Sunday, October 12, 2008

Responsibility - what's that?

The October issue of Transfusion has many interesting papers.

Two items caught my eye and got me to thinking how we view responsibility when colleagues fail to provide optimal care. We seldom get upset. Is it because we value calm, reasoned reactions over emotional responses? Because a quality system promotes a blame-free approach and focuses on systemic problems? Because we believe, "Let he who is without sin cast the first stone"?

This blog is about how we as a profession responsible for the safety of the blood system deal with physicians who order inappropriate transfusions and how that process compares with how we deal with other health professionals who do not follow established best practices. Besides the Transfusion papers, an Australian report featured the same problem:
The two Transfusion papers that caught my eye were
The first paper was a magnet - who can resist the notion of an immaculate transfusion? The authors describe the immaculate transfusion as issuing blood for transfusion when no order to transfuse exists.

After their transfusion service began to require a photocopy of a physician's order to transfuse blood before issuing products for transfusion (except when the need for rapid issue was more important than full compatibility testing), they audited blood requests for compliance to having a written physician order. As would be expected, they found various inappropriate blood requests, including some immaculate transfusions.

My guess is that they mean a virginal transfusion, since immaculate conception is a doctrine that believes that Mary, biblical mother of Jesus, was born without original sin. But
immaculate transfusion has a nice ring to it.

The second paper initially jumped out because of its Canadian authors, two of whom I know. But more significantly, I have always been interested in cryo, a wonderful discovery by Judith Pool in 1965 that gave new life to people with hemophilia. See this reminiscence by Dr. Martin Inwood:

Cryo's downfall?
Cryo became associated with the tainted blood scandal of the1980s and its legitimate uses became fewer with the development of safer alternatives (e.g., recombinant Factor VIII and desmopressin, aka DDAVP). Over time, it seemed that cryoprecipitate and its association with the AIDS tragedy
had become an outcast, a black sheep of the blood component family, admittedly nothing like the association of FVIII concentrate prepared from huge plasma pools of tens of 1000s of potentially infected donors.

Somewhat surprisingly, cryo continues to be used extensively, and inappropriately in many cases, as has been documented in this large Canadian study.
To keep with a biblical analogy, (I know it's a stretch), it is almost as if cryo was a prodigal son , a product that behaved badly yet was somehow rehabilitated.
  • If you like music, listen for the prodigal son reference as the Dubliner's sing Wild Irish Rover.
The study in the October 2008 issue of Transfusion involved a 2 month audit of cryoprecipitate use in 25 major Canadian hospitals. Cryoprecipitate transfusions were designated
  • appropriate if a fibrinogen level (taken 6 hr before/after transfusion) was not more than 1.0 g/ L
  • inappropriate if the pretransfusion fibrinogen level was more than 1.0 g/L and posttransfusion fibrinogen level was more than 1.0 g/L or not performed
4370 units of cryoprecipitate were transfused in 603 events to 453 patients representing 62% of cryoprecipitate issued to Canadian hospitals during the study.

Allowing that current recommendations for appropriate cryo transfusion are based on the opinion of TM experts, since no randomized controlled clinical trials have been done, the authors use the criterion of bleeding and significant hypofibrinogenemia (fibrinogen 0.8-1.0 g/L), which is generally considered appropriate.

Results. Using a fibrinogen level of at or below 1.0 g/L as "appropriate," (and ignoring whether patients were bleeding):
  • 24% of cryoprecipitate transfusions were considered appropriate
  • 34% were inappropriate
  • 42% - appropriateness could not be determined
The authors conclude that "A 2-month audit of cryoprecipitate use in Canada revealed that most cryoprecipitate use in Canada is not in accordance with published guidelines." They note that, "During the period of our audit, up to 72 percent of patient exposures likely only carried risk, with no perceivable benefit."

What are inappropriate transfusions?
These papers discuss unauthorized transfusions and inappropriate transfusions. If unauthorized transfusions are immaculate conceptions, what analogies apply to inappropriate transfusions? Inappropriate transfusions include
  • transfusing the wrong blood component (e.g., transfusing plasma as a blood volume expander instead of non-blood products such as saline, pentastarch, et al., which also need to be transfused appropriately but carry no infectious risks)
  • infusing the wrong dose (e.g., transfusing 4 RBC when 2 would achieve the desired clinical effect)
  • transfusing a blood component or product when transfusion is not clinically indicated
Why should we care?
Why should we care about transfusions that result in no immediate adverse event or in adverse events that are hard to document? Inappropriate transfusion implies that the treatment may
  • not do what it is supposed to do, namely improve the patient's condition or prevent further deterioration
  • potentially harm patients by exposing them to infectious disease risks and other adverse events
  • needlessly increase the cost of health care by wasting scarce blood products
According to this BMJ paper, reasons to reduce blood exposure include
  • Immunological complications
  • Red cell alloantibodies: haemolytic transfusion reaction
  • HLA antibodies: refractoriness
  • Transfusion related acute lung injury, post-transfusion purpura, transfusion associated graft versus host disease, etc.
  • Errors and "wrong blood" episodes
  • Infections (bacterial, viral, and possibly prion)
  • Immunomodulation (risk of infection or malignancy)
  • Litigation
  • Limited resource
At their worst, inappropriate transfusions constitute an unneeded treatment that is forced upon an unsuspecting victim who may or may not be harmed. At their best, who knows? Even if all they do is waste money, that is money that could be spent more effectively elsewhere.

