- Email: [email protected]
Cost savings with interventions to reduce aerosolized bronchodilator use in mechanically ventilated patients
Cost Savings with Interventions to Reduce Aerosolized Bronchodilator Use in Mechanically Ventilated Patients James M. Camamo PharmD, Kurt Weibel PharmD, Terence O’Keeffe MB ChB MSPH, FACS, Yvonne Huckleberry PharmD, Brian Kopp PharmD, Conrad Diven MD, Brian L. Erstad PharmD, FCCP, FCCM, FASHP PII: DOI: Reference:
S0883-9441(14)00217-2 doi: 10.1016/j.jcrc.2014.05.016 YJCRC 51530
To appear in:
Journal of Critical Care
Received date: Revised date: Accepted date:
14 February 2014 28 April 2014 22 May 2014
Please cite this article as: Camamo James M., Weibel Kurt, O’Keeffe Terence, Huckleberry Yvonne, Kopp Brian, Diven Conrad, Erstad Brian L., Cost Savings with Interventions to Reduce Aerosolized Bronchodilator Use in Mechanically Ventilated Patients, Journal of Critical Care (2014), doi: 10.1016/j.jcrc.2014.05.016
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Cost Savings with Interventions to Reduce Aerosolized Bronchodilator Use in Mechanically Ventilated Patients
T
James M. Camamo, PharmD, Kurt Weibel, PharmD, Terence O’Keeffe, MB ChB, MSPH,
RI P
FACS, Yvonne Huckleberry, PharmD, Brian Kopp, PharmD, Conrad Diven, MD, Brian L. Erstad, PharmD, FCCP, FCCM, FASHP
SC
University of Arizona Medical Center, Department of Pharmacy Services; The University of
NU
Arizona College of Medicine; The University of Arizona College of Pharmacy, Department of
MA
Pharmacy Practice and Science
Manuscript Communication:
ED
James Camamo, PharmD
PT
Department of Pharmacy Services
University of Arizona Medical Center
CE
1501 N Campbell Ave
AC
Tucson, AZ 85724-5009
Email: [email protected]
Word count (abstract): 198 Word count (body of text): 2030
ACCEPTED MANUSCRIPT 2 Abstract: Purpose
T
The purpose of this evaluation is to describe the cost savings associated with multimodal
RI P
interventions aimed at reducing aerosolized bronchodilator use in mechanically ventilated patients without adversely affecting costs associated with length of stay.
SC
Materials and Methods
NU
Subjects were included in the analysis if they were > 18 years of age; on mechanical ventilation in the ICU; and received aerosolized bronchodilators. Patients were excluded if they had
MA
reversible airway disease, an indication needing bronchodilator therapy. Patient data was obtained using the University Health System Consortium Clinical Data Base/Resource
ED
ManagerTM to compare outcomes during two 6 month time periods separated by a 4 month
PT
intervention phase aimed to reduce bronchodilator use. Results
CE
There were no significant differences in age, sex and lengths of stay (observed and expected)
AC
between the pre-intervention and post-intervention phases. Based on whole acquisition costs, the total cost of bronchodilators dispensed to the adult ICUs over the 6 month post-intervention phase was reduced by $56,960 compared to the 6 month pre-intervention phase ($120,562 vs. $63,602, respectively). Conclusions Multi-modal efforts to restrict aerosolized bronchodilator therapy in mechanically ventilated patients were successful and led to sustained reductions in use that was associated with substantial reductions in cost, without affecting lengths of stay. Key words: bronchodilator, mechanical ventilation, cost savings
ACCEPTED MANUSCRIPT 3 Introduction Aerosolized brochodilators, are often prescribed to mechanically ventilated patients in the
T
intensive care unit (ICU), regardless of the presence or absence of reversible airway disease.
RI P
Early advocates for this therapy cite preliminary data demonstrating that beta-2 agonists reduce pulmonary edema and accelerate alveolar fluid clearance in patients with acute lung injury.1,2
SC
Despite the purported benefits, two multicenter randomized trials involving mechanically
NU
ventilated patients in mixed medical-surgical intensive care units have demonstrated no benefit and potential harm from aerosolized beta-2 agonist therapy.3,4 This led to the recommendation
MA
that aerosolized beta-agonists be limited to patients who have demonstrated reversible airway
ED
disease.5
The studies negating the purported benefits of aerosolized beta-2 agonists prompted us to
PT
reassess our current prescribing patterns. Preliminary data at our institution indicated that
CE
aerosolized bronchodilators were commonly prescribed to mechanically ventilated patients that did not have evidence of reversible airway disease. In addition to concerns about lack of
AC
efficacy and potential harm, the frequent use of aerosolized bronchodilators in our ICUs raised resource concerns that included personnel time and drug expenditures. In light of rising health care costs, the department of pharmacy services worked closely with the medical staff and nursing to reduce waste and promote the rational use of aerosolized bronchodilators in the ICU. The purpose of this evaluation is to describe the cost savings associated with multimodal interventions aimed at reducing aerosolized bronchodilator use in mechanically ventilated patients. Prior to conducting our interventions, we hypothesized that reductions in the use of aerosolized bronchodilators in ICU patients will reduce costs without adversely affecting lengths of stay.
