Hidden and Camouflaged: HIT Fail

Manual whiteboard system superior to computer technology:

In some cases, providers noted that computer systems hid some of the information; if only three comments could be viewed per screen, they had to click to get to another screen, requiring them to search for information that might demand immediate attention.

The study also found that there were fewer visual cues with the computational system. Some providers noted that they used to be able to get a sense of the status of the emergency department just by walking through the room and visually checking the manual whiteboard. “Without that public display, providers have to sit down at the computer and check it, which can add time or reduce awareness,” said a principal investigator.

These are potentially disastrous consequences in an ED environment where patients can be highly unstable and serious events transpire rapidly and irreversibly. An investigator in the study observed that "the results provide an important case study of what can happen when new technologies are developed without sufficient understanding by designers of the nature of the work in which they will be used."

This last observation raises a more fundamental question: why does such an axiomatic, common-sense statement, especially in a domain as complex as medicine, need to even appear in print?

Clinical Software is Designed for Utopia

Flashback to 2003: It's surprising to note how little has changed...

Patient care information systems like CPOE ... can create unintended or "silent" errors, according to a separate study conducted by the same author in the Netherlands and Australia.

"Many information systems simply don't reflect the health care professional's hectic work environment with its all too frequent interruptions from phone calls, pages, colleagues and patients. Instead these are designed for people who work in calm and solitary environments. This design disconnect is the source of both types of silent errors …Some patient care information systems require data entry that is so elaborate that time spent recording patient data is significantly greater than it was with its paper predecessors," the authors wrote. "What is worse, on several occasions during our studies, overly structured data entry led to a loss of cognitive focus by the clinician."

Read more.

Scot M. Silverstein, MD on HIT

Excerpts:

...keyboards and mice were unprotected from bacterial contamination. My advice on changes to more appropriate hardware and other measures to ensure patient safety were simply overruled by IT personnel.

My counsel as postdoctoral-trained Medical Informaticist and ICU-trained clinician was simply dismissed and overruled by the IT staff and CIO on grounds that the IT staff were unfamiliar with existing, ICU-appropriate computer hardware and wouldn't support or even evaluate "nonstandard" (to them) computers in any case. Further, hospital administration sided with the IT department, a department led and staffed by business computing personnel (of “management information systems” or MIS background) totally lacking in clinical knowledge and experience. Patients remained at risk...

While clinical IT is now potentially capable of achieving many of the benefits long claimed for it such as improved medical quality and efficiency, reduced costs, better medical research and drugs, earlier disease detection, and so forth, there is a major caveat and essential precondition: the benefits will be realized only if clinical IT is done well. For if clinical IT is not done well, as often occurs in today’s environment of medical quick fixes and seemingly unquestioning exuberance about IT, the technology can be injurious to medical practice and biomedical R&D, and highly wasteful of scarce healthcare capital and resources.

By Design: CPOE Falls Short

Emphasis added:

“The problems are multiple—the presentation of information is often confusing or invisible, the pull-down menus don’t always make sense,” says Ross Koppel, PhD, who serves on the faculty of the University of Pennsylvania and has studied and written about CPOE extensively. “The alerts that come are almost always overwritten. They’re written by the legal team who don’t want to get sued for anything, so they warn of every drug interaction...”

Physicians who suffer “alert fatigue” sometimes overlook something, or the alerts miss potential problems they should have caught. The CPOE interfaces look like “the state of art, circa 1992,” in part because they were developed on an accounting software platform and their development hasn’t yet achieved that of clinical systems, Koppel says.

The user interfaces create a “hostile user environment” and are often difficult to read and easy to populate with misinformation, says Scot Silverstein MD, adjunct professor at Drexel University’s College of Information Science and Technology and a healthcare journalist. The confusing user interfaces “actually promote user error,” he says. “Some of these interfaces are guaranteed to cause user confusion and create mistakes—not just CPOE but other (IT) systems, as well...”

