Improving Single Maintenance and Reliever Therapy for Patients Admitted for Asthma Exacerbation

Abstract

Introduction:

In 2020, single maintenance and reliever therapy (SMART) became guideline-recommended care for school-age children in the United States with poorly controlled, persistent asthma. Pediatric inpatient providers are well positioned to prescribe SMART, as they often care for patients with poorly controlled asthma. Our interdisciplinary team aimed to increase the proportion of SMART prescriptions at discharge for eligible pediatric patients admitted for asthma exacerbation from 17% to 40% by September 2023, consistent across strata of payor type, race, and Child Opportunity Index (COI).

Methods:

Four primary drivers of SMART prescription at discharge were identified: familiarity, prescriber culture, decision support, and logistics. Interventions targeting these drivers, including education and clinical decision support, were implemented during 10 Plan-Do-Study-Act cycles. This quality improvement project included patients who were prescribed an inhaled controller medication on admission and had 2 or more hospitalizations and/or emergency room visits for asthma exacerbation requiring systemic corticosteroids within 12 months. The outcome measure was SMART prescription at discharge, stratified by payor type, race, and COI.

Results:

Between January 2021 and December 2023, 312 hospital encounters involving 215 unique patients occurred. SMART prescription at discharge increased from 17% at baseline to 38% and was sustained for 19 months. Similar increases in SMART prescriptions at discharge were observed among Black patients, those with government-sponsored health insurance, and those with very low COI.

Conclusions:

Using quality improvement methodology, SMART prescriptions increased at discharge for pediatric patients admitted for asthma exacerbation, including in demographic strata where disparities are often observed.

INTRODUCTION

In 2020, the National Heart, Lung, and Blood Institute (NHLBI) released the “2020 Focused Updates to the Asthma Management Guidelines” (2020 Focused Updates), which includes a recommendation for single maintenance and reliever therapy (SMART).¹ SMART involves treatment with a single inhaled corticosteroid (ICS)–formoterol combination medication used for both daily maintenance and symptom relief for children with moderate to severe persistent asthma. The 2020 Focused Updates recommendation is based on high-certainty evidence for children aged 12 and older and moderate-certainty evidence for children aged 5–11 years.^1–3 Also included in the Global Initiative for Asthma⁴ Guidelines, however, with slightly different recommendations, SMART represents a paradigm shift in asthma management within the United States. Multiple barriers to its implementation have been identified, such as patient/caregiver familiarity and the ability to obtain medication refills.⁵ Practical guidance for implementing SMART is published,⁶ including patient/caregiver education regarding self-management, a customized asthma action plan, and acknowledgement of insurance coverage; however, successful, sustained integration into pediatric clinical practice has not yet been demonstrated.

Pediatric inpatient providers are well positioned to prescribe SMART, as they frequently care for children admitted for asthma exacerbation who are potentially eligible for adjustment of their maintenance asthma medication (controller medication). Prior efforts to improve inpatient asthma care have primarily focused on standardizing acute exacerbation treatment through clinical pathways.^7–10 These initiatives led to increased use of recommended acute care medications, a shorter length of stay (LOS), and lower healthcare costs and readmissions,^7–10 as well as connections with outpatient and home-based services.^11–13 Controller medication accuracy at discharge is an area where improvement efforts are less well documented.^12,14–16

Approximately 7% of children in the United States have a diagnosis of asthma, with about 4% of these children requiring hospitalization for asthma exacerbation each year.¹⁷ Though hospitalizations for asthma exacerbation have declined over the last decade, well-described disparities regarding pediatric asthma persist.¹⁸ Compared with White children with asthma, Black children with asthma are nearly 3 times more likely to be hospitalized and 4 times more likely to die from asthma.^17,18 In addition, up to 20% of children hospitalized for asthma exacerbation are rehospitalized within a year, with Black and Hispanic children at higher risk.¹⁹ Furthermore, children from low socioeconomic households are more likely to have persistent asthma.^20,21 Recognizing disparities within pediatric healthcare, quality improvement (QI) efforts are increasingly called upon not only to improve care but also to advance equity in health.²²

Hospitalization serves as an opportunity to implement safe, timely, effective, efficient, patient-centered, and equitable guideline-concordant recommendations for the chronic management of asthma.^15,23 Eleven months after the release of the 2020 Focused Updates, at our children’s hospital, we identified that the SMART prescription rate for eligible patients admitted for asthma exacerbation was lower than the historical institutional trend for controller medication adjustment at discharge (40%). In response, our specific, measurable, achievable, relevant, timebound, inclusive, and equitable (ie, SMARTIE) aim was to increase SMART prescription at discharge for eligible pediatric patients admitted for asthma exacerbation from 17% to 40% by September 2023, with sustained increases across payor type, race, and Child Opportunity Index (COI) strata.

