Systemic Steroid and Nebulized Budesonide Combination Therapy Versus Systemic Steroid Monotherapy in Patients with Acute Exacerbation of Chronic Obstructive Pulmonary Disease in a Community Hospital: A Retrospective Cohort Study

Abstract

Purpose: To evaluate clinical outcomes and costs of inhaled corticosteroid (ICS) and systemic corticosteroid combination therapy versus systemic corticosteroid monotherapy for treatment of acute exacerbations of chronic obstructive pulmonary disease (AECOPD). Methods: Hospitalized patients aged 41 to 85 years old who received ≥40 mg/day of systemic prednisone equivalents between April 3, 2017 to July 31, 2017 and April 3, 2018 to July 31, 2018 with a primary discharge diagnosis of AECOPD. Two cohorts were identified: those who received >2 doses of ICS (combination therapy) and those who received ≤2 doses of ICS (monotherapy) while on systemic corticosteroid therapy. Primary outcomes were progression of respiratory support or ≥20% increase in daily dose of systemic corticosteroids. Secondary outcomes were hospital length of stay (LOS), COPD 30-day readmissions, in-hospital mortality, and nebulized budesonide costs. Results: One hundred twenty-eight patients met inclusion criteria. Daily corticosteroid dose increases were similar between the combination and monotherapy cohorts (4% vs. 5%, P = 0.76) as was progression in ventilatory support (12% vs. 8%, P = 0.53). In-hospital mortality (4% vs. 1%, P = 0.36) and COPD 30-day readmissions (16% vs. 9%, P = 0.22) were not significantly different, however, patients in the combination arm had longer lengths of stay (4.8 days vs. 3.9 days, P = 0.04). Total nebulized budesonide costs were $1857 with a mean of $37 per patient stay for combination therapy cohort. Conclusion: Outcomes showed no clinical difference between combination therapy and monotherapy. This study suggests monotherapy may be more cost-effective while providing similar outcomes for the treatment of hospitalized patients with AECOPD.

Introduction

Chronic obstructive pulmonary disease (COPD) is the third leading cause of death in the United States with an estimated cost of $50 billion in 2010. ¹ Acute exacerbations of COPD (AECOPD) are frequently the cause of hospital admissions and have been shown to account for 45% to 50% of the total direct costs incurred by patients with COPD. ²

Recommendations for initial treatment of AECOPD per the Global Initiative for Chronic Obstructive Lung Disease (GOLD) include a short-acting beta agonist (SABA) with or without a short-acting muscarinic antagonist (SAMA). ³ Systemic corticosteroids have been shown to improve forced expiratory volume (FEV1), improve SpO2, shorten recovery time, minimize the risk of early relapse, decrease treatment failure, and reduce hospitalization LOS. ³ Recent studies propose the use of systemic corticosteroids for AECOPD may be less effective if patients have low serum concentrations of eosinophils. ³ GOLD recommends a dose of 40 mg of prednisone per day for 5 days if indicated based on the REDUCE trial.^3,4

Nebulized budesonide alone may be used as an alternative to systemic corticosteroids, however, there are no well-designed randomized clinical trials in AECOPD to suggest using both concomitantly confers additional clinical benefit. ³ Two retrospective studies have evaluated the effectiveness of combining systemic and inhaled corticosteroids (ICS) in the treatment of AECOPD. Neither showed any improvement in LOS, rates of mechanical ventilation, inpatient mortality, or 30-day readmissions.^5,6 Another study has illustrated significant hospitalized ICS medication costs in COPD treatment. ⁷ If ICS medications can be discontinued without incurring negative effects to the patient and LOS while patient’s are on systemic steroids, a closer look into combined use of systemic and inhaled corticosteroids is warranted.

A med-use evaluation at our community hospital found that the combined use of systemic and inhaled corticosteroids occurred in approximately 60% of 164 hospitalized patients admitted with an AECOPD diagnosis between December 2017 and January 2018. A pharmacy driven protocol in which nebulized budesonide is discontinued in patients with AECOPD on concomitant systemic corticosteroids ≥40 mg prednisone equivalents per day was approved by the medical staff and implemented on April 3, 2018. The purpose of this study was to compare the effects on clinical deterioration, in-house mortality, LOS, 30-day COPD readmissions, and drug costs between combination systemic plus inhaled corticosteroid therapy versus mono systemic corticosteroid therapy for AECOPD in our hospitalized patients.

