Clinical decision support systems key in medical error prevention, quality improvement
ORLANDO—Clinical decision support (CDS) systems are not used to their full effectiveness, however, well-designed CDS improve the overall safety and quality in a clinical setting, according to Jonathan Teich, MD, PhD, from Elsevier Health Sciences and Harvard University in Cambridge, Mass., who addressed a packed audience at the 2008 HIMSS conference.

Teich said that CDS are necessary because one-third of the medication errors are significant, according to the Institute of Medicine. He added that “one of the scariest statistics I’ve ever seen” is that it takes 17 years before the average physician implements new recommendations into his or her practice.

Teich hopes CDS’ will be used to inform and remind physicians about proper and updated information and recommendation.

CDS increases guideline adherence, decreases medication errors, increases monitoring and surveillance, improves patient care, improves appropriate resource utilization and saves money—proving its worth as a proven solution, according to Teich.

He acknowledged that some of the information is confusing because “every month, there is a study that says that CPOEs prevent errors, save lives and money; and in the odd months, new articles or studies emerge saying that they cause errors and hurt patients.”

He exemplified a recent study in the Archives of Internal Medicine, which found pre-CPOE, the hospital reported more than 2 percent of medical errors (dosing), and three years after the CPOE implementation, medication errors had been reduced to 0.2 percent.

However, Teich said that despite their positive results, the conflicting news and studies are presenting a false understanding of CDS.

He said that new CDS is not just only focused on alerts, which Teich defined as often times intrusive.

There are six different types of CDS:
  • Workflow support CDS;
  • Informational CDS: answering your question;
  • Interoperable CDS between workers, applications and devices can provide support in a cross-person and cross-workflow approach;
  • Web 2.0 and other new technologies;
  • The knowledge environment; and
  • Research to practice (R2P) – this is crucial because new studies with recommendations could be downloaded instantly in practices, and instantly linked into the CDS.
For example, a stat order could bring reminders to the attention of the physician at the point of care, according to Teich.

In regards to workflow support, Teich said that CDS can assist task support through information because physicians have been proven to be hesitant to ask questions, but they are more willing to research, if it doesn’t require a great deal of research or time.

“Quick information is better than good information in our culture today,” Teich said. He said this supports having interactive references from such sources as UpToDate Online, MD Consult and Epocrates, at the fingertips of the physicians. Ideally, the physician will be able to obtain that information at some point without the search, and, if the doctor is inputting the disease, the reference would also appear as an option on the screen.

Teich said that “there are about 50 questions that we as doctors have.”

In the future, Teich said that CDS will provide “clinicians or patients with clinical
knowledge and patient-related information, intelligently filtered or presented at appropriate times, to enhance patient care.”

ONCHIT licensed a road map of real-world barriers to nationally facilitated CDS, which found that there is a lack of sharing framework.

The National CDS Roadmap lead to six strategic objectives:
1. Develop practical standard formats for representing CDS knowledge and interventions;
2. Establish standard approaches to organizing and distributing CDS;
* “And, when there is a new recommendation, it can be downloaded into your IS,” Teich said.
3. Develop solutions to policy, legal and financial barriers;
*“If the doctors know they will not get reimbursed for mistakes, they might be more cautious,” he noted.
4. Compile and disseminate best practices for usability and implementation;
5. Develop standards to collect, learn from and share national CDS experience;
* “If we share information about how it works, why can’t the developers also collect that knowledge of how it works, which can advance information,” Teich said.
6. Use EHR data systematically to advance knowledge.
To effectively implement CDS, so much of the success of these future technologies is based on communication, and Teich also suggests that declaring victory early and often is a good methodology for implementation process.

Currently, Teich said CDS is not supplied or used to full effectiveness in all settings, and can usually attribute to design and communication issues. Full effectiveness comes from usability design; best-practice implementation skills; facilitation of common elements; and group help for small practices.

He also said that quality and transparency drive implementation, and P4P is driving more demand for effective CDS. Teich concluded that CDS does improve safety and quality, and “the trend more workflow-oriented CDS is actually catching hold.”