Recent industry surveys find that a very small percentage of physicians are using CPOE at the bedside. ARRA's proposed meaningful use criteria states at least 10 percent of orders must be input via CPOE or the hospital will get dinged, which brings up the decade old question as to why MD Luddites refuse to use this great technology.
How did we get here, and what can be done to increase physician usage? To understand the core issues, we need to view things from the eyes of a typical community doctor.
Before CPOE:
Mr(s). Doc stands at Ms. Patient's bedside flips through the paper chart, scribbles a couple of orders (while conversing with the patient) then pats the patient on the shoulder and says, “You're doing fine; this new medication should help; see you tomorrow.” It all takes about one minute.
The unit clerk or nurse then transcribes the order(s) into a computer system (hopefully). Mr. Doc sees about 10 or 15 more patients at the hospital then gets in his car and drives to the office.
While in his car he gets a call from the nurse reviewing his order. The nurse says that Ms. Patient had trouble swallowing the night before, and asks if they can change the med route from oral to injection. He says it's fine.
In his office an hour later after seeing four patients and while waiting for the next, he quickly checks his e-mail. He sees one from the hospital pharmacist saying that the med ordered for Ms. Patient is no longer available as an injectable, and asks if they can substitute another brand. The Pharmacist lists three alternatives with revised doses. The Doc responds “OK, go with number two.”
At this point the order is now ‘complete.’ How long did it take? An industrial engineer would say the combined time for the physician, nurse, pharmacist, and unit clerk, with phone tag and conversation probably took about 15 minutes. But if you ask the doc he would say ‘less than a minute,’ basically the time it took to scribble the order in the chart. In his mind the time in his car does not count (it's downtime anyway) and responding to e-mails is part of his usual office routine.
It gets even more onerous if the patient has complications, there are medication/diagnostic conflicts, or multiple overlapping orders are involved. In each situation they get chased down, sorted out and preliminarily resolved by hospital staff, who then communicate with the physician to reach final order approval. But in the end it's still only a minute for the physician.
After CPOE:
At Ms. Patient's bedside he picks up his notebook, or iPod, or BlackBerry, or similar, and enters the same med order. Maybe not as fast as scribbling on paper, but not too bad.
Assuming response times are instantaneous (could be a big assumption), the system comes back and displays assessment information about the swallowing problem. The system suggests he change the med route (this is a really advanced system!). He isn't sure about this, so he scrolls back into the EMR and sees where the late shift nurse made a note about swallowing difficulty, so he cancels/changes the original order and requests an injection.
Red flags go up again when he gets a message from the Rx system that this med is no longer available (or recommended) as an injectable. The system then shows him a list of three alternatives (after a slight delay, of course). He picks one, cancels then re-enters the new order, changes the dosage, then signs off, and now he's done, except for patting the patient on the shoulder and saying, “see you tomorrow.” Total time: five or six minutes.
As noted earlier, if the situation involves more complex cases or multiple orders even more of his time gets eaten up.
How much time did it take nursing and pharmacy to complete this order request? Zero, or near zero.
If he sees 15 patients and each one takes an additional four minutes more, to him that's 60 minutes extra effort. But the hospital “saves” the hour. We all can agree it is better patient care, but who's paying for it? One hour of his time is equal to four or five patients in his office, maybe $300 to $500 in billings. If this happens two or three times a week, it quickly becomes real money and nobody is paying him a salary to hang around.
So, from where Mr. Doc sits, CPOE is great for the hospital and hopefully delivers better patient care, but he's carrying the lion's share of the time cost, and to an independent practitioner - time is money.
Where CPOE works and why
If you look under the covers at why CPOE has worked in the Mayo, Kaisers and Cleveland Clinic, it's because the attending physicians are part ‘owners’ of the hospital. They get paid a salary and bonus based on the performance of their practice, and the performance of the hospital and all owned facilities. They readily accept that less support staff will save the hospital money. But they also understand how it will result in a monetary benefit for them and improve patient care, which can lead to even more patient referrals and more revenues.
In military facilities it's even simpler. The Colonel says to get it done and the MD captain says “Yes, sir!” In regard to the VA system, a recent Wall Street Journal report dated Oct. 27, 3009 entitled “The Digital Pioneer” by Jane Zhang, stated:
To be sure, the VA's health care system isn't a perfect roadmap for the industry - since the agency is in a unique position. The VA … employs the doctors … which makes it easier to mandate performance standards. The VA has an incentive to keep patients healthier because it takes care of veterans for life and sicker patients eat up the VA's budget faster.
How'd we get here?
Today, and for the past half century, we have been in a situation where the person most responsible for ‘product definition’ and most responsible for bringing in the business is not employees of the hospital. It goes back to the establishment of the AMA and the AHA in the early 20th century. Both of these groups were focused on increasing utilization of hospital and medical services and even at that time, just as today, medical care was relatively expensive. Keep in mind that almost all doctors in the early part of the 20th century were independent practitioners and hospitals were places to be avoided. So to drum up business they both devised the idea to sell an insurance policy. Rather than work together, the AMA founded Blue Shield and the AHA started Blue Cross, each with its own similar, yet different objectives. In 1972 as the health insurance industry matured and the FTC was getting more concerned, the AMA spun off Blue Shield, and AHA split with Blue Cross. Later as the Blues saw themselves more as insurance companies than as part of the medical establishment, many of the Blues merged and eventually morphed into today's United Health, Wellpoint, etc.
