Abstract
Introduction/Objectives:
In 2022, 1 in 6 households with children experienced food insecurity (FI) in the United States. The negative impact of FI on child health is well documented and pediatric clinicians are encouraged to actively screen and intervene in clinical settings. This study aims to evaluate the feasibility and impact of a FI management protocol implemented in 2017 at a pediatric primary care health network serving patients who are Medicaid-eligible in Washington, DC.
Methods:
In 2019, an 18-item electronic survey was sent to a convenience sample of 42 pediatric clinicians within the health network to understand their knowledge, attitudes, and behaviors surrounding implementation. Both quantitative and qualitative responses were collected and analyzed. We report frequencies of the Likert-type responses, including perceived compliance with protocol components and intervention efficacy. We evaluated the relationship between FI knowledge level and rates of clinician documentation compliance by chi square and Cramer’s V statistic for effect size. Open-ended responses were reviewed, and common themes were identified and used to provide context for quantitative results.
Results and Conclusions:
Out of 42 clinicians invited to complete the survey, 35 completed responses. All respondents reported universal screening for FI (100%) at routine examinations, 80% reported frequently electronically documenting FI in medical records, and 91% of clinicians reported frequently referring families who screened positive for FI to at least one FI resource, with 24% reporting that resources met families’ needs. Open-ended responses revealed increased awareness of FI prevalence and of patient experiences in households experiencing FI, increased satisfaction with clinical management of FI, but also concerns around having limited clinical time to do the protocol and the usefulness and accessibility of referred resources. In conclusion, implementing this pilot FI protocol was feasible, but clinicians perceived limited impact of the protocol on alleviating FI and desired more robust intervention options. Further improvements include shifting the burden of performing the protocol away from the clinician, such as by streamlining the protocol or identifying a resource staff member, and establishing more accessible and effective FI interventions such as “Food as Medicine” offerings in partnership with community organizations.
Background
At the household level, food insecurity (FI) occurs when the household struggles to afford an adequate quantity of food for every person in the household to live an active, healthy life. 1 In 2022, 17.3% of United States (US) households with children experienced FI, the highest rate in more than half a decade. 1 Sub-populations at increased risk of FI include households with incomes below 185% of the federal poverty level; children and single parent households; families in urban or rural settings, as well as African American and Hispanic headed households. These racial and ethnic disparities can be associated with the structural racism and systemic injustices that have been historically perpetuated through inequitable policies and practices. 2
Far reaching and substantial consequences for child health make FI a prominent concern for primary care pediatricians. In particular, FI is associated with many negative health and behavioral effects on children, including poor nutrition and disrupted growth, obesity, increased physical illness, decreased school performance, behavioral and emotional problems, dysthymia, and suicidal ideation.3,4 Because FI is highly correlated with diet-related chronic disease, such as obesity and type 2 diabetes, FI protocols have been seen as a way to improve household food security, leading to better health outcomes by reducing the burden of diet-related chronic diseases in patients. 3 Since the 2015 American Academy of Pediatrics (AAP) policy statement on FI, many pediatric settings have begun such “screen and intervene” FI protocols.4-5 Best practices and recommendations later expanded to highlight the importance of universal screening for FI, documentation of FI and interventions in the electronic medical record (EMR), and intervening in ways such as providing emergency food assistance; connecting families to federal nutrition programs, referrals to social workers or other ancillary staff; and enrollment into clinical-community collaboratives such as produce prescription initiatives and other “food as medicine” type interventions.3-4,6 Studies evaluating clinical management of FI have reported challenges around protocol implementation, including limited use of universal screening 5 and validated screening tools, 7 lacking appropriate resources, 5 and lacking time during appointments to manage patient FI. 5
Screen, document, and intervene protocols are a growing and promising method to address FI in the pediatric clinical setting. However, it is not yet clear how such protocols can be best implemented to meaningfully alleviate patient FI. This paper seeks to assess the feasibility and impact of a FI management pilot protocol at a pediatric, primary care network, and through evaluating pediatric clinicians’ practices and attitudes, we aim to better understand the challenges around implementation and resource referral efficacy.
