Article Text
Abstract
Background Intravenous albumin has limited indications supported by randomised controlled trials, yet it is often prescribed for indications not supported by evidence.
Aim To reduce unnecessary transfusion of albumin.
Interventions Under the leadership of a multidisciplinary quality improvement team, evidence-based recommendations were disseminated in tandem with a new electronic order set, an educational strategy, qualitative interviews with prescribers and a return policy change to reduce wastage.
Implementation and evaluation Interventions were introduced in a staggered fashion. The primary outcome, appropriate use of albumin, was monitored and quantified using pre-intervention and post-intervention audits. Process measures included statistical process run charts of monthly usage of 5% and 25% albumin and wastage. Data on length of stay (hospital and intensive care), new inpatient starts on kidney replacement and mortality were collected as balancing measures.
Results Appropriate albumin usage based on indication increased from 30% to 50% (p<0.0001). There was significantly less overall albumin usage in the post-intervention period compared with the pre-intervention period (negative coefficient, p<0.0001), driven by a major reduction in the utilisation of the 5% formulation (p<0.0001). Overall albumin usage was significantly lower in the post-intervention period, decreasing from 800 to 450 vials per month. The intervention resulted in significantly less wastage (negative coefficient, p=0.017). Mortality, length of stay and new starts on kidney replacement therapy remained constant throughout the study period.
Conclusion Improved prescribing of albumin was achieved with a multifaceted approach. Substantial and sustained reductions in usage were achieved without negatively impacting patient-important outcomes. The estimated annual savings for the purchase cost of albumin was CAN $300 000. We provide a structured process for other organisations to optimise their use of albumin.
- Quality improvement
- Patient safety
- Audit and feedback
Data availability statement
No data are available.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Intravenous albumin is commonly prescribed to inpatients for non-evidence-based indications, and evidence-based quality improvement strategies are needed to align practice with evidence.
WHAT THIS STUDY ADDS
A multifaceted intervention can provide a substantial and sustained improvement in transfusion practice with marked cost savings.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Our multifaceted approach can easily be tailored to the context of other institutions to achieve improvements in albumin prescribing practice.
Introduction
Intravenous albumin is used in many clinical scenarios, either as an iso-oncotic (4–5%, 250–500 mL) or as a hyper-oncotic (20–25%, 100 mL) formulation.1 The iso-oncotic preparation is employed for volume expansion, and the hyper-oncotic formulation is used for restoration of oncotic pressure1 or mobilisation of fluid from extravascular spaces.2 Although used broadly in the hospitalised patient, albumin has only been shown in trials to have potential benefits in patients with complications of cirrhosis. Albumin reduces the incidence of paracentesis-induced circulatory dysfunction in patients undergoing large-volume paracentesis,3 although the impact on patient-important outcomes is less clear. Albumin appears to also reduce mortality and renal failure in patients with spontaneous bacterial peritonitis.4 Albumin is also commonly employed in hepatorenal syndrome,5 although no randomised trials of albumin compared with no albumin have been conducted.
Outside of patients with complications of cirrhosis, large randomised trials have compared intravenous albumin with crystalloids in critical care patients, adult and paediatric patients with sepsis and patients undergoing cardiac surgery.6–10 Collectively, these trials find patient-important outcomes with albumin no better than crystalloid but care markedly more expensive. More recent evidence also suggests that adverse consequences of albumin have been underappreciated, such as volume overload, anaphylaxis, more red cell transfusions and peripheral gangrene.6 8 11–13 These trials have called into question the common use of albumin in patients without cirrhosis.14
Published audits of albumin usage from different countries have found that albumin is regularly prescribed for non-evidence-based indications and is generally over-prescribed.15–17 Albumin utilisation doubled in Ontario, Canada, between 2000 and 2017, with only 22% of orders being transfused to patients with cirrhosis.16 To mitigate unnecessary albumin usage and wastage, we undertook a single-centre, multifaceted quality improvement project to align albumin prescribing with published evidence.
Methods
Context
At the time of project initiation, our institution was working towards the Choosing Wisely Hospital Designation, creating a generally receptive environment for improving utilisation. As such, this quality improvement (QI) project was identified by hospital administrative leads as a key organisational priority on the hospital’s quality improvement plan (QIP) to ensure alignment with the Choosing Wisely principles across the organisation.
