Overview

This project at several academic medical center-affiliated pediatric primary care practices utilized an electronic health record screening tool for sugar-sweetened beverage and 100% fruit juice intake to inform overall pediatric obesity population health management efforts. 

Clinic

Wake Forest Baptist Medical Center (WFBH)

Collaborators

Joseph A. Skelton, MD, MS
Fang-Chi Hsu, PhD
Pascaline Ezouah, BS
Elsie M. Taveras, MD, MPH
Jason P. Block, MD

Status/Stage of Development

Completed

Funding Sources

The research was funded through the Robert Wood Johnson Foundation, through their Healthy Eating Research program.

Practice Setting

Academic Medical Center

National/Policy Context

  • In 2014, the Institute of Medicine (now the National Academies of Sciences, Engineering, and Medicine) published a report that established standards for tracking health determinants in electronic health records (EHRs). Such social determinants of health include factors such as physical activity and alcohol use.
  • Pediatric practice guidelines recommend avoiding sugar-sweetened beverages and limiting 100% fruit juice intake. 
  • Unlike other clinical information such as allergies, medications, and weight, sugar-sweetened beverage consumption does not have a searchable, trackable field in EHRs, with many providers entering free text that is not standardized. 
    • This makes screening, conducting population-level research, and intervening more challenging.

Local/Organizational Context

  • The pediatric patient population served by WFBH has a high prevalence of overweight and obesity. As of 2017, 13.1% of 10-17 year olds in NC had obesity, and 15.4% of high school students. Earlier data on pre-school-aged WIC participants showed that 15% of 2-4 year olds had obesity.  
  • WFBH is transitioning to a Learning Health System, and investing in population health management strategies to better serve its patient population and surrounding communities. Because SSB consumption is a proven risk factor for childhood obesity, the health system supported the use of the EMR to conduct more systematic screening efforts around sugary drink consumption.

Patient Population Served and Payor Information

  • Wake Forest Baptist Health is situated in Forsyth County, which as of 2013 was 12.4% Hispanic or Latino, 2% identifying with two or more races, 2.1% Asian, 27.1% Black/African American, and 58% White and not Hispanic or Latino.
  • In this study, screened individuals ranged from 6 months to 17 years, with 51% male and 49% female children. 40% of children were non-Hispanic White, 23% were African American, 27.6% were Hispanic, and 9.2% identified as Other.

Leadership

  • Aside from the research team members, local institutional champions for this project included:
    • Adam Moses and Traci Kirkner, Senior Applications Analysts
    • Bethany Howell, Patient Education Systems Coordinator
    • Richard Lord, MD, Vice President of Clinical Operations – Population Health
    • Clinic Managers at all participating practices 
  • Ongoing project leadership is provided by the Principal Investigator, Dr. Kristina H. Lewis

Project Research + Planning

  • The authors created a single-item screening question based on wording used in the 2013-2014 NHANES (National Health and Nutrition Examination Survey) dietary screener (see “Tools or Products” section below). 
  • Researchers trained medical assistants and nursing staff to read the screening question.
  • The intervention utilized existing medical assistants and nursing staff in the practices to perform screening during patient rooming and thus did not require added personnel or major workflow disruption.

Tools or Products Developed

 

  • Sugar-Sweetened Beverage Screening Tool: This screening tool was developed based on the 2013-2014 NHANES (National Health and Nutrition Examination Survey) questions on sugar-sweetened beverages and was built by the informatics team at Wake Forest Baptist Medical Center. 

 

      • The screening question was built into the EHR as a “best practice alert” (BPA), and set to fire automatically after vital signs were entered for an eligible child.
      • The BPA was enabled for all in-person pediatric primary care visits for children 6 months through 17 years of age.
      • The BPA is repeated over time for each child. If the EHR detects that >90 days have elapsed since screening data was last entered for a child, it will fire again.

 

  • Ordinal categorical response choices for frequency of consumption are displayed within the BPA as clickable checkboxes, to minimize data-entry error.

 

The answer choices represented frequencies from “never” to “4 or more times per day.” Additional detail on the tool and text used can be found here.

Training

  • Multiple training sessions led by research staff were held before the project roll-out for medical assistants and nurses involved in the rooming process. 
    • During these 15-minute in-person trainings, research staff discussed the health impact of sugar-sweetened beverage consumption, trained staff on how to perform the screen properly, and trained staff on the proposed workflow. 
  • After the BPA was activated in participating clinics, research staff checked in once a week with clinical staff members for the first month of the roll out, and once per month after the first month of the project had passed. Clinical practice managers were given reports of their clinic’s performance, relative to a benchmark goal of screening 70% of eligible children.