Methods

This was a single-center retrospective analysis of patients with AECOPD aged 41 to 85 years old who received ≥40 mg/day of prednisone equivalents between April 3, 2017 to July 31, 2017 and the same timeframe in 2018. The patient selection dates for each cohort were chosen in consideration of the protocol implementation date of April 3, 2018. Since concomitant use of ICS with systemic corticosteroids was no longer a common occurrence post-implementation, a seasonally adjusted comparator arm was needed from the pre-implementation period. Since our facility is based in Arizona, it is impacted by seasonal visitors and the subsequent effect on patient volumes, hospital census, and patient demographics could potentially influence results if not considered. Identification of patients was based on an ICD-10-CM code of J 44.1 admitted to Banner Desert Medical Center. Subjects were excluded if they had a documented history of lobectomy, lung cancer, primary pulmonary hypertension, sarcoidosis, lung transplant, interstitial lung disease, lupus, idiopathic pulmonary fibrosis and/or intubated in the field prior to admit to decrease confounders from other pulmonary disease states. Two cohorts were identified: those who received >2 doses of ICS during admission (combination therapy) and those who received ≤2 doses of ICS while on systemic corticosteroid therapy (monotherapy). The two-dose cut-off was chosen due to the fact that some patients may have received limited doses prior to auto-discontinuation of budesonide per protocol. It was felt that 2 or less doses during a typical 4 to 6 day length of stay for AECOPD would likely have limited impact on the overall treatment outcome. Data collection was completed by retrospective chart review by the principle investigator. As such, there was no risk to or any direct involvements with the participants. Our system institutional review board (IRB) approved a waiver of consent for this study. Unique patient identifiers were linked to nonspecific study identifiers on a separate electronic data sheet that were destroyed at the end of the study to ensure patient confidentiality and compliance with HIPAA standards. No unique patient identifiers or nonspecific study identifiers are included in presented data. Data collection included patient demographics with home COPD medications, admit location, BAP-65 criteria (BUN, heart rate, age, and mental status) to assess severity on admit, and COPD related pharmacotherapy during hospitalization. In addition to scoring tools, other markers of worsening AECOPD were assessed by patients’ needs for non-invasive ventilation (NIV), mechanical ventilation (MV), and escalation of corticosteroid dosing. Hospital and ICU LOS, in-hospital mortality, and 30-day COPD respiratory readmission data was also collected. IRB approved by Banner Health Phoenix Panel, iRIS reference #019279.

Statistical Analysis

Sample size was determined by the number of patients included within the time frames provided and not designed to target specific statistical power. Descriptive statistics were used to report continuous variables as means and standard deviations and categorical variables as percentages. Any difference between cohorts for absolute variables was determined with a Student’s t-test. Fischer’s exact test was used for binary data. A Chi-squared test was used to determine differences in proportions between cohorts. Respectively, linear regression and logistic regression was used to identify for correlations between continuous and categorical variables. For all tests, P ≤ .05 was considered significant. Missing data was excluded from intended statistical analysis.

Results

Study Characteristics

One hundred forty-eight patients were identified during the study timeframe with 128 meeting inclusion criteria. Fifty patients received combination therapy and 78 patients received monotherapy (Figure 1).

Figure 1.

Study inclusion.

Patient demographics were not different between cohorts. Preadmission COPD medication regimens did not significantly differ (Table 1), except fewer patients in the monotherapy cohort were on home short-acting beta-2 agonist therapy (59% vs. 76%, P = .048) and inhaled corticosteroid therapy (45% vs. 62%, P = .05) versus the combination therapy cohort.

Table 1.

Patient Demographics.

Demographics	Combination therapy (n = 50)	Monotherapy (n = 78)	P-value
Mean age—years (IQR)	64.74 (9.3)	64.56 (9.2)	.91
Male sex (%)	22 (44)	34 (44)	.96
Tobaccoism (%)	41 (82)	67 (86)	.55
Mean weight—kg (IQR)	87.05 (32.1)	88.81 (34.5)	.77
BMI
<20 (%)	5 (10)	10 (13)	.78
20-29 (%)	27 (54)	33 (42)	.21
≥30 (%)	18 (36)	35 (45)	.36
Home COPD Therapy
No home COPD meds (%)	6 (12)	18 (36)	.16
Home SABA (%)	38 (76)	46 (59)	.04
Home SAMA (%)	12 (24)	13 (17)	.30
Home LABA (%)	26 (52)	32 (41)	.22
Home LAMA (%)	15 (30)	25 (32)	.80
Home ICS (%)	31 (62)	35 (45)	.05
Home systemic steroids (%)	0 (0)	2 (3)	.52
Home O2 (%)	14 (28)	28 (36)	.35

Severity on Admit

BAP-65 class on admission and need for non-invasive (NIV) or mechanical ventilation (MV) within 24 hours of admission were not statistically significant between either cohort (Figure 2). Direct admission to the ICU occurred more frequently with combination therapy. (8% vs. 0%, P = .021) (Table 2)

Figure 2.

BAP-65 classification.

Table 2.

Admit Severity.

