Early Peanut Introduction in Infants: Improving Guideline Adherence With EMR Standardization

This quality improvement (QI) initiative targeted all infants seen for routine care at 4, 6, and 9-month WCC encounters. The intervention occurred in an off-site, academic, residency continuity clinic at UNC serving several counties in North Carolina. The clinic serves a large population of patients on Medicaid or self-pay (66%) and patients experiencing food insecurity. Spanish is the preferred language for approximately a quarter of the population.

During the baseline period, the clinic conducted 134 WCC encounters for 4, 6, and 9-month-old infants between January 1, 2022, and March 31, 2022. The average in-room time for patients was 63 minutes, and diphtheria-tetanus-acellular pertussis (DTaP) vaccination rates for the 6-month WCC encounters were 100%. Providers documented EPI guidance during 13.9% of these visits. Home peanut introduction handouts and smart lists were unavailable during baseline data collection. There was no documentation of reported peanut consumption in any patients’ EMR during chart review at baseline.

This QI initiative lasted from April 2022 to August 2022, when a leadership handoff occurred. The project leader collected sustainability data through December 2022. Using QI methodology, the QI team developed a standardized work protocol and CDS tools within the EMR, including smart lists (a predefined list of text choices), visit templates, and patient education handouts for home peanut in English and Spanish. The team executed modifications through plan-do-study-act (PDSA) cycles to improve guideline adherence.

The project lead identified a firm commitment from the clinic director regarding the importance of EPI and standardizing an approach to implement the guidelines. Other stakeholders for the project included the clinic staff, pediatric residents, and other clinical faculty in practice.

Primary care practices often lack the resources to invest in infrastructure and training. Practice facilitation increases the likelihood of success in QI initiatives, increases provider adherence to evidence-based guidelines, and improves care quality metrics in many clinical settings.²⁴  The project lead served as the practice facilitator and engaged in QI activities, such as kickoff meetings, goal setting, maintaining momentum, and planning for leadership handoff.

The team streamlined and standardized eczema classification based on physical exam findings, body surface area affected, and topical steroid use. Thereafter, we directed the provider to follow the work protocol for the appropriate EPI guidance. The eczema classification and the EPI guidance resulted in 1 unified standard work protocol for providers (Fig 1).

The QI team determined that the instructions in the addendum guidelines, as written, were complex and might deter caregivers. The addendum guidelines offer 3 different types of peanut-containing foods.¹  The team felt using peanut butter was most practical and accessible to families, and thus we removed other forms of peanut-containing foods from our handout. PORCH, a community organization collecting food donations, supplies the clinic with food, including peanut butter, ensuring access to those families with food insecurity.

Given the site-specific needs assessment, the project lead simplified the instructions for home introduction of peanuts. The new handout captures similar safety guidance and simplifies feeding directions (Fig 2). Providers electronically inserted the handout in the patients’ after-visit summary. The clinic used UNC translation services for a Spanish-translated version of the handout to serve the clinic’s Spanish-speaking families.

There were 3 changes to the clinic’s provider templates for 4, 6, and 9-month WCC encounters. Members of the QI team adapted the note templates for these WCC encounters in the EMR to remove redundant elements in the existing note templates. The sections of the note template addressing peanut introduction were designed to be easy and quick to use.

Second, the QI lead included an anticipatory guidance section for EPI screening for each 4, 6, and 9-month WCC visit template for the associated visit. The EPI screening tool involved multiselect drop-down lists for solid food introduction, peanut introduction, and assessment of other risk factors for the development of peanut allergy. The template then prompted the provider to select a low or high-risk level for the infant related to the development of peanut allergy. Based on the risk stratification, the provider chose from another drop-down list to guide the family toward home introduction or direct the provider to order a serum immunoglobulin E to peanut. In rare cases, the work protocol prompted the provider to recommend peanut avoidance when there is a history of allergic reactions consistent with the diagnosis of peanut allergy. At 6 and 9-month WCC visits, the templates prompted providers to ask about peanut consumption since the infant’s last visit.

Third, we modified the physical skin exam findings on the EMR template to a specific detailed skin assessment drop-down list detailing the presence or absence of typical morphologic features of eczema on specified areas of the infant’s body. This focus on standardizing and improving eczema classification is essential to adequately stratify infants into the correct risk category for the development of peanut allergy. Screenshots of EMR changes appear in Fig 3.

