Experts See Continued Gaps In Measuring Ambulatory Safety

Major gaps persist in understanding ambulatory safety and virtually no credible studies have shown how to improve it. This was the salient finding of a recent American Medical Association (AMA) Center for Patient Safety/CSC report.

More than 10 years ago, in December 2000, the Agency for Healthcare Research and Quality (AHRQ) called together a group of experts to review the state of ambulatory patient safety research and to establish an agenda for the future. Last year, the AMA Center for Patient Safety surveyed the progress of ambulatory safety over the past decade and contracted the Falls Church, Va.-based CSC to assemble all available research about ambulatory care between 2000 and 2010 in a report, “Research in Ambulatory Patient Safety 2000-2010: A Ten-Year Review”.

Some of the key findings from the report include:
•    Diagnostic errors, including missed, delayed and incorrect diagnoses, are some of the most common threats to patient safety in ambulatory care settings.
•    The lack of follow-up on abnormal test results is a major cause of harm in ambulatory care, as is in office-based surgery.
•    The lack of a standard taxonomy for errors in ambulatory settings, noted by the AHRQ in 2000, still exists and continues to pose challenges in comparing results across studies.

Erica Drazen, Sc.D., managing director and report author, in the Waltham, Mass.-based Global Institute for Emerging Healthcare Practices at CSC, was surprised that there hasn’t been more definitive research done on adverse events in ambulatory practices, and especially in ambulatory surgery, but says the lack of funding in this area is the primary cause. She notes that two-thirds of the studies cited in the report come from only three institutions: Brigham and Women’s Hospital (Boston), Michael E. DeBakey VA Medical Center (Houston), and Regenstrief Institute (Indianapolis).

Another challenge in analyzing ambulatory data is that it is extrapolated from institutions using electronic health records (EHRs), while many ambulatory practices have yet to go electronic. Drazen says that it was difficult to determine what the industry’s biggest problems were with the lack of ambulatory studies, and the only way her team could identify patient safety errors was to look at malpractice settlement data, which only encompasses a subset of errors made in the ambulatory setting.

One of the report recommendations was to fund the development of a virtual research center that would allocate resources to the design, execution, and funding of a study among a group of typical primary care practices and solo practitioners. Another of the report’s recommendations was to further research to support the role of information technology to improve ambulatory patient safety, including computerized physician order entry (CPOE) and EHRs, and that these technologies should be evaluated within the larger contexts where they are implemented.

Diagnostic Errors
Drazen says that health IT can better support ambulatory patient safety by capturing diagnostic errors within an EHR. She says that the recent Institute of Medicine (IOM) report, “Health IT and Patient Safety: Building Safer Systems for Better Care,”  highlighted the need for clinical information systems to be able to capture information about errors and adverse events. “There are techniques now for doing surveillance based on paper records to identify events that signal that an error might have occurred,” Drazen adds. “So they [the IOM] are proposing that those be built into the software so that they would automatically be captured.”

To improve diagnostic error surveillance, Drazen says, capabilities should be added to record when certain medical interventions (i.e. when an antidote for a medication reaction is ordered, when a sudden change in lab value occurs, or when a switch in medication within the same class occurs) happen. Once these interventions are captured in electronic form, it makes it much easier to submit the information electronically for reporting purposes, Drazen says.

Drazen also recommends that EHRs allow providers to provide instant feedback to IT staff to improve safety of clinical information systems, so this information can be consolidated and sent to a central place. She also believes that EHR vendors should be required to report patient safety incidents, and encourage, not inhibit end users from sharing feedback on safety issues and screenshots of what they’ve done to improve safety. Many vendor contracts prohibit their customers from reporting safety issues beyond the vendor, says Drazen; but to the extent feasible, there should be certification requirements to mandate incident reporting. Federal law requires medication tampering and medical device failure to be reported, so Drazen thinks that EHR safety reporting might not be such a stretch.

Abnormal Test Results
Another area for patient safety improvement that the report identified is the tracking of the notification of abnormal test results. “If you look within those diagnostic errors within ambulatory care, the ones that lead to the greatest harm as evidenced by malpractice settlements are missing cancer diagnoses,” says Drazen.

“Here’s my one wish,” she adds: “that there should be a requirement in the certification criteria that required systems to flag and notify providers of abnormal test results that had not been viewed within a  week of the time they were available.” She also recommends that for imaging procedures, the system should continue to remind a provider until there is confirmation of follow-up.

Drazen says that the Veteran’s Administration did studies to improve this problem and sent electronic notifications of a test result not only to the ordering provider, but also to the primary care provider. In these studies, the follow-up of lab results actually declined because each provider assumed the other was looking at the result and notifying the patient. “We need to do experiments to find out what the effects are because sometimes your gut feeling is wrong,” she says. “But secondly, by building capability to look to see if someone opened the result, and give them the ability to escalate those results that weren’t opened, and to then do research to see what interventions were successful.”