April 2008
Feature Story

Karen Wagner

It began 10 months ago with a team of approximately 50 employees, from which was drawn a working group of about 25 laboratorians and administrators, who selected 20 candidates, from whom they picked two finalists. Now Henry Ford Health System is about to choose between those two finalists in its nearly year-long search for document-management software.

This, however, is not a tale of winners and losers, but rather a look at the process that one health care system has gone through to find the software program best suited to its needs.

Pathologist J. Mark Tuthill, MD, and fellow employees at Henry Ford Health System, Detroit, have worked methodically since last June to find software to create, edit, publish, and electronically distribute laboratory policy and procedure manuals across the health care system.

“That was a relatively tall order, but we were able to meet most of those needs,” says Dr. Tuthill, division head for pathology informatics at Henry Ford, which includes a main laboratory, 27 ancillary lab sites, three hospitals, and more than 600 laboratory employees.

Dr. Tuthill began his software search by assembling a team to create a wish list of what laboratory staff wanted in a document-management system. The initial team, he says, consisted of about 50 employees and included physicians, medical technologists, administrators, and laboratory division heads and supervisors.

One wish was for more consistency in the policy and procedure manuals so they would be easier to understand and navigate, Dr. Tut­hill says. Each of Henry Ford’s manuals differs slightly from the others because they were developed at the individual lab level. Supervisors also wanted to be able to acknowledge, through an electronic signature, that employees read the manuals, Dr. Tuthill says. “People were really frustrated with the need for manual sign-off, particularly in our medical centers, where the same policies and procedures are used in 27 different sites and had to get signed 27 different times by four people,” he explains. “Personally, one day I remember signing 400 procedures and policies.”

It is also important for the manuals to be available electronically, in part, so medical personnel outside the lab can access them, Dr. Tuthill says. He cites as an example nurses who conduct point-of-care lab tests and therefore need to be aware of laboratory policies and procedures.

Dr. Tuthill and a working group of about 25 laboratorians and admini­strators, drawn from the larger team of 50, initially identified 20 software vendor candidates, primarily through online ­research. The group sought recommendations via pathology and laboratory medicine electronic mailing lists, including those operated by MedLab and the Association for Pathology Informatics.

“A leadership team of four individuals did much of the analysis and presented the information back to the work group,” Dr. Tuthill says. Work group members also offered feedback on select software demonstrations they attended. And Dr. Tuthill gathered additional research about a few software products from colleagues familiar with the systems.

Choosing the best terminology was one of the group’s first challenges, Dr. Tuthill recalls. An online search for “document management” will turn up hundreds of software applications, “90 percent of which have nothing to do with what a lab is trying to do—manage the life cycle of a document,” he explains.

The team had better luck with the phrase “content management.” Document management, Dr. Tuthill says, includes simple systems that will scan a document and then allow it to be saved electronically. The term does not necessarily describe a system that has more sophisticated functionality, such as workflow management, publishing capability, or electronic sign-off, he adds.

After narrowing its search to 20 vendors, the work group further researched the candidates’ products online and then trimmed the selection to 12 vendors.

Using the initial wish list, the leadership team put the desired features and functions into a formal spreadsheet and from that spreadsheet created a request for proposal that it mailed to the 12 candidates. The RFP asked for system and company information on about 55 elements, each given a different value, in six major categories, Dr. Tuthill says. At the same time, the working group created a spreadsheet to evaluate the responses. The group eventually presented the results to the larger lab team for its appraisal.

After extensive review, the larger team made several suggestions, among them that the software have an assessment function, something Dr. Tuthill hadn’t considered the last time he produced policy and procedure manuals 10 years earlier. Proving, through something resembling a quiz, that users not only read the manuals, but also understand them, is an important capability, he says. The team also made suggestions about licensing models and electronic signature capabilities.

At that point, Dr. Tuthill says, the working group excluded six of the 12 candidates because they ­didn’t have all the desired functionality, such as assessment capabilities or the ability to publish to the Internet. Then the leadership team viewed software demonstrations by the six finalists and shared its findings with the working group and senior leadership.