ACCEPTED MANUSCRIPT 4 Methods This retrospective review was conducted at the University of Arizona Medical Center, a 420-bed
RI P
T
tertiary care facility with medical and surgical intensive care units in Tucson, Arizona. The project was reviewed by the IRB and considered exempt from further review.
SC
Subjects were included in the analysis if they were > 18 years of age; on mechanical ventilation
NU
in the ICU; and received aerosolized formulations of albuterol sulfate, ipratropium bromide or albuterol sulfate/ipratropium bromide. Patients were excluded if they had reversible airway
MA
disease, an indication where bronchodilator therapy is beneficial. Patient specific data was obtained using the University Health System Consortium (UHC)
ED
Clinical Data Base/Resource ManagerTM (CDBRM) to compare outcomes during two time
PT
periods separated by a period where multimodal interventions aimed at reducing utilization of aerosolized bronchodilators occurred. The multimodal interventions were implemented over a 4
CE
month time period beginning February 15, 2012. During this time, aerosolized bronchodilators
AC
were removed from ICU order sets. Additionally, waste reduction strategies were implemented to reduce lost or misplaced inhalers. These included labeling the inhaler with patient identifiers rather than the box; working with nursing and transport staff to insure inhalers stayed with the patient when transferred to different levels of care; and storing inhalers in consistent areas located in the ICU medication rooms. Finally, the clinical staff pharmacists partnered with physician attendings to educate residents on the appropriate use of bronchodilators in the ICU. Data from the pre-intervention phase was collected between July 1st, 2011 and December 31st, 2011 and data during the post-intervention phase was collected between July 1st, 2012 and
ACCEPTED MANUSCRIPT 5 December 31st, 2012. The time period for the post-intervention phase took place after all interventions were fully implemented.
RI P
T
The CDBRM is an online database maintained by UHC which contains discharge and line-item data from more than 200 UHC members and affiliate hospitals. For this project, the CDBRM
SC
was utilized to obtain patient demographics, inpatient length of stay, and intensive care unit (ICU) length of stay (LOS). Additionally, the expected values for LOS were collected using
NU
CDBRM, which makes adjustments for risks using a logistic regression model in combination
MA
with the 3M Health Information Systems, Agency for Healthcare Research and Quality comorbidity software and the UHC complication profiler to assign illness severity levels. In
ED
addition to patient demographics and outcomes, the CDBRM was used to extract data on direct costs, which estimates the cost of patient care using a cost-to-charge ratio based on costs and
PT
revenues from the Centers for Medicare & Medicaid Services (CMS). In addition to direct costs
CE
(e.g supplies, labor and medication), the total cost of all aerosolized bronchodilators during the two evaluation periods was determined using whole acquisition costs and the number of doses
Data Analysis
AC
dispensed to the adult intensive care units.
Data were initially entered into Microsoft Excel® and subsequently transferred to STATA® version 13 (StataCorp, College Station, Tx) for statistical analysis. Two sample Student’s t-tests and ANOVA were used for comparisons of continuous data, while the Fisher’s exact or chisquared test was used for comparisons of nominal data. These comparisons were performed for the pre- and post-intervention periods, as well as for bronchodilator versus no bronchodilator
ACCEPTED MANUSCRIPT 6 groups in the post-intervention period. Statistical significance was defined as p<0.05 for all testing.
SC
Pre-Intervention vs. Post-Intervention Comparisons
RI P
T
Results
There were no significant differences in age or sex between the historical control group
NU
comprised of consecutive ICU patients and the more recent cohort of patients evaluated after the multi-modal interventions to reduce bronchodilator use had been implemented (Table 1).
MA
Similarly, there were no significant differences in the lengths of expected or observed lengths of
intervention groups (Table 2).