Methods
In response to the 2015 AAP statement 4 and a prior assessment of FI management in the broader metropolitan region (DC, Maryland, and Virginia) which showed that 66% of pediatric clinicians reported screening for FI and only 13% used a standardized screening tool, 8 a large urban pediatric primary care health network in Washington, DC implemented a FI management protocol in 2017. This health network is composed of 6 clinics and collectively serves >40 000 individual patients yearly. Approximately 80% of these patients identify as African American, 90% primarily speak English, 5% identify as Hispanic, and >80% are enrolled in Medicaid. In January 2017, clinicians were educated on how to screen for FI, document in EMR, and intervene using referrals to local and federal resources through 3 optional training sessions: (1) a live, center-wide, didactic training seminar on FI; (2) a self-directed, written guide that gave an overview of the protocol; and (3) a pre-recorded intervention toolkit discussing federal and local resources to address FI. They began formally screening and intervening to address FI in March 2017. Under the new protocol, the Hunger Vital Sign (HVS) 9 was given to all families in the waiting room by the front desk staff prior to routine Well Child Checks (WCCs) for ages 0, 6, and 12 months and yearly until age 21. According to the protocol, the completed screener was given to the clinician in the examination room to record in the patient’s EMR. Following a positive screen for FI, clinicians provided resource lists to families and/or connected families to ancillary staff or clinical partners. Clinicians were instructed to record an ICD-10 code for FI, add FI to a patient’s problem list (SNOMED-CT), and document resource referrals in the patient’s EMR.
Two years after the protocol was implemented, an electronic survey using the Research Electronic Data Capture (REDCap) tool was sent out in April 2019 to all 42 primary care pediatric clinicians in the network to examine clinical knowledge, attitudes, and behaviors regarding the FI protocol implemented in 2017. Inclusion criteria was current employment as a pediatric clinician in the health network in April 2019 and consent to participate. The author’s institution provided regulatory oversight and the work was IRB-approved.
The 18-question survey focused on knowledge, attitudes, and behaviors with surrounding FI screening, documentation, resource referral/intervention. It was tested with outside faculty, psychologists, and trainees to establish face validity before being used with the pediatric clinician cohort. Quantitative measures included self-reported perceived rates of screening, documenting, intervening, and completing training before protocol implementation. In addition, the survey assessed clinician perception of patient-reported resource use and perceived resolution of FI in patients. Quantitative response categories included 5-point Likert-type scales (“never/did not ask,” “rarely,” “sometimes,” “often,” and “always”), “select all that apply,” and frequency categories. To facilitate interpretation of the data, the 5-point Likert-type scale questions were simplified into 2 category responses: yes (“always”/“often”), and no (“sometimes”/“rarely”/“never/did not ask”). Qualitative responses consisted of open-ended responses: clinicians were asked to provide explanations for why they did not document FI in the problem list upon a positive FI screening, how performing the protocol affected their understanding of FI prevalence in their patients, and what changes they would make to improve the clinical management of FI based on their experience with the training and protocol. The purpose of the quantitative analysis was to measure feasibility via self-reported rates of compliance with protocol components, and to measure impact through clinician-perceived rate of alleviation of FI in patients. The purpose of the qualitative analysis was to assess feasibility through perceived barriers and facilitators to protocol implementation and assess impact through perceptions of efficacy through clinician resource referrals.
We report frequencies of the 2 category responses which summarized the Likert-type scale questions, and category percentages were reported to the nearest integer. Cross tabulations were used to evaluate relationships between FI knowledge level and Likert-type responses to survey questions. Exploratory associations between FI knowledge level and the 2-category simplification of Likert-type responses to documentation practices were evaluated via chi square testing for statistical significance (P < .05) and Cramer’s V statistic for effect size. Open-ended responses were reviewed, and common themes were identified and used to provide context for survey results. Results were collected and managed using the REDCap tool hosted at the authors’ institution.
Results
The survey was completed by 35 out of the 42 (83% response rate) primary care clinicians. Forty-six percent of clinicians reported accessing or participating in at least one of the FI protocol training opportunities offered prior to implementing the protocol.
All clinicians (100%) reported screening at WCCs and 43% identified triggering factors for screening outside of WCCs. These triggers were related to concerns about the child’s weight or growth and the family’s financial/psychosocial stressors (eg, housing concerns such as homelessness).
Around 34% of clinicians reported they “always/often” documented in all 3 ways: (1) using an ICD-10 code, (2) adding FI to a patient’s problem list, and (3) documenting resource referrals. A large majority (80%) of clinicians reported “always/often” documenting in at least 1 of 3 methods.
