Achieving better blood bank QA with a transfusion form

Achieving better blood bank QA with a transfusion form – quality assurance

Rich Scholten

Achieving better blood bank QA with a transfusion form

Devising a comprehensive transfusion form helped the blood bank of this 400-bed hospital meet mandated quality assurance requirements for blood utilization.

About four years ago, we realized that paperwork at our blood bank had gotten out of hand. There seemed to be a different form for everything. The use of these forms was so inconsistent that an occassional dangerous error was creeping in. Once, a unit of blood had been left out too long and had to be discarded. Another unit was wasted because it had been hastily stored in an unregulated refrigerator. Practically speaking, the current short supply of blood components leaves little room for error.

Of greater concern was the inappropriate transfusion that had once taken place at our institution when a patient was given blood of the wrong type because a nurse gave a unit of blood to the wrong patient. Fortunately, no harm was done. Nevertheless, we wanted to make sure such serious errors never happened again. We also hoped to achieve the 100 per cent review of transfusion appropriateness that is mandated by the JCAHO, which had become increasingly stringent about the matter.

According to the JCAHO manual, blood usage reviews must be done quarterly. The appropriateness of all instances in which blood and blood components have been given is to be reviewed at that time. The JCAHO also requires an evaluation of all confirmed transfusion reactions and of ordering practices for blood and components. Screening mechanisms are monitored as well, to be sure they incorporate clinically valid criteria. Finally, reports of all these activities must show conclusions, actions taken, and outcome. Collecting and organizing this ocean of data are extremely time-consuming for a transfusion service like ours, with no laboratory information system to help us crunch the numbers.

We became determined to concentrate our efforts. Going beyond quality control to quality assurance was our goal. * How we proceeded. The hospital’s blood utilization review committee–consisting of physicians, pathologists, nurses, and myself–meets bimonthly. One of our earlier tasks had been to write criteria for the use of packed red cells, platelet concentrates, and fresh frozen plasma. Follow-up, however, was restricted to spot checks for appropriateness of transfusions in selected patients. Not surprisingly, such random follow-up failed to bring us the perfect JCAHO review we wanted.

In planning our strategy, we realized that simplifying our procedure and paperwork would help. We had been using three different forms for each transfusion: one that came with the unit of blood, one for any reactions noted during the transfusion, and one containing general information about the patient. We decided to consolidate these into a single form, to be used throughout the hospital, that would help the staff understand and abide by the criteria we had established.

The results were dramatic. The next time the JCAHO reviewed our facility, we passed with flying colors. We attribute this achievement to our form–which we named the Blodgett Report of Blood Transfusion, shown at right–and, of course, to the cooperative and concerned hospital staff members who quickly learned to use and appreciate it. * What the form contains. The transfusion form combines patient information, laboratory results, and nursing instructions. It organizes data so that any signs of a potentially inappropriate transfusion will alert the nurse immediately. It even tells the nurse what to do about it.

The nurse takes the form when going to pick up a unit of blood. The form already contains patient identification, obtained from the hospital computer and imprinted in the upper right corner of the form with the patient’s plastic hospital identification card, which is pushed through a machine like those used in department stores. The nurse then enters the indication for transfusion and notes the type of blood component needed.

If the patient is to receive packed red cells, the nurse fills in either the estimated amount of blood loss or the patient’s most recent hemoglobin test result. If platelet concentrates have been ordered, the nurse enters the patient’s most up-to-date platelet count. If fresh frozen plasma is to be given, the nurse records the patient’s latest PT and PTT results.

Having all this information in one place, the technologist obtaining the blood component for the nurse can easily see whether the criteria set by our committee are being met. If the MT observes any serious discrepancy, he or she calls the matter to the attention of our blood bank pathologist immediately. An order that is merely questionable will cause the MT to issue the blood but to ask the committee to review the patient’s chart to make sure the request was appropriate. About 98 per cent of the time, it is. Otherwise, the head of the committee writes a letter to the patient’s physician as an educational tool, explaining the hospital’s criteria and suggesting that giving this type of transfusion again might not be a good idea. * The bad old days. Before we created our form, nurses occasionally picked up units of blood before starting IVs. If an IV was hard to start, the nurse would either return the blood to the blood bank–thus leaving it out of a monitored refrigerator beyond the 30-minute limit set by the American Association of Blood Banks, at which point it could no longer be re-refrigerated–or store it in an unregulated refrigerator on the nursing unit.

