Who should draw blood specimens for compatibility testing?

A member of the e-Network is interested in discovering what others are doing regarding the personnel who perform phlebotomy. Several years ago at the e-Network member's facility, venipuncture was decentralized, and the nursing staff assumed the major responsibility for blood drawing. Since that time that institution has experienced a dramatic rise in the number of mislabeled blood samples. Despite repeated efforts aimed at educating the staff about the importance of proper patient identification, mislabeling incidents continued to occur at an unacceptable rate. Consequently, that institution is considering a return to the previous model in which a staff of trained phlebotomists perform the majority of the blood draws. The questions for the e-network members are as follows:

Who is responsible for drawing blood in your facility?

If you have recently decentralized venipuncture, have you noticed an increased number of mislabeled samples?

Are there any facilities that have done away with decentralized venipuncture and returned blood drawing responsibities to a phlebotomy department due to an increase in mislabeled samples?

Is there any good benchmark data available on mislabled blood sample rates?

The following opinions are paraphrased and are being presented without attribution. They do not represent an official opinion or position of the CBBS.

1. Although the above questions were intended for hospital transfusion services, some blood centers responded. For example, the first response was from a blood collection center in San Diego, where all blood collections are performed by Nursing personnel - RNs, LVNs, and Phlebotomy Techs. This has been the case since the center opened, although originally only RNs were used. That facility's procedure includes, at the conclusion of the draw and before the donor is permitted to leave the chair, the reading aloud of the numbers on the donor record, blood bag and tubes, and this is verified by 2 people - the individual who is discontinuing the phlebotomy and a nursing assistant person who is picking up the unit to seal and ready it for the lab. Although not perfect, phlebotomy mislabeling errors are rare.
   
   
2. A second blood center, also located in Southern California, commented that RN's, LVN's and certified medical assistants are the technicians responsible for drawing blood at the member's facilities. The collection staff are required to label all tubes and the collection bag concurrently. All samples are then bundled together with the whole blood unit using rubber bands. The member's facilities have not noticed an increase in the number of mislabeled samples. In the past 12 months, there have been very few incidents involving mislabeled samples. The member did point out, however, that the facility has no available data on the rare occurrence of mislabeled samples.
   
   
3. Another member commented that at a private Medical Center in Los Angeles, the lab has "partial" responsibility for phlebotomy, overseeing some lab phlebotomists. These phlebotomists are responsible for collecting between 5 am to 9 am. The rest of the time, phlebotomy is performed by RNs or Patient Service Technicians (PSTs). The PSTs are part of the nursing team on the floors. This facility does have a problem with specimen collection and labeling and tracks these problems with their monthly "Unusual Occurrence Reports". They have had a measure of success with their "non-lab" phlebotomists because they conduct the phlebotomy training for the PST course. Their main problem at this time is the RNs who draw the specimens - and mainly on the graveyard shift. This facility has already identified that some retraining is necessary and will be trying to coordinate that with their Education and Training department. The e-network member reporting this answer thinks that it helps for the lab "to be in charge" of the training and tracking of the collection and labeling errors.
   
   
4. Another member reported that he/she works at a hospital in Southern California, which is a Level I Trauma Center and a teaching hospital (not my place!!!) at which most blood bank specimens are drawn by nurses and residents. Four years ago that facility launched a continuous quality improvement emphasis that focused on rejected blood bank specimens, and the rate fluctuated according to extensive in-services given to both nurses and residents (who have a high turnover rate). ANY deviation from the acceptability criteria for blood bank accessioning (no initials, date, time, name, ID# wrong/illegible) was cause to summarily reject a specimen. The rejection rate ran about 4-6%. The member points out that the benchmark for ALL pathology laboratory specimens, not just blood bank, according to a CAP QA Probe study (and published) is 1.5%. The member's hospital lost phlebotomists during a downsizing several years ago, but the rate before and since cannot be directly compared as data were not gathered before as now.
   
   
5. A fine institution, (almost as nice a place as USC) presented data at the AABB meeting back in 1982 (Transfusion 22:461; abstract) that, unfortunately, was never published as a full article. If you check the reference you will discover this data was collected at UCLA. The UCLA laboratory required that a new sample be collected from every patient who had not been typed by them previously (18% of total). Samples from the wrong patient (i.e., mislabeled), were submitted to the blood bank 1 per 1100 units or 1 in 300 patients transfused (not in the abstract, but attested to by a high-ranking blood banker at that facility); 80% of the errors were from non-blood bank phlebotomists (i.e., nurses/doctors in OR, or ER).
   
   
6. Finally, a member commented that at his/her hospitals, most of the phlebotomies are performed by the lab staff. However, Labor and Delivery, the ER and the OR draw their own samples. The error rates in the Labor and Delivery are alarmingly high, and error rates in the ER and OR are also higher than those for the rest of the hospital, which is covered by lab hlebotomists. Education is the next step planned at this member's hospitals.

ADDENDA Nov. 11, 1999

7. The following additional opinion has been submitted, after I had distributed the initial set of answers that were offered in response to the question that a member of the e-network asked, regarding the personnel who perform phlebotomy. If you recall, the member said that several years ago at his/her facility, venipuncture was decentralized, and the nursing staff assumed the major responsibility for blood drawing. Since that time that institution has experienced a dramatic rise in the number of mislabeled blood samples. Despite repeated efforts aimed at educating the staff about the importance of proper patient identification, mislabeling incidents continued to occur at an unacceptable rate. Consequently, that institution is considering a return to the previous model in which a staff of trained phlebotomists perform the majority of the blood draws.
   
   In this additional reply, the individual states that years ago at a Midwestern university hospital, his hospital had a policy of requiring a "check" sample (a second blood bank tube) for patients at the time they were actually transfused. This was done for all patients and was a second draw. A rapid slide ABO was done before issuing the blood. His memory is that the error rate was similar to what was stated to be found at UCLA. [Editor'sr NOTE: see the issue of Transfusion that was printed for the 1982 AABB meeting, vol 22 page 461, abstract]
   
   More recently at another such Midwestern hospital his hospital tracked all labeling errors, from simple incomplete to actual mislabeling or double labeling. Total errors ran about 2%. This was statistically lower than the baseline and the change coincided with a rigid rejection policy for any incomplete or mislabeled tubes. This total data was used as a surrogate for major mislabeling (wrong patient, double labels, etc). The major mislabeling also decreased, although the numbers are too small to evaluate statistically. Of note is that although laboratory phlebotomists drew the majority of samples, nursingor doctors drew all the major errors.
   
   Hopefully this information is valuable for those of you who are trying to address the risk of using non-lab personnel for drawing specimens that are used for blood bank testing.

Submit comments to the e-Network Forum at enetworkforum@cbbsweb.org

Ira A. Shulman, MD
CBBS e-Network Forum Senior Editor & Moderator

W. Tait Stevens, MD
CBBS e-Network Forum Editor & Moderator

Elizabeth M. St. Lezin, MD
CBBS e-Network Forum Associate Editor & Moderator