When asked about their frequency of resource referral, 91% of clinicians reported “always or often” referring families who screened positive for FI to at least 1 of 3 resource types (emergency resources, federal nutrition programs, or referrals to ancillary staff). The least commonly used referral type included referrals to ancillary staff (such as social workers, dieticians, or medical legal partnership). The most commonly referred programs were food pantries/soup kitchens (54%), WIC (51%), SNAP (49%), and the Regional Food Bank (40%). In addition, clinicians reported that these commonly referred programs were thought to be used most frequently by patients, albeit in a different order: SNAP (66%), WIC (63%), food pantries (26%), and the Regional Food Bank (17%).
When asked about experiences with following up on FI concerns, 24% of clinicians reported that families “always or often” said that their needs were met by the previously referred resources and only 14% reported that patients who had previously screened positive for FI were “always or often” no longer experiencing FI upon a subsequent screening. Table 1 provides a summary of the quantitative data.
Summary of Pediatric Clinicians’ Self-Reported Quantitative Responses on Knowledge, Attitudes, and Behaviors Regarding the Food Insecurity (FI) Management Protocol.
One incomplete response, such that n = 34.
Regarding qualitative responses, 29 clinicians described how the training and protocol affected their understanding of FI. The most common responses described gaining an increased awareness of the prevalence of FI, increased empathy and knowledge of families’ situations with FI and recognition of the limitations around available resources, and increased ability to provide better or more support to families in need of FI resources. Multiple clinicians found that performing screening for every patient at WCCs (a form of universal screening) as part of the protocol was “eye-opening” to them because of how frequently patients screened positive for FI. They also reported a better understanding of a family’s daily struggles with FI, such as the fact that for some families, their “WIC, TANF, SNAP [were] not enough.”
Twenty-six clinicians provided feedback to improve FI screening, documenting, and/or intervening. Two themes were developed from clinician responses. First, clinicians were often concerned that the available resources did not seem adequate or easily accessible to patients due to communication/referral gaps and unknown barriers. For example, a clinician had concerns about families’ ability to access referred resources, saying, “I still feel like resources available are hard to recommend in an easy-to-use way for families.” Responses often alluded to the feeling of being unable to effectively help patients who screened positive for FI, and although the resource list or existence of federal resources was seen as helpful, clinicians did not see it as always sufficient. Additionally, while clinicians preferred providing warm handoffs to ancillary staff (such as a social worker) over giving referrals to programs, such handoffs seemed to have limited availability. As a result, clinicians felt challenged to address FI with their available resources, saying, “There’s little to do. . . which feels frustrating.” Another clinician explained that although they now better understood the challenge of addressing FI after performing the FI protocol, it was “still hard to solve/offer ‘good’ solutions [to families].”
In response to this perceived lack of effective/accessible resources, clinicians requested a general improvement and expansion of provided resources, asking for “more follow up. . . more outreach of other available options, more continuity of services offered.” They also suggested establishing partnerships with local community organizations, for example asking for “better specific connections with emergency food resource providers” and “partner[ship]s with more community groups.” One clinician wanted to educate families outside of the visit, requesting “common information resources for families.” Another wanted a way to follow up with the family after giving them resources “to see how things are going.”
Secondly, clinicians expressed concerns about the time burden of the protocol. Some felt it would be more appropriate to have alternative staff handle more or all of the FI protocol-related tasks, especially those related to resource referrals, stating “it’s not realistic for the medical provider.” One clinician explained how it felt like FI “resources are constantly changing” and that it was easier “to just have SW (social worker) talk to the families about the availability, eligibility, and enrollment in resources and programs.”
Regarding specific screening, documentation, and intervention practices, clinicians were concerned that the screening process might stigmatize or miss families, and several asked for improvements to inefficiencies in the EMR and to clarify EMR documentation practices. Regarding practices related to resource referrals, clinicians wanted a more streamlined referral process, including “automatic referral to social work” and “integration of resource referral into EMR.” Table 2 summarizes the qualitative responses.
Summary of Pediatric Clinicians’ Qualitative Responses on Knowledge, Attitudes, and Behaviors Regarding the Food Insecurity (FI) Management Protocol.