METHODS

Context

This QI project took place at a large, university-affiliated, tertiary care children’s hospital with an associated pediatric emergency department (ED) and a broad catchment area that includes southeastern Pennsylvania, New Jersey, and Delaware. The hospital has more than 2,500 asthma hospitalizations annually. Children hospitalized for an asthma exacerbation are admitted from the hospital’s ED or transferred from another hospital. They may be admitted to a general pediatrics, pulmonary, or allergy service, or pediatric intensive care unit. Patients are cared for by a multidisciplinary team including physicians (ie, resident physicians [residents], attending physicians), advanced practice providers, nurses, respiratory therapists, child life specialists, case managers, clinical pharmacists, and social workers. The institution uses an externally available Inpatient Clinical Pathway for Children with Acute Asthma Exacerbation (inpatient asthma clinical pathway) to standardize care.²⁴

QI Population

From December 2021 to December 2023, the QI effort included children 5–18 years old admitted for asthma exacerbation if prescribed a controller medication according to the following steps of the 2020 NHLBI Focused Updates¹: step 2 (daily low-dose ICS and as-needed short-acting beta 2-agonist) or step 3 (daily and as-needed combination low-dose ICS–formoterol). A manual chart review identified eligible patients. Therefore, QI efforts focused on children with a history of frequent healthcare use, which we defined as 2 or more hospitalizations and/or ED visits for asthma exacerbation requiring systemic corticosteroids within 12 months of the admission, including the current admission. We used 2 courses of systemic corticosteroids within 12 months for asthma exacerbation as a proxy for poor asthma control. All included patients were admitted on step 2 or 3 therapy with poor asthma control; therefore, they were eligible for SMART according to the 2020 NHLBI Focused Updates.¹ We included multiple hospital encounters for individual patients meeting these criteria within the time frame. Only hospitalizations and/or ED visits for asthma exacerbation at our institution were included. This initiative excluded patients receiving biologic therapy, those who left the hospital against medical advice, and patients whose controller medication was adjusted within 1 week of admission or for whom a controller medication prescription could not be confirmed within the electronic health record (EHR).

Interventions

Physicians, informaticians, and a respiratory therapist formed an interdisciplinary QI team and applied the Model for Improvement framework throughout the project.²⁵ The team identified barriers to prescribing SMART at discharge for eligible patients admitted for asthma exacerbation through gathering stakeholder feedback with the institution’s Asthma Population Health Workgroup, an established multidisciplinary group including pulmonologists, allergists, social workers, and respiratory therapists dedicated to improving the care of children with asthma. Additionally, we held a focus group with Pediatric Hospital Medicine attendings and solicited feedback from clinical pharmacists, nurses, and residents on perceptions of prescribing SMART. Input from these groups informed the creation of a driver diagram (Fig. 1).

Fig. 1.

Driver diagram with primary and secondary drivers for single maintenance and reliver therapy (SMART) prescription at discharge for eligible patients admitted for asthma exacerbation with associated change ideas and assigned impact/effort. COI, Childhood Opportunity Index; PA, Prior Authorization; RN, registered nurse; RT, respiratory therapist.