Using tailored EMR reports and manual chart reviews, the project lead compiled data from baseline through 6 PDSA cycles (Table 1) and 5 months of sustainability after the project for all 4, 6, and 9-month WCCs in the clinic.

This project included 2 primary outcome measures, 1 focused on EPI documentation and the other on caregiver-reported peanut ingestion. An additional outcome and single process measure targeted provider behaviors in response to our QI initiative. Two balancing measures tracked the project’s impact on vaccination rates and patient in-room time. The QI team used control charts for the measures analyses, observing the Shewhart Rules in determining clinical and statistical significance (Table 2). The project lead used QI macros for Excel (KnowWare International Inc., Denver, Colorado; Version 2018) to generate statistical process control charts (SPCCs) and analyze data. We collected and reported data for all infants seen for their routine 4, 6, and 9-month WCC encounters over a period of 12 consecutive months.

The UNC institutional review board (IRB) found this QI initiative was an internal improvement project and exempt from IRB oversight. IRB #22-0559.

The average number of 4, 6, and 9-month infant visits in each biweekly period was 19 (ranging from 8 to 28). The workflow prompted providers to insert the home peanut introduction handout for infants with no or mild-moderate eczema at 4, 6, and 9-month WCCs. The home introduction handout was not expected for infants not yet developmentally ready for solids or infants already eating peanuts. The distribution rate of the home peanut introduction handout into the patients’ after-visit summary (AVS) increased from 5.2% at baseline to 54.1% during sustainability (Fig 4).

The EPI guidance documentation increased from 13.9% at baseline to 83.5% during sustainability (Fig 5). Reported peanut consumption by caregivers during 6 and 9-month WCCs increased from 0% at baseline to 34.6% during sustainability (Fig 6).

The adoption rate of the EMR templates, smart lists, and phrases increased from 2% at baseline to 73% during sustainability (Fig 7).

There were no significant changes to balancing measures during the project. DTaP vaccination rates remained at 100%. There was no significant shift in the patient time in-room of 65 minutes during the project (Fig 8).

We significantly improved measures to increase early peanut introduction in infants using QI methodology and interprofessional collaboration. Using a standardized EMR bundle of CDS tools, engagement, facilitation, and feedback to providers conducting the targeted WCC encounters helped achieve project aims.

Before this project, discussing EPI guidance between providers and caregivers under the 2017 and updated 2019 guidelines was a reasonable expectation in the primary care setting. Given the rationale for our project was that implementation of these guidelines was low, this particular outcome measure’s improvement from 13.9% to 83.5% throughout the project and, into 5 months of sustainability, shows the powerful impact of our intervention.

For the other measures analyzed with SPCCs, we exceeded our goals of 50% for the documentation of EPI guidance and distribution of the handout in the AVS. We fell just short of our goal for provider adoption of the templates and smart lists, but the measure improved from baseline with a statistically significant shift in the mean.

We tracked caregiver-reported peanut consumption at 6 and 9-month WCC visits through all project phases. We observed shifts in the percentage of reported ingestion between baseline, implementation, and sustainability as template use gained momentum. Most infants had not yet introduced solids by 4 months, but EPI guidance is still important during these early infant WCCs. We did not find any documentation of reported peanut consumption during our baseline review of charts. Though we did not reach our goal of 50% peanut consumption at 6 and 9-month WCCs during the project, 34.6% reported consumption in this age group by the end of sustainability speaks to the efficacy of our CDS bundle on patient-centered outcomes. Caregivers who reported no peanut consumption at the 6 and 9-month WCC visits allowed providers to reinforce EPI guidance.

Identifying and tracking balancing measures remain essential parts of QI work. Overall, we made significant changes to the context of these WCC visits with minimal disruption to these other visit aspects. Although in-room time did not significantly change throughout the project, we observed the beginnings of a data shift above the average in-room time during early sustainability, which coincided with the start of a new academic year and a surge of influenza, coronavirus disease 2019, and respiratory syncytial virus in the Fall of 2022. The clinic experienced increased visits during this time and had challenges with staffing because of illness. In addition, we suspect that many new resident providers forget to document when the patient visit is complete, which alters the recorded in-room time.

