College of American Pathologists
Printable Version

  Feature Story


cap today

New tools for netting POC connectivity
Spectrum of connectibility

February 2002
Anne Paxton

If it were easy, anyone could do it. Perhaps point-of-care testing coordinators who are trying to implement POC connectivity can take comfort in that thought. For despite the rampant enthusiasm about linking instruments with data-management systems, even the vendors concede that the course to connectivity runs anything but smoothly. Glitches, kinks, snafus, and snags go with the territory.

A much-anticipated step to ease the way was taken last November, when NCCLS approved an industry-wide, vendor-neutral standard, including standard messages, protocols, and technologies, that will permit POC instruments made by different manufacturers to talk to laboratory and hospital information systems.

Under the standard, "new devices should seamlessly link into your existing data management system without additional expense" for wiring, computers, or software, if the vendor is CIC- or NCCLS POCT1-A-compliant, according to the Connectivity Industry Consortium, which formed, hammered out the standard, and dissolved itself all within 14 months.

Christopher Fetters, president and founder of point-of-care consulting firm Nextivity, in York, Pa., served as secretary to the CIC, and is pleased that the consortium was able to accomplish so much. "We never intended to have a 100 percent solution—but I never thought we’d get this close."

"I think it will move the whole industry," says Jay B. Jones, PhD, director of regional laboratories for Geisinger Medical Center, part of a large rural health maintenance organization in Danville, Pa. Some software vendors, in fact, are already advertising their products as "CIC-compliant."

Why was the connectivity consortium important? Until it was set up, the companies invested in POC data-management systems, and their customers, could only assume that their separate proprietary standards would continue to compete, and continue to be essentially incompatible.

Abbott Laboratories, Roche Diagnostics, and Johnson & Johnson all developed proprietary connectivity platforms that would communicate their own glucose instruments’ data to an LIS, but in the last year Abbott and Roche began offering "open architecture" that would handle non-glucose POC tests as well as other vendors’ devices. Two smaller vendors, Medical Automation Systems and Telcor Inc., do not sell devices or disposables for testing and advertise their platform as completely vendor-neutral.

In CAP TODAY interviews, customers of these five POC connectivity companies discussed the benefits they gained and pitfalls they encountered in implementing a data-management system for their POC instruments.

Stepping up from "sneaker-net"

The two hospitals of the Via Christi Regional Medical Center in Wichita, Kan., started wiring their DataCare system about three years ago and piloted the system for manufacturer Roche Diagnostics in the fall of 2000. But the laboratory, which has 120 meters performing 800 to 1,000 glucoses a day, decided to postpone the actual startup, says Carolyn Strunk, MT(ASCP), point-of-care coordinator.

"When we first started looking at systems, nurses would have to connect a cable from the PC to the glucose meter, power up the glucose meter, perform a couple of keystrokes, tell the meter to download, and then clear the meter’s memory, which was password-protected, or we’d have to go up and download each meter ourselves because the meter would only store 1,000 results. We were concerned," Strunk says, "that at this time the system did not have very many levels of protection, so we decided to postpone installation until we could obtain a system that was less ’hands-on.’ Now the downloading port is also a recharger for the meter’s battery; when the meter is placed in the downloading/recharger port, everything transfers automatically."

The DataCare system allows the laboratory to see when meters were last downloaded. "We set a time frame of 24 hours, and every morning we check this function to know which meters haven’t been downloaded," Strunk explains. Although she tries to remain visible on the units, she is happy not to have to rely on the "sneaker-net" to find an instrument that was left in a patient’s room. "It’s made my life a lot easier in finding instruments. Now if I have a meter that hasn’t been downloaded, we’re on it in 24 hours, or 48 on the weekends."

Their test volume has not increased since DataCare was installed, she says, but there has been a big change in the number of tests billed. Previously, "our orders would come down on a triplicate form. We usually received the second copy. Sometimes they had placed another sheet on top of the form and began writing, and we didn’t know if these were stray marks from the carbon. Unfortunately, we were underbilling because the forms weren’t always legible. We also knew we were underbilling because when the patient was dismissed, the forms that had not been sent to the laboratory went to medical records; then they were routed to us and we may not have received these for several days. This was compromising our billing process."

