Printer Friendly

Intensive care: the next level for IT.

Intensive patient care areas like the emergency room, surgical suites and intensive care units have their own challenges, radically different from other inpatient units. Automating these areas requires a different level of IT product, design and development.

Michael Schneider, M.D., a practicing anesthesiologist and a medical informatics expert, recalls the time about a year and a half ago when a thirty-ish young woman came to see him for the first time. She was about to have surgery and had a rare medical condition called Klippel-Trenaunay syndrome. That unusual malady, the same one that has afflicted famous golfer Casey Martin, who is obliged to use a golf cart to get around golf courses, affects the body's blood vessels and can lead to dangerous blood clots in multiple places.

"This patient was presented to me one day when I was working the outpatient surgery center at the University of California-Irvine Medical Center," Schneider says. "While she was waiting to be seen, I was able to quickly research the anesthesia implications of that syndrome. I did an Internet search in order to make sure the anesthesia plan was adequate for her condition. Here was a case in which I as a clinician had an immediate need and the patient in question got good care because I had the information available to give that care. I relate any number of instances in which, in contrast, we had delays in surgery, and used medicines that could have been avoided because we didn't have access to information. And care was probably compromised because of that lack of information."

It's just that sort of contrast -- between immediately available and useful information, and a lack thereof -- that argues so convincingly for the need for strong, useful information technology solutions for the areas of highest intensity in today's health care organizations: the emergency department, the surgery area and the intensive care units. Yet, ironically, only a small number of hospitals and health systems have made the successful leap into automation of these crucial patient-care areas. Why? Experts cite several reasons:

* For the most part, the standard software products available from vendors have not until recently made clinicians' lives easier or made work processes noticeably faster, though that is beginning to change;

* The products available for each of these somewhat distinctive areas -- ED, surgery, and ICUs -- tend not to be ideal for all three areas at once, and tend not to integrate well across entire enterprises, complicating the dissemination of automation across patient care organizations;

* Specifically, many of the installations developed for each of these three areas have not been dovetailed with electronic medical records (EMRs)/computer-based patient records (CPRs);

* Many of the broader clinical automation products available for general inpatient care are difficult to optimize for these data-intensive, high-speed, heavily process-oriented areas of the patient care organization;

* Even within this broader category, there are different levels of data intensity: Surgery is at times considerably more data-intensive than the ED or ICU, and different ICUs have different levels of data intensity, while all these areas are far more data-intensive than the typical medical/surgical floor in the inpatient hospital;

* There has been a lack of involvement of clinicians in the development of these products, and a lack of acceptance among clinicians of many of the products already on the market.

The conundrum of whether to install a system that works ideally for an ED, surgery department or ICU, but doesn't dovetail well with the rest of a patient care organization's systems, or vice-versa, is a troubling one for clinicians, executives and IT people alike. "That problem speaks to the question of whether we should go to `best of breed' or `best of fit,' with `best of breed' meeting specific departmental needs, but not working together with other systems, and `best of fit' not meeting all the needs," says Margret Amatayakul, vice president, Sheldon Dorenfest & Associates, a Chicago-based consulting firm.

"We've been approaching this from a core or enterprise vendor approach, where you've put into place a core set of systems that are best of fit, and then you add on your best of breed in certain niche areas that make sense," continues Amatayakul. "But obviously, your selection of those best of breeds has to be based on their ability to integrate with the core. I think that the emerging Web-enabled technology will help achieve that balance, but you have to be very cautious."

As a result, the typical situation is one in which there is something of a mismatch between systems, and hospital-based organizations are struggling to make it all work, says Schneider. He recently joined the Long Beach, Calif. office of First Consulting Group, but continues to practice anesthesiology at UC-Irvine a few days a month.

