Labs add safety net to critical values procedure

Valerie Neff Newitt

March 2019—A hypercritical value notification policy at Northwell Health in New York State authorizes the laboratory alone to activate a hospital’s rapid response team when necessary. This is the newest step in a critical value notification process that had been working well—but not quite well enough.

A rare communications breakdown at a Northwell Health hospital nine years ago involved a critical lab value that delayed a patient care intervention, which eventually resulted in an untoward sentinel event.

In response, Northwell Health leaders devised and implemented a hypercritical value notification policy to reduce notification time and expedite patient evaluation and treatment. The new policy also empowered laboratory professionals to send the hospital’s rapid response team to a patient’s room when a licensed clinical caregiver is not available to be at the bedside.

“In late 2010 a patient in one of our EDs was determined to have succumbed due to a low hemoglobin count,” says Joseph Castagnaro, vice president, laboratory services integration and operations, Northwell Health Laboratories. “The patient was in the ED where there were numerous caregivers looking after many patients. Sadly, when the one call came in with the critical hemoglobin report, it didn’t get to the right person.”

From the tragedy came the question: Can this be prevented in the future?

With laboratory and clinical leaders from across the health system weighing in, it was determined that the following laboratory values, if not acted on, could lead to serious patient harm:

• Glucose ≤ 30 mg/dL
• Potassium ≤ 2.0 mEq/L
• Potassium ≥ 7.0 mEq/L
• Hemoglobin ≤ 5.0 g/dL

Now when laboratory staff report on these values, they call the patient’s unit and follow a standardized script, which begins with the lab staffer saying, “Do not hang up the phone. I’m calling with a hypercritical lab value.”

The rest of the process goes as follows:

• Read back and document the test result and patient information.
• Confirm the patient’s location.
• Clarify whether a rapid response team or physician is evaluating the patient.
• Communicate to the licensed staff member that the lab will call a rapid response team to address the patient’s status if a physician or RRT is not already there.
• Call the operator to request activation of a rapid response team to the patient’s bedside.

“The clinical leadership had a problem to solve,” says Dwayne Breining, MD, executive director of Northwell Health Laboratories. “We needed to prevent this small percentage—so small, it wouldn’t even show up on the radar screen—of mishaps from ever happening again. Because, when it did happen, there was a horribly negative outcome that was potentially preventable.”

May Kam Tso, BS, MT(ASCP), senior administrative director of the clinical labs, hospital lab operations, says Northwell Health, which consists of 23 hospitals, already had a “nearly perfect” communications system. “We already had a small number of critical values of which we notified floors on a timely basis, yet sometimes clinical staff did not apply the appropriate interventions in the timeframe we expected.”

In a systemwide review of patient outcomes related to critical laboratory values, Northwell Health leaders found a small number of cases in which the appropriate floor personnel were notified in a timely manner but the appropriate clinical intervention did not take place quickly enough to affect the outcome. In some of these cases, Dr. Breining says, it seemed as though more rapid escalation could have potentially led to more timely and effective clinical intervention, particularly in some cases in which laboratory test results were significantly worse than the critical value threshold.

Division chiefs from Northwell Health hospitals met for six months to brainstorm. “After many brainstorming sessions, we realized a few things,” Dr. Breining says. “One, that the critical value notification system we already had in place was pretty effective, getting the right result to the right place in the right time period virtually 99.9 percent of the time. Two, any significant revamping of that system was as likely—probably more likely—to result in a decline in performance as it was to cause improvement. And, three, engineering an additional system to get at this significant but very small percentage of cases, which could occur anywhere in any of our facilities, proved extremely challenging and could easily result in an operation significantly more resource-intensive than was feasible.”

The solution, then, was to not disturb the existing functions that almost always performed well, but to add a layer to catch the few cases in which things moved too slowly.

Tso says the goal became to strengthen “the existing infrastructure and communication system to implement a safety net for a large and complex health system.”

Rapid response teams in general have been shown to be effective in improving outcomes in the hospital setting, Dr. Breining says. “What is unique about our new process is that in specific instances the RRT is initiated directly from the laboratory.”

Once they had a plan, they also had a little pushback. “No one initially wanted to champion it,” Castagnaro says. “No one wanted to stand up and say, ‘I’ll take it and run with it.’ There was reluctance because some people felt this was overkill perhaps, in reaction to a situation that might never occur again. Others said, ‘It’s not my area’ or ‘It’s not my responsibility.’ Everyone already had plenty to do.”

Pushback came primarily from the clinicians who were already part of the rapid response teams, Dr. Breining says, noting the RRT and medical leaders in the hospitals were concerned about false alarms.

“But the lab has solid, hard, objective data,” Dr. Breining says. “We went through the data in our LIS, set the parameters on the four values we thought were worthy of this RRT activation, and could see how many times the RRT would have been triggered, based on that data. We could tell maximums per day and per week. And even in those cases, an alert wouldn’t necessarily have been activated, assuming clinicians were already on the floor. So everyone became more comfortable seeing reasonable numbers and knowing this was not something that would be encouraging 12 false alarms a day.”

They also knew there would be less activation in some hospitals than in others. Some ICUs and critical care units always have a care team at the bedside, Dr. Breining says. Some areas were excluded from the initiative in some hospitals. “In our post-anesthesia care units, for example, there is always a caregiver at the bedside monitoring a patient.”

There was laboratory pushback too. “It was very unnerving for our lab folks,” Castagnaro says. “Unless you’re a blood banker, it’s rare that lab folks get that close to the patient.”