Criteria	Combination therapy (n = 50)	Monotherapy (n = 78)	P-value
ICU admit (%)	4 (8)	0 (0)	.02
NIV on admit (%)	8 (16)	11 (14)	.80
MV on admit (%)	2 (4)	0 (0)	.15

Inpatient treatment

Use of SABAs or antibiotics, mean duration of corticosteroid therapy, or patients with corticosteroid use >5 days was not significantly different between cohorts (Table 3).

Table 3.

Inpatient Treatment.

Inpatient treatment	Combination therapy (n = 50)	Monotherapy (n = 78)	P-value
SABA therapy (%)	43 (86)	66 (85)	.83
Antibiotic use (%)	44 (88)	67 (86)	.73
Mean systemic corticosteroid length of therapy—days (IQR)	4.3 (2.5)	3.5 (1.9)	.52
>5 d of systemic corticosteroid use	9	13	.84
Day 1 average daily systemic steroid dose—mg (IQR)	207 (80)	208 (80)	.68
Day 2 average daily systemic steroid dose—mg (IQR)	153 (68)	139 (62)	.48
Day 3 average daily systemic steroid dose—mg (IQR)	127 (77)	114 (55)	.76
Day 4 average daily systemic steroid dose—mg (IQR)	108 (108)	102 (64)	.23
Day 5 average daily systemic steroid dose—mg (IQR)	104 (139)	88 (80)	.26

Clinical Outcomes

An increase in daily systemic corticosteroid doses ≥20%, progression to NIV or MV, inpatient in-hospital mortality, any increased level of care, 30-day respiratory COPD readmission rates, and mean ICU LOS were not statistically different between the 2 cohorts (Table 4). The mean hospital LOS, however, was significantly longer in the combination therapy cohort (4.8 vs. 3.9 days, P = .041).

Table 4.

Clinical Outcomes.

Clinical outcomes	Combination therapy (n = 50)	Monotherapy (n = 78)	P-value
Increase in systemic corticosteroids ≥20% per day (%)	2 (4)	4 (5)	.76
Progression to NIV or MV (%)	6 (12)	6 (8)	.53
Increased level of care (%)	1 (2)	3 (4)	.65
In-hospital mortality (%)	2 (4)	1 (1)	.36
30-d COPD readmission (%)	8 (16)	7 (9)	.22
Mean ICU LOS—days (IQR)	5.8 (5.9)	5 (4.6)	.92
Mean hospital LOS—days (IQR)	4.8 (2.4)	3.9 (2.2)	.04

Costs

Total nebulized budesonide medication costs were $1857 in the combination therapy cohort versus $133 in the monotherapy cohort. The average budesonide cost per patient was $37 versus $2, respectively (Table 5). Average systemic steroid use per day was not statistically different between the 2 cohorts (Table 3, Figure 3). An estimated cost of $9437 for methylprednisolone in all study participants was seen (Table 5). Only 1 patient out of 128 patients was initiated on oral prednisone for treatment, the rest were started on intravenous (IV) methylprednisolone. Many patients were continued on systemic IV corticosteroids during the majority their stay and ultimately discharged with oral systemic corticosteroids.

Table 5.

ICS and Intravenous Steroid Costs ($). ^a

Inpatient treatment	Combination therapy overall cost (n = 50)	Combination therapy cost per admit	Monotherapy overall cost (n = 78)	Monotherapy cost per admit
Nebulized budesonide expenditure (total doses) b	1857 (376)	37	133 (27)	2
Est. methylprednisolone expenditure (total days) c	4154 (192)	83	5283 (245)	67

^aCost rounded to the nearest whole U.S. dollar.

^b$4.94 per nebulized budesonide dose.

^c$7.18 per methylprednisolone 125 mg IV vial. Average of 3 doses per day.

Figure 3.

Systemic corticosteroid use.

Discussion

This retrospective study evaluated inhaled and systemic corticosteroid combination therapy and systemic corticosteroid monotherapy for AECOPD. The results showed no significant difference in clinical outcomes for both cohorts except the hospital LOS in the monotherapy arm was significantly lower by 0.9 days. The purpose of the study was to validate non-inferiority in that the addition of an ICS to recommended systemic corticosteroids would confer limited benefit during an AECOPD admission. Interestingly, these results are similar to what has been reported with previous retrospective studies.^5,6

Possible factors that could affect LOS for each cohort were evaluated. Spirometry is not routinely done on admitted patients to properly stage severity at the study site, so the risk of inpatient mortality and COPD severity in the study patients was stratified using the BAP-65 scoring tool, which has been validated in patients with AECOPD greater than 40 years of age. ⁸ This scoring tool uses patient specific parameters including blood urea nitrogen (BUN), any sign or symptom of altered mental status, heart rate, and age. The scoring places patients in classes I-V with increased in-hospital mortality and mechanical ventilation risks as scores escalate. However, there was no significant difference between cohorts in severity of COPD on admit based on the BAP-65 scores (Figure 2). Inpatient treatment regimens (Table 3) recommended by the GOLD Report were also not significantly different between cohorts. ³ It is worth noting that 4 (8%) of the combination therapy patients transferred to the ICU on admission versus none in the monotherapy cohort. On closer evaluation, however, patients who were admitted into the ICU (either on direct admit or due to an increased level of care transfer) did not suggest a significant LOS difference in each arm that could significantly impact the overall results.

