vertigo are designed by people who haven’t seen it.
- Vendors should spend the money to hire medical proofreaders to look at their products. Nothing screams “we rushed this to market” like misspelling medical words. I saw this more than once at HIMSS in the nether reaches of the exhibit hall. Once you lose providers’ trust with something like that, it’s extremely difficult to get it back.
- Allowing existing users to beta-test your product, if they’re willing to donate the time (or if you’re willing to pay for it) can make a huge difference in the quality of releases. This allows for detection of those issues that are particular to the way individual clients use a system. Otherwise, there’s no place like production.
- Vendors need to disclose their defects. More and more are doing this in a meaningful way. On one hand, I understand the call to have the FDA regulate systems. On the other hand, as a physician, I know that the FDA is, in many ways, a joke. Deserving products can’t get to market because of the complexity of the process, manufacturers with deep pockets continue to slipstream their products, and the vitamin/supplement industry got themselves classified in a way so they can say what they want, produce what they want, and generally get away with it.
- Hospitals and practices need to thoroughly test their systems prior to go-live, regardless of what the vendor says or how many clients they have live. Your staff will find shortcuts and features that no one has seen and will rapidly become dependent on these, much to the consternation of your call center, trainers, analysts, and the vendor. The only way to trust the system is to test the living daylights out of it before taking it live.
Unless physicians and their IT teams understand these points, they aren’t ready to leave paper. Those who do understand need to play an active role pushing vendors to improve and vendors need to address core functionality.
If progress isn’t made, the likelihood of a federal mandate increases. We’ve already seen what Meaningful Use has done to the industry. As Scotty once said, “She can’t take much more of this, Captain.”