Interventions
Using the Plan, Do, Study, Act cycle QI methodology,18 we undertook the interventions outlined in table 1. Due to the number of interventions and lack of control over the date of approval, the implementation was staggered for when an intervention was ready for release. For example, order set release dates were under the jurisdiction of the order set committee.
Working group
Stakeholders from across the hospital system were invited to participate in this QI project. These stakeholders included physicians, nurses, medical laboratory technologists, perfusionists, administrators and trainees from across the hospital system. The individuals comprising the working group reflected departments with high albumin usage to ensure that viewpoints on guidelines and strategies for the reduction of albumin usage were reflective of end users. The QI project working group also modelled optimal prescribing practice within their respective areas.
Guidelines
Clinical practice recommendations were drafted, circulated, modified and then posted. These recommendations were based on the most current published randomised trials of albumin usage. The new clinical practice guidelines included 13 evidence-based recommendations (see online supplemental material). The clinical practice guidelines were posted on the internal hospital website on 28 May 2021 after seeking input from the QI project working group and obtaining approval from the Transfusion Medicine Advisory Group and the hospital Medical Advisory Committee.
Supplemental material
Order sets
A computerised order entry system was developed on EntryPoint (Think Research, Toronto, Ontario, Canada) to standardise albumin ordering and to improve the tracking of usage metrics. Institutional procedures for creating novel order sets were followed. The predefined criteria for ordering albumin were created with best practices and evidence-based indications as per the draft International Collaborative Transfusion Medicine Guidelines (ICTMG) at the time of release for community consultation.19 The new order set was implemented on 16 September 2021, and its use was voluntary. Physicians could continue to order albumin on a paper physician order. This was done to avoid encroaching on individual physicians’ autonomy and respecting variations in workflow. Simultaneously, the existing paracentesis order set was revised to align with the dosing recommendation of the new albumin order set. The cardiac surgery admission order set for the post-cardiac surgery intensive care unit (ICU) order set was also revised to ensure concordance. Utilisation of the order set by the patient could not be monitored as the software could not be interrogated for this data.
Educational modules
Based on the new clinical practice guidelines, brief (5–10 min) educational modules were created via narrated PowerPoint. One module was created for each hospital department that had recorded albumin usage (gastroenterology, cardiac surgery, critical care, etc), specifying what was and was not indicated within their area of specialty. The modules were released in January 2022 on the hospital intranet. The use of the modules could not be tracked electronically to monitor uptake.
Return policy change
The existing institutional return policy for albumin stated that dispensed but unused IV albumin must be returned to the blood bank within 1 hour of issue. This ‘1 hour rule’ was derived from Health Canada return requirements for red blood cell units and was not intended for room temperature-stored products such as albumin. The return policy was updated to provide physicians and nurses 24 hours for return to the blood bank before being considered waste. The new return policy took effect on 13 April 2021.
Study of interventions
The primary outcome of this study was the appropriateness of albumin prescribing (as determined by indication, dose and formulation) by pre-intervention and post-intervention medical chart audits of albumin usage. The pre-intervention and post-intervention audits included two groups of 100 consecutive patients between 1 April to 22 April 2021 and 13 January to 17 February 2022, respectively. Audit data was collected by one of the authors (CMF).
Pre-intervention audit
A baseline medical record audit was completed for 100 consecutive patients receiving albumin transfusions from 1 April to 22 April 2021. This was done to first understand both patient and prescriber characteristics, such that our interventions could be targeted and tailor-made. Study data were collected and managed using the REDCap electronic data capture tool. REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies, providing (1) an intuitive interface for validated data capture; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages and (4) procedures for data integration and interoperability with external sources. Information collected in the audit included patient demographics (ie, age, sex, reason for admission, length of stay, final disposition), prescriber demographics (ie, department, level of training) and prescription characteristics (ie, volume, formulation, indication, appropriateness of transfusion).
Qualitative interviews
Eight physicians were identified from the pre-intervention audit as heavy albumin prescribers and were contacted to participate in individual 30 min qualitative, structured interviews. These interviews consisted of 15–20 open-ended questions (see online supplemental material) to gain a deeper understanding of physicians’ thought processes when ordering and prescribing albumin, as well as to provide insight into institutional attitudes towards albumin usage, and potential strategies that could be implemented to align usage with guidelines. Interviews were recorded after obtaining consent, and a thematic analysis was performed to identify major themes surrounding albumin usage from the interviews. Interviews were performed until saturation was achieved.