ED
stay, or in the number of cases of ventilator-associated pneumonia between the pre- and post-
PT
Based on whole acquisition costs, the total cost of aerosolized bronchodilators dispensed to the
CE
adult ICUs over the 6 month post-intervention phase was reduced by $56,960 compared to the 6 month pre-intervention phase ($120,562 vs. $63,602, respectively). The annualized cost savings
AC
associated with multi-modal interventions designed to optimize aerosolized bronchodilator use in the adult ICUs is reported to be approximately $113,920. There was no significant change in the total direct costs between the pre-intervention and post-intervention phases ($53,483 + 97,300 vs. $47,000 + 49,650 p=0.196, respectively). Bronchodilator vs. No Bronchodilator Comparisons There were 315 patients who received aerosolized bronchodilator therapy and 168 patients who did not receive aerosolized bronchodilator therapy after implementation of the multi-modal interventions to reduce inappropriate prescribing. The patients who received a bronchodilator
ACCEPTED MANUSCRIPT 7 were older than those who did not (57.5 + 16.3 years vs. 52.9 + 18.1 years, p=0.004), respectively). There were 42% females in the bronchodilator group and 38% in the non-
T
bronchodilator group (p=0.379). The expected length of hospital stay was significantly longer in
RI P
the group of patients that received bronchodilator therapy (15 + 10.5 vs. 10 + 8.3, p<0.001, respectively). Similarly, the actual observed lengths of both hospital (20 + 16.6 vs. 11 + 15.9,
SC
p<0.001, respectively) and ICU (11 + 11.2 vs. 5 + 7.6, p<0.001, respectively) stay were
NU
significantly longer in the group of patients that received bronchodilator therapy. There was no significant difference in the number of cases of ventilator-associated pneumonia between the
MA
bronchodilator groups and no-bronchodilator group in the pre-intervention phase (14 vs. 2, p=0.057, respectively) and post-intervention phase (12 vs. 5, p=0.636, respectively). The total
ED
direct costs post-intervention were significantly greater in the group of patients who received
PT
bronchodilator therapy compared to those who did not ($56,966 + 54,086 vs. $28,313 + 32,801,
AC
Discussion
CE
p<0.001).
The major finding of this trial is that substantial cost reductions can be achieved through interventions aimed at curtailing bronchodilator use in mechanically ventilated ICU patients. Further, the concomitant reductions in bronchodilator use and cost were sustained and these cost reductions accrued without affecting lengths of stay. Up until recently, aerosolized bronchodilators (especially beta-2 agonists), were promoted in mechanically ventilated patients, regardless of the presence or absence of any reversible airway disease. Advocates for such therapy argued that beta-2 agonists work on many pulmonary cellular pathways thought to be associated with the pathophysiology of acute lung injury
ACCEPTED MANUSCRIPT 8 (ALI).1,2 Several preliminary studies have demonstrated that beta-agonists improve alveolar fluid clearance, which is often impaired in patients with acute respiratory distress syndrome,
T
especially in more severe cases.1,2,5 Perkins et al conducted a single-centered, double-blind,
RI P
randomized trial comparing intravenous salbutamol (15mcg/kg/hr) to placebo in 40 patients with ALI or acute respiratory distress syndrome (ARDS). This proof of concept study demonstrated
SC
that extravascular lung water on day 7 (primary endpoint) was lower in the salbutamol-group
NU
compared with the placebo-group (9.2 vs 13.2 ml/kg, respectively, p=0.04).2
MA
Despite reports of a potential benefit, inhaled bronchodilators, particularly beta-2 agonists, have recently been the subject of increasing scrutiny after the publication of two multicenter,
ED
randomized controlled trials that found no clinical benefit associated with their use. The first trial randomized 282 patients on mechanical ventilation with ARDS (bilateral infiltrates and an
PT
arterial oxygen tension to fraction of inspired oxygen ratio of 300 or less) to receive either
CE
aerosolized albuterol at a dose of 5 mg or saline placebo every 4 hours for 24 hours after extubation or up to 10 days.3 There was no significant difference between the albuterol and
AC
placebo groups for the primary outcome of ventilator-free days (14.4 days vs. 16.6 days, p=0.087, respectively). The patients receiving albuterol had higher heart rates (e.g., approximately 4 beats per minute), but this did not translate into more episodes of cardiac arrhythmias. The second trial randomized 326 patients on mechanical ventilation with ARDS (bilateral infiltrates and an arterial oxygen tension to fraction of inspired oxygen ratio of 200 or less) to receive either intravenous salbutamol at a dose of 15 ug/kg of ideal body weight each hour or placebo for up to 7 days.4 The trial was stopped early because there was higher 28-day mortality (primary study endpoint) in the salbutamol compared to the placebo group (34% vs. 23%, RR = 1.47, 95% CI 1.03 to 2.08, respectively). The authors reported that salbutamol was
ACCEPTED MANUSCRIPT 9 poorly tolerated because of tachycardia, lactic acidosis and arrhythmias; however, the underlying cause for the increased mortality remained unclear.4 Although questions were later raised about
T
sample size and analysis issues,6 the results led others to believe a change in practice was
RI P
warranted with beta-agonist use limited to patients with demonstrated reversible airway obstruction.5 Finally, aerosolized bronchodilator therapy has been shown to be an independent
SC
predictor of ventilator-associated pneumonia.7 In our study, no significant differences were
NU
observed in the number of cases of ventilator-associated pneumonia between the bronchodilator group and no-bronchodilator group in the pre-intervention phase (14 vs. 2, p=0.057, respectively)
MA
and post-intervention phase (12 vs. 5, p=0.636, respectively).
