TRADITIONALLY, MANAGED CARE administrators focused on retrospective data analysis--using utilization, cost and claims data to see how their organizations are doing financially and how they stack up against other organizations. While this form of decision-making still exists, there is a second type of decision support tool that also could help payors improve quality and keep an eye on costs. Instead of using data collected after the patient goes home, some vendors are creating software that allows physicians to make real-time decisions at the point of care.
It remains unclear whether a managed care organization can directly provide real-time decision support to providers (that’s probably the responsibility of physician groups), says Cynthia Burghard, a research analyst with GartnerGroup, Stamford, Conn. But by helping to improve the quality of care and outcomes, real-time decision support could benefit the payor community almost as much as retrospective analysis.
However, deciding which type of decision support system is best suited to an organization’s needs can be challenging. David Aquilina, vice president of marketing for Object Products, Minneapolis, offers this advice: "You have to ask a couple of questions: Do you want a solution for retrospective analysis and reporting or do you want real-time decision support?" He continues, "Then ask yourself whose decisions are you seeking to support [external medical managers, case managers, medical directors] or physicians at the point of care?"
Two classes of vendors provide decision support based on past data: those whose primary purpose is decision making and those based in a niche area (lab systems, for example) that assist in the decision-making process, Burghard says. She believes that users should look at four capabilities when choosing which retrospective decision support system to purchase.
Database design. Decision support systems based on retrospective analysis must begin with clean, edited and mapped data. Systems also must be able to pull together data from disparate systems.
In the case of Performance Measurement Systems, Neptune, N.J., the system gathers previously reported claims data out of QuadraMed’s (the parent company) EDI database, says Guy Evans, vice president of Performance Measurement Systems.
If an organization uses QuadraMed’s EDI capabilities to report claims, the data is automatically cleaned, edited and formatted to address decision support queries--including severity adjustment.
Data enhancement. Detailed data must be rolled up into meaningful units of analysis that include medical services, clinical conditions, surgical procedures and definitions of different populations.
For example, the best way to examine utilization and other cost considerations is to develop custom queries for the software. "You want to be able to look at where your utilization is today and where it will be in the future," says Lane Nelson, marketing director for Pilot Software, Cambridge, Mass. "That way you’ll know where those problems are going to be so that you can be prepared."
External data for analysis. Benchmarking data internally and among similar, outside organizations is a key to data analysis.
Reporting capabilities.Retrospective decision support systems must provide pre-formatted reports and graphics and still allow for ad hoc reporting capabilities.
"Some decision support systems are what you call ’canned systems.’ You only get out of those systems the reports that are put in," explains David Girolamo, managed care coordinator at St. Peter’s Medical Center, New Jersey. "But there’s things that you need that are completely out of the realm of possibility." He says systems such as Performance Measurement System’s QuanTIM provide that flexibility.
Although real-time decision support software is an "embryonic industry," according to Burghard, Tom Handler, senior research analyst at GartnerGroup, says it’s not even that large yet.
Recently, Handler began investigating real-time clinical decision support--something he says will prosper as a result of the growing importance of computer-based patient records. "I believe very strongly that point-of-care decision support is very important to these systems," he says.
Object Products is working on tying real-time decision support to physician practice management systems--an area that could directly affect managed care organizations. "The prevalent emphasis is still on solutions for aggregate data that come from the managed care organization. That’s the source of basic information a managed care organization is going to use for a lot of its reporting needs," Aquilina says. "HEDIS has started to provide emphasis for managed care organizations to move beyond financial and claims data. Some of the measures in HEDIS require clinical data."
Handler agrees that managed care administrators need to think in terms of clinical data and actual patient care. "Our vision [at GartnerGroup] says that managed care will have to interact with the patient record. Clinical decision support should reside in the clinical systems.
Aquilina offers three system capabilities for users choosing a real-time clinical decision support tool:
Real-time use of patient-reported and clinical outcomes data. By synthesizing outcomes data, patient-reported data and good analytical tools, physicians will be able to query a clinical information system at the point of care and find out immediately what course of action will yield the best outcome for a specific patient.
Clinical guidelines. "If administrators have some clinical activities going on [on their decision support systems], clinical guidelines allow physicians to start to see how their behavior patterns impact the organization," says Alec Karys, vice president of software development and technology services at InterQual, Marlborough, Mass. "If you’re strictly looking at the financial data you’re missing part of the puzzle."
InterQual recently signed agreements with Rothenberg Associates (part of QuadraMed), Woodland Hills; Health Systems Technologies, Seattle; and OAO Healthcare Systems, Chatsworth, Calif., to integrate clinical guidelines into these vendors’ financially-based systems.
Watch duplicate data entry. Data results must automatically drive clinical decision support. These systems must not result in extra work for the physicians or they will not use them.
Handler agrees that physician buy-in is critical to getting real-time clinical decision support systems off the ground: "Until you get the physicians actively using the systems, they will continue taking notes manually and then the clerks are really the ones using the system [for data entry]."
Lisa Paul is senior editor at Healthcare Informatics.