Tech Involved
  • Electronic medical record
  • Telephone
Team Members Involved
  • Administrator
  • MAs
  • Researcher
  • RNs
  • Software Engineers/ IT
Workflow Steps

  • The child arrives to his or her clinic visit and is roomed by a medical assistant or a nurse.
  • During the rooming process, the medical assistant or nurse logs into the EHR and enters vital signs. After vitals are entered, a “Best Practices Alert” box pops up onto the screen and reads: “In the past month, how often did (child’s name) drink a sugar-sweetened beverage or 100% fruit juice? Sugar-sweetened beverages include things like fruit-flavored drinks, juice from concentrate, punch, Kool Aid, soda, sports drinks, sweet tea or flavored milks.” 
  • The medical assistant or nurse reads this prompt to the child directly if he/she is 13-17 years old or to the parent/caregiver and child if he/she is 12 years or younger. If the family or child is Spanish-speaking, they are provided a written version of this prompt in Spanish.
  • The medical assistant or nurse clicks a box indicating frequency of the child’s sugar-sweetened beverage consumption, ranging from “never” to “4 or more times per day”. 
  • If the response given by the patient or their caregiver exceeds sugary beverage recommendations for that child’s age group (i.e. any sugar-sweetened beverage or juice intake for a child younger than 12 months), an automatic message generates in the after visit summary (AVS) that patients and caregivers receive.
    • The message is tailored to the child’s age group and describes the recommendations for limiting consumption of sugar-sweetened beverages, provides clinical reasoning for the recommendations, and provides suggestions for alternatives to sugar-sweetened beverages such as water. 
    • The AVS paragraph is printed in English or Spanish depending on the language preference documented in the patient’s chart.
    • Screening is repeated 90 days or later after the initial screening at any subsequent in-person clinic visit to gather longitudinal data.
  • During the initial year of implementation, front-line clinical staff were rewarded with a $10 gift card quarterly if their clinic screening rates meet or exceed 70% of screening-eligible visits.

Outcomes

 

  • Screening compliance rates: The rates of screening in participating clinics throughout the 8 months of the intervention exceeded the goal of 70%, with a monthly low of 80% and monthly high of 92%. 
  • Total patients screened: 91% of patients (24,873 individuals )in the clinics were screened at least once, with 21% having 2 or more screens. 
  • Diversity of screened patients: 35% of screened patients self-identified as non-Hispanic White, 30% as Hispanic, 26% as African-American, and 9% as Other race/ethnicity. 
  • Willingness of patients/families to be screened: 0.3% of patients/families refused screening, revealing a high willingness to participate in screening measures. 

 

      • In a smaller validation study of 200 patients, 92% were “somewhat or extremely comfortable” with being screened at the clinic visit. 

 

  • Standardization of data: Only 3% of screened patients had a free-text entry in their charts documenting the screen (rather than the desired standardized response generated from clicking the boxes on the EHR alert).
  • Data on Sugary Beverage Consumption: Through the screen, the authors determined that 41% of patients consumed at least one sugar-sweetened beverage or fruit juice drink per day in the month prior to the screen. 
  • Demographic disparities in screened patients: In both infant and non-infant groups, rates of sugar-sweetened beverage and fruit juice consumption were higher among African American patients compared to non-African American patients (28% vs 15% of infants, 34% vs 20% of children older than 1 year). 
  • Impact of screen on patient behaviors: In a smaller validation study of 200 patients, 40% of patients/caregivers reported that the screen had motivated them to change their behaviors or their child’s behaviors around sugar-sweetened beverage consumption. 

 

    • In the validation study, of the 56% of patients who remembered receiving information about sugar-sweetened beverages in the after visit summary, 64% revealed that the information provided had motivated them to change their or their child’s consumption. 

