Background
Albumin has historically been used for a wide variety of indications. However, prescribing albumin for many of these indications is not supported by current evidence or is controversial or only weakly supported by the available literature.[1-4] 

Standardized use of al­bumin is necessary to ensure maximal benefit to patients while minimizing unnecessary exposure to blood products and their associated risks (allergic reactions, circulatory overload, and transmission of bloodborne infectious agents). At $2.60 per gram (Can­adian dollars),[5] albumin is substantial­ly more expensive than the alternatives, which include crystalloids (such as normal saline or Ringer lactate) and synthetic colloids (such as the hydroxyethyl starches), entailing a significant financial burden for a product of questionable clinical benefit.

In 2003, the BC Transfusion Medicine Advisory Group (TMAG) audited albumin prescribing practices in British Columbia and found that albumin was being used for indications that were considered unsupportable by literature-based guidelines in other jurisdictions.[6] 

In 2004, the SAFE study,[7] a randomized controlled trial of albumin versus normal saline for fluid resuscitation in critical care pa­tients, found no significant difference between the two groups with respect to 28-day mortality, duration of mech­anical ventilation, or hospital length of stay. 

These studies provided the stimulus for the development of the Guidelines for Albumin Use in Adults in British Columbia, a joint effort of the BC Provincial Blood Coordin­ating Office, TMAG, and a group of acute care physicians. The guidelines[8] were released in October 2007 and are available online at www.pbco.ca [13].

Given that numerous studies have shown that albumin is used inappropriately most of the time,[9-11] it was hoped that the introduction of locally developed guidelines would result in a change in prescribing practices. 

There were early indications that this might be the case: in the calendar year following release of the guidelines, the amount of albumin issued to BC hospitals by the Canadian Blood Services decreased by 9.7%, while albumin use across Canada (excluding Quebec) rose by 3.6% (oral communication with J. Unrau, hospital liaison specialist, Canadian Blood Services BC and Yukon, 2 December 2011). 

In order to examine the effect that the introduction of guidelines for the use of albumin has had on BC prescribing practices, we performed another province-wide audit of albumin use in 2009. We hypothesized that the proportion of albumin used inappropriately (i.e., for indications not named in the guidelines or those categorized in the guidelines as “Indications for which albumin should not be used”) would decrease compared with the baseline established in 2003.

Methods
After the study design was approved by the research ethics board and institutional review board of each participating site, the audit began.
Twelve of the thirteen hospitals that participated in the original 2003 audit were asked to participate in the 2009 audit. The Children’s and Wo­men’s Health Centre of BC was ex­cluded because the 2007 guidelines are for adult patients only, and most of the albumin at that site is used for pediatric patients. 

The hospital transfusion services at all sites were asked to complete a form for each albumin order filled during the study period: 1 October to 1 December 2009. The form included fields for the hospital, the type and amount of albumin ordered, the specialty of the ordering physician or service, and the clinical indication for the product. To participate, hospitals were required to collect data for at least 1 month, and were then asked to collect data for a full 2 months. 

At the end of the study period, the anonymized completed forms were returned to the Provincial Blood Coordinating Office for entry into a database and further analysis. The specialty of the ordering physician and in­­dications for use were then compared with data from the 2003 audit, which were reorganized to include only the hospitals that participated in both audits. 

Finally, transfusion medicine direc­tors and chief technologists at each site were asked to describe how the Guidelines for Albumin Use in Adults in British Columbia document was dis­­tributed within their respective hospitals and whether additional education about the Guidelines was provided.

Results
Nine hospitals in four of BC’s five health regions agreed to participate in the study. All hospitals began data collection on 1 October 2009. Eight hospitals collected data for 2 months, and one hospital collected data for 1 month. A total of 315 forms were re­ceived. 

Eight of the hospitals reported a capture rate of 90% to 100% of their albumin use; one hospital reported capture of approximately 50% of use. As this hospital accounted for less than 2% of provincial albumin use and for only 1% of the albumin used in the study, the incomplete data capture does not affect the conclusions.

The total amount of albumin or­der­ed during the 2009 audit was 67277.5 grams versus 47032.3 grams for the same hospitals during the 2003 audit. Because hos­pitals collected for different lengths of time in the first study, these amounts are not directly comparable. The amounts can, however, be used as denominators for determining the proportion of appropriate albumin use in each year.

Use by indication
The indications for albumin use ac­cording to the guidelines are shown in Table 1 [14]. The 2003 values have been reformulated to include only the hospitals that participated in both studies. Indications have been divided into three categories: 

(1) Appropriate use (i.e., indications are listed in the 2007 guidelines under the heading “Indications for which albumin may be used”); 

(2) Inappropriate use—specified (i.e., indications are listed in the guidelines under the heading “Indications for which albumin should not be used”); 

(3) Inappropriate use—not specified (i.e., indications are not specifically mentioned in the guidelines because no evidence supporting the use of albumin for them exists). 