ED
Weighing all data, experts in our institution agreed that the nonselective, routine use of aerosolized bronchodilators was no longer justifiable in mechanically ventilated patients. The
PT
sustained reductions in aerosolized bronchodilator therapy in our mechanically ventilated
CE
patients were achieved by a multi-modal approach that began with discussions of the topic with local experts and opinion leaders. Our P & T committee subsequently became involved in the
AC
effort by requiring the removal of bronchodilators from order sets in our computerized prescriber order entry software. Throughout the process, various educational efforts were undertaken and steps to reduce wastage were implemented with nursing involvement. Finally, the initial success of the program was shared with other physician leadership and hospital administrators who recognized the importance of this program and gave their backing for ongoing evaluation. There are potential limitations associated with this evaluation including its retrospective design with the potential for unknown confounders. Additional limitations include restriction of the use of all bronchodilators including ipratropium, not just beta-agonists, and the use of a historical control. With respect to the use of an historical control, the reductions in bronchodilator use and
ACCEPTED MANUSCRIPT 10 associated costs were substantial and did not appear to be a function of other practice-related changes during the pre- and post-evaluation periods. Another limitation is that an accurate
T
record of the duration of mechanical ventilation was not available, so it is possible that there
RI P
were differences in time to extubation between groups, albeit without concomitant changes in hospital or ICU lengths of stay. Similarly, validated criteria for diagnosing ARDS and assessing
SC
severity of illness were not utilized and there may have been underlying differences in morbidity
NU
between groups.
MA
In conclusion, multi-modal efforts to restrict aerosolized bronchodilator therapy in mechanically ventilated patients were successful and led to sustained reductions in use that was associated with
ED
substantial reductions in cost that did not affect outcomes related to length of stay. We recommend hospitals evaluate their use of aerosolized bronchodilators in mechanically ventilated
PT
patients and limit prescribing to appropriate patients (e.g those with obstructive pulmonary
AC
CE
disease).
ACCEPTED MANUSCRIPT 11 References 1. Perkins GD, McAuley DF, Thicket DR, et al. The beta-agonist lung injury trial
T
(BALTI): a randomized placebo-controlled clinical trial. Am J Respir Crit Care Med
RI P
2006;173:281-287.
2. Matthay MA. Beta-adrenergic agonist therapy as a potential treatment for acute lung
SC
injury. Am J Respir Crit Care Med 2006;173:254-255.
3. The National Heart, Lung, and Blood Institute Acute Respiratory Distress (ARDS)
NU
Clinical Trials Network: Randomized, placebo-controlled clinical trial of an aerosolized
MA
β2-Agonist for treatment of acute lung injury. Am J Respir Crit Care Med 2011;184:5618.
ED
4. Smith FG, Perkins GD, Young D, et al: Effect of intravenous β-2 agonist treatment on clinical outcomes in respiratory distress syndrome (BALTI-2): a multicenter, randomized
PT
controlled trial. Lancet 2012;379:229-35.
CE
5. Thompson BT: Β-agonists for ARDS: the dark side of adrenergic stimulation. Lancet 2012;379(comment):196-8.
AC
6. Abroug F, Besbes LO, Ouanes I, et al: Intravenous salbutamol in ARDS and increased mortality. Lancet 2012;379(letter):1875. 7. Jaillette E, Nseir S: Relationship between inhaled β2-agonists and ventilator-associated pneumonia: A cohort study. Crit Care Med 2011;39:725-30.
ACCEPTED MANUSCRIPT 12
T
Table 1. Patient Demographics in the Pre-and Post-Intervention Groups. Post-Intervention
(n=449)
(n=483)
Sex (% male)
64
59
Age (years + SD)
55.4 + 17.66
AC
CE
PT
ED
MA
NU
SC
RI P
Pre-Intervention
Variable
55.9 + 17.08
P-value
0.160 0.606
ACCEPTED MANUSCRIPT 13
Table 2. Length of stay and Pneumonia in the Pre- and Post-Intervention Groups.