Frankly, if I were to be transfused inappropriately with any blood product and contacted a transmissible disease (admittedly a low risk, but not zero) or suffer any adverse event, I would not think kindly on the physician who prescribed the transfusion. Even without any harm being done, I would resent the needless exposure to potential risks. Is this reasonable accountability? I would also wonder about how the inappropriate order got through the transfusion system, which presumably has processes to monitor and prevent inappropriate transfusions (more on this later).

Is this playing a non-productive blame game? Should I just say, "What the heck. My doctors were doing what they thought best, even though they were ignorant of good transfusion practice." Is ignorance a valid excuse?

How to detect, identify, and prevent inappropriate transfusion
The first step in prevention is to detect problems and classify them. With transfusion-associated adverse events, this is accomplished by hemovigilance systems.
With inappropriate transfusions, identification is harder. Such transfusions may not result in immediate harm or may result in adverse events that are difficult to link directly to transfusion as a significant cause. As well, the harm may be impossible to identify except by large studies, which, even then, cannot prove causation definitively.

With inappropriate transfusions, the identification process is usually done by developing standards against which ordering behavior is measured, then auditing performance for compliance with defined best practices and standards.

All of this takes resources and expertise, both of which are increasingly scarce.

We are in the infancy stage but have developed several tools.

Clinical indications
One tool is to include clinical indications on the transfusion request form, along with relevant hematogic and related criteria. Some examples:

  • RBC: Hb less than 70g/L or Hct. less than 21%
  • Platelets: less than 20 x 109/L
  • Cryoprecipitate: Bleeding and fibrinogen less than 1.0 g/L
Presumably, the inclusion of such guidelines on the form gives clinicians the clue that these are accepted indications for transfusion of a particular product and that they should carefully consider transfusion if their patient's situation falls outside the guidelines, realizing that a patient's overall clinical condition is paramount in decisions to transfuse.

Another blood utilization tool is the maximum surgical blood order schedule (MSBOS), which has been around for more than 30 years. See this example.

Transfusion service as gatekeeper
Transfusion service clinical laboratory scientists/technologists are expected to act as gatekeepers when clinical indications and data are not completed on forms or seem inappropriate, or when orders exceed the MSBOS. This puts them in an awkward position, for physicians generally do not like their decisions questioned by anyone, especially by non-physicians. TM labs generally contact the medical director or designate and have them deal with clinicians on a peer-to-peer basis.

However, sometimes circumstances dictate direct contact with clinicians. Ask any experienced technologist or nurse who has ever questioned a physician's order, usually indirectly, in the most respectful, even subservient manner, what the experience is like. In many cases, the experience is not pleasant, and sometimes down-right abusive.

From my experience surveying technologists across Canada who work in transfusion service laboratories about what happens when clinicians order questionable or clearly inappropriate transfusions, there are several possibilities:

  • Nothing at all happens. The inappropriateness may go unnoticed, may be ignored because staff are too busy, or may be ignored because a culture of questioning a doctor's orders is not encouraged.
  • If the blood order is drastically inappropriate and would clearly cause immediate harm, the TS medical director is contacted, intervenes, and manages to protect patient safety.
  • If the blood order does not correspond to current best practices and is inappropriate according to the TS's criteria for clinical use of a particular product, the TS medical director may or may not intervene.
The bottom line is that, if the clinician persists, the technologist is often told,
  • "Just give the blood."
Presumably, this happens because ultimate responsibility for patient treatment, including transfusion, rests with the attending physician AND, as would be expected, TS staff do not like having a policing role.

Other tools
Many other strategies and tools exist to promote safe and effective blood utilization but are beyond the scope of this blog. They include hospital transfusion committees, targeted education by Transfusion Safety Officers, hospital grand rounds, educational conferences, wallet cards with reminders of transfusion guidelines, websites with transfusion guidelines, etc.

Clinician education on appropriate blood use remains a significant challenge. An earlier blog a year ago focused on this topic:

The blog discusses a paper (Dzik S. Use of a computer-assisted system for blood utilization review. Transfusion 2007;47(s2): 142S) dealing with computerized physician order entry and computer-assisted blood utilization review and feedback.