Post-intervention audit
A repeat medical record audit of 100 consecutive patients of Kingston Health Sciences Centre receiving albumin transfusions was completed post-implementation of interventions, from 13 January to 17 February 2022, using an identical methodology as the pre-intervention audit. The data collected from the repeat audit were used to assess the primary outcome. Studies have shown that a minimum of 50 red cell transfusions must be audited at an institution to accurately determine appropriateness.20 Since albumin is transfused for a broader number of clinical scenarios, compared with red blood cells, we selected 100 audits to ensure the capture of less common indications.
Additional data collection
Following the completion of the post-intervention audit, data on existing parameters continued for repeat auditing and longitudinal assessment of the QI project’s success. Data for our process measures were obtained from the Canada Institute for Health Information Discharge Abstract Database.
Measures
Process measures included overall combined monthly usage of 5% and 25% formulations, monthly 5% formulation usage, monthly 25% formulation usage, monthly usage of the novel order set and monthly wastage. To ensure our interventions did not cause inadvertent harm to patients, we collected data on balancing measures, including intensive care length of stay, hospital length of stay, new initiations on dialysis and mortality during the pre-intervention and post-intervention periods. These process and balancing measures were collected from January 2017 to December 2022.
Analysis
Descriptive analysis was conducted by pre-intervention and post-intervention using mean, SD for age, frequencies and proportions for other categorical demographics and clinical characteristics. Statistical process run charts were used to track monthly usage of the 5% and 25% albumin formulations and total albumin wastage. χ2 or Fisher exact tests were used, as appropriate, for comparing the proportions of appropriate albumin usage based on indication between the pre-intervention and the post-intervention periods. To search for significant time trends of albumin usage (ie, total, 25% and 5% formulations) and albumin wastage, linear regression analysis was conducted over months in the pre-intervention period and the post-intervention period, respectively. We also compared the albumin usage and wastage between post-intervention and pre-intervention using linear regression models. Albumin usage and wastage were graphed monthly in the pre-intervention and post-intervention periods. Generalised linear regression models (Poisson distribution with log link function) were used to assess balance measures. The fitted regression lines and 95% CIs for the predicted values were also displayed on the graphs. P values <0.05 were considered statistically significant. All analyses were conducted using Statistical Analysis Software (V.9.4) and R package (V.4.3.0).
Ethical considerations
No conflicts of interest or ethical concerns were identified. The study did not require Research Ethics Board approval, but was registered as a QI project at the institution.
Results
Albumin usage audits
The pre-intervention and post-intervention audits each included 100 consecutive patients receiving albumin (table 2). Appropriateness of use based on indication increased from 30% to 50% (p<0.0001). The appropriateness of the 5% and 25% formulations increased from 21% to 30% (p=0.24) and 46% to 66% (p=0.019), respectively.
Monthly albumin usage—interrupted time series analysis
There was no significant time trend for total albumin usage during the pre-intervention period. However, there was a significant decreasing trend in total albumin usage during the post-intervention period (slope=−0.32; p=0.046). There was significantly less overall albumin usage in the post-intervention period compared with the pre-intervention period (negative coefficient, p<0.0001) (figure 1A).
There was no significant time trend in the usage of the 25% formulation during the pre-intervention or post-intervention periods. There was also no significant difference in the usage of the 25% formulation between the pre-intervention and post-intervention periods (p=0.13) (figure 1B).
There was no significant time trend in the usage of the 5% formulation during the pre-intervention period. However, there was a significant decreasing trend in 5% formulation usage during the post-intervention period (slope=−0.32; p=0.011). There was significantly less 5% formulation usage in the post-intervention period compared with the pre-intervention period (negative coefficient, p<0.0001) (figure 1C).
Albumin wastage
There was no significant time trend in total albumin wastage during the pre-intervention period. However, there was a significant decreasing trend in albumin wastage during the post-intervention period (slope=−0.07; p=0.0006). There was significantly less wastage during the post-intervention period compared with the pre-intervention period (negative coefficient, p=0.017) (figure 1D).
Qualitative interviews
Targeted interviews with users of albumin yielded four key themes surrounding plausible contributors to overprescribing at our institution (table 3).
Balancing measures
There were no significant time trends for the proportion of deaths (p=0.5448), hospital length of stay (p=0.6814) or ICU length of stay (p=0.7576) (figure 2). Additionally, we did not find any significant time trends for new inpatient dialysis starts (p=0.21) (figure 3).
