Driving Away Docs | [node:field-byline] | Healthcare Blogs Skip to content Skip to navigation

Driving Away Docs

September 17, 2009
by aguerra
| Reprints

Before you get militant about CPOE, make sure you’re not the only one.

Let’s start with a few basic premises. Most hospitals in this country are “staffed” with physicians that don’t technically work for them. These independent doctors run their own practices; with their patients constituting an often formidable “book of business” that can be directed to the acute-care institution of choice. In all but the most rural areas, physicians have that choice, exercising it by where they obtain admitting privileges. This reality makes for an often lopsided power relationship in hospital/physician affairs.

Doctors, let’s be honest, usually run the show. Physicians who generate significant revenues are treated like royalty, especially at non-elite medical institutions (where reputation can tip the power relationship back in favor of the health system). But, alas, that is not where the majority of medicine in this country is practiced. For most acute-care organizations, it’s about how much they can do (ital) for (end ital) the doctors, not to them.

At least in its current iteration, CPOE is something doctors largely feel is being done to them. Perhaps this is because of immature user interfaces, perhaps because the methodologies used for obtaining clinician buy-in are still amateurish. (Though well intentioned, I’m not sure if buying a physician dinner constitutes advanced implementation methodology.)

For the latter, there are consultancies that claim to have perfected the rollout.

To address the former, CIOs try switching vendor systems (the grass is always greener), upgrading with the same vendor (the devil you know) or implementing add-ons (Web-native applications that sit between physicians and core clinical systems). In all cases, the goal is identical — giving physicians a Web-like CPOE experience, with IT tying everything together behind the scenes.

While KLAS confirms that the much-anticipated HITECH buying frenzy has yet to materialize, many are investigating the upgrade path. One can imagine the discussions CIOs are having with their vendors:

“My doctors absolutely hate your CPOE functionality.”

“How about upgrading to version X.2?”
“Is it any better?”


“But we just spent a few million getting on X.1 and have written all the interfaces between it and our ancillaries.”

“Oh, yes. That’s true.”

“And going to X.2 requires me to reconfigure my whole server environment.”
“Yes. I didn’t think of that.”

“And I’d have to retrain all my clinicians, which they certainly won’t like.”

“They sure can be ornery.”

“So what am I supposed to do?”
“Why don’t you buy the doctors dinner and convince them to use X.1?”

If, at that dinner, the local cardio-thoracic superstar surgeon (CTSS) repeats that he “didn’t go through 10-plus years of education to do data entry,” what will you do? What will your organization do? Will you banish his millions to the hospital across town which might be far more sympathetic to his position?

Perhaps that facility is playing the HITECH game more adroitly. You see, the initial definition of meaningful use only requires that hospitals have 10 percent of their orders going in through CPOE by 2011. Will slick hospitals get to that 10 percent on the backs of their junior (lower revenue generating) docs, while leaving Mr. CTSS alone as long as they can? Will crafty hospital administrators lure Mr. CTSS to the “more understanding” environment of their facility?

“Of course we don’t expect you to do data entry. You’ll find a home here.”

The games will get even more interesting next year, when CMS informs the industry exactly how hospitals are to prove their compliance with HITECH’s meaningful use and certification requirements. Will it be a simple attestation signed by the CIO, CEO and CFO (or any combination)? Will there be some inspection process for everyone (though I can’t imagine CMS has such manpower) or will reporting be like filing taxes, done on the honor system with the threat of audits as a looming stick?

Whatever is ultimately decided, there will be wiggle room in both achieving compliance with the objectives (getting to 10 percent) and in the reporting (we’re at 9 percent, they’ll never audit us). As usual, the uber-scrupulous may suffer for their veracity, while those with a more flexible interpretation of reality could find that approach yielding fatter margins.

It’s incumbent upon HHS, CMS and ONCHIT to scenario-play their meaningful use certification and reporting requirements all the way down the line to find the wiggle room many exploit. With some special interest prodding (see HIMSS), the government has started this ball rolling, so it has an obligation to ensure the honest aren’t left in fiscal disarray as a reward for their veracity. When there’s so much money on the line, someone will figure out how to make Mr. CTSS happy. When it’s time to talk to him about CPOE, tread lightly, or someone else will.