(n=449)
(n=483)
13.3 + 19.14
13.2 + 10.01
15.5 + 18.60
MA
observed (days) 8.8 + 14.98
Pneumonia
16
CE
PT
ED
ICU stay (days)
AC
NU
expected (days) Hospital stay
P-value
T
Post-Intervention
RI P
Hospital stay
Pre-Intervention
0.783
SC
Variable
16.4 + 16.89
0.435
9.2 + 10.48
0.647
17
0.971
S0883-9441(14)00217-2 doi: 10.1016/j.jcrc.2014.05.016 YJCRC 51530
To appear in:
Journal of Critical Care
Received date: Revised date: Accepted date:
14 February 2014 28 April 2014 22 May 2014
Please cite this article as: Camamo James M., Weibel Kurt, O’Keeffe Terence, Huckleberry Yvonne, Kopp Brian, Diven Conrad, Erstad Brian L., Cost Savings with Interventions to Reduce Aerosolized Bronchodilator Use in Mechanically Ventilated Patients, Journal of Critical Care (2014), doi: 10.1016/j.jcrc.2014.05.016
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Cost Savings with Interventions to Reduce Aerosolized Bronchodilator Use in Mechanically Ventilated Patients
T
James M. Camamo, PharmD, Kurt Weibel, PharmD, Terence O’Keeffe, MB ChB, MSPH,
RI P
FACS, Yvonne Huckleberry, PharmD, Brian Kopp, PharmD, Conrad Diven, MD, Brian L. Erstad, PharmD, FCCP, FCCM, FASHP
SC
University of Arizona Medical Center, Department of Pharmacy Services; The University of
NU
Arizona College of Medicine; The University of Arizona College of Pharmacy, Department of
MA
Pharmacy Practice and Science
Manuscript Communication:
ED
James Camamo, PharmD
PT
Department of Pharmacy Services
University of Arizona Medical Center
CE
1501 N Campbell Ave
AC
Tucson, AZ 85724-5009
Email: [email protected]
Word count (abstract): 198 Word count (body of text): 2030
ACCEPTED MANUSCRIPT 2 Abstract: Purpose
T
The purpose of this evaluation is to describe the cost savings associated with multimodal
RI P
interventions aimed at reducing aerosolized bronchodilator use in mechanically ventilated patients without adversely affecting costs associated with length of stay.
SC
Materials and Methods
NU
Subjects were included in the analysis if they were > 18 years of age; on mechanical ventilation in the ICU; and received aerosolized bronchodilators. Patients were excluded if they had
MA
reversible airway disease, an indication needing bronchodilator therapy. Patient data was obtained using the University Health System Consortium Clinical Data Base/Resource
ED
ManagerTM to compare outcomes during two 6 month time periods separated by a 4 month
PT
intervention phase aimed to reduce bronchodilator use. Results
CE
There were no significant differences in age, sex and lengths of stay (observed and expected)
AC
between the pre-intervention and post-intervention phases. Based on whole acquisition costs, the total cost of bronchodilators dispensed to the adult ICUs over the 6 month post-intervention phase was reduced by $56,960 compared to the 6 month pre-intervention phase ($120,562 vs. $63,602, respectively). Conclusions Multi-modal efforts to restrict aerosolized bronchodilator therapy in mechanically ventilated patients were successful and led to sustained reductions in use that was associated with substantial reductions in cost, without affecting lengths of stay. Key words: bronchodilator, mechanical ventilation, cost savings
ACCEPTED MANUSCRIPT 3 Introduction Aerosolized brochodilators, are often prescribed to mechanically ventilated patients in the
T
intensive care unit (ICU), regardless of the presence or absence of reversible airway disease.
RI P
Early advocates for this therapy cite preliminary data demonstrating that beta-2 agonists reduce pulmonary edema and accelerate alveolar fluid clearance in patients with acute lung injury.1,2
SC
Despite the purported benefits, two multicenter randomized trials involving mechanically
NU
ventilated patients in mixed medical-surgical intensive care units have demonstrated no benefit and potential harm from aerosolized beta-2 agonist therapy.3,4 This led to the recommendation
MA
that aerosolized beta-agonists be limited to patients who have demonstrated reversible airway
ED
disease.5
The studies negating the purported benefits of aerosolized beta-2 agonists prompted us to
PT
reassess our current prescribing patterns. Preliminary data at our institution indicated that
CE
aerosolized bronchodilators were commonly prescribed to mechanically ventilated patients that did not have evidence of reversible airway disease. In addition to concerns about lack of
AC
efficacy and potential harm, the frequent use of aerosolized bronchodilators in our ICUs raised resource concerns that included personnel time and drug expenditures. In light of rising health care costs, the department of pharmacy services worked closely with the medical staff and nursing to reduce waste and promote the rational use of aerosolized bronchodilators in the ICU. The purpose of this evaluation is to describe the cost savings associated with multimodal interventions aimed at reducing aerosolized bronchodilator use in mechanically ventilated patients. Prior to conducting our interventions, we hypothesized that reductions in the use of aerosolized bronchodilators in ICU patients will reduce costs without adversely affecting lengths of stay.