Four primary drivers of SMART prescription at discharge were identified: familiarity, prescriber culture, decision support, and logistics. The QI team developed change ideas and prioritized them based on anticipated impact and effort (Fig. 1). The first driver, familiarity, centered on the care team’s difficulty in identifying SMART-eligible patients and challenges in educating patients/caregivers about SMART. This driver is addressed through multidisciplinary education and revisions to visual aids (ie, asthma flipbook). Prescriber culture, the second driver, is related to prescribers’ inconsistent assessment and documentation of asthma control, tendency to defer adjustment of controller medication to outpatient providers, and the perception that SMART was difficult to prescribe. We addressed this driver through education, revisions to the asthma admission note template with prompts to assess and document asthma control, and the creation of an EHR-based inpatient SMART order set. Subsequently, the order set was updated to include a combined prescription template for both daily and as-needed medications, and the asthma discharge instructions were revised to include SMART-specific instructions. Interventions also included positive feedback communicated via a monthly email to all providers who prescribed SMART. The third driver, decision support, described a lack of alignment between the 2020 Focused Updates and the existing inpatient asthma clinical pathway and asthma admission note template. This driver is addressed by revising the pathway and note template to include the 2020 Focused Updates. Logistics, the final driver, referenced insurance barriers: the limits necessary to facilitate SMART were not included in the state’s Medicaid Formulary, and there was an initial concern that the institution’s outpatient pharmacy did not have an adequate supply of SMART-appropriate inhalers. In partnership with case managers and pharmacists, we included language to facilitate SMART identification and prior authorization approval in the prescription “Pharmacy Comments” within the EHR-based inpatient SMART order set. Prior authorization typically lasted 6–12 months, depending on the pharmacy benefit manager. Further, the institution’s outpatient pharmacy closely monitored the SMART-appropriate inhaler supply. These interventions were implemented and refined in a series of Plan-Do-Study-Act (PDSA) cycles (Table 1). Notably, PDSA cycle 4 included multiple interventions implemented simultaneously as a “SMART bundle.”

Table 1.

Description of 10 PDSA Cycles to Increase SMART Prescription at Discharge for Eligible Patients Admitted for Asthma Exacerbation

PDSA Cycle	Date	Title	Description
1	December 2, 2021	Asthma Population Health Workgroup Education & Focus Group	Presented baseline data, provided SMART education (case-based discussion regarding SMART eligibility and evidence), and offered feedback regarding inpatient barriers/facilitators to prescribing SMART at the monthly meeting
2	March 16, 2022	PHM Education & Focus Group	Presented baseline data, provided SMART education (case-based discussion regarding SMART eligibility and evidence), and offered feedback regarding inpatient barriers/facilitators to prescribing SMART at the monthly Section of Pediatric Hospital Medicine Quality Improvement (QI) Meeting
3	April 27, 2022	Nurse (RN) Education	Provided SMART education (case-based discussion regarding SMART eligibility and evidence) at the monthly hospital-wide nursing Shared Governance Meeting
4	June 9, 2022	SMART Bundle	1. Revised inpatient asthma clinical pathway to include CDS for (1) assessment and documentation of asthma control and (2) 2020 Focused Updates 2. Created an EHR-based inpatient SMART order set 3. Revised visual aids (ie, Asthma Binder) for patient/caregiver education 4. Shared example of Asthma Action Plan that included SMART with PHM attendings and pediatric residents 5. Displayed SMART Tip Sheet in care team workrooms 6. Advertised revision of inpatient asthma clinical pathway and EHR-based inpatient SMART order set via inpatient screen savers 7. Partnered with the pharmacy to ensure adequate stock of SMART-appropriate inhalers
5	October 17, 2022	SMART Feedback	Monthly email to attendings who prescribed SMART at discharge in the prior month to (1) provide thanks and (2) request feedback
6	October 25, 2022	Respiratory Therapy Education	Provided SMART education (case-based discussion regarding SMART eligibility and evidence) at the monthly hospital-wide respiratory therapy Town Hall Meeting
7	November 7, 2022	Link Daily and As Needed (PRN) Prescription (Rx)	Linked prescription for daily and as-needed prescription for ICS–formoterol medications in the EHR and EHR-based inpatient SMART order set
8	February 13, 2023	Revise Asthma Admission Note Template	Revised the asthma admission note template to include CDS for (1) assessment and documentation of asthma control and (2) 2020 Focused Updates
9	April 28, 2023	Revise Discharge (DC) Instructions	Revised EHR status asthmaticus discharge instructions to include SMART-specific instructions
10	September 26, 2023	Allergy & Pulmonary Education & Focus Group	Presented QI project, including SMART education, and obtained feedback at the Division of Pulmonary and Sleep Medicine’s monthly Grand Rounds. The Division of Allergy and Immunology was invited to attend the presentation

Description includes date, title, and description of the PDSA cycle.