"One of our clients now is Cedars-Sinai Medical Center," says Schneider, referring to the famous Los Angeles tertiary care center. "They have a very sophisticated ICU information system and have just installed an ED system. Their OR system, however, is strictly for scheduling and intra-operative recording of notes, and is 10 years old and needs to be replaced. I've followed physicians on rounds at Cedars and I can tell you that their surgical ICU system is very sophisticated. All the physicians and nurses use it for charting and you can get instant lab results and medication histories. But right next to these beautiful 21-inch screens are huge paper charts because the rest of the hospital doesn't have the same automated charting capability and isn't integrated with the surgical ICU system." That's more or less the state of the art, he notes.

Challenges remain in large part because the ED, ICU and surgery suites are fundamentally different from other parts of the inpatient hospital, says Debra Sly, MN, R.N., senior manager of First Consulting Group's Seattle office and a consultant with a background as a criticalcare nurse.

"In these intense environments," Sly notes, "the volume of information and data being gathered requires an information system that can manage that kind of data. When you step down into the less intensive areas, the technology required to manage that data is not as great. What we're finding is that people no longer want just a niche system for the ED or ICU, but one that can integrate across environments. It's difficult to find a system that can scale up and down appropriately and accomplish the whole continuum.

"So what you get right now," Sly continues, "is either a vendor with primarily a generic acute-care or med-surg system that scales up to some degree to accommodate the increased data in critical care but isn't completely sufficient for that environment; or, vice-versa, a system designed for critical care that is too dense and expensive, and which is difficult to scale down for use in the less-acute environments."

At the moment, ICUs have the greatest penetration of IT automation among the three specialized areas, says Sly, but that should change in the next few years. In the meantime, she acknowledges, there are very, very few organizations where appropriate automation is fully diffused across the ED, ICU and surgery departments, and linked to enterprise-wide clinical systems.

A practical, workable solution at UCLA

In the meantime, there are excellent examples of individual systems that are showing what can be accomplished. One such system is located just across town from Cedars-Sinai at the University of California at Los Angeles (UCLA) Medical Center in the Westwood neighborhood of Los Angeles, where for four-and-a-half years clinicians in the medical ICU have enjoyed capabilities few of their peers can boast. "We have the ability to do all of the charting for almost everybody who works in our unit--the nurses, the doctors, the respiratory therapists, the social workers, dietitians, physical and occupational therapists, using our automated system," reports Paul Bellamy, M.D., the medical director of UCLA's medical ICU, one of seven critical care units at UCLA and the one that so far has been fully automated.

When shopping for a system over four years ago, Bellamy notes, only three vendors seemed to have the level of sophistication that he and his colleagues were looking for, and only one system, from the San Diego-based CliniComp, had all the features they wanted. It's turned out well and satisfies one of the core needs he and his colleagues see every day.

"In the ICU," says Bellamy, "for simple tasks like charting by nurses, vital signs are taken much more frequently then on a general medical floor where they might be taken anywhere from once every hour or two to maybe every six or eight hours. In the ICU, the minimum frequency is usually an hour or two apart, and as often as every five or 10 minutes. A whole different level of functionality is required at that density. Plus, there's much more being recorded at those intervals than on a patient on a floor. The floor patients are normally going to be the normal traditional vital signs and maybe pulse oximetry, whereas we'll have invasive lines for measuring such things as blood pressure, heart function and intracranial pressure. In addition, our patients have much more complicated medical problems requiring many more medications that need to be charted and tracked.

"In selecting the system we did, our feeling was that a system that worked in the ICU could easily be scaled down to work for the floor, whereas a system designed for the floor wouldn't necessarily very easily scale up to the ICU."

This is an important consideration in a unit with the level of traffic his has, notes Bellamy. The medical ICU at UCLA is an eight-bed unit with a nursing staff of over 30. At any given time, there will be at least one attending physician in the ICU assisted by two residents, three interns and a medical student, two nursing assistants and two respiratory therapists, all of whom are charting on the system.