ACCEPTED MANUSCRIPT 4 Methods This retrospective review was conducted at the University of Arizona Medical Center, a 420-bed
RI P
T
tertiary care facility with medical and surgical intensive care units in Tucson, Arizona. The project was reviewed by the IRB and considered exempt from further review.
SC
Subjects were included in the analysis if they were > 18 years of age; on mechanical ventilation
NU
in the ICU; and received aerosolized formulations of albuterol sulfate, ipratropium bromide or albuterol sulfate/ipratropium bromide. Patients were excluded if they had reversible airway
MA
disease, an indication where bronchodilator therapy is beneficial. Patient specific data was obtained using the University Health System Consortium (UHC)
ED
Clinical Data Base/Resource ManagerTM (CDBRM) to compare outcomes during two time
PT
periods separated by a period where multimodal interventions aimed at reducing utilization of aerosolized bronchodilators occurred. The multimodal interventions were implemented over a 4
CE
month time period beginning February 15, 2012. During this time, aerosolized bronchodilators
AC
were removed from ICU order sets. Additionally, waste reduction strategies were implemented to reduce lost or misplaced inhalers. These included labeling the inhaler with patient identifiers rather than the box; working with nursing and transport staff to insure inhalers stayed with the patient when transferred to different levels of care; and storing inhalers in consistent areas located in the ICU medication rooms. Finally, the clinical staff pharmacists partnered with physician attendings to educate residents on the appropriate use of bronchodilators in the ICU. Data from the pre-intervention phase was collected between July 1st, 2011 and December 31st, 2011 and data during the post-intervention phase was collected between July 1st, 2012 and
ACCEPTED MANUSCRIPT 5 December 31st, 2012. The time period for the post-intervention phase took place after all interventions were fully implemented.
RI P
T
The CDBRM is an online database maintained by UHC which contains discharge and line-item data from more than 200 UHC members and affiliate hospitals. For this project, the CDBRM
SC
was utilized to obtain patient demographics, inpatient length of stay, and intensive care unit (ICU) length of stay (LOS). Additionally, the expected values for LOS were collected using
NU
CDBRM, which makes adjustments for risks using a logistic regression model in combination
MA
with the 3M Health Information Systems, Agency for Healthcare Research and Quality comorbidity software and the UHC complication profiler to assign illness severity levels. In
ED
addition to patient demographics and outcomes, the CDBRM was used to extract data on direct costs, which estimates the cost of patient care using a cost-to-charge ratio based on costs and
PT
revenues from the Centers for Medicare & Medicaid Services (CMS). In addition to direct costs
CE
(e.g supplies, labor and medication), the total cost of all aerosolized bronchodilators during the two evaluation periods was determined using whole acquisition costs and the number of doses
Data Analysis
AC
dispensed to the adult intensive care units.
Data were initially entered into Microsoft Excel® and subsequently transferred to STATA® version 13 (StataCorp, College Station, Tx) for statistical analysis. Two sample Student’s t-tests and ANOVA were used for comparisons of continuous data, while the Fisher’s exact or chisquared test was used for comparisons of nominal data. These comparisons were performed for the pre- and post-intervention periods, as well as for bronchodilator versus no bronchodilator
ACCEPTED MANUSCRIPT 6 groups in the post-intervention period. Statistical significance was defined as p<0.05 for all testing.
SC
Pre-Intervention vs. Post-Intervention Comparisons
RI P
T
Results
There were no significant differences in age or sex between the historical control group
NU
comprised of consecutive ICU patients and the more recent cohort of patients evaluated after the multi-modal interventions to reduce bronchodilator use had been implemented (Table 1).
MA
Similarly, there were no significant differences in the lengths of expected or observed lengths of
intervention groups (Table 2).
