Home » Time Capsule » Currently Reading:

Time Capsule: Hospitals Want Software to Do The Dirty Work of Changing Physician Behavior

September 4, 2011 Time Capsule 4 Comments

I wrote weekly editorials for a boutique industry newsletter for several years, anxious for both audience and income. I learned a lot about coming up with ideas for the weekly grind, trying to be simultaneously opinionated and entertaining in a few hundred words, and not sleeping much because I was working all the time. They’re fun to read as a look back at what was important then (and often still important now).

I wrote this piece in July 2006.

Hospitals Want Software to Do The Dirty Work of Changing Physician Behavior
By Mr. HIStalk


An editorial in the latest American Journal of Managed Care titled “Defending Computerized Physician Order Entry From Its Supporters” stresses that physician order entry (CPOE) and clinical decision support systems (DSS) are separate entities, despite popular perception. Ross Koppel, a sociologist and Penn professor, says that sloppy terminology has confused the respective benefits and shortcomings of CPOE and DSS.

Ross’s sociologist view is interesting. We’re expecting a lot from immature CPOE and DSS systems that most hospital executives can’t define, even when they’re plunking down hard-earned capital dollars to purchase them.

CPOE is a smart typewriter that, standing alone, has little ability to improve patient outcomes. It prevents transcription errors, although those seldom harm patients because they’re usually caught anyway. CPOE makes it easy to choose common order defaults instead of “winging it.”

Beyond that, the benefits (both clinical and financial) come from DSS, not CPOE. (I’ve often joked that hospitals could use e-mail as a poor man’s CPOE system – just let the doc free-text whatever he or she wants and send it to the appropriate department, thereby eliminating transcription and turnaround time errors for free.)

Commercially available DSS systems are, unfortunately, mostly frightfully immature, even more so than CPOE. Early adopters share war stories of nagging alerting, inflexible third-party rules, the inability to customize and personalize, and problems with performance-sapping rules engines incapable of delivering alerts of any more sophistication than the old hard-coded screen edits. No wonder doctors have been underwhelmed.

Still, the real problem is right down Ross’s alley. Hospitals usually buy CPOE and DSS because they’ve failed to control physician behavior otherwise — often euphemized as “reducing practice variation” or “practicing evidence-based medicine.” They want software to do the dirty work that they can’t or won’t: telling physicians that they’re wrong and demanding that they change. When docs don’t follow the electronic cookbook’s rules any better than the paper ones it replaced, systems and vendors are blamed.

I’ve been involved in two CPOE/DSS implementations, both involving large IDNs and well-known vendors. In both cases, hospital administrators ill-advisedly shot their patient safety technology wad on CPOE, confident that it would improve patient care better than any other investment (despite ample contradicting studies). Physician adoption was universal in one, minimal in the other, but one element was common: 90% of the expected DSS benefit never materialized. Pre-implementation enthusiasm gave way to the grim reality that the system wasn’t going to be much help in changing practice patterns. We purchased DSS, but implemented a smart typewriter.

No software contains a switch that turns resistant physicians into docile, rule-following sheep who make better decisions under the watchful eye of Big Brother’s can’t-miss medical guidelines. Displaying a few dumbed-down alerts won’t convince them they need to change. But if your hospital has already spent a few million on CPOE and DSS thinking that was the case, you’ve learned that already.

Maybe physicians will recognize the next generation of systems as their ally, not their enemy. After all, they want the best outcomes for their patients, too. Where they disagree is that we have the answer right now with these first-generation CPOE and DSS applications that we can’t even define.

HIStalk Featured Sponsors


Currently there are "4 comments" on this Article:

  1. Thanks to Mr.H for reviving this rather flattering reference to my work and words. I have two responses to this 5 year old column.
    1. CPOE is more than an email program, fancy typewriter or input device. And it’s a lot more than the input front end for generating CDS. Even without CDS, CPOE guides choices by its listed options and pull down menus, it speeds information to the pharmacy and eMAR, it requires specific doses and schedules, it provides order sets that also help clinicians perceive many of their related choices, etc., etc. Yes, as I’ve written in my articles and books, it has its several faults, but these should not be confused with its useful functions
    2. CDS is limited by many factors: a. clinical trials are based on pts with one illness taking one drug, but most pts are taking 11-14 drugs and have many comorbidities; b. most real pts also have compromised liver, heart and kidney functions that cause MDs to have to make serious considerations beyond the CDS algorithms; c. no one knows the drug-drug interactions among the 4,500 drugs in the formulary and the, say, 12 drugs the patient is on; d. pts are usual old and always sick, and thus often surprising in their reactions; e. As shown by Metzger et al, CDS misses about 85% of pregnancies and age-related issues; f. too many alerts are written by lawyers and not by clinicians; g… I could go on but I won’t.
    The most important take home lesson is that clinicians override or ignore about 90-95% of most CDS alerts. How can we then call CDS the major reason for CPOE’s existence?

  2. Interesting contrast in the comments of Frank Poggio and Ross Koppel. Still, after 35 years of emergency medicine practice and 16 months now as CMIO, I’d say Ross is on target and Mr. Poggio doesn’t get it. Unfortunately, too many hospitals are listening to Mr. Poggio and his consulting crowd instead of investing in physician leadership so that the challenges Ross Koppel cites lead to sensible efforts to engage physicians in addressing them (the challenges RK cites).

  3. RK states: “Yes, as I’ve written in my articles and books, it has its several faults, but these should not be confused with its useful functions.”

    What useful functions, exactly? Several hospitals resorted to vacuum tubes and runners to improve throughput for years after the CPOE switch was turned on.

    RK must be aware that the enormity of the CPOE faults are unknown, as are the adverse outcomes and near misses which emanate from a counter intuitive workflow that CPOE rigidly enforces. Errors resulting in deaths occur across the continuum of CPOE and its bidirectional hub like links with ancillary services througout hospitals.

    An accounting of the size of the abyss has not been contemplated (nor wanted by the HIT industry). CDS is rife with defective directives that amplify the immediate work flow dysfunctions of CPOE.

    HIT industry take heart, RK has shaded the color of his feathers.

Founding Sponsors


Platinum Sponsors




























































Gold Sponsors











Reader Comments

  • Eddie T. Head: Re: Cerner subreddit - the subreddit became private over the weekend. No more leaks, I guess....
  • IANAL: How many virtual primary care startups can the market support?...
  • Ernie Riddle: Good insight and a note that there is a lot of change ahead. Many technology platforms are active in Healthcare and the...
  • Quynh Tran: Appreciate the shift in language from "follow the science" to "evidenced-based." The latter may be less intimidating a...
  • Matthew Holt: I'm at HLTH The testing on site was smooth--once you go into it. The Clear app to get into the testing was not.... ...

Sponsor Quick Links