A side benefit is the ability to track testing by operator in a timely manner. "What we have found out—and we never had a true handle on—was the amount of testing by employees on themselves. We do give them a little freedom to do that, but now we can see how much is performed, and we just go to these operators and educate them."

Several of the instruments in the POC program are non-Roche, she says. "Respiratory care and the laboratory partnered up to purchase the connectivity system. One reason we purchased our blood gas instruments was the availability of the connectivity system. But we have instruments that are not produced by the connectivity manufacturer. We also have a couple of instruments that need to be replaced because of their age and support. I will evaluate those instruments produced by the manufacturer of my connectivity system, but because my system is an open system, I have the flexibility to choose the product that best meets my facility’s needs."

Influencing instrument choice

The system at Henry Ford Hospital in Detroit, the RALS-Plus, was put in place two months ago in conjunction with a new meter, Roche’s Inform, that has more than 200 download locations across four hospitals. While RALS-Plus is not interfaced with the hospital mainframe, the interface with the LIS is being tested now. "Once that’s complete, we can go ahead with connectivity to the LIS," said Karen Bourlier, MS, MT(ASCP), alternate-site testing coordinator.

She expects that the RALS-Plus, made by Medical Automation Systems, will progressively eliminate the need for other connectivity solutions, as her hospitals upgrade their devices. "As long as your data-management system can hook up to a variety of vendors, you need to run off one main system. It’s absolutely critical, because if you had to check in five different desktop systems to make your system run appropriately, there’s no way."

"Connectivity is now driving a much larger part of the choice process," she adds. "RALS is new to us, but any systems we bring up we will say, hey, we’ve got RALS now, we can maybe move this higher up the food chain."

To her, being able to track volume for the first time is one of the chief attractions of connectivity. "We know what supplies we use, but we find that’s a really squirrelly number. People hoard supplies, they order too many based on volume, or supplies disappear."

"I had a chairman who always used to tell me, consultants say you can cut costs by cutting volume on your glucose tests. I’d say that’s probably true, but I can’t give you a testing volume." With more precise tracking, she hopes the laboratory can develop protocols with caregivers that might, for example, provide for stopping glucose tests on patients who are stabilized and normal after two or three days.

So far, Bourlier says, there has been little progress in getting non-instrumented results recorded. "The companies come at you and say you can enter results in a desktop computer. But a nurse is never going to do that. They’re used to manual documentation, and the only way it will happen is when they either have a bedside solution, like a bedside keyboard, or a handheld device they can pop into a download station."

The Henry Ford transition was particularly stressful because bar-coded armbands and glucose meters were brought up at the same time. "We’re having a lot of problems at my institution with bar-coded armbands. They’re not working out the way we thought. If we don’t put in the correct information, the results won’t go to the LIS. They will sit there until I personally go to the workstation and attempt to resolve the problem," Bourlier says.

What kind of problems are there? "If there is any dirt or solution on them they won’t work; if they’re not flat, the laser won’t read them. People were not used to using the bar-code wands and would say they didn’t work. If you try to scan too close to the band, it won’t read; you have to scan straight across about six to eight inches away. If any iodine or alcohol drips on them, they’re no good. You can only print the bar code so small before it becomes unreadable." She adds, only half-jokingly, "What we really need is an identification chip implanted in our arm."

"The other day, only 30 percent of patient armbands were being scanned," she says. "I could tell from the way the patient records were coming across. Have the nurses given up? Are the armbands not on the patient? I don’t know, but it means that 70 percent of the POC glucose results are going to fail to get into the system. And they have a chance of putting in a wrong number that actually belongs to another patient. So we need to stabilize it."

The complaint level, however, has not been as high as she expected. "The nurses have been very patient because they love the glucose meter. They like the ease of use. It’s easier to get the patient sample on the strip, it’s more stable, there’s a rechargeable battery. They’re just so thrilled with the step-up in technology that they’re not screaming, because I would expect them to scream having to punch in a 12-digit number."

But, before upgrading POC technology, "I would highly recommend having an operator identification system and a patient identification system in place and very stable."