ED
stay, or in the number of cases of ventilator-associated pneumonia between the pre- and post-
PT
Based on whole acquisition costs, the total cost of aerosolized bronchodilators dispensed to the
CE
adult ICUs over the 6 month post-intervention phase was reduced by $56,960 compared to the 6 month pre-intervention phase ($120,562 vs. $63,602, respectively). The annualized cost savings
AC
associated with multi-modal interventions designed to optimize aerosolized bronchodilator use in the adult ICUs is reported to be approximately $113,920. There was no significant change in the total direct costs between the pre-intervention and post-intervention phases ($53,483 + 97,300 vs. $47,000 + 49,650 p=0.196, respectively). Bronchodilator vs. No Bronchodilator Comparisons There were 315 patients who received aerosolized bronchodilator therapy and 168 patients who did not receive aerosolized bronchodilator therapy after implementation of the multi-modal interventions to reduce inappropriate prescribing. The patients who received a bronchodilator
ACCEPTED MANUSCRIPT 7 were older than those who did not (57.5 + 16.3 years vs. 52.9 + 18.1 years, p=0.004), respectively). There were 42% females in the bronchodilator group and 38% in the non-
T
bronchodilator group (p=0.379). The expected length of hospital stay was significantly longer in
RI P
the group of patients that received bronchodilator therapy (15 + 10.5 vs. 10 + 8.3, p<0.001, respectively). Similarly, the actual observed lengths of both hospital (20 + 16.6 vs. 11 + 15.9,
SC
p<0.001, respectively) and ICU (11 + 11.2 vs. 5 + 7.6, p<0.001, respectively) stay were
NU
significantly longer in the group of patients that received bronchodilator therapy. There was no significant difference in the number of cases of ventilator-associated pneumonia between the
MA
bronchodilator groups and no-bronchodilator group in the pre-intervention phase (14 vs. 2, p=0.057, respectively) and post-intervention phase (12 vs. 5, p=0.636, respectively). The total
ED
direct costs post-intervention were significantly greater in the group of patients who received
PT
bronchodilator therapy compared to those who did not ($56,966 + 54,086 vs. $28,313 + 32,801,
AC
Discussion
CE
p<0.001).
The major finding of this trial is that substantial cost reductions can be achieved through interventions aimed at curtailing bronchodilator use in mechanically ventilated ICU patients. Further, the concomitant reductions in bronchodilator use and cost were sustained and these cost reductions accrued without affecting lengths of stay. Up until recently, aerosolized bronchodilators (especially beta-2 agonists), were promoted in mechanically ventilated patients, regardless of the presence or absence of any reversible airway disease. Advocates for such therapy argued that beta-2 agonists work on many pulmonary cellular pathways thought to be associated with the pathophysiology of acute lung injury
ACCEPTED MANUSCRIPT 8 (ALI).1,2 Several preliminary studies have demonstrated that beta-agonists improve alveolar fluid clearance, which is often impaired in patients with acute respiratory distress syndrome,
T
especially in more severe cases.1,2,5 Perkins et al conducted a single-centered, double-blind,
RI P
randomized trial comparing intravenous salbutamol (15mcg/kg/hr) to placebo in 40 patients with ALI or acute respiratory distress syndrome (ARDS). This proof of concept study demonstrated
SC
that extravascular lung water on day 7 (primary endpoint) was lower in the salbutamol-group
NU
compared with the placebo-group (9.2 vs 13.2 ml/kg, respectively, p=0.04).2
MA
Despite reports of a potential benefit, inhaled bronchodilators, particularly beta-2 agonists, have recently been the subject of increasing scrutiny after the publication of two multicenter,
ED
randomized controlled trials that found no clinical benefit associated with their use. The first trial randomized 282 patients on mechanical ventilation with ARDS (bilateral infiltrates and an
PT
arterial oxygen tension to fraction of inspired oxygen ratio of 300 or less) to receive either
CE
aerosolized albuterol at a dose of 5 mg or saline placebo every 4 hours for 24 hours after extubation or up to 10 days.3 There was no significant difference between the albuterol and
AC
placebo groups for the primary outcome of ventilator-free days (14.4 days vs. 16.6 days, p=0.087, respectively). The patients receiving albuterol had higher heart rates (e.g., approximately 4 beats per minute), but this did not translate into more episodes of cardiac arrhythmias. The second trial randomized 326 patients on mechanical ventilation with ARDS (bilateral infiltrates and an arterial oxygen tension to fraction of inspired oxygen ratio of 200 or less) to receive either intravenous salbutamol at a dose of 15 ug/kg of ideal body weight each hour or placebo for up to 7 days.4 The trial was stopped early because there was higher 28-day mortality (primary study endpoint) in the salbutamol compared to the placebo group (34% vs. 23%, RR = 1.47, 95% CI 1.03 to 2.08, respectively). The authors reported that salbutamol was
ACCEPTED MANUSCRIPT 9 poorly tolerated because of tachycardia, lactic acidosis and arrhythmias; however, the underlying cause for the increased mortality remained unclear.4 Although questions were later raised about
T
sample size and analysis issues,6 the results led others to believe a change in practice was
RI P
warranted with beta-agonist use limited to patients with demonstrated reversible airway obstruction.5 Finally, aerosolized bronchodilator therapy has been shown to be an independent
SC
predictor of ventilator-associated pneumonia.7 In our study, no significant differences were
NU
observed in the number of cases of ventilator-associated pneumonia between the bronchodilator group and no-bronchodilator group in the pre-intervention phase (14 vs. 2, p=0.057, respectively)
MA
and post-intervention phase (12 vs. 5, p=0.636, respectively).