After four-and-a-half years on the system, the contrast between before and after automation is very clear, Bellamy continues. "We did some before-and-after evaluations and found that our charting had become more complete and accurate. It was clearly more legible, which was a big plus, and our charting was always available to us. And, any chart from the last four years can be un-archived in about three to four minutes, so we're not dependent on the medical records department to get old records at midnight on a Friday night. They're always accessible to us. Those kinds of things make a very big difference," he adds.

One of the biggest benefits, though, is the fact that "We also chart our procedures on the system," Bellamy says. "That allows us to create a flow sheet that tracks the consequences of the procedures. Related to that, we also do direct data-sharing, so we don't reproduce each other's work."

Bellamy cites a classic example of the need for automation to improve clinical processes in the ICU. "In the typical paper-based ICU process, you have respiratory therapists creating flow sheets or series of chronically laid-out tables to chart patients' ventilator settings," he says. "At the same time, you have the nurses replicating that process on separate sheets of paper because both groups need to know what's going on. And, the nurses want to track the same information as the doctors do. Yet the nurses typically will not chart as accurately as the respiratory therapists, and in the process of recharting, things sometimes get recorded wrong. Now that we're automated, we simply have the respiratory therapists doing the charting. The electronic version is available for both the RTs and the nurses for their records and charts. The nurses have learned to trust the RTs on this -- and we've eliminated duplication of effort and the added potential for errors."

Of course, there's always a need for staff adjustment to the system. In the case of the medical ICU at UCLA, "We went through a train-the-trainers period for coming up on the system," Bellamy notes. "We actually customized a lot of our flow sheets from each profession so that they all had ownership of their parts or their portions of the system. Then we did the troubleshooting on all that software before we went live. It took us a long period to do our customization. We were victims of the typical university approach that says that if it's not perfect, it's not good enough. We were a little perfectionistic."

Operating rooms as data monsters

The kind of data density found in an ICU is intensified in an operating room, with important implications for data analysis and process improvement, notes Schneider.

"During surgery," he says, "there are a lot of people putting their hands in and creating data about the patient. The surgeon is actually not the most data-intensive. The surgeon's role is really to manage the patient. His or her tasks include the dictation of the procedure and assuring that the history and physical are completed -- which actually may or may not be done by the surgeon himself, and, in any case, is a small subset of the information.

"In fact," says Schneider, "it's the anesthesiologist who is supervising the largest amounts of data entry during surgery, both through his direct, continuous monitoring of vital signs and an array of medical status monitoring, and via his supervision of the nurses' monitoring and data entry.

"Automation can be a very big help in the surgical suite, especially for the anesthesiologist and nurses," continues Schneider, "because during surgery they're documenting so much detailed clinical information that any automation of that documentation process will help save them time. The challenge is that it's been very difficult to find systems that automate but don't get into clinicians' way."

Not surprisingly, however, it's physicians who tend to be most resistant, initially, to new ways of doing things -- that includes automation. "There are anesthesiologists," concedes Schneider, "who hate the idea of an automated record-keeper because they think it'll give ammunition to lawyers. In fact, often times, moment-to-moment changes in vital signs can be artifactual, meaning that they don't actually stand for significant status changes in a patient. For instance, the surgeon may lean on the blood pressure cuff in a way that causes abnormal blood pressure to be recorded, or a monitor may be jostled, leading to a momentary blip in vital signs. But my feeling is that documentation will help because it'll be so obvious you have this one abnormal item of data, and if you have continuous data entry, you can better manage care."

The real practical challenge, Schneider adds, is that "You have to say, `I'm willing to bend a little bit to accommodate the installation of these new systems.' And the benefits are tremendous. There's real potential for using automation of data to support far better outcomes analysis--analysis of clinical, cost and procedural data, that would be very difficult to get at otherwise."

If it can be done in Ghana ...

Perhaps one of the most dramatic examples of the use of relatively simple automation to support clinical care processes in a data-intensive environment is one involving telecommunications connections between Boston and the west African nation of Ghana. It's a case study that highlights the benefits of relatively simple IT/telecom solutions for emergency/urgent care.

