Clinical decision support: Can hospitals afford to do without?

A lack of money may be costing hospitals money

Clinical decision support systems can improve the quality of care, reduce costs, and improve patient satisfaction. Sounds great. But if that’s so, then why aren’t more health care facilities using them?

"The big health systems are generally not buying anything, not even the clinical systems," says Jane Metzger, vice president of First Consulting Group, a Lexington, MA, provider of information-based consulting and integration and management services.

"The reality is that at this point everyone is talking about this and is looking into it, but no one is buying, renting, or subscribing to the systems," says Alan Portela, vice president of sales and marketing at Clinteligent, an application service provider and distribution arm of parent company CliniComp in San Diego. "The market is slow because people are worrying about the Balanced Budget Act of 1997 and about HIPAA [the Health Insurance Portability and Accountability Act of 1996]. We have CIOs who are interested and have good ideas on what they are going to implement, but they are moving slowly."

Either way, you pay

Although technology has become less expensive and more flexible, the industry is in the midst of one of the worst cost/profitability crises it has seen, Metzger says. "That makes it harder to come up with funds. At the same time, some of the industry pressures on quality and errors are mounting. How it all will play out is difficult to predict."

One problem for providers is that many have various definitions of what clinical decision support (CDS) is.

"Some people refer to outcomes measurement. Some say disease management. Other people just talk about good access to clinical data," Portela says.

The award-winning article "A Practical Framework for Understanding Clinical Decision Support," which Metzger co-wrote, classifies CDS systems in four general categories. These categories are:

Access to information: This is the most basic aspect of decision support and provides easy access to general or patient-specific clinical data and information. "CDS capabilities in this category are aimed at delivering the right information at the right time and place to enable information clinical decision making," the authors say.

"It is essential that hospitals start automating their clinical documentation, at least at this first level," Portela says.

Guided choice: Provides a second level of support for clinical decision making. CDS capabilities in this category are designed to make it easy for care providers to make the right choices among available options.

"First you have to have the clinical documentation," Portela explains. "Then you start throwing in all the software modules that help you with the guided choices."

Knowledge-based prompting: This level of CDS is designed to assist clinicians in making the right care decisions. "These CDS tools help to determine the correct clinical diagnosis in planning therapy or by alerting the clinician to relevant important information," the authors say.

Understanding clinical practice: CDS capabilities in this category are aimed at achieving overall improvement in care delivery. These systems are population-focused and are used retrospectively to identify patterns and trends, which in turn can be used to guide future decisions, identify current best practices, and evaluate and refine clinical guidelines, the authors say.

Most hospitals were too busy last year dealing with Y2K issues to focus on CDS, Portela says. "They only implemented core functionality — ancillary systems, financial systems, and order entry systems. Now they are all trying to get to the decision support level right away. Without doing the first clinical documentation phases, they are not going to be able to analyze the data at the level they want."

Almost everyone will play catch-up

Only about 10% of the hospitals and physician groups have their clinical data automated, Portela says. By the time others catch up, it may be too late.

"It will take CIOs an average of 12 to 18 months to find the [clinical decision support] products and install them," he says. "That will be too late. They will not be able to compete."

Application service providers (ASPs) offer one option that allows health care facilities to outsource the automation.

"ASP models offer an opportunity to have someone else other than the hospital take care of its information technology," explains Bruce Fried, JD, partner at the Washington, DC-based law firm Shaw Pittman and chair of its health law group. "For many community hospitals, an outsourcing ASP strategy makes sense, both in terms of economics and the challenge to keep technicians and up-to-date systems."

ASP allows people to implement systems quickly without going through those long selection processes, Portela says. "It also allows the CIO to throw the risk of failure to the vendor."

An outside company, for example, can offer a clinical data repository (CDR) via remote connectivity.

"In essence, all your data will be [in the CDR], and the vendor will start applying third-party products on top of the database to do the analysis — the clinical decision support, outcomes measurement, disease management," Portela explains. "Not only is the vendor going to look retrospectively at the data that you have within your institution, but with the hospital’s permission, the vendor will be able to provide comparative data of how you are performing compared to other institutions."

The ASP model, however, just addresses some of the upfront technology cost, Metzger says. It doesn’t solve the whole problem.

"You still have to make a lot of change in the practice. You still have to train physicians and you still have to give them time to adjust."

Since many providers are challenged with the first phase of clinical decision support, Portela recommends they first begin automating the clinical areas that are high in data collection, such as the intensive care units, the emergency departments, and the labor and delivery departments. Then they or an outside vendor should collect the data from those units and from areas such as labs and radiology. Once those data are in one central, clinical repository, they can start adding to the data repository to start providing the type of reports that they need. "After you do those basic steps, the rest is not that hard."