PHM, Pediatric Hospital Medicine.

Study of the Intervention

To study the impact of our interventions, we compared the rate of SMART prescriptions at discharge during the baseline period (January–November 2021) with the intervention period (December 2021–December 2023). Patients were identified through the institution’s asthma report (Epic Systems, Verona, WI), which includes admissions in an “observation” or “inpatient” status with (1) an asthma pathway order set ordered and/or (2) a final coded diagnostic-related group for asthma exacerbation. This report is filtered to include patients aged 5–18 years with 2 or more hospitalizations and/or ED visits for asthma exacerbation. It is subsequently manually reviewed to identify SMART eligibility and a SMART prescription at discharge. Manual chart review also identified the patient’s number of asthma exacerbation hospitalizations and ED visits in the last 12 months at the time of admission. We randomly selected 50 hospital encounters for independent review by 2 physicians to identify SMART eligibility and SMART prescription at discharge. This process resulted in 100% agreement between physicians. Additional data elements (eg, payor group, race, COI, age at admission, legal sex, 90-day revisitation, and LOS) were queried via an institutional data warehouse (DBeaver Community, https://dbeaver.io) for each hospital encounter. We reported race by the patient/caregiver during the health system registration process. Institutional EHR-based inpatient SMART order set use was obtained via an EHR workflow tool, Phrase Health (Philadelphia, PA). An EHR-based inpatient order set was available for all patients admitted to the institution for asthma exacerbation. Statistical process control (SPC) charts were created using commercial Excel software (QI Macros, Denver, CO) and were available to the QI team throughout the project.

Measures

The outcome measure was the proportion of hospital encounters in which SMART was prescribed at discharge, defined as a daily and as-needed ICS–formoterol prescription. The process measure was the institutional use of the EHR-based inpatient SMART order set. Balancing measures included LOS (in h) and the proportion of patients who experienced revisitation, defined as an ED visit or hospitalization for an asthma exacerbation within 90 days of discharge.

Analysis

SPC charts were used to analyze each measure. Specifically, we used p-charts to assess the impact of improvement efforts for SMART prescriptions at discharge and 90-day revisitation, an XmR chart to display LOS, and a c-chart to display institutional use of the EHR-based inpatient SMART order set. Due to seasonality²⁶ and the impact of the COVID-19 pandemic²⁵ on asthma exacerbation ED visits and hospitalizations, rational subgrouping of 5 or 10 consecutive hospital encounters, depending on sample size, was used to analyze outcomes and balancing measures. Rational subgrouping is a process that organizes data generated under similar conditions into groups to measure variation between subgroups.^27,28 We applied the standard Associates in Process Improvement rules²⁸ to evaluate for special-cause variation across all SPC charts. We further stratified the outcome measure by payor type, race, and COI. In addition, we examined the outcome measure after excluding patients who were prescribed SMART as their controller medication at admission. This article was written according to the Standards for Quality Improvement Reporting Excellence (SQUIRE) 2.0 guidelines.²⁹

Ethical Considerations

Our work includes analysis of well-described disparities in pediatric asthma, including the social construct of race, as well as payor type and COI, to examine whether interventions narrowed, maintained, or widened disparities.^19–21,30 Due to the limited number of hospital encounters, we are only able to assess dichotomous strata in our analyses: government/nongovernment payor type, Black/non-Black race, and very low/nonvery low COI. Previously described disparities inform subgroups.^18,20,21,30 This QI initiative does not meet the criteria for human subjects research as determined by the hospital’s institutional review board.

RESULTS

Our project involved 312 hospital encounters for asthma exacerbations among 215 unique patients. Of these, 52 encounters occurred during the baseline period (January–November 2021) and 260 occurred during the intervention period (December 2021–December 2023) (Table 2).

Table 2.