Integration’s benefits

With a reputation as one of the best-wired health systems in the country, Geisinger Medical Center has managed to dodge some of the more routine difficulties in adjusting to new technology. Its 600 salaried physicians, for example, know it is a condition of employment that they don’t order a nurse to order a test like PSA and handle the paperwork; for outpatients they can order the test themselves on an EpicCare terminal, and in the near future they’ll be able to do the same for inpatients.

"We’ve gone paperless, in essence," says Dr. Jones. "The physician has to click on a screen and go through the interface with the computer that includes inputting ICD-9 codes for medical necessity, and doing what is viewed as a clerical task. It created consternation at the beginning. There was the typical grousing and foot-dragging." But since they can now go to any terminal and call up patients’ electronic medical records, "they have instantaneous access, and they can’t live without it."

Geisinger, a large diffuse medical practice that is highly integrated, includes two medical centers and 60 practice sites, and thinks of itself as a "mini-Mayo Clinic." For the last three or four years, Geisinger has used a POC workstation now called Quick Multi Link, or QML, developed by Telcor to directly interface its LifeScan glucose meters to the Sunquest LIS.

The LifeScan connectivity has been live since 1998 and there is a whole line of instruments Geisinger hopes eventually to connect to the Telcor workstations, Dr. Jones says. "We’re connecting the Medtronics ACT devices through Telcor, and we’re looking at connecting HemoCue through QML, too." i-Stat devices are likely to be consolidated under QML within a year, and down the line are connectivity plans for other POC devices including prothrombin time instruments.

"What will really impact the laboratory is the complexity and variety of connectivity, not so much the sheer volume," says Dr. Jones, who estimates he’s gone from 60 technicians and technologists in the chemistry laboratory to 3,000 testing personnel. "It will be a skill set that laboratories have to get used to: dealing with complex issues of integration at the

Nailing down the numbers

James Aguanno, PhD, director of the core laboratory and director of point-of-care services for Baylor Health Care System in Dallas, says connectivity was one of the main considerations when his system chose Abbott’s Precision Net. It was implemented across six hospitals for glucose in 1999, and the handheld i-Stat was added in 2001.

His hospital system does more than a million POC tests per year, including 1,000 glucoses a day run on 120 meters. Deficiencies found in a Joint Commission inspection led the hospital to look for "solutions to clean up our point-of-care testing," he says. And connectivity was the deciding factor among the different vendors.

"Here in this hospital alone, we have 1,500 users of glucose and i-Stat, so we have a lot of volume." But the actual figures were vague until the data-management system was implemented. "Right off the bat, we were able to get accurate volumes on what we were doing. Previous to Abbott, we knew how many strips we were buying, but we had no idea how many we were actually using."

"This is a big hospital—a thousand beds spread out over more than a city block, and one feature I thought was an absolute necessity is bidirectional interface. I thought if we’re going forward in the future, my point-of-care coordinator should never have to touch a glucose meter out on the floor," Dr. Aguanno says. "If we have to set up new users, put new lot numbers of strips in place, or we have software updates, that should all be able to be performed through the network."

"Running around this place is very unproductive time. You’ve got to manage things centrally; otherwise you’re dead. In a small hospital, you can kind of brute-force some things. But here, if you’re not automated, point-of-care will swamp you."

Since Precision Net was installed, "we’ve been through two Joint Commission and two CAP inspections without a single deficiency in point-of-care testing." But he’s also happy about the improvement in record-keeping. "There were lots of results not being recorded before," Dr. Aguanno says. "That was a big issue with the physicians. They felt fairly vulnerable because they were potentially writing orders against values that never really existed in anything. That always made them nervous. Now all those values are in our data manager."

In his view, bringing in a third-party network solution can be problematic. "That makes it now a four-way problem, because you have the information systems hospital network, you have the customer, the vendor, and this third-party group who may have supplied a networking solution that is not a workable solution. You wind up with a lot of finger-pointing, as always. The vendor’s going to say, ’my part’s fine,’ and you’ll never get anywhere."

"We’re not exactly where we want to be," he says. "We did the LIS interface first at one of our smaller hospitals. We haven’t put it into operation here yet. I did that because the smaller place is a lot more manageable, and I wanted to make sure it all works well there. But we’ll be installing it over here in the big hospital fairly soon."