The availability of the coronavirus disease 2019 vaccine for infants in our target population, new residents starting with the academic year, and surges of respiratory illness coincided with a decrease in our measures tracked with SPCCs. These findings were not unexpected, but revamping toward education, engagement, and facilitation showed these measures rebounded quickly, and during the second half of sustainability, in-room time showed recovery, evidenced by consecutive decreases in the final 3 months of the project.

A particular strength of this QI project was the use of the EMR to make counseling about EPI easier for providers. The prompt in the note template reminded providers to counsel about this topic in our clinic, where many resident physicians rotate through without consistent presence to learn new processes. In addition, the patient education materials were easily inserted into the after-visit materials for patients, allowing a tool to guide a conversation with families and materials for the family to take home. Lastly, our process-driven CDS tools captured reported peanut consumption by caregivers at 6 and 9-month WCCs.

Other published QI work aiming to improve EPI guideline adherence discusses projects that used education sessions and pre and postassessments but did not incorporate CDS tools.^13,25  Although these projects documented increased provider awareness and knowledge about EPI guidelines, the projects lacked supporting data that increased knowledge translated into practice changes during WCC encounters. One QI project used emails to providers, small group education sessions, reminder cards at workstations, home introduction sheets, and onsite assistance by an allergist to improve guideline adherence.²⁶  Guideline adherence did not exceed 17% during the intervention cycles. Although these interventions represent CDS tools, they were neither standardized nor inclusive of EMR templates or smart phrases.

The most similar effort to improve EPI guideline adherence is the iREACH program, a CDS bundle utilizing EMR features and handouts to aid providers in EPI in primary care settings. In a sample of 143 WCC encounters at 4 and 6 months, results showed better adherence to guidelines (52.4%) with the use of the iREACH bundle compared with the control clinic (14.1%) without the bundle (P < .001).²⁰ 

Koplin et al showed that guideline adherence might only prevent up to 44% of peanut allergy diagnoses if providers restrict interventions to high-risk infants.¹¹  Therefore, continuing to use the home peanut introduction handout in infants with no eczema may be beneficial in preventing unnecessary peanut avoidance. PCPs are most likely to interface with otherwise healthy infants, and assessing infants for early atopy offers a unique opportunity to practice primary and secondary prevention. Most infants observed during this QI project met the criteria for immediate home introduction of peanuts and were able to avoid unnecessary allergy referrals. PCPs should confidently promote broad EPI and diet diversification in patients showing developmental readiness for solid food introduction.

Although this project at a single site showed improvements in early peanut introduction, the goal of generalizability of this QI initiative to the broader population remains an unmet need. Moving the needle on this initiative will require working with more primary care clinics. One barrier to generalizability is that some offices do not use EMR systems, making the reproducibility of these interventions more challenging. However, the handouts can serve as stand-alone documents for printing and distribution to families. Additionally, CDS tools could include modifying prompts about EPI guidance on paper charting systems in place of EMR accessibility.

Another limitation of our study is that there are significant anticipatory guidance recommendations for WCCs. Our study examined several balancing measures but could not capture the full complexity of WCCs and aspects of the visit that may have been lost because of adding in this counseling. Future projects should assess caregiver understanding and knowledge of EPI guidelines after visits, as this was beyond the scope of our project.

Peanut allergy in children is a population health problem affecting families and healthcare systems. Research shows EPI can reduce the incidence of peanut allergy, but adoption of this practice remains low in primary care settings. QI methodology improved documentation of EPI guidance at routine WCC encounters and reported rates of peanut consumption without impacting other measures at our site. Broader PCP use of CDS tools and EMR standardization could improve guideline adherence to prevent peanut allergy in infants.

Thank you to the providers and staff at the UNC Children’s Primary and Specialty Care Clinic for engaging with this project. Thank you also to Dr. Edwin Kim for his mentorship.

AAP

American Academy of Pediatrics

AVS

after visit summary

CDS

clinical decision support

EMR

electronic medical record

EPI

early peanut introduction

IRB

Institutional Review Board

LEAP

Learning Early About Peanuts

PCP

primary care provider

PDSA

Plan Do Study Act

QI

quality improvement

SPCC

statistical process control chart

UNC

University of North Carolina

WCC

well child check