Characteristics of Hospital Encounters (n = 312) and Unique Patients as per Their First Discharge Record (n = 215) Included in the QI Project Population

Characteristic	Hospital Encounters, Total (n = 312)	Patient, First Discharge Record (n = 215)
Age at Admission, y
Range	5.0–18.5	5.0–18.5
Median	9.7	9.8
Mean, SD	10.0 ± 3.3	10.2 ± 3.5
IQR	7.4–11.9	7.4–12.3
Legal sex, n (%)
Female	133 (42.6)	94 (43.7)
Male	179 (57.4)	121 (56.3)
Payor type, n (%)
Commercial	61 (19.6)	43 (20.0)
Government	241 (77.2)	168 (78.1)
Other	2 (0.6)	1 (0.5)
Not listed	8 (2.6)	3 (1.4)
Race, n (%)
Black or African American	245 (78.5)	160 (74.4)
White	28 (9.0)	21 (9.8)
Other	25 (8.0)	22 (10.2)
Multiracial	11 (3.5)	9 (4.2)
Asian	2 (0.6)	2 (0.9)
Refused	1 (0.3)	1 (0.5)
COI, n (%)
Not reported	4 (1.3)	3 (1.4)
Very low	236 (75.6)	153 (71.2)
Low	22 (7.1)	17 (7.9)
Moderate	23 (7.4)	18 (8.4)
High	6 (1.9)	6 (2.8)
Very high	21 (6.7)	18 (8.4)
Hospitalizations and ED visits within the last 12 mo for asthma exacerbation
Range	2.0–13.0	2.0–13.0
Median	3.0	3.0
Mean, SD	3.6 ± 1.8	3.4 ± 1.6
IQR	2.0–4.0	2.0–4.0
ICU exposure, n (%)
No	250 (80.1)	177 (82)
Yes	62 (19.9)	38 (18)
Discharge service, n (%)
General pediatrics	263 (84.3)	186 (86.5)
Critical care	18 (5.8)	12 (5.6)
Pulmonary	18 (5.8)	10 (4.7)
Allergy	5 (1.6)	4 (1.9)
Not listed	3 (1.0)	3 (1.4)

ICU, intensive care unit; IQR, interquartile range.

Outcome Measure

SMART prescription at discharge increased from a baseline value of 17%–38%, demonstrating special-cause variation after PDSA cycles 3 and 4 with sustained improvement for 19 months (Fig. 2). SMART prescription at discharge for children with a government payor increased from 14% to 37%, demonstrating improvement consistent with special-cause variation after PDSA cycle 4. For children with a nongovernment payor type, the proportion of SMART prescriptions at discharge was 32% and did not demonstrate improvement (Fig. 3). For Black children, the SMART prescription rate at discharge increased from 15% to 35%, whereas for non-Black children, the proportion remained at 26%, unchanged during the intervention period. (See figure 1A, Supplemental Digital Content 1, which displays outcome measure of SMART prescription at discharge for eligible patients admitted for asthma exacerbation, stratified by race [A] and COI [B] demonstrating improvement consistent with special-cause variation for children who are Black [labeled at 35%] or for children with very low COI [labeled at 51%], respectively. Children who are non-Black [labeled at 26%] or who have nonvery low COI [labeled at 24%] do not demonstrate improvement. Data are displayed using rational subgrouping via p-charts. Each data point represents 5 hospital encounters [* indicates a subgroup with fewer than 5 hospital encounters]. The aim line is labeled at 40%. For race, the UCL is 99% for children who are Black and 80% for non-Black children. For COI, the UCL is 100% for children with very low COI and 76% for children with nonvery low COI. The LCL for all p-charts is 0%. Direction of improvement indicated by an upward arrow. LCL, lower control limit; UCL, upper control limit, https://links.lww.com/PQ9/A736.) The greatest degree of improvement occurred for children with a very low COI, where the SMART prescription rate at discharge increased from 18% to 51%. This finding demonstrates special-cause variation after PDSA cycle 2 and again after PDSA cycle 9, with sustained improvement. For children with nonvery low COI, SMART prescription at discharge was 24% at baseline and did not demonstrate improvement (Supplemental Digital Content 1, https://links.lww.com/PQ9/A736). After removing from the analysis patients prescribed SMART as their controller medication at the time of admission, 260 hospital encounters for asthma exacerbation remained (48 and 212 encounters occurring during the baseline and intervention periods, respectively). Based on these data, the SMART prescription rate at discharge was 10%, which improved to 31% after PDSA cycles 3 and 4, with sustained improvement. (See figure 2, Supplemental Digital Content 2, which displays outcome measure of SMART prescription at discharge for eligible patients, excluding patients with SMART prescribed as their controller medication at the time of admission. These data demonstrate improvement consistent with special-cause variation labeled at 31%. Data are displayed using rational subgrouping via a p-chart. Each data point represents 10 hospital encounters. The aim line is labeled at 40%. The UCL is 75%, and the LCL is 0%. Direction of improvement indicated as an upward arrow. LCL, lower control limit; UCL, upper control limit, https://links.lww.com/PQ9/A737.)