More workstation consolidation is what the laboratory world needs, he notes. "If you look at POC testing today, you’ll find lots of instruments and tests, but essentially it’s one test/one instrument." He finds the prospect of training 3,000 operators on a new instrument a nightmare. By contrast, he says, "When creatinine came out on the i-Stat in 1999, we had several places around the hospital that needed it, and it meant nothing more than getting a cartridge and running it. Once the system is in place and new tests come out, it’s very, very easy to add new tests."

The bonus of billing

Providence Saint Joseph Medical Center in Burbank, Calif., chose the LifeScan DataLink POC Solution, made by Johnson & Johnson "That data-management system, which has been up about two years, handles patient results and quality control and is now interfaced to our LIS, which is Misys [Sunquest]," says Mark Barglowski, MT(ASCP), MBA, director of laboratory and respiratory care services. The nurses at Providence Saint Joseph are happy with the system, which has meant reduced time tracking quality control and less time transferring results, because the glucose meters download through a docking mechanism.

The system has also made billing progressively easier. "We just instituted billing for glucose last month. And I’m working on a second site, Providence Holy Cross Medical Center. As soon as I get billing set up there then I’ll pursue urine dipsticks, ACTs, and occult bloods at both sites. But I figured there are more dollars out there for glucose meter testing than for other low-volume tests, in terms of billing."

However, like most other connectivity implementations, Providence Saint Joseph hasn’t been problem-free. "We instituted bar-code stickers for operator ID, but it hasn’t always proved effective. The reasons could be anything from the bar codes aren’t picked up well by the meters themselves, or they are worn out very easily on the back side of the nurses’ identification badge. Where I’ve heard they actually work is when the bar code is part of the identification badge for hospital staff." As for patient identification bar codes, "That’s something I’d like to pursue with our hospital," he says.

Barglowski welcomes the CIC connectivity standard. "We’ve followed the whole process since its inception," he says, noting that connectivity will let him choose a "best of breed" instrument and not be tied to one particular vendor or data-management system. In addition to the LifeScan glucose meters, "we have Quidel for our POC pregnancy, Hemochron for our ACTs, and our dipsticks are Bayer Multistick, so we have quite a few vendors."

But the standard’s impact hinges on where hospitals are in the instrument selection process, as well as how well vendors’ data-management systems function to pick up nonautomated POC, he points out. "That would be a real selling point, because you have to be able to show a financial advantage in today’s hospital marketplace to get approval right now."

"People are getting to the point where they’re looking at connectivity as the means to rationalize the cost of a data-management system. So many of our peers, at least in the Los Angeles area, are not billing for POC, and I think that’s the next step in the evolution. We’ll see more people billing for POC, and they’re only able to do that because their data-management systems are interfaced to the LIS."

"That’s where you justify the cost," he says. "Initially, POC was put in place and the justification was better patient outcomes as well as better financial outcomes, with heavy emphasis on patient outcomes." But once data management is implemented, the financial savings become more appealing. "Hospital administrators can look at the reduction of time spent downloading meters and collecting quality control data," Barglowski says. Eventually, he hopes POC data will be used to produce long-term trending, especially in glucose, and perhaps with good diabetes management affect the outcomes of patient care.

Is wireless in your future?

"In the next five years," Baylor’s Dr. Aguanno forecasts, "more and more tests that are now visual will go to instruments. I know Abbott has plans to allow entry of results from non-instrument-based tests. Even though the test won’t be performed on their instrument, they will add the capability of entering it on their platform and transmitting it through their network," Dr. Aguanno says. "It’s clear if you talk to any customer, if the test is not on an instrument, it’s so difficult to manage that once given an option, we’re going to head that way—especially if that instrument is one we’re already using like an i-Stat."

But the CIC standard stops short of making wireless connectivity uniform. "We didn’t address true radio-frequency (RF) wireless," Fetters explains. "We did ’wireless,’ in that the standard accommodates infrared, or what we call ’point and squirt.’ You can walk up and download data, so it’s technically wireless, but not in the true sense where you could download from any place on the floor."