“Pretty quickly, we realized that there were only two or three candidates that were truly viable,” Dr. Tuthill says. These vendors had the highest scores on the objective assessments based on the RFP responses. Dr. Tuthill called five references for each of the remaining two finalists and spent about 15 minutes talking to each. “I would highly recommend folks do that—it’s very telling,” he says. The two finalists then made presentations to the working group.

Dr. Tuthill hoped to make his final selection at CAP TODAY press time. The determining factors rested on cost, licensing arrangements, and potential contract issues.

The investment will be between $80,000 and $110,000 for licensing, implementation, and maintenance costs over a five-year period, Dr. Tuthill says. Training costs will add $50,000 to $75,000. “One of the caveats is that there are costs that are hidden, in addition to the cost of the license or the cost of the annual maintenance over five years,” he explains, referring to the cost of database software and hardware, such as a server. Henry Ford’s lab, however, already has the software and hardware necessary to support a document-management system, Dr. Tuthill says.

Whether a laboratory buys a system or has the vendor host it also affects the price. The hosting option would cost more on an annual basis, “but you don’t have to buy hardware and software, and you don’t have to support it and do back-ups because they take care of all that,” Dr. Tuthill says. At Henry Ford, the buy or “rent” decision has yet to be made. “I think our penchant is we’re probably going to host it ourselves, because we can, and it’s not going to cost us any more.”

Deciding whether to use a per-user or per-seat licensing model has proved an easier task. Under the per-user model, every user has a license to use the software—that equates to 600 licenses for 600 users. Under the per-seat model, the buyer purchases a certain number of licenses, say 50, which means that up to 50 employees can use the software at once, Dr. Tuthill explains. If the system is used all day by everyone, the per-seat model is insufficient. But, he adds, a document-management system is used sporadically, so a per-seat model would work and is favored by the Henry Ford working group.

“Per-seat is easier to manage and is typically less costly,” Dr. Tuthill says, adding that while the cost for each license under a per-seat model may be more, the overall cost for the licenses will be less because fewer licenses are needed. Typically, he explains, the per-seat model is set up according to a 10:1 ratio, so 50 licenses should support 500 users.

Lastly, Henry Ford must tackle contractual issues, Dr. Tut­hill says, such as who will be allowed to use the software. “When you get down to the contracting phase,” he adds, “you better have somebody on the back burner because you may find that your frontrunner has some problems at the legal level that you just can’t get by.”

When all is said and done, Dr. Tuthill’s final decision will not center on differences in product features, but on these other factors. “Features and functions almost all look the same,” he says. “Cost is different. Licensure is different. The contract is different. And then, their philosophy of doing business is different.”

After weathering this long process, Dr. Tuthill says there are few elements he would have changed. However, he points out that he initially wasn’t going to use an RFP, but rather pick two vendors that appeared to be a fit and choose one. “I think we could have done that and we probably wouldn’t have suffered,” he says. “I think we may have made our job a little bit more complex. That said, I think we’ve done an extremely thorough job.”

As Dr. Tuthill and his team have been searching for the right software, the lab has been busy preparing for its arrival. Staff have been organizing data and other information in a manner that allows it to be easily input it into a document-management system. “Because if you haven’t got your content in a clean, orderly fashion, you may have an application, but it’s going to be a bear to ... get your content in,” Dr. Tuthill says.

The working group hopes to have the document-management system in place by July, Dr. Tuthill says. Some divisions of the lab should be fully functional with the new system by September, and the entire lab should be up to speed with it by June 2009. “So, all told, I would say this is probably a two-year life cycle in an organization that’s as complex as ours, with as many laboratories and customers as we have to support,” he says.

Overall, Dr. Tuthill has been pleased with the process. “We really did this right,” he says. “We did it like you would have selected an LIS. But you know how it goes in life—the harder you plan, the better off you are.”