Sustainability
Continued monitoring of albumin usage per month after completion of the quality improvement project in December of 2022 found sustained improvement in utilisation (online supplemental figure 1), without return to baseline utilisation.
Supplemental material
Discussion
In this study, we demonstrated that substantial and sustained reductions in unnecessary albumin transfusions can be achieved at a tertiary care centre using a multifaceted QI framework. Through our multifaceted intervention, we improved appropriateness, optimised prescribing (dose and formulation), reduced overall utilisation and decreased wastage, without impacting patient-important outcomes (mortality, length of stay, rates of kidney replacement therapy). Our study finds that it is possible and safe to markedly de-prescribe albumin in the inpatient population.
While both global albumin use and albumin product prices are rising, along with international shortages,21 demand has only increased, with the mean global per capita consumption of albumin products increasing by 7% from 2012 to 2016.22 Canada has followed this global trend, with Ontario, Canada’s most populous province, seeing a more than doubling of albumin usage over the past two decades.16
In addition to the growing concerns regarding the rising clinical usage and cost of albumin, there also exist ethical concerns surrounding its unnecessary use. Canada, like many other developed countries, relies heavily on manufactured blood products from US-paid donors.23 Paid donation centres in the USA often target low-income individuals, with collection centres being commonly located in low-income areas and along the Mexican border.24 Additionally, the financial incentive of paid donation creates the risk of donors misleading clinics about their eligibility. Indeed, one plasma donation centre in Ohio found that 13% of donors reported having misled clinic workers about their health to donate to receive compensation.25 These concerns are further compounded by the risk of adverse reactions from donation, with female donors being at higher risk than male donors.26
It is therefore essential that hospital use of albumin remains in-line with guidelines and evidence. This study may serve as a foundation for other hospitals to reduce adverse transfusion events, achieve significant cost savings, reduce unnecessary prescribing and align clinical practice with the evidence. Our efforts in this QI project resulted in an estimated annual cost savings of CAN $300 000 just for the purchase cost of albumin. Additional unmeasured savings include decreased technologist time, decreased nursing time, decreased porter runs to the blood bank and fewer infusion sets.
Using the Plan, Do, Study, Act cycle method, our QI project incorporated several distinct interventions to address unnecessary albumin orders. These interventions were derived from an initial 100-patient audit, in conjunction with prescriber interviews, which were used to understand the scope of the problem and plan areas in which targeted interventions could be deployed. Characterising usage patterns at our centre was of critical importance, as previous QI studies looking at department-level albumin use have demonstrated varying causes of overuse.16 17 Both our audit and qualitative interviews suggested that the primary drivers of unnecessary usage were gaps in knowledge on the use of crystalloids versus colloids, a deeply ingrained culture of liberal albumin prescribing and a lack of knowledge about albumin-associated adverse effects. As such, our intervention centred around two main strategies: knowledge translation via educational seminars and readily accessible ordering guidelines. Following the formal introduction of department-specific teaching modules and a hospital-wide ground rounds teaching session, we saw a significant decline in overall albumin usage, as well as a significant increase in appropriate ordering.
Our study also showed that the rollout of albumin usage guidelines contributed to a reduction in overall usage, and an increase in appropriate ordering, in the period between the pre-intervention and post-intervention audits and beyond. This finding is consistent with a study by Martelli et al from 2003,27 in which the authors assessed the impact of formal albumin usage guidelines on overall ordering at three Italian hospitals. The authors found that albumin ordering declined in the two hospitals that had adopted the guidelines and increased in the hospitals that had opted out of adopting the guidelines. Conversely, a recent study by Udeh et al in 202028 which used a variety of interventions to reduce albumin usage in critical care units (including cost transparency, provider education, individualised audits and provision of feedback) found success despite foregoing usage guidelines. Interestingly, the authors note that this may have been due to the baseline attitudes of their centre’s prescribers, who were interested in, and highly receptive to, curtailing albumin usage. This is in contrast to the findings of our qualitative interviews with a subset of prescribers at our centre, who were more resistant to changing their ordering habits in spite of our guidelines and the evidence base from randomised controlled trials. Our results are consistent with the conclusions drawn by Udeh et al28 that the implementation of formal guidelines may be of greater benefit at centres where prescribers are less receptive to efforts to change albumin ordering. Furthermore, our QI project found that additional educational materials were required for departments that were highly resistant to change (posters and huddles for data feedback to nursing staff in critical care areas where there was a pro-albumin culture). These additional materials were ultimately successful in changing usage patterns, and centres hoping to use our methods should be aware of the potential need for tailored interventions for each department. Of particular benefit in these situations is the QI project working group. Leveraging existing relationships and receiving real-time insight from areas of high usage allowed us to adjust department-specific interventions to meet the needs of prescribers, as well as to provide bedside education to nurses and prescribers. The QIP working group functioned as a conduit for bilateral communication; that is, information flows into the group from individual members acting as representatives from separate departments, and information from the working group can also be disseminated to individual department representative members. Other studies have found success in improving albumin usage through the use of a multidisciplinary working group, and our results are consistent with these studies.29
An additional simple intervention that we included in our QI project was to update our unused product return policy. In addition, nurses in the cardiac critical care unit, who had previously been unaware that albumin not used by the return time was discarded, were educated on the return policy. By aligning our return policy with that of peer centres, we were able to achieve a reduction in wasted products. Adjusting existing return policies at individual centres to align with best practices is a simple and cost-effective way of reducing overall albumin utilisation.