In brief, the educational intervention in the paper is as follows:

  • A TM physician reviews daily reports of transfusion orders and clinicians who made questionable decisions to transfuse are targeted for education.
  • A non-judgmental e-mail is sent to the clinician within 24 hours of the decision to transfuse.The e-mail displays the pre- and post-transfusion lab data; provides the criteria for the review process; and links to an in-house educational site.
  • On the website, for each blood component the physiology of blood use is summarized and articles on the clinical use of blood are listed. Each article is linked to a summary of the paper's findings and to the published paper.
  • The e-mail invites the physician to reply if there are questions or concerns.
The author's innovative strategy uses a non-judgmental approach to educating physicians, which fits with the quality systems approach of being non-punitive and staying away from the blame game.

Other studies have shown varying degrees of effectiveness of educational interventions:

What is the responsibility of the TS medical director in preventing inappropriate and potentially harmful transfusions? What is the responsibility of ordering physicians? What is the responsibility of lab technologists and nurses who may notice that a doctor's transfusion order seems questionable?

The fact is that we all have varying degrees of responsibility.

My experience is that transfusion services have dedicated professionals who fight the good fight and achieve steady nano-progress with educating clinicians who order transfusions. Most clinicians have little transfusion education in their undergraduate programs - sometimes it amounts to only a few hours. This leaves a huge gap for transfusion services AND the physician to bridge.

What happens to physicians personally when they order inappropriate transfusions? In the absence of immediate patient harm, apparently, not much. The current approach is to improve a clinician's transfusion practice through gentle face-saving educational interventions designed to motivate voluntary changes in behavior. Maybe this is all that can be expected.

But how much responsibility do clinicians share for being knowledgeable about a treatment they provide? Decisions about transfusing patients are difficult and often made under stressful conditions. That said, a hallmark of educated professionals is to know what they do not know, to actively pursue continuing education, and to seek help when needed.

If technologists and nurses were to consistently break accepted best practices, and such behavior put patients at risk, what would happen? Would they receive nothing but educational interventions? Perhaps. For how long would repeated deviations be tolerated? In some regards,
technologists and nurses are in a different position, because best practices and guidelines are often documented in written SOPs used in the training of new staff and these SOPs must be followed.

blood safety standards specify that blood transfusion training is mandatory for all staff involved in transfusion practice. But apparently such training is not mandatory for physicians who prescribe blood transfusion.

Are physicians treated differently because of their status in the health care system or are they treated the same as other professions? Other than offering multiple voluntary educational opportunities and providing gentle education after the fact for those who order inappropriate transfusions, what strategies and tactics can improve physician transfusion practices? These are questions to ponder.

Is the current situation - ongoing inappropriate transfusions for multiple blood components and products - the natural result of what happens when there are no consequences for inappropriate behavior? Transfusions have the ability to do much good and to save lives but carry risks.


Today, blood utilization tends to focus on cost control of expensive plasma protein products such as albumin and IVIG. For example, in Canada for the year ending Mar. 31, 2008, IVIG was the largest single expense at $137.4 million, almost 16% of the total budget that neared $900 million.

See utilization management for some of the projects that have been done in this area.

Cost control is essential to sustain the entire blood system, but how can the focus on the most expensive products be better balanced with an overall utilization strategy? One idea is to use a variation of the original Pareto Principle (80-20 rule that 80% of the wealth was owned by 20% of the population) - namely, that 80% of consequences stem from 20% of the causes.

Should we try to prioritize? To identify the 10, 15, or 20% of the physicians or medical services that cause most (60, 70, or 80%?) of the inappropriate transfusions, regardless of cost, and target education at them?

We also need to balance the need for blame-free interventions with processes that better motivate clinicians to assume responsibility for becoming knowledgeable about transfusion practice. Assigning blame causes resentment and anger; it tends to drive people underground and undermine error prevention strategies.

However, should not blaming people equate with having no consequences, especially where patient safety is involved? This issue has been examined in

The author writes:
  • In my experience as both a regulator and safety exponent, systems issues usually accompany breaches of professional responsibility (weak regulations, reporting requirements, or inadequate training). It depends how you look and where.
  • A root cause analysis would nearly always identify systems problems and rarely individuals. Systems failures may also mitigate the level of responsibility for the individuals.
  • Where and how professional responsibility fits into the ‘‘no blame’’ culture is unclear. How can we make it clearer?
Whether immaculate, virginal, or something more "sinful," inappropriate transfusions remain an ongoing challenge around the globe.

One of the key issues is responsibility.

Take a peek at this rendition of Responsibility by the punk band MxPx. All docs ordering transfusions - Tx Rx?
- should take note of their behavior, as in the MxPx lyrics:
  • Responsibility? What's that?
  • Responsibility? Not quite yet.
  • Responsibility? What's that?
  • I don't want to think about it; we'd be better off without it...
Comments are most welcome BUT, due to excessive spam,  please e-mail me personally or use the address in the newsletter notice. .