Future Outcomes

  • In order to improve the accuracy of the screening and obtain results comparable to more sophisticated screening methods, such as those used in NHANES, the authors have recently implemented a 2-question screening tool that separately assesses SSB and 100% fruit juice. The wording for those questions is as follows: 
    •  “On a usual day in the last month, how often did _______ (child’s name) drink 100% pure fruit juice like orange, apple or grape juice? Don’t include fruit-flavored drinks like Kool-Aid or lemonade.”
    • “On a usual day in the last month, how often did _____(child’s name) drink regular sodas, fruit-flavored drinks like Kool-Aid or lemonade, sports drinks like Gatorade, sweet tea or any other sugary drinks?”
    • Response options for both questions are: Never, Sometimes but not every day, 1x per day, 2 x per day, 3 or more x per day, and Refused
    • A repeat validation study of this new 2-item measure (not published) showed improved correlation with NHANES-item-derived SSB and FJ consumption: Spearmann Correlation Coefficient for SSB consumption (comparing new screening question to NHANES questions: 0.77 p<0.0001), and for FJ consumption (comparing new screening question to NHANES questions 0.59 p<0.0001), and for total SSB+FJ consumption 0.82 (p<0.0001).
  • The authors are working to develop a series of automated, technology-enabled interventions to pair with the screener to address overconsumption of SSB and fruit juice in a way that does not further burden clinical staff and would be scalable and generalizable to other health systems, still leveraging the EHR to both enable screening and facilitate delivery of intervention components. 

Benefits

    • The clinics achieved a very high rate of screening, demonstrating proof-of-concept that a sugar-sweetened beverage screen is a feasible tool used in primary care clinics. 
      • This will allow for population-scale, data-driven interventions in the future based on longitudinal patterns. 

 

  • The screened population was highly diverse, suggesting that the large-scale data collection methods maintained the diversity of respondents and that the data collected may be reflective of experiences of the greater population served.

 

  • Clinic staff were able to execute the screening protocol properly the vast majority of the time, producing data that is easily searchable and extractable in the EHR due to its standardized format. 
  • The workflow did not disrupt clinical practice, did not require added staff, and easily integrated data collection into the patient rooming process, which likely contributed to the high levels of screening achieved.

Intervention-Specific Challenges

  • Screening rates fell during times of the year in which other public-health interventions occurred, such as flu shot administration. 
    • Therefore, it may be challenging to maintain high levels of sugar-sweetened beverage consumption screening amid having clinicians respond to other high-priority demands, as clinic staff have time limitations.
  • Over time, as more alerts are built into EHRs to encourage collection of other data points, providers may become prone to alert-fatigue and may dismiss the alert boxes that arise. 
  • The results of the single-question screen did not correlate strongly to the results obtained in the validation mini-study that utilized NHANES sugar-sweetened beverage questions. 
    • Although the authors reached an impressive number of patients, the data that is being collected may not be accurate when compared to the data obtained using methods from a large national study.
    • The authors cite a prior paper that showed that when individuals are asked separately about multiple categories of beverages, they often report higher intake, as patients may underestimate consumption when all forms of beverage are combined in a single question. 
  • In the validation study, many respondents indicated they had not read the after visit notice on sugar-sweetened beverages that was generated, demonstrating that another patient education method may be superior to attempt to change behaviors. 
  • There is currently no process validation tool in place to make sure that clinic staff are reading the alert box verbatim, which could lead to variability in the ways that patients are screened.

Glossary

  • NHANES (National Health and Nutrition Examination Survey): an annual survey assessing diet, health behaviors, and patient demographics through interviews, physical exams, and lab tests. 

Sources

  • Heyman MB, Abrams SA; SECTION ON GASTROENTEROLOGY, HEPATOLOGY, AND NUTRITION; COMMITTEE ON NUTRITION. Fruit Juice in Infants, Children, and Adolescents: Current Recommendations. Pediatrics. 2017 Jun;139(6). pii: e20170967. doi:10.1542/peds.2017-0967. PubMed PMID: 28562300.
  • Lewis KH, Skelton JA, Hsu FC, Ezouah P, Taveras EM, Block JP. Implementing a novel electronic health record approach to track child sugar-sweetened beverage consumption. Prev Med Rep. 2018 Jun 19;11:169-175. doi: 10.1016/j.pmedr.2018.06.007. eCollection 2018 Sep. PubMed PMID: 29988772; PubMed Central PMCID: PMC6031146.
  • Institute of Medicine. (2014). Capturing Social and Behavioral Domains and Measures in Electronic Health Records: Phase 2. Washington, DC: The National Academies Press. Retrieved from http://www.nationalacademies.org/hmd/Reports/2014/EHRdomains2.aspx.

Innovators
  • Jason Perry Block, MD, MPH
Talk to the Innovators