Fig. 2.

Outcome measure of SMART prescription at discharge for eligible patients admitted for asthma exacerbation, demonstrating improvement consistent with special-cause variation labeled at 38%. Data are displayed using rational subgrouping in a p-chart. Each data point represents 10 hospital encounters (* indicates a subgroup with fewer than 10 hospital encounters). The aim line is labeled at 40%. The UCL is 82%, and the LCL is 0%. Direction of improvement indicated as an upward arrow. LCL, lower control limit; PRN Rx, as-needed prescription; UCL, upper control limit.

Fig. 3.

Outcome measure of SMART prescription at discharge for eligible patients admitted for asthma exacerbation stratified by payor group. Patients with the government as their payor (37%) demonstrate improvement consistent with special-cause variation. Patients with a nongovernment payor (33%) do not demonstrate improvement. Data are displayed using rational subgrouping via a p-chart. Each data point represents 5 hospital encounters (* indicates a subgroup with fewer than 5 hospital encounters). The aim line is labeled at 40%. The UCL is 100% for patients with the government as their payor and 91% for patients with a nongovernment payor. The LCL is 0% for both. Direction of improvement indicated as an upward arrow. LCL, lower control limit; UCL, upper control limit.

Process Measures

Institutional use of the EHR-based inpatient SMART order set demonstrated stable uptake following its implementation in June 2022 as part of PDSA cycle 4. (See figure 3, Supplemental Digital Content 3, which displays monthly EHR-based inpatient SMART order set use for all patients admitted for asthma exacerbation, with stable uptake without improvement during the intervention period, resulting in 9.263 counts. The UCL is 18.294 counts, and the LCL is 0.133 counts. Direction of improvement indicated as an upward arrow. LCL, lower control limit; UCL, upper control limit, https://links.lww.com/PQ9/A738.)

Balancing Measures

Ninety-day revisitation for patients with a SMART prescription at discharge was 5% and 13% for patients without a SMART prescription at discharge. (See figure 4A, Supplemental Digital Content 4, which displays that 90-day revisitation for patients with (labeled at 5%) and without [labeled at 13%] a SMART prescription at discharge does not demonstrate improvement [A]. The UCL for revisitation among patients with a SMART prescription at discharge is 26%, and the LCL is 0%. The revisitation UCL for patients without a SMART prescription at discharge is 44%, and the LCL is 0%. LOS for patients with [labeled at 38.870 d] and without [labeled at 41.940 d] a SMART prescription at discharge does not demonstrate sustained improvement [B]. The LOS UCL for patients with a SMART prescription at discharge is 89.637 days, and the LCL is 0 days. The LOS UCL for patients without a SMART prescription at discharge is 83.113 days, and the LCL is 0 days. Data are displayed via rational subgrouping. Each data point represents 10 hospital encounters [* indicates a subgroup with fewer than 10 hospital encounters]. LCL, lower control limit; UCL, upper control limit, https://links.lww.com/PQ9/A739.) Neither met the criteria for special-cause variation. Patients with a SMART prescription at discharge had an LOS of 38.9 hours, whereas patients without a SMART prescription had an LOS of 42.0 hours (Supplemental Digital Content 4, https://links.lww.com/PQ9/A739). Neither group demonstrated sustained improvement in LOS. Data were unavailable for 9 hospital encounters.

DISCUSSION

We significantly increased the number of SMART prescriptions at discharge for pediatric patients from 17% to 38%, which is just slightly below our goal. This improvement occurred within 16 months of the publication of the 2020 Focused Updates¹ and has been sustained for 19 months. Importantly, our effort demonstrated an increase in SMART prescriptions at discharge among children from demographic groups associated with asthma morbidity, including children who identify as Black, those with government as their payor group, and those from lower opportunity neighborhoods (ie, very low COI).²² This highlights the opportunity for QI methodology to improve both guideline-concordant care and health equity; however, further efforts need to examine why our interventions did not increase SMART prescription at discharge for children not in these groups (nongovernment payor type, non-Black, nonvery low COI) to ensure that a disparity is not introduced within the QI effort.