Thanks for connecting the dots. Your formulation certainly reflects a (perhaps "the") common motif pretty accurately. Let me add some complementary details.

In my experience, almost all CTSS's have physician assistants who write their notes and enter their orders (whether on paper or CPOE). Today, in hospitals who have rolled out CPOE that I've visited, the PA for the CTSS is the rule. The non-CTSS Docs, which is pretty much everyone else who's not a busy surgeon, don't have PAs.

The physicians who are hospital-based (and not surgeons), including hospitalists and intensivists are an important, next class of docs essential to the 10%-or-orders criteria. These docs don't work in the shadow of a CTSS. They are the attending-of-record for their patients, and their patients are always pretty sick. They either work directly for the hospital/health system, or they work for a group that the hospital has contracted. In terms of pareto charts of physician involved in hospital admission volume, the ED staffing is close-by. Their total encounter numbers and order volumes, of course, are much higher.

When looked at from the revenue-vs-costs lens that you're employing, whether employed by the hospital or contracted, physicians are almost always subsidized by the hospital (in terms of their hospital based work.) Inpatient is generally mission (money losing) outpatient is generally margin.

Most CEO/CIOs physician strategy for initial roll-out has an arm focused on these employed in-patient docs.

Anthony, I'm a little disappointed in you. You haven't picked up the term and concept of "Evaluation" (of live CPOE systems). This represents the work of our clinical friends at CSC (formerly FCG) and dubbed the Leapfrog Flight Simulator. It recognizes that HCIT isn't just theoretically a 2-edged sword in terms of negative, unintended consequences. Dr Dave Classen and colleagues have observed that live CPOE systems need to be assessed. And, that the familiar rapid implementation programs (all vendors) that are needed to meet the ARRA/HITECH provision timelines make pre-go-live and recurrent post-go-live, perhaps annual system testing extremely important.

So, building on your topic "CPOE is something doctors largely feel is being done to them," there is a shared imperative for hospital executives including CIOs. If society is mandating CPOE (which, of course it is), then we're all being forced to use it. It's being done not just to docs, but to all of us, including hosptial executives. And, physicians aren't the only end-users, although our/their diagnostic and therapeutic role is generally unique, as is their use of CPOE.

Where does this dialogue lead? It's the imperative of the CEO and Board of Directors to commit, to the physicians and the entire broader community, that we're all going to do CPOE right. That means safety. And that means Evaluation of CPOE. Before and after the physicians go-live. Anthony, Evaluation is rarely planned or discussed today, either in public, or behind boardroom doors. I'll be happy to get you the brief background data to back this up.

Your reaction?


Thanks for your comment. I obviously couldn't agree with you more.

Of course, if you'd like to study something more absurd than the current state of CPOE user interfaces, try reviewing the HIE landscape and business plans (more grants please!).

As part of any rollout it is critical "rehearse" nurse's and clerk's responses to physician requests to place orders for them. We would respond to an aggressive physician: "I will do it this time, but I am mandated to report the event." It is cause for dismissal for anyone other than the physician to place an order within a session that the physician has used a password to sign onto. Though it may seem Orwellian, it is also easy to track ordering by physician, location, and time if it ever comes down to the need for such confirmation. Human behavior is pretty predicable. It is much better to prepare for resistance than it is to fight it.
Now, there are many causes of resistance. Many are substantial and defensible http://www.feldbaum.com/images/Sept_06_Healthcare_IT_News.bmp . The "clunky" interfaces that can make even routine rounding painful are embarrassingly primitive and counterintuitive. If it was a web site no one would ever return for a visit.