Since this study was completed to assess any potential clinical outcomes impact of the approved nebulized budesonide discontinuation protocol implemented on April 3, 2018, patient inclusion from a comparable seasonally adjusted prior time frame in 2017 was necessary in order to locate an adequate number of patients who could meet the inclusion criteria set forth in the combination therapy cohort. Additional analysis was completed by further segregating each cohort by admit year to assess effects on LOS due to difference in practice between years. Segregating patients into their separate years showed the combined hospital LOS to be 4.0 days in 2017 and 4.6 days in 2018. Respiratory-specific COPD 30-day readmission rates were also higher in the combination therapy cohort (16% vs. 9%, P = .228). While not statistically significant, this is likely due to our study being underpowered as most readers would consider this a meaningful difference. This could also suggest patients in the combination therapy arm had more complicated diseases that were not detected using BAP-65.

This study also collected data on systemic corticosteroid prescribing patterns. Average first day systemic corticosteroid doses were above 200 mg prednisone equivalents, because the majority of patients were initiated on higher dose IV methylprednisolone. At day 5 of systemic corticosteroid therapy the daily average dose was approximately 100 mg prednisone equivalents. Only 1 patient was started near the GOLD-recommended daily dose of 40 mg prednisone equivalents. The increase in systemic corticosteroid injection costs surmounted to approximately $73 per patient admission. With COPD admissions ranging from 600 to 900 per year for this hospital, similar prescribing practices on all patients would amount to $43 800 to $65 700 of intravenous corticosteroid costs. In comparison, hospital acquisition costs of prednisone 40 mg are only $0.28 per dose. If this lower dose oral regimen were followed more routinely, the drug cost of treatment for a similar number of patients for 5 days would be significantly lower at a range of $840 to $1260 in systemic steroid therapy costs per year.

Additionally, although the stated nebulized budesonide cost reduction in this study between comparative cohorts would not be considered large (Table 5), these study patients represent only a small percentage of the overall respiratory population treated in our institution annually. As stated previously, our institution and health system delivers respiratory medications preferentially via the nebulized route to control inhaler drug costs and waste except for limited populations in which inhalers are still used during hospital admission. The significant reduction in respiratory drug costs have previously been reported secondary to the implementation of our inhaler to nebulization protocol as well as the high percentage of overall nebulized budesonide costs throughout the health system. ⁷ Our institution also observed annual nebulized budesonide expenditures of $175 924, accounting for 52% of our overall inhaler and nebulized medication drug costs of $338 532 in 2018. The significant occurrence of duplicative therapy of ICS and systemic corticosteroids in our respiratory patients ultimately led to the development, approval, and implementation of the automatic budesonide discontinuation protocol, as stated. In the 12 months post-implementation, our annual nebulized budesonide costs dramatically reduced to $102 583, which is a reduction of $73 341 (41.7%). In addition to this study, an attempt to further evaluate any potential negative impact of this protocol on our overall institutional COPD outcomes was done by utilizing data from the medical information data analysis system (MIDAS). This also revealed no negative trends in COPD average LOS, mortality rates, or all-cause 30-day readmissions (Table 6).

Table 6.

MIDAS COPD Outcomes.

Clinical outcomes	Pre-implementation (4.17-3.18)	Post-implementation (4.18-3.19)
Average LOS (days)	5.53 (n = 864)	5.36 (n = 606)
Mortality rate (%)—Overall	1.85 (n = 864)	1.49 (n = 606)
Mortality rate (%)—Medicare	2.82 (n = 213)	2.89 (n = 173)
30-d readmissions (%) (Overall)	18.04 (n = 848)	14.91 (n = 597)
30-d readmissions (%) (>64 y)	17.67 (n = 464)	15.17 (n = 323)

Conclusion

This study does not suggest any clinical benefit of adding an ICS to existing systemic corticosteroid therapy in the treatment of hospitalized patients with AECOPD. The adoption of a similar protocol which minimizes concomitant ICS therapy may be a consideration for hospitals to reduce institutional drug costs. A prospective, randomized controlled trial evaluating the combined use of systemic and ICS versus mono-systemic therapy in AECOPD remains warranted.