Autoverification Implementation

Vol. 20 • Issue 3 • Page 20


Imagine a lab where you, the director, are able to take care of management tasks, follow up on difficult cases and address purchasing and quality issues without having to constantly worry about workflow in your lab. Laboratorians are devoted to reviewing only those test results that require a second look and have adequate time to dedicate to cases that may need more attention, because your laboratory information system (LIS) is able to immediately release many test results into the patient record with no interference from lab staff.

For many labs, this dream is already a reality. Autoverification-the automatic release of results received from a lab instrument without laboratorian review based on pre-determined rules-frees up valuable time to focus on flagged results and quality control. It’s an increasingly available and embraced tool to enhance the lab’s hectic workflow.

Autoverification has been a lifesaver for managers like Raffick Bowen, PhD, MT(CSMLS), DCIChem, FCACB, DABCC, MHA(cand). Dr. Bowen is an assistant professor of Pathology and associate director of the Clinical Chemistry and Immunology Laboratory at the Stanford University School of Medicine in Palo Alto, CA. Five years ago, his lab was fielding regular complaints about the 90-minute turnaround time (TAT) for troponins and basic metabolic panels taking up to three hours.

A root-cause analysis of the lab’s dragging TATs exposed the need for manual review of each and every specimen coming into the LIS report as the culprit. Dr. Bowen and his colleagues immediately began looking at numbers and preparing a budget and timeline to implement autoverification.

The implementation process, he admits, was an exhausting one. But the payoffs have made the cost and effort well worth it.

The setting of the parameters that determine whether results can be automatically released is the most time- and energy-consuming part of the implementation process. Dr. Bowen explains that he started with extensive research. “I spent a lot of time looking at literature, talking to colleagues, going to conferences and seminars and sitting down with other medical staff to figure out the best way to approach this,” he says. If you aren’t already, you must become intimately familiar with procedures, policies and test codes as you begin writing your rules.

Where to Start?

It may sound obvious, but it bears repeating: The first tests you should autoverify are those that will have the biggest impact on your lab. Look at tests with the highest volumes and the slowest TAT. Dr. Bowen started with routine chemistry tests, including both the basic and comprehensive metabolic panels. “After we implemented those, we tweaked the rules for about a year; when we were comfortable with that, we moved on to Phase 2 of implementation, which was drugs-of-abuse testing, therapeutic drug monitoring tests and urine chemistry tests,” Dr. Bowen explains.

Be prepared to tweak your rules upon implementation. When you first start autoverification, you’ll still be looking at the results prior to release. This is the time to look for any problems or rules that might need to be changed. Autoverification criteria should be validated using simulated data and clinical specimens, and each rule should be tested individually and in combination with other rules.

Dr. Bowen points out that labs must also know what not to autoverify. Chemistry and hematology tests are usually the simplest to autoverify, “but we don’t do body fluids like peritoneal, pericardial, pleural and cerebrospinal fluids, beta HCG, ethanol and a few other tests,” Dr. Bowen says. “Body fluids are very delicate specimens, and we want to have human intervention there. For beta HCG, we were worried about heterophile antibody interferences or reporting erroneous results, which can have devastating impact on a patient. As for ethanol, most of the time, it is ordered with an osmolar gap, so we need to wait for the plasma osmolality results and calculate the osmolar gap before releasing the results to the patient record. Drugs like tricyclic antidepressants aren’t automated in our lab, so we can’t autoverify that,” he explains. “Also, we do not autoverify positive urine drug of abuse results.”

Rules and Regulations

Labs that decide to make the leap to autoverification should familiarize themselves with a number of regulations and guidelines. Be sure you can answer these questions from the College of American Pathologists Laboratory General Checklist:

• Gen. 43360: “If the laboratory employs autoverification procedures for tests performed on any instrumentation, is there a signed policy by the laboratory director approving this action?”

• Gen. 43370: “Is there documentation that the laboratory has validated the autoverification process against the rules prescribed by the laboratory director before implementation?”

• Gen. 43380: “Is there documentation that the autoverification process has been tested at least annually, and whenever there has been a change to the system?”

• Gen. 43450: “Is there documentation that calculations performed on patient data by the computer are periodically reviewed?”

• Gen. 43600: “Are result entries checked against a defined range of expected results to detect absurd values before reporting?”

You’ll also want to familiarize yourself with the Clinical and Laboratory Standards Institute document AUTO10-A, “Autoverification of Clinical Laboratory Test Results; Approved Guideline.” This document provides a framework to help individual labs design, implement, validate and customize rules for autoverification.

Technical Notes

Your information technology staff will be your best friends throughout the implementation and rules-writing process, Dr. Bowen confirms. Build a solid relationship with the IT person or team within your facility, as well as the service and support people at your LIS or middleware company. Get people on the same page early on, and keep them there with regular updates and meetings.

Most importantly, Dr. Bowen notes, don’t be afraid to ask your existing IT vendors for the functionality you’re looking for. Some LISs already have autoverification capabilities built in; other labs will purchase middleware and set up autoverification that way. If you’re not sure if your LIS is capable of performing a particular task, don’t hesitate to ask. “One lesson we learned was not to underestimate the capability of the LIS,” Dr. Bowen says. Often, labs are only using a fraction of the functionality their LIS can provide.

Fringe Benefits

Many lab managers share a common concern in implementing such an impactful change: getting staff buy-in and support. Dr. Bowen says he made sure to include his laboratorians from the beginning and ask for their feedback at different steps of the process. Give staff a chance to test the rules before finalizing them and consider their input. Remind them of the many benefits they’ll reap after autoverification is up and running-fewer angry phone calls from physicians and more time to focus on quality and other projects.

Dr. Bowen notes that his staff is “thrilled to be able to think and act like scientists now that they are no longer required to spend the bulk of their time reviewing normal results.” Polls have shown an increase in employee morale, and that results in better relationships with lab customers and an increase in patient safety and satisfaction, he adds.

Once autoverification is up and running, labs can expect to enjoy the trifecta of lab benefits: improved patient safety, decreased TAT and cost savings. Beyond that, the improved employee morale (and less sick time) and uptick in quality and documentation in the lab makes the implementation process a road worth navigating.

Kelly J. Graham is associate editor.