Survey Says: Patient identification practices

Last month we asked our readers and visitors to our web site all kinds of questions on their policies on patient identification, and if they follow them.

First, let's establish how patients should be identified. The newly released CLSI venipuncture standard requires those who draw blood specimens to ask patients to state their full name and birth date, and to spell their first and last names. The information provided must then be compared with the patient's ID band. If the patient cannot provide the information (e.g., language barrier, cognitive disability, unconscious, sedated, etc.), a family member or caregiver must verbalize the information on the patient's behalf. Outpatients without ID bands must provide another form of identification. 

We understand not every facility has obtained a copy of the revised venipuncture standard, and those who have may not yet have implemented the new provisions on patient ID. But we were amazed at how many of those who participated in our survey have policies that reflect the most current revision. We're not so amazed by those who admit they don't always follow the policy. Here's the stats on the five steps:

• state full name: 77 percent;
• state birth date: 98 percent;
• spell first and last name: 7 percent;
• seek a third-party when patient is unable to provide the required information: 86 percent;
• compare information provided with ID band and test request: 89 percent.

We're thrilled that 98 percent of those responding have a policy on requesting the patient's birth date. (The other 2 percent responded they are required to ask for two patient-specific identifiers, onsider: A healthcare system in Houston looked into their database of all patients and found that out of 3.5 million patients, two percent of them had the same first name, last name and birth date of which may or may not be the birth date.) However, only 77 percent said their policy is to only request the patient's first and last names. Here's what the remaining 23 percent needs to cone as at least one other person in the database. That means 7,000 patients in Houston are at risk of being identified as someone else if only the first name, last name, and birth date are requested.  

We're not so thrilled that so many survey participants admitted they don't always follow their facility's policy. Perhaps that's why there are over 160,000 adverse patient events in the U.S. every year caused by those who draw blood samples but do not properly identify their patients or samples. Here's the percentages of those who know their facility's policies on the five steps, but don't follow them every time:

• have the patient state his/her full name: 14 percent don't always comply;
• have the patient state his/her birth date: 5 percent don't always comply;
• have the patient spell his/her first and last name: 48 percent don't always comply;
• seek a third-party when patient is unable to provide the required information: 11 percent don't always comply;
• compare information provided with ID band and test request: 7 percent don't always comply.

These statistics are especially concerning since 48 percent of those responding admit to having found an identification bracelet attached to the wrong patient.

Among those who admitted they don't always ask patients to state their name, seven percent said they only ask if they don't know them. Here's why that will result in patient death or serious medical consequences someday. All the patients a healthcare professional might draw blood from during the course of any given day will fall into one of three categories: 1) patients who are total strangers, 2) patients they know with solid confidence (regular patients, friends, family, etc.) and 3) those with whom they are somewhat familiar. The patients in Group 1 are not at significant risk because most healthcare professionals are going to follow protocol for patients they've never seen before. Patients in Group 2 are not at great risk, either, because of the high level of familiarity. Make no mistake, there is some risk for both groups. Some people make exceptions to the policies even for strangers, and some have a false sense of familiarity with patients, and make exceptions as well. 

However, those in the third group, patients who are somewhat familiar, are at the highest risk of being misidentified and suffer potentially tragic consequences. That's because the healthcare professional's familiarity with the patient is not certain. Depending on other variables, collectors may talk themselves into being sure when they really aren't for the sake of expediency, simplicity, or ill-conceived rationale. 

One survey respondent commented "We will check spelling of unusual names." This strategy, however, won't prevent Eve Snyder from being transfused with blood intended for Eve Schneider. Nor will it save John Smith from being mistaken for Jon Smith, John Smith for John Smythe, and Jon Smith for Jon Smythe. However, if the standard protocol reflects the CLSI standard and is applied to every patient, whether they appear familiar or not, nobody gets misidentified. Because patient identification cannot be automated, humans have to automate themselves by consistently applying the standard protocol in every regard without exception.

What these survey results tell us is that a significant portion of patients are at risk of transfusion-related death, medication error, misdiagnosis, and general mismanagement due to lax adherence to facility policy, and facility policy that does not reflect the standard by which all facilities are being held  as of April when the GP41 revision came out.

This month's survey: This month, we're asking supervisors what single attribute would they like to see their staff as a whole improve upon? Putting the shoe on the other foot, we're also asking non-management personnel what single attribute they'd like to see their supervisors improve upon. Need to vent? Now's your chance.

Take the survey.