By stating on our form that neither practice is acceptable, we have eliminated both contingencies. In fact, not a single unit of blood has been wasted due to inadequate storage in the four years since we started to use the form. * No more wrong recipients. Among the most serious errors a hospital transfusion service can make is for a unit of packed red cells to be given to the wrong patient. Serious hemolytic transfusion reactions can result. At our facility, two nurses must now compare the identification information on the unit of blood with identification on the patient’s blood bank wristband. Before the transfusion may be given, both nurses must sign a statement (in the middle of the first page of the form) affirming that they have taken this step. Not one unit of blood has been given to the wrong patient at our institution since the form entered our protocol. * Less paperwork. The form not only improved our quality assurance record but also saved us many hours of tedious clerical tasks. Lacking a computerized patient information service, we label all blood components manually. Staff members previously had to fill out numerous forms for each transfusion, writing the same information several times.

Since we did not have access to a patient database containing the latest laboratory results, busy nurses often failed to double-check the appropriateness of all transfusions before beginning them. We hoped to improve accuracy by inserting the following information into the form:

[P]A box in which the blood bank technologist checks off the type of blood filter to use. With this procedure, we have eliminated a considerable amount of confusion.

[P]A note reminding nurses that AABB requirements call for each unit of blood to be completed or stopped within four hours after removal from the blood bank. Because the time at which the blood unit was removed from the blood bank is logged on the form, the nurse can quickly determine how long the blood will be able to be used.

[P]A section in which the nurse who is about to administer the blood component records the patient’s temperature, pulse, and blood pressure.

[P]A checkmark indicating that the patient signed a consent form before the procedure took place.

After the transfusion has been completed, the top page of the form is placed in the patient’s chart. A copy made automatically on pressure-sensitive paper is filed at the blood bank. * More troubleshooting. While we were working on the form, we noticed an anomaly. Although an average of 600 units of packed red cells was being transfused per month, only three or four febrile reactions were being reported from the nursing units. That number seemed oddly low.

For the next three months, as we received completed transfusion reports from the nursing units, we looked over the temperatures recorded during transfusions. We found 35 cases in which no transfusion reaction had been reported to the blood bank despite a temperature elevation of 1 C. We asked the nursing quality assurance department to find out why such a small proportion of temperature elevations was being reported. The group discovered several reasons for the situation:

1. Patient warming. In some nursing units, especially those in the oncology department, nurses customarily kept patients comfortable–consequently raising their temperatures–by providing extra blankets and warm fluids during transfusions.

2. Negative feedback. Certain physicians routinely instructed nurses to ignore small temperature fluctuations, which they considered trivial. Some of these doctors were particularly disturbed when called in the middle of the night about an elevation of 1 C.

3. Nursing judgment. Many nurses, knowing that certain patients were normally febrile or that a temperature elevation could occur for other reasons, might dismiss the possibility of a transfusion reaction and fail to report the change.

With this new understanding, our committee decided to alter hospital policy for reporting suspected transfusion reactions. If an elevated temperature is the only symptom, nurses may ask the blood bank to perform a Stat transfusion reaction workup directly, without having to notify the physician first.

The established procedure for a suspected transfusion reaction, once carried on a separate form, was printed on the back of the comprehensive form for quick reference. Formerly, precious time had been wasted in hunting for the list of instructions back at the nursing station. Any nurse who suspects that a transfusion reaction is occurring now fills out the form and sends it to the blood bank. There, our pathologist checks the results of the reaction workup and writes comments and conclusions on the same form. One copy of the completed form is attached to the patient’s chart; the other goes to the blood bank.

If a clerical check shows any discrepancy, the direct antiglobulin test is positive, or there is evidence of red cell hemolysis, the blood bank must immediately notify the pathologist and attending physician, to permit prompt treatment. This procedural change has increased the number of reported febrile reactions by about 25 per cent. Even though most such reactions are benign, we wouldn’t want to miss a hemolytic reaction because the characteristic early sign, a temperature elevation, had been ignored. * Dramatic results. A bonus of our streamlined procedures is the approbation we have received from other departments in the hospital. The medical records department finds the full-page form easier to photocopy and microfiche than the smaller forms used in the past. Our business office finds the new form handy for answering questions from patients’ insurance companies. The nurses, who wondered at first whether we had created more paperwork for them rather than less, soon adopted the form enthusiastically.

Best of all is the ease with which we now process and track transfusions. We have progressed from a 10 to 20 per cent spot check, done after transfusions were completed, to a 100 per cent check done beforehand. The form has thus not only simplified the entire transfusion process for the hospital staff but also greatly enhanced the quality of the patient care we are able to provide.

COPYRIGHT 1989 Nelson Publishing

COPYRIGHT 2004 Gale Group