Of note, other centres have employed the use of financial incentives for prescribers as a means of reducing albumin usage29; however, this strategy can pose additional ethical challenges, potentially discouraging the use of albumin in situations when clinically indicated. Several studies have also assessed the efficacy of pharmacist-led interventions for reducing inappropriate albumin usage.17 These interventions include pharmacist review for all orders and direct communication between prescribers and pharmacists for all albumin orders. While the majority of these studies have found success with these interventions, they are not readily translatable to every centre. In Canada, blood products are stored and dispensed by the blood bank rather than the hospital pharmacy. As such, we were unable to incorporate this proven, although labour-intensive, intervention into our QI project.
There are some limitations to our study. Our methodology was designed based on our institution’s high baseline usage and therefore may not be generalisable to centres with higher compliance with guidelines at baseline. Additionally, our methodology defined the appropriate use of albumin in the cardiac surgery department as having first failed 3000 mL of crystalloid or more, yet we remain uncertain regarding if and when albumin should be used as a resuscitation fluid in cardiac surgery.6 We also faced challenges in promoting widespread use of our new order set, as our centre does not yet have a hospital-wide integrated electronic medical record system, requiring physicians to print all albumin orders from the electronic system. As such, prescribers experienced frustration with unreliable printers requiring escalation of printer failures to the corporate information technology department. Prescribers were also resistant to using the order set when entering patient allergies was mandatory. This requirement was removed after receiving feedback, and an additional requirement to include the infusion rate was added at the request of the nursing staff. In order to facilitate buy-in, the QI project team needed to remain open to all suggestions and feedback put forward by the clinical teams to continuously improve our interventions.
We have demonstrated through this study that improving albumin use at a systems level can be done through the implementation of a site-specific QI framework. Our results highlight the utility of baseline auditing to characterise the scope of the problem and the importance of engaging end-users in designing, implementing and assessing interventions. Such engagement centred on qualitative interviews and the formation of a multidisciplinary working group to ensure that interventions were targeted and appropriate for our centre. A multifaceted approach was required to achieve marked deprescribing across all hospital locations, with additional targeted interventions in areas more refractory to change.
Data availability statement
No data are available.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants but an Ethics Committee(s) or Institutional Board(s) exempted this study. Quality Improvement project exempt from written informed consent by IRB.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
X @DrGDigby, @jeanniecallum
Contributors All authors contributed to the design of the quality improvement project, assisted with its deployment, reviewed the manuscript and approved the manuscript. CMF, SH, GD, SAS and JC drafted the manuscript. Data collection was performed by CMF, SH, SA and DM. JC, guarantor.
Funding Canadian Blood Services (Transfusion Medicine Research Program Support Award), funded by the federal government (Health Canada) and the provincial and territorial ministries of health.
Disclaimer The views herein do not necessary reflect the views of Canadian Blood Services or the federal, provincial or territorial governments of Canada. The funding sources had no role in the design of this study and will not have any role during its execution, analyses, interpretation of the data or decision to submit results.
Competing interests JC has received research support from Canadian Blood Services and Octapharma. SAS has received honorarium from AstraZeneca, Novo Nordisk, Otsuka and KVR Pharmaceuticals. GD has received research funding from MaRS/Merck & Co and from Pfizer, as well as a honorarium from Merck & Co for a speaking engagement and from AstraZeneca for participation in a working group.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.