Verifying blood product orders in the laboratory

A pathologist in California comments that in his experience, the process of communicating blood orders to the laboratory has always been a source of error, in the common case where the physician order is taken off a chart by staff and then manually entered into the computer. The intended blood can be ordered for the wrong patient, the wrong component can be ordered for the right patient, or the wrong number of units can be ordered -- and of course these are not mutually exclusive. He has seen all of these variations in his hospital practice. He points out that one potential solution is to require a copy of the actual physician order to be scrutinized by transfusion service lab staff prior to issuing units. Of course one would have to exempt certain critical care areas. He states that a process that requires reviewing original physician orders is in place for many hospital pharmacies, for example, to minimize error. He would like to know if any institution follows such a practice (or something analogous to it) for verifying blood product orders. If so, what are the blood product volumes and what areas are exempt from the requirement?

The following comments have been received.

ADDENDA Sept. 5, 2008

1. A Blood Bank Section Chief located at a hospital in North Carolina reports that in her institution, a nurse brings a copy of each physician blood product order to the Blood Bank for review when blood products are issued. The ER and OR are exempt from this requirement, as well as any trauma. They transfuse about 300-400 units monthly.

2. A colleague located a community hospital in Arkansas reports that her hospital requires a copy of the physicians’ orders when special blood products, such as irradiated, CMV negative, etc. are requested. Her hospital's laboratory does not stock those products routinely and have experienced times when nurses in the Oncology unit just “have a feeling” and will enter orders for the products when in fact, they were never requested by the physician. They do not exempt any area from this requirement unless an emergency release protocol is initiated. They transfuse approximately 300 units of RBC per month.

3. A colleague in Maine reports that for several years prior to implementing computerized physician order entry, his institution had a transfusion specific provider order form. Once completed, and prior to dispensing product, the order form was faxed to the blood bank for review and confirmation of the order. The provider had to specify the product and reason for transfusion. Their written orders all required addressograph identification in the upper right hand corner. The physician was also required to write the patient’s name on the first line of the transfusion order form. Part of the order confirmation was to make sure the name in the two locations was identical. ER and OR were exempt. At the time, their transfusion volume was approximately 600-700 red cell units per month.
   
   
4. A very experienced blood bank technologist writes that she has always thought that hospitals hired individuals who were underqualified to work as 'ward secretaries' or 'ward clerks'. In her estimation, that job should be high paying and demand at least a BA or AA college degree because these are the people who must read the physician orders and enter them for the various departments. She adds: "No wonder there are always errors, the people doing the job haven't had sufficient education and training for it." In the responding technologist's former position as a transfusion service supervisor, where they transfused monthly approximately 600 units of RBC's with additional FFP and platelets, they sometimes had clerks with initiative who faxed the order to the blood bank for interpretation. In her opinion, that was great, but certainly not consistently done and not for the majority of times. At the hospital where the responding technologist is now employed, the nurses and ward secretaries call the blood bank and read what the doctor ordered and ask how to enter it. Unfortunately, this wouldn't work in a larger setting. The responding technologist foresees that the only way to cut down on order misinterpretation is for the doctor to directly enter the orders into the order management computer system, bypassing the ward clerk interpretation step, so that his/her orders as "written" go directly to the transfusion service. Many physicians use Blackberries, iPhones, laptops, etc. so entering the orders directly shouldn't cause too much of a learning curve for those physicians.

5. A very experienced transfusion medicine physician whose laboratory issues 3000 components monthly, reports that she has great empathy for the pathologist who initiated this discussion about the process for verifying blood product orders in the laboratory. She derives some comfort from this discussion because she now knows that she is not alone! She reports that at her facility 'call card' errors have been increasing as staffing levels deteriorate, as registry personnel are assigned in larger numbers to work on wards, and as clerical errors are more tolerated by an institution that is trying to retain personnel, particularly nursing staff. She adds: "For years, Nursing has attempted to shift this call card error problem to the Blood Bank by suggesting that a copy of the physician’s order be used for blood issue. The blood bankers have taken the position that if the nurses can’t read or interpret the physician’s order and bad handwriting, how were we supposed to do so?
   
   However, since call card errors continued to increase, the hospital’s medical director demanded that the physician’s order become the call card. This apparently was a more expedient fix than holding the nursing staff accountable for the increasing number of errors. Thus, a triplicate form was developed that provides patient identification, clinical information, a listing of available blood products and a place for the infusion rate as well as premedication orders and what post-transfusion lab tests should be sent. The best part about the new form is that between checking boxes and listing numerical data on the request, the need for physicians to actually write anything has been minimized so that misinterpretation errors will hopefully also decrease. Also, the form will facilitate utilization review and other quality improvement activities. Since plasma and platelet orders were often confused with one another, we separated plasma and platelets on the form and listed them as “PLATElets” and “plaSMA” to emphasize that these are different products. The only anticipated problems that should be resolved as the medical staff gets used to using the new transfusion request form is that it is rather detailed, but more importantly, only two blood order issues can be made on a form. The top copy becomes the chart copy. The two copies become the first and second blood bank call slip. For trauma and other hemorrhaging patients who require multiple blood products, having to fill out multiple forms may be cumbersome and could delay blood availability." The responding Physician concludes saying, "We shall soon see what the resulting benefits and problems are."

6. An assistant chief technologist at a hospital transfusion service in Florida reports that their 300 bed hospital and cancer center transfuses approximately 1100 products a month. Approximately 55% of the transfused products go to patients with special needs such as the need to receive irradiated and leukocyte reduced products. They require all physician orders to be faxed to the blood bank at the time the order is placed into the hospital information system (HIS). The blood bank technologist uses this order to determine if any special needs need to be added to the patient permanent blood bank record. When they first started HIS order entry for blood bank products, approximately 15% of the orders were incorrectly entered. After education of the staff that enters orders, they have reduced this to approximately 8%. Variation in physician ordering practices (terminology) as well as the interpretation of handwritten orders by support staff were the major contributing factors for incorrect order entry. They are in the process of implementing a physician check box order form to eliminate these types of errors. They will still require the order form be faxed to the blood bank when the order is placed into the HIS. They will not require this form be used in trauma situations, but will require the ER to use it for routine transfusions. No decision for the OR has been made yet. They also track when the blood bank technologist does not enter the special need into the patient blood bank computer system. Fortunately 99% of the time the blood bank technologist does update the patient record. However, with the implementation of the physician check box order form, they are going to require this be used as the pick up slip. At the time of issue the RN picking up the blood and the blood bank technologist will verify any special needs have been entered into the computer.

7. A respected transfusionist in the VA system writes: Do institutions distinguish between (1) the process of asking the Blood Bank or Transfusion Service to set up blood products for transfusion; and, (2) the process of ordering the nurse to administer the blood. Are two separate written orders required? If not, are verbal orders acceptable for the blood administration?

8. A colleague in Arkansas reports that the physicians at her hospital will write an order to "transfuse two units PRBCs", and that order serves as both the order to crossmatch two RBC units as well as transfuse those two units. The respondant says that her hospital does not require seperate orders to crossmatch and then transfuse the RBCs. However, if a written order has been issued to prepare two units "to hold", then another order is required to transfuse those units. Verbal orders to transfuse RBCs can be accepted, but such orders must be followed up with written orders.

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

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

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