When we explored the association between FI knowledge level and documentation practices, we found that 67% percent of clinicians who reported being “moderately/very” knowledgeable reported adding FI to the problem list, whereas only 10% of clinicians who reported being “somewhat/not” knowledgeable reported adding FI to the problem list (p = .051). The association had a large effect size (V = 0.516). There was a similar relationship between knowledge level and the documentation practice of using the ICD-10 code. “Moderately/very” knowledgeable clinicians reported a higher rate of ICD-10 documentation versus “somewhat/not” knowledgeable clinicians (70.8% compared to 27.3%, respectively p = .01). This association had a medium effect size (V = 0.409). Table 3 summarizes this analysis.
Rate of Pediatric Clinician Compliance with Documentation by Self-Reported Food Insecurity (FI) Knowledge Level.
One response was “don’t know,” so it was excluded from analysis, such that n = 34.
Significant at the .01 level.
Indicating a medium effect size.
Indicating a large effect size.
Fourteen clinicians provided qualitative responses for why they did not add FI to a patient’s problem. The most common reasons were forgetting to do so, not knowing how, and/or because doing so was not “highly relevant to the visit.” In addition, a few clinicians expressed concerns about patient sensitivity to having “this [FI] diagnosis showing up on their problem list.” Others felt that the problem list was not updated frequently enough for clinical management of FI. One clinician expressed doubt that a patient “receiving WIC resources need[ed] this [FI diagnosis] followed as a problem.”
Discussion and Conclusion
After implementation of a FI management protocol at a large urban pediatric primary care health network, clinicians self-reported higher rates of FI screening at regular visits (100%) in comparison to regional averages before implementation (13%), 8 along with favorable documentation rates (80%) and consistent referrals during appointments to various resources (91%). Furthermore, 89% percent of clinicians reported an increased understanding of the prevalence of FI in their patient population after protocol implementation. Qualitative responses gave unique insights into each component of the protocol and revealed improvements in clinician understanding of FI prevalence and management. Both quantitative and qualitative responses highlighted the need to address the limited time clinicians experienced while trying to complete the protocol during appointments, and how patients who screened positive for FI were perceived to not often receive accessible or useful resources.
The screening component was very feasible. FI screening appeared to greatly increase upon comparing pre-protocol implementation rates in the DC region (66% of clinicians reported infrequently screening, 13% reported using a standardized screener 8 ) to post-protocol implementation rates reported in the specific regional cohort (100% reported screening at WCCs with a validated screening tool established in all clinical sites). This improvement in screening rates suggests that a well-structured FI protocol can improve standardized screening beyond national averages, which was around 33% for pediatricians according to one 2021 survey. 5 The success of the protocol in increasing screening rates appears to reflect the general consensus from pediatric clinicians that screening for FI is important and acceptable to perform in the clinic.5,10,11
Notably, the protocol required a short, validated screener (HVS) to be administered to all patients without discrimination, a process known as universal screening. This process of universal screening helped clinicians to develop an increased awareness of the extent of FI in their patient population, which they had previously underestimated. While the HVS screening tool was unable to explore the degree of FI experienced by the family, such evaluation would require a more extensive FI screening tool that may add to survey fatigue and be of limited benefit in clinical settings. Furthermore, although clinicians were required to screen universally during WCC visits, they were also given autonomy to screen and probe spontaneously, including during non-WCCs, whenever they felt it was appropriate. Such autonomy appeared to complement the universal screening process, allowing clinicians to use key subjective findings to respond with resources to support families. Overall, these results suggest that universal screening with a short, validated screener increases clinical awareness, improves identification of FI in clinic, and may fit within a clinical workflow of primary care practices. 3
While the screening component of the protocol was very feasible, qualitative responses provided insights to improve the screening process. One clinician highlighted that the screening process might stigmatize patients. This concern around screening is consistent with existing literature on FI screening in clinical settings, although in previous analyses, patients have responded positively when screened by clinicians, suggesting that proper screening demonstrates clinic/provider care and empathy, especially when the purpose of screening is made clear.10,12 The concern also re-emphasizes the importance of mitigating stigma through practices such as universal screening. 13 Providing additional clinician training on the potential for bias when assessing and addressing social risk may also help to raise clinician awareness and attenuate harm done to patients. 14 Lastly, concerns around patient confidentiality were brought up in the qualitative data, which is a legitimate consideration given the potential risks associated with storing and sharing social needs data through EHR. Health systems should follow best practices to develop responsible social needs data management strategies that minimize potential harm. 14 While administering a short FI screener appeared very feasible for our study, FI screening protocols in similar clinical settings may benefit from developing standards of screening for social risk which account for the potential for stigma to negatively influence patient care and limitations of EHR in handling patient confidentiality.
