Auditing Intraoperative Transfusions to Promote High-Value Perioperative Care

Anesthesiologists play an important role in perioperative quality improvement efforts, and the oversight of blood utilization and transfusion practice is a prime example.1 Given that blood transfusion is the most common procedure performed in US hospitals,2 and one of the top 5 overused procedures identified at a Joint Commission Overuse Summit,3 avoiding unnecessary transfusions is an obvious opportunity to reduce risks and costs for our patients. A primary method for promoting quality is the process of auditing medical records, with the aim of giving feedback to the medical care team, to encourage guideline compliance. The audit, in our opinion, is akin to one big psychology experiment. It often comes down to using “the carrot or the stick” approach, or some method of “policing” medical practice to improve guideline compliance. Some providers take feedback well and consider this process to be a learning opportunity, while others consider the audit to be a confrontation, questioning their clinical judgment. Despite these different views, the overriding important issue is to pick your battles carefully, by choosing the right cases that are most appropriate for auditing. In the current issue of Anesthesia & Analgesia, Dexter et al4 describe a novel method for selecting cases for auditing, with the focus on appropriateness of intraoperative red blood cell transfusions. The authors point out that audits with feedback are recommended in the AABB (formerly the American Association of Blood Banks) Patient Blood Management Standards5; however, the standards do not tell how to audit or which cases should be chosen for audit. The primary findings in the study by Dexter et al4 are that cases were more likely to be chosen for audit for a missing hemoglobin (Hb) measurement than for a Hb concentration higher than the designated threshold (>9 g/dL). In addition, the authors found that cases with a median intraoperative blood loss <500 mL (not ≥500 mL) were more likely to be chosen for audit. The authors also conclude that the decision to audit is more complex for intraoperative transfusions than for transfusions performed outside the operating room, and this is consistent with other studies where the intraoperative period is ignored when assessing and encouraging transfusion guideline compliance.6 The issues that make the intraoperative setting tricky for auditing, especially for assessing blood utilization, are multifactorial and are summarized in the Table. First is the aspect of “attribution,” which is who made the decision to transfuse—the anesthesiologist or the surgeon? This question arises all the time from people who are not familiar with the operating room. The standard answer is that this is a joint decision, which is supposed to result after a discussion. In our experience, when there is a hole in the vena cava, surgeons are not in the mood for having a discussion, especially about the latest randomized trial on transfusion triggers. The other tricky business that occurs in the operating room is that intravascular volume is in flux, hemodynamic instability is common, and there is interaction among the anesthetic depth, vital sign abnormalities, and vasopressor and fluid administration. In fact, acute blood loss without fluid administration results in little or no change in the Hb concentration. To add complexity, intravascular volume can be difficult to assess, and we are often treating the whole patient rather than an isolated Hb value when we decide to transfuse. Furthermore, to audit intraoperative transfusions retrospectively is challenging, since the auditor was not there on the scene when the decision to transfuse was made.Table.: Challenges of Auditing Intraoperative TransfusionsAn important conclusion by Dexter et al4 about the missing Hb measurement as a reason to audit cases happens to be a recurring theme with intraoperative transfusions. This occurs, eg, in many cases where an indwelling arterial or central venous catheter is not present for drawing blood samples. Drawing blood proximal to an intravenous, or through an intravenous, usually results in a diluted sample, which can predispose toward a falsely low Hb level and unnecessary transfusions. Besides, often the extremities are not even accessible during surgery. A true point of care test requiring only a drop of blood is ideal, and such devices exist. They are, however, often rationed or not allowed by the laboratory, due to expense, loss of quality control, difficulty interfacing results with the medical records, or difficulty in billing for such tests. Even when a blood sample is properly obtained by phlebotomy, the turnaround time can be too slow during a blood loss scenario, and ideally a “stat laboratory” is located close to the operating room to provide faster results. The other main finding in the study by Dexter et al,4 that cases with median blood loss <500 mL are more likely to be audited, makes sense because cases with substantial bleeding (≥500 mL) often do not need auditing, they need transfusions. For these cases, however, other methods such as autologous salvaged (cell saver) blood may prevent the need for allogeneic transfusion,7,8 which supports the need to audit cases individually to see if such blood conservation methods were appropriately deployed. Another reason that intraoperative transfusions are harder to audit and are often ignored in blood management efforts is that all but 1 of the 8 randomized trials on which the recent transfusion guidelines are based9 enrolled either postoperative or intensive care unit patients, for whom active bleeding was not an issue. None of these trials assessed transfusion triggers during the intraoperative period. Goodnough et al,6,10 who run a very successful blood management program with an overall decrease in blood utilization by 25%, routinely excluded intraoperative transfusions as well as patients who were actively bleeding when they designed their clinician decision support and best practice advisories. Although active bleeding is a valid reason to transfuse at higher Hb levels, all patients in the operating room are not actively bleeding. When a blood management program is run out of the blood bank, by blood bankers, these clinicians can be unfamiliar with the operating room environment and hesitant to question the clinical judgment of their surgical and anesthesia colleagues. When selecting cases for audit, for any quality improvement program, this implies at some point that manual chart reviews will be done. The manual chart review, however, is a whole new level of involvement, requiring someone with clinical knowledge, usually at a “higher pay grade,” who has time to dig through the medical records in detail. Caution is needed to prevent too many charts from being chosen for audit, which increases time and cost, as well as potentially evoking more confrontation from providers who think their clinical judgment is being unnecessarily questioned. If too few charts are chosen for audit, then the quality initiative might be unsuccessful, allowing guideline noncompliance to prevail. In conclusion, Dexter et al4 should be commended for tackling the intraoperative period, an area where some have feared to tread in terms of assessing and addressing the appropriateness of blood utilization. By doing this kind of quality work, anesthesiologists can reduce risks and costs while maintaining optimal outcomes, thus promoting high-value perioperative care. DISCLOSURES Name: Robert A. Sikorski, MD. Contribution: This author helped draft and edit the manuscript. Conflicts of Interest: None. Name: Nadia B. Hensley, MD. Contribution: This author helped draft and edit the manuscript. Conflicts of Interest: None. Name: Steven M. Frank, MD. Contribution: This author helped write the majority of the manuscript. Conflicts of Interest: S. M. Frank has received consulting fees from Haemonetics and Medtronic. This manuscript was handled by: Marisa B. Marques, MD.