A blood banker in Ohio asks if there are benchmark data for how many times per 50,000 units of blood issued that a transfusion committee should be sending out a letter which questions a physician's decision to transfuse a blood product. For example, how many letters are normally sent to members of the medical staff at "Expert U. Hospital", if 50,000 transfusions are given per year at that facility? What number of transfusions would you expect to fail an initial screening procedure and therefore go to the Peer Review Process? What number of transfusions would fail a Peer Review Process and result in a letter/e-mail/etc. being sent to an ordering physician for clarification of the transfusion decision?

The inquiring blood banker has provided an outline of their review process.

1. A report is printed from the Laboratory information (LIS) system that lists every transfusion (by facility) and the patient's most recent Hb & Hct, Platelet Count, PT/PTT/Fibrinogen (based on the product given) before the transfusion as well as the same information post transfusion.

2. This report is reviewed by a member of the Transfusion Service against a standard set of criteria, a copy of which can be seen by clicking on the link to a WORD document.

3. Any transfusion that falls outside of the initial screening criteria is referred to the Medical Director of the Transfusion Service. The Medical Director reviews that patient's chart and presents the information gathered at a Peer Review Meeting (Transfusion Committee).

4. The Peer Review Committee has 2 options: Transfusion decision was sound based upon information gathered from the patient's chart, or sending a letter to the ordering physician for clarification of the decision to transfuse.

The Peer Review Committee at the inquiring blood banker's largest facility (50,000 transfusion per year) meets monthly and over the course of a year may issue 10-12 letters for clarification.

However, at the second largest facility (22,000 transfusions per year), there has not been a request for clarification of the transfusion decision in the last 3 years.

The inquiring blood banker is looking for a "benchmark" that would indicate that either one facility may be looking too close (she thinks probably not) or that the other facility needs to look closer (she thinks probably yes).

In response to the above, the following replies have been submitted.

ADDENDA April 4, 2002

1. A Texas blood banker wonders if the number of letters sent out by the Transfusion Committee would be physician-dependent. She questions if the hospital with 22,000 transfusions is staffed with physicians who are more familiar with the transfusion triggers and adjust their orders accordingly. Perhaps the hospital with 50,000 transfusions/year hospital is staffed with physicians who are in training or just not as familiar with the transfusion criteria. She suggests that it might be a valuable research project to poll as many hospitals as possible to collect this type of information, and then perform a statistical analysis of those data.

2. A blood banker from Kansas is astounded by the number of peer review audits that the inquiring blood banker's institutions perform. The Kansas blood banker's institution audits 5% of transfusions retrospectively. Their transfusion volume is 7,000 units/year. They also require a reason for each transfusion to be coded electronically in the order entry system prospectively. The transfusion service reviews the reason, but does not check laboratory data on every transfusion prospectively. Their guidelines for transfusion are similar, but more conservative for some components. Their medical director reviews all audited transfusions not meeting criteria and writes a letter to the physician to explain his reason for ordering the transfusion. Approximately 2 letters are written each quarter. The transfusion committee reviews the audit summary and gives the final determination on the outliers after hearing the physician's response. The system works well but there are some places where philosophical differences remain and they are at impasse (cardiovascular surgery). The determination of the Transfusion Committee also goes to the Medical Staff Quality Review Committee and to the medical staff office for reappointment purposes.

ADDENDA April 5, 2002

3. A blood banker from California reports that his facility completes a review of all transfusions, similar to that of the inquiring blood banker. They refer questionable cases to a Pathologist for review and then on to the Blood and Tissue Review committee, if the Pathologist cannot find good reasoning for the transfusion. Typically they send 3-5 cases for Pathologist review each month - of those only 3 or 4 a year go on for further review. They typically transfuse 15,000 units a year (all products).

4. A blood banker at a hospital in a rural, 'non-teaching' facility in south central Virginia reports that they had approximately 13,000 discharges for 2001. Their blood utilization review process is very similar to that described by the "blood banker in Ohio". 

• A monthly report from the Laboratory Information System is generated at the beginning of each month, (prior month's data). The report list all blood products transfused, location (in patient or out patient), ordering physician, and specified lab tests (Hb, PLT, PT, PTT). 
• This data is screened, by an RN review analyst, against the hospital's current set of criteria, which is very similar to that used by the "blood banker in Ohio". 
• When a case falls out of the initial screening, the RN reviews the patient's medical record, for documentation of clinical indicators what would justify the transfusion of the blood products. 
• A case that cannot be justified based on data obtained from this review is discussed with the Director of Laboratories. He will then review the patient's medical record. If after his review of the medical record, the transfusion cannot be justified; he will send a letter to the ordering physician requesting clarification of his/her decision to transfuse the blood product(s). 
• Once the physician responds to the Director's letter, the case may be closed or presented for discussion by the Transfusion Committee. After the committee discusses the case, it will be closed or sent to the ordering physician's Department chair with recommendations from the committee. 
• The hospital transfused 6,591 units of blood products last year (2001). The Medical Director of Laboratories sent only one (1) letter to a physician in 2001. During the year 2000, they transfused 6,155 units and four (4) letters were sent to physicians. 

5. A blood banker in the Midwest applauds the efforts to improve and ensure appropriate use of blood and blood components by peer review. However, she would be interested to hear whether institutions that send out letters to ordering physicians really feel that this process improves transfusion practices in their facilities. She reports performing utilization review at a large teaching institution, and has discontinued performing retrospective reviews and sending letters to physicians because she feels that, in the long run, it is a waste of time and does not result in improving transfusion practice. Retrospective reviews CAN improve transfusion practice if they are used to identify a trend or group of physicians who are not following the institution's normal guidelines.) The process of identifying a possible inappropriate transfusion by screening labs and orders, having the Medical Director review the case, sending a letter to the ordering physician, waiting for a response from the physician (which sometimes never came), and reporting to the Transfusion Committee (which meets quarterly) required several months to complete. Usually, at the end of the process, she found that the ordering physician, who was sometimes irate about receiving such a letter (and may in fact not even remember the transfusion in question), still felt the transfusion was appropriate. The Transfusion Committee generally gave the ordering physician the "benefit of the doubt", and no further action was taken. Performing prospective or concurrent reviews seems to be more effective. Like the Kansas blood banker, she reports that her institution requires that all requests for blood components have a reason for transfusion checked on the form. Transfusion Service staff reviews corresponding lab values (prospective review) to affirm that the reason for transfusion is valid (suffice to say that if the Transfusion Service is very busy at the time, perhaps not every order is scrutinized). Staff will contact the blood bank physician to review orders that appear inappropriate and contact the ordering physician for clarification. If the transfusion is ordered stat, or during the evening or night shifts, Transfusion Service staff can choose to fill the order, but have the order reviewed by the Blood Bank physician the next morning (concurrent review). The Blood Bank physician, when reviewing orders the following morning, usually reviews the patient's chart and speaks with the patient's physician about the order in question. This facility is a teaching institution, so verbal, face-to-face communication between the ordering physician, who may be a resident or fellow, appears to be the most effective. (I think it's better to take advantage of that "teachable moment" instead of taking the transfusion in question through committees and generating a letter weeks or months later.) Granted, no system for utilization review is perfect. It depends on what level of technical and physician coverage the Transfusion Service has, and what kind of rapport the Medical Director has with house staff. But in my opinion, less time should be spent on sending letters and talking about isolated cases in Transfusion Committee meetings and more time spent on "real-time" communication with clinicians that might improve transfusion practice.

Please submit comments to the e-Network Forum.

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