Prior literature provides recommendations for implementing SMART, and a recent pharmacist-led initiative demonstrated increased SMART prescribing at an adult primary care practice through education and clinical decision support (CDS).³¹ Similarly, our QI project used these interventions, and we observed improvement after PDSA cycles 3 and 4, which included education and the SMART bundle. Historically, education has not consistently led to sustained improvement; therefore, interventions targeting CDS are most likely to support sustained improvement. Two CDS interventions (ie, the inpatient asthma clinical pathway and the EHR-based inpatient SMART order set) were included in PDSA cycle 4, as part of the SMART bundle. Institutional use of the inpatient asthma clinical pathway is not measured because it is publicly available. Use of the EHR-based inpatient SMART order set was consistent following its introduction and coincided with sustained improvement in the outcome measure. The EHR-based inpatient SMART order set required the provider to manually enter the order set name, which may have limited uptake.

The implementation of clinical practice guidelines has faced various barriers, including physician attitudes, knowledge, and the environment.³² Our project benefited from in-depth stakeholder feedback, enabling the easy integration of interventions tailored to the local context into clinical practice. Specifically, the QI team strategically created a bundle of interventions that simultaneously improved the ability to prescribe (eg, inpatient asthma clinical pathway revision, EHR-based inpatient SMART order set creation) and discharge (eg, visual aid revision, asthma action plan examples) a patient on SMART.²⁶ The analysis, including only those patients who were not prescribed SMART as their controller medication at the time of admission, also demonstrated an improvement from 10% to sustained improvement of 31%, reinforcing our belief that our efforts facilitated SMART prescription at discharge.

Results of one of the balancing measures revealed a promising signal: the 90-day revisit rate for patients with a SMART prescription at discharge was 5%, compared with 13% for those without a SMART prescription. This finding may not be surprising, given evidence that SMART reduces the risk of severe asthma exacerbations.^3,33 Further investigation is required to better understand the relationship between SMART prescription at discharge and revisitation.

The project has several limitations. Regarding project design, not every patient admitted for asthma exacerbation should have their controller medication adjusted; patients not prescribed SMART were not assessed. For example, poor asthma control may be secondary to controller medication misuse, with education serving as the best intervention. Additionally, controller medication adjustment is a shared decision; we did not assess patients/caregivers who declined SMART, nor did we follow up with patients/caregivers who agreed to SMART to confirm adherence. Finally, alternative options to SMART are included in the 2020 Focused Updates, but we did not specify when those alternatives were selected. Regarding data collection, we did not include hospitalizations and/or ED visits for asthma exacerbation that may have occurred at other institutions.

Additionally, it is possible that systemic corticosteroid use within the last 12 months was underreported, as we did not include visits to outside institutions, ambulatory care, or urgent care. The EHR-based inpatient SMART order set includes all patients admitted for asthma exacerbation at our institution; therefore, it does not reflect only the patients included in this QI project. Finally, regarding generalizability, this QI effort includes patients with a history of frequent healthcare use; given the limited number of hospital encounters, we were only able to assess dichotomous strata in the analysis of our outcome measure. Therefore, conclusions may not be generalizable, and it is difficult to detect meaningful differences between groups.

CONCLUSIONS

The QI methodology increased the proportion of SMART prescriptions at discharge for pediatric patients admitted for asthma exacerbation, including in demographic strata where disparities are often observed (eg, race, payer type, and COI).

ACKNOWLEDGMENTS

For assistance with the study, we thank Michael Posencheg, MD; Bridget Rauch, BS; Honey Pezzimenti; Elizabeth Brooks, MPH, MSSP; Samantha Horn, MD; Eric Shelov, MD, MBI; David Anderson, MD; and Daniel Hyman, MD, MMM. For editing assistance, we thank Kristin McNaughton, MHS.

Supplementary Material

pqs-11-e867-s001.pdf ^{(260.4KB, pdf)}

pqs-11-e867-s002.pdf ^{(76.2KB, pdf)}

pqs-11-e867-s003.pdf ^{(66.4KB, pdf)}

pqs-11-e867-s004.pdf ^{(184.5KB, pdf)}