In 1966, along came Medicare. If you go back and study the legislation of the day, you will find that physicians fought Medicare with a vengeance and wanted no part of the government or the institutional side of the package. Today if you tried to take Medicare away you'd have a rebellion - and not just from seniors. So, Medicare in 1966 solidified the arms'-length relationship of doctors and hospitals by creating Medicare Part A (hospital) and part B (physician) payment systems.
The structure we have today, that of full physician independence, has been around a very long time and is fortified through separate provider and piece work based payment systems. Again, to the doc time is money, but some challenge this by saying it should be his (her) professional responsibility. If we really want professionalism to lead, then an overhaul of the organizational structure and payment systems will have to happen first.
How do we fix it?
How can we get more of those independent physicians to use CPOE? From my point-of-view there are five possible ways to address this issue. None of which are perfect, most of which are fraught with more challenges. They are:
1) Require all physicians to be hospital employees la the military and Mayo model.
OK, maybe it's the real solution but not at all feasible, not realistic and how can you change 100 years of precedent in only a few, impossible. Well not completely impossible, this actually has been happening very, very slowly over the last decade or two.
As more and more medical school graduates are either female and/or foreign students the desire to be a ‘lone ranger’ doc is less and less appealing. Female and particularly foreign docs are more accepting of working as a team in large organizations. Being on call 24x7, dealing with constantly changing payment rules, and all the hassles of malpractice is not their cup of tea. They'd rather spend more time with patients and have a predictable schedule. So, if we can wait long enough, way beyond the ARRA target dates, we might just get there.
2) Expand Hospitalists Programs.
Hospitalist programs got their start about 10 years ago. Hospitalist programs are similar to physician residency programs in that the physician is assigned to specific inpatients 24x7. The major difference between a residency program and hospitalist program is that a resident has three roles, medical education, clinical research and patient care. Sometimes these are very conflicted goals. Meanwhile the hospitalist is only focused on inpatient care. Expanding hospitalists programs is really a derivative of number one. Hospitalists are usually employees of the hospital, or a contractor to the hospital. Use of CPOE tools by hospitalists is very high and their operational objectives are almost 100 percent in line with the institution.
Since hospitalists tend to be newer graduates and frequently foreign doctors, they readily accept information technologies and understand the downstream benefits of source data capturing. They are on the patient floor eight or 10 hours a day and routinely see the benefits of an EMR and CPOE. Unfortunately mandating hospitalists programs has its political hurdles. Attendings may not accept them for fear of ‘losing’ the patient. The biggest hurdle is cost, 24x7 medical coverage is not cheap.
The Institute of Medicine and other organizations such as The Leapfrog Group have made strong positive statements about hospitalist programs, particularly in ICU settings. But alas, they haven't been willing to foot the bill.
Here's a wild idea. When a Cerner or Epic or Eclipsys sells an EMR-CPOE system, usually for $20 million or more, why not include a hospitalist program with it? Maybe they should team up with one of the growing hospitalist contractor companies. Then the system vendor can easily guarantee 100 percent CPOE utilization.
3) Pay attending physicians for their time.
If it takes a doc an extra five minutes to basically do the data entry, why not pay him/her for it? I have heard several physicians suggest this. Again it's a cost issue, and Medicare and the insurance companies won't cover it. Additionally the hourly rate for some physicians, such as surgeons and specialties, could be quite steep.
4) Make better use of physician assistants and other clinical staff.
Some have suggested that to reduce physician time we allow physician assistants, pharmacists and selected nurses to initiate, change, modify, or cancel orders when conflict and other issues arise. This has been done on a limited scale and usually after the fact. But as our CPOE and EMR systems get more and more sophisticated and can bring up more and more real time quality issues it becomes hard to see how physician time will decrease. Actually ‘dumbing down’ the systems might work better, but in the end that would be very short-sighted and foolish. If we choose to introduce greater support staff involvement we will also have to re-engineer many of the work flow processes and install more tools to monitor, control and verify all the clinical hand-offs.
5) Improve the CPOE systems.
Faster, faster and faster, get to sub-nanosecond response times, that's the ticket. Faster response times are always better, but the reality is the faster our systems get, the more we want the system to do. We are a long way from building a system that incorporates all the algorithms, judgments and expertise for even one moderately complex medical order. The faster we can handle the order and its component parts the more we'll build into the editing, checking, and auditing for quality goals. Ultimately this brings more and more information to the physician's screen, all of which would have previously happened after the fact, meaning more physician time in the patient room. Faster systems will improve care quality, but I don't think speed alone will get more physicians to use CPOE.
More important than faster systems is building in more work flow flexibility and re-engineering core processes such as hand-offs, and better case management and care coordination. All of these have more to do with work flow and process re-engineering than technology.
My conclusion is we have to stop trying to kid, cajole, arm twist, embarrass, and penalize community physicians into using CPOE. As things are structured today it is not in their self interest. What with the government threatening everyday to cut physician payments, most physicians are scrambling for any way to increase their volume by eliminating or off-loading tasks, not doing more for organizations that don't pay them.
In my experience the best strategy will be a combination of 1, 2, 4, and the workflow of 5. All of which will take more time than ARRA will probably allow.
CONTINUE THE CONVERSATION
Wiki-fy this story at http://www.healthcare-informatics.com by posting comments, listing relevant resources and linking to associated events.
Frank Poggio is President of The Kelzon Group, a consulting firm focused on HIS systems. He also has served as the CFO and CIO at the University Wisconsin Hospital and was Senior HIS Consultant with KPMG. Healthcare Informatics 2010 March;27(3):46-50