Documenting for FI was moderately feasible, as 80% of clinicians frequently documented FI in at least one way, and 34% of clinicians reported frequently documenting FI in all 3 ways. Reasons for the somewhat lower rates of documentation in all 3 ways may be gleaned from our qualitative data. Multiple clinicians expressed that the EMR documentation process was inefficient. Furthermore, knowledge level of FI may have also been a factor, as clinicians who reported being “very/moderately” knowledgeable about FI reported documenting FI as a standardized code more frequently than those who reported being “somewhat” knowledgeable about FI. It is possible that clinicians with less knowledge of FI may not have documented FI because they perceived it as a low value priority. In comparison, clinicians with greater knowledge of FI may have found increased value in documenting because they had an increased awareness of the threat of FI to their patients’ and families’ health. Next, one clinician reported that infrequent updates to the problem list was one reason they did not use the problem list (SNOMED-CT) to document FI. As FI status of a household often waxes and wanes and may not be constant throughout the year, 1 the problem list approach to documenting FI may not be capable of accurately indicating whether a family is currently experiencing the negative effects of FI. Clinicians also expressed that there was some patient sensitivity regarding documentation of FI. This supports the ongoing call in the broader literature to set up documentation procedures which ethically manage patient social needs data. 14 Clinics and organizations should consider creating documentation and data use policies that securely store sensitive data, respect a patient’s right to withhold such data, and educate clinicians about how viewing such data may exacerbate their biases. 11
Overall, documentation practices are not as well explored as screening in the broader literature.11,12 Our results suggest that careful planning of efficient electronic documentation procedures may reduce issues around patient privacy and clinicians facing limited time to perform documentation, enabling the success of FI protocols in the pediatric, clinical setting.
Providing resource referrals was feasible for clinicians. Ninety-one percent of clinicians reported referring patients with a positive FI screen to at least 1 of the 3 resource types, suggesting that clinicians were able to consistently provide resource referrals. These results add to the current literature that pediatric FI protocols successfully enable pediatric clinicians to intervene for FI to some capacity. The most common resources offered were emergency food interventions (i.e., food banks and pantries) and federal nutrition programs (i.e., SNAP, WIC, and school meals), both of which are evidence-based approaches to combat FI.3,4 Notably, we found that clinicians were less likely to refer patients to “ancillary staff,” such as social workers, dieticians, or the medical legal partnerships. Qualitative responses referenced the limited availability of these resources, especially concerning social workers. This suggests that providing consistent referrals to ancillary staff (such as a social worker) would likely require hiring and/or outsourcing to additional staff beyond the typical workforce at a pediatric clinic, in agreement with previous findings.3,12
Although providing resource referrals appeared feasible, the referrals had limited perceived impact on alleviating patient FI. Few families were described to use or access the offered resources, and only 24% of clinicians perceived that the referred resources frequently met a family’s needs upon follow up visits. In addition, qualitative responses described a low rate of perceived patient use of and access to resources. These findings are consistent with other protocols that found the limited effectiveness of resources for families may be a common problem faced by pediatric and general social needs protocols.5,12 Although only 14% of clinicians reported that patients were no longer food insecure upon a subsequent FI screening, the protocol was not intended to entirely resolve a family’s FI.
While root causes of FI are often difficult to address through clinical protocols, there are resources that have been traditionally used to address FI. Federal nutrition programs have been shown to decrease FI, poverty, and improve overall health outcomes for those who successfully enroll. 15 Emergency food interventions are useful as a supplement to federal nutrition programs or for families unable to meet the qualifications of or disinterested in federal nutrition programs, and can provide meaningful short-term assistance. However, they alone are often not enough to adequately meet under-resourced families’ food needs.16,17 “Food as Medicine” interventions, including medically tailored meals/groceries and produce prescriptions, are frequently integrated into clinical systems and are an emerging and promising method to address not only FI but also nutrition security.18,19,20
In qualitative responses, clinicians highlighted difficulties connecting families to useful and accessible resources and suggested methods beyond providing resource referrals. Most interventions under this pilot FI protocol consisted of non-facilitated referrals to emergency resources and federal programs. Although these programs themselves can be useful, such referrals often fail to address the difficulty of accessing the programs due to complex application processes, income requirements, transportation barriers, and limited hours in which community programs may operate. 19 To illustrate, one clinician perceived that when families were ineligible for federal nutrition programs due to their income level, merely providing an emergency resource referral was often insufficient. To overcome these stated barriers, future protocols may benefit from establishing a food pantry within the clinic, or creating partnerships with community organizations that can facilitate direct access to tangible food resources. Examples include federal nutrition and cash assistance programs and food/produce deliveries and vouchers. 6 Overall, clinician responses touched upon the limitations of solely providing paper resource referrals, and suggested new methods to more directly provide meaningful resources to patients.