Daniel Carlin, M.D., a board-certified emergency physician, has had a lot of life adventures, ranging from being a ship's medical officer in the U.S. Navy for a few years to being a medical volunteer abroad. Now, Carlin has a dual title as CEO of the World Clinic and director of the Global Telemergency Center at the Division of Infectious Disease and Geographic Medicine at the New England Medical Center (NEMC). What does all that mean?

Since April of this year, it's meant that Carlin, from his office at NEMC in Boston, has helped to telemedically direct the delivery of health care services to nearly 2,000 industrial workers who are building a thermal energy plant in the Takoradi River Valley in western Ghana.

Through connections Carlin was brought on board as director of medical services for those workers, whose ranks include Americans, Europeans and some Ghanaians and other west Africans, by the large engineering firm Stone and Webster. The company began to see a loss of worker productivity and high medical evacuation costs early on in the project's development, as the workers fell ill with malaria, dysentery and other diseases and were injured in industrial accidents. Based on a cost-effectiveness calculation, the decision was made to post a nurse practitioner in a customized trailer infirmary at the construction site and to attempt the development of a telemedicine support structure back at NEMC, with Carlin as the medical director.

The technology infrastructure began in utmost simplicity, with the establishment of phone, fax and Internet connections between NEMC and the Ghana construction site. In addition, the American nurse practitioner, John Dornbusch, had an Epson Photo PC-600 digital camera and tripod that allowed him to take photos for chest X-rays and the like. Soon after opening the clinic, Dornbusch began using a new portable EKG device called the Valentine System PC/ECG, made by the Torrance, Calif.-based Brentwood Medical Products. The device allows the clinician to use the processor from the personal computer onsite to do the EKG monitoring and converts the signals into a digital signal that can be processed and either downloaded via the Internet to NEMC as the remote site, or printed out and faxed.

That particular equipment installation was extremely fortuitous, Carlin notes. One week later a patient case came in that would test the viability of the remote-ED system. "We had a 46-year-old American worker present with chest pain," he recalls. "His family had a history of heart disease. We would have activated the emergency evacuation protocol had we not had the EKG device installed. Because it had been, we were able to do EKGs on the worker in the field. Dornbusch put him on IV and nitroglycerin and started faxing me the EKGs from Ghana, and essentially acting like an ER doctor onsite."

Dornbusch could have sent the EKG readings via the secure Internet connection, Carlin says, but the smallish bandwidth would have slowed the speed needed. As it happened, all the EKGs looked normal and the patient was put on beta blockers to slow his heart rate. A presumptive diagnosis of esophageal spasms was made and the patient was flown on a regular commercial flight to NEMC for a full cardiac workup.

Since then, the worker has been on Prilosec and is doing well. Not incidentally, having the equipment, nurse practitioner and consultative arrangement in place saved the employer between $60,000 and $70,000 over the cost of immediate medical evacuation, Carlin notes. And, more importantly, the test case proved the viability of the telemedicine program.

Stone and Webster is in negotiations with NEMC to establish two more linkages of the type now up in Ghana, though Carlin says he can't reveal where in the world those will be. But for Carlin, the success of the clinic so far speaks for itself. To date the clinic has seen more than 5,000 patients and done over 300 teleconsults, with Internet-support telecommunication of images in 25 cases, and has begun to do simple teleradiology. Carlin and Dornbusch, with the support of specialist consults from NEMC, have taken care of malaria, dysentery, common flu and hand injury cases, among others, for the Ghana-based workers and their families.

What's more, the email setup is becoming more sophisticated. The current arrangement is that an email from Dornbusch in Ghana arrives as an email message not only to Carlin, but also to a network of appropriate physicians at NEMC. The email arrives on a server with a specific identification number and Carlin makes the decision as to which specialist(s) to send it to for further consult as needed. The specialist is simply requested to point his or her browser to the URL and use his password to penetrate the firewall, read the consult message, to which a medical history and possibly a bitmap or jpeg graphic, sound or video file has been attached. The specialist can type into a box on the screen or dictate directly by voice into the system. The voice message can be listened to by another clinician at any time.