ED
Weighing all data, experts in our institution agreed that the nonselective, routine use of aerosolized bronchodilators was no longer justifiable in mechanically ventilated patients. The
PT
sustained reductions in aerosolized bronchodilator therapy in our mechanically ventilated
CE
patients were achieved by a multi-modal approach that began with discussions of the topic with local experts and opinion leaders. Our P & T committee subsequently became involved in the
AC
effort by requiring the removal of bronchodilators from order sets in our computerized prescriber order entry software. Throughout the process, various educational efforts were undertaken and steps to reduce wastage were implemented with nursing involvement. Finally, the initial success of the program was shared with other physician leadership and hospital administrators who recognized the importance of this program and gave their backing for ongoing evaluation. There are potential limitations associated with this evaluation including its retrospective design with the potential for unknown confounders. Additional limitations include restriction of the use of all bronchodilators including ipratropium, not just beta-agonists, and the use of a historical control. With respect to the use of an historical control, the reductions in bronchodilator use and
ACCEPTED MANUSCRIPT 10 associated costs were substantial and did not appear to be a function of other practice-related changes during the pre- and post-evaluation periods. Another limitation is that an accurate
T
record of the duration of mechanical ventilation was not available, so it is possible that there
RI P
were differences in time to extubation between groups, albeit without concomitant changes in hospital or ICU lengths of stay. Similarly, validated criteria for diagnosing ARDS and assessing
SC
severity of illness were not utilized and there may have been underlying differences in morbidity
NU
between groups.
MA
In conclusion, multi-modal efforts to restrict aerosolized bronchodilator therapy in mechanically ventilated patients were successful and led to sustained reductions in use that was associated with
ED
substantial reductions in cost that did not affect outcomes related to length of stay. We recommend hospitals evaluate their use of aerosolized bronchodilators in mechanically ventilated
PT
patients and limit prescribing to appropriate patients (e.g those with obstructive pulmonary
AC
CE
disease).
ACCEPTED MANUSCRIPT 11 References 1. Perkins GD, McAuley DF, Thicket DR, et al. The beta-agonist lung injury trial
T
(BALTI): a randomized placebo-controlled clinical trial. Am J Respir Crit Care Med
RI P
2006;173:281-287.
2. Matthay MA. Beta-adrenergic agonist therapy as a potential treatment for acute lung
SC
injury. Am J Respir Crit Care Med 2006;173:254-255.
3. The National Heart, Lung, and Blood Institute Acute Respiratory Distress (ARDS)
NU
Clinical Trials Network: Randomized, placebo-controlled clinical trial of an aerosolized
MA
β2-Agonist for treatment of acute lung injury. Am J Respir Crit Care Med 2011;184:5618.
ED
4. Smith FG, Perkins GD, Young D, et al: Effect of intravenous β-2 agonist treatment on clinical outcomes in respiratory distress syndrome (BALTI-2): a multicenter, randomized
PT
controlled trial. Lancet 2012;379:229-35.
CE
5. Thompson BT: Β-agonists for ARDS: the dark side of adrenergic stimulation. Lancet 2012;379(comment):196-8.
AC
6. Abroug F, Besbes LO, Ouanes I, et al: Intravenous salbutamol in ARDS and increased mortality. Lancet 2012;379(letter):1875. 7. Jaillette E, Nseir S: Relationship between inhaled β2-agonists and ventilator-associated pneumonia: A cohort study. Crit Care Med 2011;39:725-30.
ACCEPTED MANUSCRIPT 12
T
Table 1. Patient Demographics in the Pre-and Post-Intervention Groups. Post-Intervention
(n=449)
(n=483)
Sex (% male)
64
59
Age (years + SD)
55.4 + 17.66
AC
CE
PT
ED
MA
NU
SC
RI P
Pre-Intervention
Variable
55.9 + 17.08
P-value
0.160 0.606
ACCEPTED MANUSCRIPT 13
Table 2. Length of stay and Pneumonia in the Pre- and Post-Intervention Groups.
(n=449)
(n=483)
13.3 + 19.14
13.2 + 10.01
15.5 + 18.60
MA
observed (days) 8.8 + 14.98
Pneumonia
16
CE
PT
ED
ICU stay (days)
AC
NU
expected (days) Hospital stay
P-value
T
Post-Intervention
RI P
Hospital stay
Pre-Intervention
0.783
SC
Variable
16.4 + 16.89
0.435
9.2 + 10.48
0.647
17
0.971