Hospital Pharmacy - June 2020 - 160

160	

Hospital Pharmacy 55(3)

Figure 4.  Count of pleural effusions diagnoses during changes in albumin utilization

baseline adherence is low.25 We had a sizable opportunity
with annual albumin costs over $1million. There were no
significant changes in monthly census during the project, and
our acquisition cost of albumin remained the same. As such
the decreased albumin expenditure can be attributed to a
change in usage patterns. Admittedly, we had a receptive
audience with broad agreement on the need to control costs,
indicated by the response from clinicians volunteering to
participate in the task force.36 This suggests that overt consensus on the overarching enterprise strategy is crucial for
such projects.36-38
Our cost transparency intervention involved informing
our clinicians of our albumin unit acquisition and total annual
costs. This was necessary as clinicians' estimates of drug and
health care service costs are inaccurate, as we too had found
in our informal surveys.39 In addition, we shared unit-level
monthly data on albumin expenditure. This approach is distinct from typical cost transparency interventions that merely
expose clinicians to the cost of drugs or tests at the point of
ordering them. Such information is devoid of context:
Whereas any individual prescription may have a modest or
high cost, the clinician remains unaware of the cumulative
budget impact of their and others' prescriptions to the institution, or even to the patient. We speculate that the initial
downtrend in albumin costs from April through June 2013
was influenced by clinicians becoming aware of just how
much was being spent monthly on albumin. This is concordant with other reports of cost transparency, which have
shown mixed results.17,20,24,40 Arguably, the initial downtrend
we observed could have been due to common cause variation
and not necessarily the cost transparency intervention.
However, as subsequent months showed, this downtrend

marked the beginning of a run of sustained decreases in albumin utilization.
The educational sessions aimed to challenge prior
assumptions, impart new knowledge, drive change, and consolidate new behavior. We also sought to achieve consensus
around the evolving literature on albumin-clarifying
absence of evidence of benefit, even if no proven harms, yet
higher cost. Contemporaneously, European and American
regulatory agencies enacted limitations and warnings about
hetastarches (for increased bleeding and renal injury); this
may have bolstered the plausibility of harm from 25% albumin.41 Viewed in isolation, only a relatively smaller decrease
in usage was seen after these sessions, consistent with the
finding that knowledge transfer does not guarantee behavior
change.15 However, the education occurred against a background of ongoing cost transparency. We can therefore infer
that the additional reduction reflected a response to repetition (albeit differently) and reinforcement of the message to
rationalize albumin usage.42 This is buttressed by the fact
that monthly costs from April to October exhibited a steady
downward trend, and all were below our baseline average of
$78 000. This trend of 7 or more consecutive data points all
below the baseline average is one of the control chart rulesof-thumb that indicate statistically significant special cause
variation-in this case attributable to the combination of cost
transparency and education.43
By late 2013, our albumin costs reached a nadir and seemingly plateaued. Perhaps message extinction, inertia, residual
uncertainty, or persistent unawareness hindered change especially for outlier clinicians. Alternatively, the concept of
using albumin sparingly was so discordant with their values
and preferences, revealed in prior prescribing patterns, as to
Hospital Pharmacy - June 2020

Table of Contents for the Digital Edition of Hospital Pharmacy - June 2020