And while it may sound exotic to hear of telemergency connections between Ghana and Boston, the reality, says Carlin, is that the same kinds of connections could be made between any two points in the health care world. It could happen between an ambulance on the plains of North Dakota and a tertiary care center in Minnesota, or even between two sites in the same land-based health system. In the end, he says, what's most important should also be the clearest. "You have to fundamentally decide what your medical mission is and what its parameters are, and then support that mission technologically. And you should keep it simple. We're successful because we're using phone, fax and the Internet, and everybody knows how to use those tools, particularly doctors. If you're going to require four or five parties to work together, give them a common tool, and chances are it's going to be successful."

Time to move forward on all fronts

Given the immense potential for automation of the ED, ICU and surgery areas in patient care facilities, a person not familiar with the situation down on the ground might assume that such automation has already made strong headway in all three areas. That is not the case at this point in time.

In fact, implementations remain oddly limited, notes First Consulting Group's Sly. "In the ED, most organizations have patient-tracking or patient-locator systems, and automated patient-education materials for when people leave the ED," she says. "But as far as gathering clinical data while people are in the ED, that area remains rather lean. Of the three clinical areas, the one with the most penetration in terms of clinical content to date is probably the critical care unit or ICU, as opposed to surgery or the ED.

"The challenge in the ED," says Sly, "is that you have a high-acuity patient with intensive care needs, and that, combined with the volume of information needed to support care, means that it's been hard to find an expeditious data entry mode.

"They've done more work in the ED setting than most with voice-recognition tools," she continues, "but the technology is still not there. Even if you have voice recognition up to a 96-97 percent accuracy level, that three to four percent inaccuracy level is not tolerable in the health care environment.

"Meanwhile, you're busy delivering care in the ED, and your hands are tied up and there's no way to get the data entered if you're the one entering it," says Sly. "Clinicians are still the ones who have to make sure data is being entered correctly, even if it's via an automated system for certain pieces of equipment."

Sly says that integration of data remains the key challenge going forward in all three settings, and that means both integration of data from automated devices, and more broadly, integration of data across systems and provider organizations. "I'm working at a place now," she says, "where respiratory therapy keeps all their information on a clipboard, not even accessible to other disciplines. In addition, we need to talk about the documentation of clinical outcomes following the implementation of standards and protocols of care, and automation has a great potential to assist in that process. In the critical care environment, improving efficiency of care is very important, both in terms of decreasing overall health care costs, since those are some of the most expensive environments, as well as improving outcomes over time via feedback from outcomes systems."

In the process, says Sly, it will be crucial to integrate information systems across the continuum and to be able to "scale up and scale down" across systems for data intensity while "expediting data entry in those environments in which people don't have a lot of time or free hands."

Making progress on those fronts, Schneider says, will require a far higher level of clinician input into IT product design and development, but such an investment of time and energy will be worth it. "All of this," he concludes, "speaks to the value of having vital information at your fingertips. It doesn't have to be about rare diseases, either. It could be about finding out about someone's hemoglobin level or pregnancy test when they come into the pre-operative holding area, so that you don't waste hours waiting for valuable information."

In the end, he adds, "I think usability will be critical. The users have to see that the vendors are making a real effort to see how these intensive patient care areas work. And if they see that the vendors are making an effort to do this, clinicians will be willing to adjust their clinical processes to accommodate the information Systems they need to improve care processes."

Mark Hagland is a Chicago-based independent journalist and public speaker in health care.
COPYRIGHT 1998 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1998 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Industry Trend or Event
Publication:Health Management Technology
Date:Dec 1, 1998
Previous Article:Emerging technologies: Has their time come?
Next Article:Evaluating IT investments.

Related Articles
The Long Haul.
Intervenability: another measure of health risk: by coupling predictive modeling with evidence-based medicine, health plans can identify patients who...
Trends in POCT.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters