Raymond D. Aller, MD
The need for every hospital and laboratory to adopt foolproof positive patient identification measures is crystal clear. Yet the rate of adoption for such technology varies markedly, occurring at a lethargic pace in some areas and a snappier pace in others.
Positive identification of patients and specimens should be a strict three-stage process that comprises the following:
- Registration. Hospital admitting departments and outreach laboratories should identify patients upon arrival using definitive measures, such as biometric devices.
- Matching. Laboratories should ensure that patients are linked to their specimens using bar coding or radio-frequency identification, or both.
- Tracking. Specimens should be tracked electronically during every step of their journey through the laboratory to final test result.
Identification at registration. Hospital registration departments and outreach laboratories should distinguish patients using identifiers that cannot be changed or imper-sonated. The solution: biometric identification using fingerprint, electronic face recognition, iris pattern, palm vein, hand configuration, or similar means. Unlike procedures for verifying name, birth date, driver’s license, and other text-based identifiers, biometric processes can distinguish an individual at a one-in-a-million or better accuracy rate if engineered properly.
Biometric tools increasingly are being used for patient identification in hospital admitting departments (see “Biometric products finding a home in health care,” page 118). While slow to catch on in the laboratory, they are equally effective for identifying patients presenting for outpatient phlebotomy. In such settings, where a single interaction typically occurs, it’s generally not feasible to apply a barcode wristband.
Before implementing its biometric palm scanners last year, Harris County Hospital District, Houston, Tex., looked at its database of 3.5 million registered patients and found that in nearly 70,000 instances, at least two patients shared the same first name, last name, and date of birth.
But even when names don’t overlap, mix-ups occur. When re-registering at a medical facility, patients are sometimes given new medical record numbers because staff members do not, for whatever reason, access the patient’s previous record. Consequently, the records must be merged, increasing the opportunity for identification errors. On the flipside, trying to “unmerge” two patients who were mistakenly morphed into one can be extremely difficult. LAC+USC Medical Center, Los Angeles, spent hundreds of thousands of dollars merging duplicate records and then had to devote money and staff resources to “unmerge” those that were combined erroneously.
Not all mix-ups are so innocent. Harris County Hospital District purchased its palm scanner in part because it found that some patients were presenting fraudulent identification so they could receive free health care for which they were ineligible.
Matching patients with specimens. National studies conducted in recent years suggest a wrong-blood-in-tube error rate of between 0.05 percent and one percent. These patients will be diagnosed and treated on the basis of blood that belongs to another patient. Similar errors occur in medication administration and every other instance in which caregivers interact with patients.
Recognizing that manual systems of verification can’t eliminate these situations, laboratories increasingly are electronically matching patients to specimens, most commonly using bar-code wristbands and bar-code specimen labels, although use of radio-frequency identification is on the upswing.
A growing number of lab information system vendors and other software vendors are delving into this area of the positive patient identification marketplace in an effort to respond to concerns about patient safety, yet identification matching at the point of care is not yet universal.
In-lab positive specimen tracking. Laboratorians have, for many years, done a satisfactory job of ensuring that specimens are handled properly as they move through the lab. Automated identification of specimens via bar coding is now the norm in clinical pathology and is rapidly becoming standard in surgical pathology and cytology. Nevertheless, laboratories should examine their processes for ensuring that the patient identified on the tube is securely linked to the results as verified in the laboratory information system. In laboratories that have manual procedures in place or batch processing without machine transmission of identifiers, lab staff should scan bar codes on specimen tubes as results are entered.
In conclusion, automated positive patient identification at every step of the patient care process should be mandatory. Are we there yet? No, but we’re making progress.
Dr. Aller is director of informatics and clinical professor in the Department of Pathology, University of Southern California, Los Angeles. He can be reached at firstname.lastname@example.org.