Implementation of a Critical Asthma Protocol in a Pediatric ICU

Patients

After approval from the institutional review board of the University of Minnesota, a pre-intervention cohort age 2–18 y was identified by problem list and discharge diagnosis codes of critical asthma, status asthmaticus, asthma with exacerbation, and asthma without status asthmaticus using a patient database (Virtual Pediatric Systems, Los Angeles, CA). The Virtual Pediatric Systems database is a large, multi-institutional database that allows for patients to be identified in a variety of ways, in this instance by diagnosis codes. This cohort timeframe of January 1, 2015, to December 31, 2016, was intentionally selected to be outside of an outbreak of enterovirus D68 from mid-August 2014 to January 2015 that led to severe respiratory symptoms in children as described by many, including the Centers for Disease Control and Prevention (https://www.cdc.gov/non-polio-enterovirus/about/ev-d68.html, Accessed September 2, 2019).^15-16 Patients were then reviewed by the authors to ensure inclusion and exclusion criteria were met. Subjects were included if they were 2–18 y old when presenting and symptoms were thought to be consistent with an obstructive respiratory pattern due to asthma exacerbation. Patients were excluded if they had a condition causing wheezing or obstruction that was not felt to be due to asthma per the emergency department and PICU physicians, or if they had any of the following comorbidities: congenital or acquired heart disease, congenital airway anomalies, severe developmental delay, neuromuscular disease, severe bronchopulmonary dysplasia, pulmonary hypertension, primary immunodeficiency, or cystic fibrosis. Demographics and data pertaining to primary and secondary end points along with other medication administration information were gathered using the hospital electronic medical record. The Respiratory Assessment Score (RAS) was also collected at multiple points including initial emergency department score, subsequent emergency department score, and score upon admission to PICU. The RAS was adapted for use at University of Minnesota Medical Center from the Pulmonary Score, which has been validated for use in pediatric asthma patients aged 5–17 y in the emergency department, and is the standard assessment by which respiratory status is serially monitored across our institution.¹⁷ The RAS score reflects patient’s symptoms including age-dependent breathing frequency, auscultatory findings, and retractions. On a scale of 1–9, we defined 1–3 as mild, 4–6 as moderate, and 7–9 as severe (Fig. 1). Study data were collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools hosted at the University of Minnesota. REDCap is a secure, web-based application designed to support data capture for research studies, providing an intuitive interface for validated data entry, audit trails for tracking data manipulation and export procedures, automated export procedures for seamless data downloads to common statistical packages, and procedures for importing data from external sources.

• Download figure
• Open in new tab
• Download powerpoint

Fig.1.

Respiratory Assessment Score (RAS) scoring system, initial therapies, and inclusion/exclusion criteria. RT = respiratory therapist; NPO = nil per oral; PO = per oral (by mouth); I/O = input/output; IV = intravenous; QID = 4 times daily; QnH = every n hours; pMDI = pressurized metered-dose inhaler; ICS = inhaled corticosteroids; BPD = bronchopulmonary dysplasia; RAS = Respiratory Assessment Score.

Asthma Protocol

A standardized critical asthma protocol, the Pediatric Severe Asthma Pathway (PSAP, see Fig. 2), was developed using results of a provider survey, a literature review, and knowledge of unit practices. Goals of the PSAP include both rapid escalation and de-escalation of interventions, starting with those that are the least invasive and best tolerated. Patients are placed into categories based on RAS scores, with initial treatments often starting in the pediatric emergency department or outside hospital. The PSAP standardizes the frequency of monitoring, assessments, and administration of continuous albuterol, intravenous steroids, and intermittent ipratropium inhalation for all patients admitted to the PICU with severe asthma (ie, initial RAS scores of 7–9, which have traditionally been started on continuous albuterol by the institution’s pediatric emergency department and thus require admission to the PICU). While critical asthma patients are typically assessed hourly in our institution, a 2-h interval was chosen to determine improvement, deterioration, or no change to allow for an effect to be seen. A RAS score of 7 was used as the cutoff for escalation or de-escalation, as this represents the minimum score our institution classifies as severe asthma. RAS scores of 7 or above progressed stepwise along the protocol, while scores 6 or below would de-escalate stepwise in reverse order. In addition to the escalation and de-escalation protocol, all medication usage was standardized in terms of dosage, frequency, and delivery. Medication administration was tracked throughout the study by recording dosages given and whether boluses were given if indicated, and medication administration was documented as given correctly or incorrectly in a binary fashion. As part of the PSAP, respiratory therapists document RAS score hourly, as well as changes in management and rationale for deviation from the protocol using a paper form. PSAP teaching sessions were conducted for staff physicians, fellows, residents, and respiratory therapists. A copy of the PSAP is included at the bedside of every asthma patient in the PICU. Continuous albuterol is administered with an Aerogen Solo (Aerogen, Galway, Ireland) using a 25-mL syringe setup through a standard face mask or through a heated system if on high-flow nasal cannula. In our institution, this is run through the Aerogen box or through a Servo-i ventilator (Getinge, Wayne, NJ) using a T-piece.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Escalation/de-escalation protocol. RAS = Respiratory Assessment Score; IV = intravenous; HFNC = high-frequency nasal cannula; VBG = venous blood gas; BMP = basic metabolic panel; US = ultrasound; BPAP = bi-level positive airway pressure; Q2H = every 2 h; ECMO = extracorporeal membrane oxyenation.

Outcome Measures

A 22-month post-intervention cohort was reviewed from August 1, 2017, to May 5, 2019, through the Virtual Pediatric Systems database with queries using the same inclusion and exclusion criteria as the pre-intervention cohort. As with the pre-intervention cohort, patient charts were reviewed to ensure inclusion and exclusion criteria were met, and all data were taken directly from the institution’s electronic medical record. The primary end point was time on continuous albuterol, with secondary end points of hospital LOS, PICU LOS, and time from end of continuous albuterol to transfer out of ICU. The latter end point is used as a marker for possible confounders to PICU LOS. All post-intervention subjects were analyzed for adherence to protocol. Subject treatment was deemed not protocol-adherent if escalation or de-escalation was not taken every 2 h as dictated by the RAS score at that time. Medication dosing and administration were also reviewed and deemed appropriate or off-protocol per the schedule defined by the protocol. After preliminary results, a separate cohort, labeled the protocol-adherent cohort, was identified post hoc if adherence was completely maintained, and results were analyzed as a subset to evaluate the protocol.

Statistical Analysis

Demographic distribution differences between the cohorts were evaluated with the chi-square test for gender and with the Fisher exact test for race due to small numbers for some categories. A 2-sample t test was used to test for differences in mean age. Unadjusted distribution differences in time on continuous albuterol, time off albuterol prior to discharge, PICU LOS (in hours) and hospital LOS (in days) were evaluated with the Wilcoxon rank-sum test, a non-parametric test appropriate for skewed distributions.

Generalized linear models for each outcome measure were run to evaluate differences between pre- and post-intervention cohort’s mean values for time on continuous albuterol, time off albuterol prior to PICU discharge, PICU LOS, and hospital LOS adjusted for age and gender. Sensitivity analyses included unadjusted analysis excluding 4 subjects with outlier values in one or more outcomes (> 100 h on continuous albuterol, PICU LOS > 100 h, or hospital LOS > 15 d) and generalized linear models for each outcome adjusted for race in addition to age and gender.