In addition to concerns around the effectiveness of referred resources, many clinicians also expressed that the overall FI protocol was too time consuming. Clinicians elsewhere have similarly reported time burdens as a significant obstacle to addressing social needs like FI. 12 Regarding documentation specifically, clinicians requested an increase in the automation of documentation processes, and for partnerships with community organizations to assist with social needs documentation. Pursuing strategies that streamline EMR workflow and incorporate input and buy-in from clinical practice leaders may make FI screen and intervene protocols less burdensome. 14 Furthermore, we found that clinicians desired and used extra assistance to do the protocol, such as having additional staff. This finding is common among existing literature reviewing the impact of social needs protocols. 12 For example, the AAP recommends that a dedicated resource staff could handle resource referral after a visit, minimizing clinician time spent on the protocol during the visit. 3 A survey of US children’s hospitals reported that social workers and nurses most frequently screened patients for FI in the inpatient settings, providing additional evidence that FI screening does not require clinician time. 5 Our data strongly supports addressing clinician time burden in FI protocols and adds to the growing body of evidence that the limited time availability of clinicians in clinic must be carefully accounted for when establishing FI protocols in the pediatric clinical setting.
This study had limitations. The sample size was small, and all data was self-reported by the clinician. Furthermore, the convenience sample was composed of all clinicians within the pediatric network who chose to respond, and demographics were not collected for this sample. Thus, these results are not necessarily widely generalizable to all other pediatric clinical settings. However, while not explicitly recorded, the great majority of the workforce that was asked to participate in the study were pediatricians and nurse practitioners, suggesting that our study findings are applicable in pediatric clinical settings with a similar healthcare provider workforce and patient populations. Although we did not control for potentially significant secular trends that may have occurred during the 2-year period between protocol implementation and evaluation, we are not aware of any major events during the period that would have influenced clinician ability to manage FI. Although the survey was not validated, it was tested on a similar population for face validity before use in this study. Lastly, clinicians may have entered the clinical workforce after the 2017 training was offered, resulting in an underreporting of 2017 training use. Despite these limitations, our study had a strong response rate, and the protocol was successfully implemented and evaluated as a new social needs protocol in the pediatric clinical setting. Our results highlight areas where healthcare settings can make improvements to FI management protocols, have led to ongoing community-clinical partnerships in the District of Columbia and its surrounding metropolitan areas. 21
In conclusion, implementing this pilot FI protocol was feasible, but clinicians perceived that the protocol had limited impact on alleviating FI. We provide insights on good protocol practices, such as screening with a short, validated screener, and on the limitations of the protocol. Qualitative and quantitative data illuminated two issues that negatively affected the feasibility and impact of the protocol: a lack of clinician time to perform the protocol and a lack of accessible or useful interventions to provide to patients. Although these are common issues that arise in implementing social needs protocols, our analysis of qualitative responses provide insight on methods to address these issues. In particular, our analysis suggests that most aspects of the protocol should be integrated with EHR, or diverted to another clinical staff member or external partner when possible, so as to not overburden the primary care clinician. Furthermore, to address the lack of usable or accessible FI resources for patients, clinics should focus on establishing effective interventions in partnership with community organizations as part of FI protocols, such as “Food as Medicine” programs that directly provide tailored food assistance and nutrition education.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethical Approval
Institutional Review Board approval was obtained from Children’s National Hospital (protocol number Pro00007307 on 6/9/2016). All research adhered to the tenets of the Declaration of Helsinki and was HIPPA compliant. Clinicians signed consent forms to be a part of the study prior to completing the survey.