"There’s the option for continuously connected devices, and even within the standard you could implement wireless connectivity," he adds. "But the devices would have to be a lot smarter and more expensive, and they wouldn’t necessarily be standard. In other words, one vendor could do one type of wireless connectivity, another could do another, and it leaves users in kind of a bind."

The consortium was reluctant to push one wireless standard out in favor of another, Fetters notes. "There are a number of different competing wireless technologies in the clinical space, but there was really no clear winner," Fetters says. Other obstacles, too, will hinder wireless, at least over the near term, says Geisinger’s Dr. Jones. Despite the extensive electronic integration at his medical center, all of its connectivity is wired. "We’re holding back on wireless because we need to be part of a very broad program to do it. The laboratory alone can’t afford everything needed in telemetry to do our own thing with wireless; we will have to wait until the hospital has multiple telemetry devices."

The prospects of seeing existing instruments retrofitted to become CIC-compliant are slim, Fetters believes. "In a year and a half or two years, all the manufacturers should have devices that are compliant," he predicts. The ideal, in his view, would be for them to update their firmware and make all their current devices CIC-compliant. "We specifically wrote the standard so legacy devices could be brought up to speed. But I don’t think they’ll do it. They’re always into doing the next new best thing."

How realistic a goal is "plug-and-play" capability?

"I really think we will have plug-and-play if the manufacturers are smart and really adhere to the stuff you can do with the standard," Fetters says. "The problem before was that when you connected a device to what we called an access point—a terminal server that lets the device download through the network—you had to tell the receiving software where the port was and what was going to be coming down. The infrared transfer section of the CIC connectivity standard is the same one that PDAs, laptops, and printers use to send data back and forth. Built into the first communication greeting is an announcement of what device you are, the version you are, and where you’re downloading from."

He would like to see the plug-and-play technology commonly seen for computer peripherals adapted for POC instruments, usually including a compact disc with drivers. "If the POC device manufacturers can include drivers for a ubiquitous system like Windows, then the computer doesn’t inherently need to know how to run the instrument; device manufacturers supply a CD with all the information for the data-management system to interpret the data that goes back and forth," Fetters says.

For the time being, Deanna Bogner, MS, MT(ASCP) says, there are quite a few material obstacles in the way, in her opinion. As point-of-care testing coordinator for Christus Santa Rosa Health Care in San Antonio, she praises the work of the CIC, but wonders how much help the standard will be for POC coordinators. "I have three instruments capable of downloads and all of them use a separate and distinct system to transfer the data," she says. "I have three phone lines in some areas just on the possibility that all instruments will eventually be able to transfer into one computer. The CIC allows for that, but what about a universal way of transferring the data to the computer system that the coordinator uses to monitor instrumentation?" If one cradle size fit all instruments, she argues, it would save space on the units as well as money spent on phone drops and cabling.

Dr. Jones is skeptical: "I have enough experience to see it as ’plug-and-work’" as opposed to plug-and-play. "It has to be integrated, you need close-to-real-time data feeds, you have to implement billing systems, you need to train operators. It’s somewhat misleading to say it will be simple and a panacea. There will be an evolutionary process of people ’plugging and working’ for a couple of years."

He warns that laboratories could become "disintermediated."

"When the laboratory is no longer the intermediary of testing results, that’s a bad thing," Dr. Jones says. "We’re sitting in our factory-centralized core laboratory, and tests are being done at the point-of-care we don’t know about. You could have diabetologists running a whole disease-management program with very little participation of the laboratory. The results are being stored at somebody’s Internet-site data warehouse. Over time, there’s a
real threat."

Instead, he believes laboratories can view point-of-care testing as the most active growth point for what is being called the "distributed laboratory"—a single site that owns all the CLIA licenses across the system and integrates remote test results, whether from hospital floors, clinics, or, increasingly, private homes, into the same LIS. As other health systems move steadily closer to an entirely electronic medical record, he predicts, "I think it’s a model that’s going to emerge more and more."

Anne Paxton is a writer in Seattle.

To order a copy of the universal standard for POC connectivity (under order code POCT1-A, Point of Care Connectivity: Approved Standard), visit the NCCLS Web site at