[CPOE vendor contracts require physicians to] sign non-disclosure agreements prohibiting them from publicly discussing problems with the software. Physicians can tell a colleague about a problem they have been having with CPOE, but contractually they cannot publicly present that information at conferences or bring it up in online discussion groups...

via HIT Exchange

Emphasis added:

HEALTH CARE INFORMATION TECHNOLOGY (HIT) vendors enjoy a contractual and legal structure that renders them virtually liability free—“hold harmless” is the term of art—even when their proprietary products may be implicated in adverse events involving patients. This contractual and legal device shifts liability and remedial burdens to physicians, nurses, hospitals, and clinics, even when these HIT users are strictly following vendor instructions. Vendors avoid liability by relying on the legal doctrine known as “learned intermediaries” and on warranties prohibiting claims against their own products’ fitness. According to this doctrine and legal language, HIT vendors are not responsible for errors their systems introduce in patient treatment, because physicians, nurses, pharmacists, and health care technicians should be able to identify—and correct—any errors generated by software faults...

Vendors retain company confidential knowledge about designs, faults, software operations, and glitches. Their counsel have crafted contractual terms that absolve them of liability and other punitive strictures, while compelling users’ nondisclosure of their systems’ problematic, even disastrous, software faults...

~ Ross Koppel

Emphasis added:

Leapfrog’s Binder says hospitals “tend to be prehistoric in their thinking” when it comes to healthcare IT, and their vendors have not put interoperability in the forefront of product design. Other industries embrace technology and allow it to work across national and international systems, but some CPOE applications can only talk to themselves...

...scholars say the “hold no harm” clause should be removed, a move that would quickly improve the software, and the government could prohibit vendors from requiring physicians to sign confidentiality oaths. Providers should be able to share best practices, problems, and data with others using the same software across the country, not just with colleagues down the hall.

Problems With Cerner

via Health Care Renewal:

Despite extensive nursing informatics support (approx. 7000 hrs per month for the first six months after implementation), our error rate went from approx. 176 medication errors per day one month after implementation to approx. 100 medication errors per day currently. It has been stable at this level for at least the past 3 months. Many of these errors involve high risk medications (e.g., heparin, morphine). In order to understand the etiology of the errors, i need to explain how cerner processes CPOE orders.

With Cerner, orders are grouped into plans (powerplans). These plans can include sub-plans (phases) and sub-sub-plans (sub phases). A typical orders display screen appears as follows: (b)(4). Under the plans is an order tab which displays other orders.

This design allows two distinct sources of error.

First, there is no consistent way to view orders for medications. A medication order can display either in a plan, sub-plan, or under the orders tab...

This facilitates duplication of both medication and non-medication orders (there is a medication checker which is so poorly designed and does little to aid the pharmacists in detecting duplicate medications).

Second and more dangerous, high risk intravenous medications can be run either inside or outside of the plan...

Read the rest.

Original document.

CPOE Systems Have Plenty of Design Flaws

CPOE failures:

CONSIDER IT A DILEMMA of the computerization of health care. While electronic medical records need physician input via computerized provider-order entry (CPOE), doctors are having a tough time figuring out how to navigate systems that are being rushed to market.
A case in point: A recent study revealed that introducing a CPOE system with clinical decision-support actually increased the number of duplicate medication orders. The irony is that CPOE is supposed to prevent that exact problem.
The study, which was published online in July by the Journal of the American Medical Informatics Association, found that after CPOE was implemented, duplicate medication orders jumped from 2.6% of all orders to 8.1%. The study looked at drug orders in two ICUs (adult and cardiac) at a 400-bed community teaching hospital... 

What finding was most disturbing? 

Many duplicate orders were placed within an hour of each other, sometimes by the same provider, sometimes by different providers. 

Why didn't the warning systems work? 

There was a problem with how the algorithms were designed. Duplicate order checking is based on identifying similar medications, which includes both the medication name and the route of administration. If I said, "Replace potassium," but one person ordered it by IV and one ordered it orally, the system doesn't issue a duplicate med alert because the routes of administration are not the same. That needs to be changed.
The alert picked up three different kinds of duplicates: the same order, the same medication and the same drug class. Many duplicate alerts for drugs in the same class are false positives because we use a lot of meds in the same therapeutic class. Plus, we found that the content of the alerts that popped up was very complex.

Read the rest...

With paper charts, orders can be reviewed in chronological order, an option not necessarily available in CPOE systems.