The amount of albumin used in 2009 considered to be appropriate according to the 2007 guidelines in­creased by nearly 35% when compared with the amount used in 2003. Indeed, the top three indications in 2009 were all considered appropriate: 

(1) postoperative volume expansion for cardiac surgery; 

(2) plasmapheresis; and 

(3) paracentesis and ascites. 

However, the majority of orders submitted for albumin in 2009 were still for “inappropriate” indications and only 23.5% of all orders in 2009 were for appropriate indications.

Use by specialty
The specialties of the physicians who ordered albumin are shown in Table 2 [15]. The only major change from 2003 was a large increase in use by internal medicine. However, this is almost completely offset by a large decrease in use by nephrology, and the changes may be the result of coding the same practitioners differently. 

As in 2003, the major users of albumin were internal medicine, cardiac surgery, general surgery, and hematology. However, it is important to note that cardiac surgery use moved from primarily inappropriate use in 2003 (cardiac surgery: pump prime) to appropriate use in 2009.

Educational survey
Seven of the nine hospitals completed the survey about disseminating the 2007 albumin guidelines and providing additional education. Five sites distributed the guidelines via hard copy, one site distributed the guidelines via e-mail, and one site distributed the guidelines via multiple methods (e-mail, hard copy, posted on intranet, and integrated into a preprinted order form). 

Two hospitals distributed the guidelines to all physicians, three hospitals distributed only to the specialties that typically use albumin, one hospital only to nephrologists and intensivists, and another hospital only to nephrologists. Two hospitals provided additional education about the guidelines: one held grand rounds for all physicians and the other held resident rounds.

Conclusions
Albumin has been shown to have uncertain or no benefit over conventional intravenous fluids for many of the indications for which it is most commonly used, yet the overall use of albumin continues to rise in Canada. The amount of albumin issued to Canadian hospitals (excluding Quebec) grew 51.8% between 2003 and 2009 (oral communication with J. Un­rau, hospital liaison specialist, Canadian Blood Services BC and Yukon, 2 December 2011).

Two years after the introduction of evidence-based guidelines for albumin use in BC, the results are both heartening and discouraging. On the positive side, the total amount of albumin being used for appropriate indications has risen dramatically, from 22% to 56%. The top three indications are now in the appropriate category, in contrast to 2003 when the top three indications were in the inappropriate category. This could be interpreted as evidence that practice is changing toward more evidence-based use of albumin.

However, looking at the number of orders received for each indication in the 2009 audit, it is discouraging to see that 39.4% of orders were for indications specifically listed in the guidelines as “Indications for which albumin should not be used” and an additional 37.1% of orders were for indications that were inappropriate, although they were not specifically named in the guidelines. 

It should be noted that this is a “best-case scenario,” as the ordering clinicians were given the benefit of the doubt when the data were analyzed. For example, the indication “hepatic resection” was considered appropriate even though the actual indication in the guidelines is “post-operative liver resection pa­tients with concomitant ascites.”

The apparently suboptimal impact of the albumin guidelines may have been due to ineffective strategies for implementation or dissemination of the information. 

The delay between the 2007 release of the guidelines and the 2009 audit may mean that a temporary improvement in albumin use was not captured: there are hints of this in a 9.7% decrease in the amount of albumin issued to BC hospitals in 2008 followed by a 0.8% increase in 2009 (oral communication with J. Un­rau, hospital liaison specialist, Canadian Blood Services BC and Yukon, 2 December 2011). It is known that clinical education strategies require frequent reinforcement to continue altering practice.[12]

In summary, the introduction of evidence-based guidelines in British Columbia has led to mixed results. Although the amount of albumin be­ing used for appropriate indications has increased since the introduction of provincial guidelines, the majority of orders for albumin continue to be for inappropriate indications, thus highlighting the importance of continuing education regarding the costs and risks of albumin use and the need to follow evidence-based guidelines.

Competing interests
None declared.

Acknowledgments
The authors would like to acknowledge the technical and medical staff of all the hospital transfusion services who contributed data. As well, the authors acknowledge the support and counsel of the BC Transfusion Medicine Advisory Group members in the planning of both audits and in the development of albumin guidelines. Finally, thanks go to the staff of the BC Provincial Blood Coordinating Office for their help with study design, data entry, data analysis, and manuscript preparation.

Dr King is a staff anesthesiologist at Royal Jubilee Hospital and Victoria General Hospital in Victoria, BC, and a clinical instructor in the Department of Anesthesiology, Pharmacology, and Therapeutics at the University of British Columbia. Dr Roland is a staff hematopathologist at Vancouver General Hospital in Vancouver, BC. She is also an assistant professor in the Department of Pathology and Laboratory Medicine at UBC. Ms Selin is manager of utilization management for the British Columbia Pro­vincial Blood Coordinating Office in Vancouver. Dr Chipperfield is regional medical leader for Blood Transfusion Medicine, Vancouver Coastal Health, and a staff hema­topathologist at Vancouver General Hospital. She is also an assistant professor in the Department of Pathology and Laboratory Medicine at UBC. Dr Morrison is medical director of Transfusion Medicine for Fraser Health and chair of the Transfusion Medicine Advisory Committee of BC. He is also an assistant professor in the Department of Pathology at UBC.