Wow, what a great discussion. In my experience you are both right-on. Historically CPOE really was a shifting of data entry onto the back of physicians. I can even recall how we originally framed CPOE as the cure for poor handwriting and incomplete and non-compliant order format. What will provide the most "meaningful" leverage will be when we can document significant outcome (specific to vendor used) improvement with CPOE. This will, of course, require sophisticated actionable clinical decision support that fits smoothly into physician workflow patterns. Physicians will endure (if not embrace) technology that has proven benefits to patient care — particularly outcome.

First off, let me thank both of you for your comments.


Regarding CTSS that have their PA's enter orders, that will not be good enough in the current environment, as the goal is not simply to make sure the pharmacist or lab tech can read the order, but that the physician has entered it themselves (think of the game "telephone" where people pass on a message from one to the next, eventually amazed at how much the original message changed over only a few transmissions).

For example, many fear that nurses will be put upon by physicians to "just enter this for me. I don't know how to work that damn thing, and I've got six more patients to see at two more hospitals." I know there are hospitals around the country that are bracing for that fight. (reference to it is made in this interview). So what's the different between a nurse and a PA entering an order, not much. It's got to be the physician putting in the order themselves or an important point of error creation is left in the system.

To your second point, as I tried to make clear in my post, I was only discussing independent physicians that have admitting privileges. I believe that the CPOE paradigm is much different, and more simple, when referring to hospitalists, as they are employees.

Regarding the role of evaluation and continuous improvement, I agree that there is no way to do CPOE/EMR successfully without it. This is simply something too complicated to get right "out of the box." The problem is that the majority of hospitals in this country are small, just as the majority of physician practices are small, and this means the resources (read the above referenced interview) and skill sets are not there. I wonder how many IT Directors at hospitals under 100 beds know about the Leapfrog Flight Simulator. Without any disrespect intended, they probably assume it's something their kids play on the Wii.

My overall point would be that this is tough stuff. Not impossible given the resources (both financial and human), but far tougher than those with a superficial presence in this industry (oohhh, HITECH!) know.


In my recent interviews and discussions around CPOE, it's become more and more apparent that it's all about the user interface. Doctors use the Web just like the rest of us, and they will embrace the CPOE experience if it has a similar feel. If they can zip around Amazon and the Yahoo homepage, they want the same experience when entering orders. The killer app will be the one that makes them feel like they never left the Internet (and in some cases, they won't have to).

Thanks Jim. I totally agree that nurses must be given the tools (i.e., training, scenario playing, etc) to resist when asked to enter the order. If they are not prepared, it will be the administration that failed the process, not the nurses.

let's get real here. I have done CPOE for over 18 years. I have seen the worst and some good. First, it presumes that physicians have keyboarding skills. Wrong. I have had two partners who can't find the space bar. Second, it takes longer to do CPOE. I spend too much time with my Cerner system now, when I hit the enter button, watching the server upload the orders. I blame Cerner for some really stupid software functions. That hourglass icon plagues me when I have other work to do. third, in CPOE as I see it daily, the windows are fouled up, so that ordering is cumbersome. did any designer actually beta-test this with a doctor? you want docs to fail to use the system, just design a lemon! I see poor design right and left. Example this week: it is 9 p.m. and I am trying to put orders in the system, only to be blocked because a nurse is accessing the file. How stupid not to design a hierarchy in the system, where an attending physician trumps everyone in order to get the orders in the machine! The attitude seems to be what GM did in the 1970's: rush a system into the marketplace, and let the users do the fixing, rather than doing what the Japanese manufacturers learned, which is that your product must have high quality from the outset. fourth: do not presume that every doctor will have a scribe to do the orders. in the middle of the night, there is none. ALL doctors must learn the system. fifth: hospitals invariably do not want to pay (emphasize pay, pay for time away from the office) for the docs to be educated in the system. they expect use of CPOE as a physician duty. education receives shortcuts. web training modules stink. immediate help, context related, when one hits a roadblock, is invariably poor (my experience here is with Cerner and Meditech).

sixth and most important, when a physician calls the IT people about a problem with the system, stop dissing the doc. That happened to me. how about presuming that the doctor wants the system to be successful and wants to make it better. if so, then the person calling with a problem that is clearly a system failure is your friend, not your enemy