Rapid Diet Assessment Screening Tools for Cardiovascular Disease Risk Reduction Across Healthcare Settings: A Scientific Statement From the American Heart Association

Poor dietary quality has surpassed all other mortality risk factors, accounting for 11 million deaths and ≈50% of cardiovascular disease (CVD) deaths globally.¹ Paralleling this trend, US healthcare expenditures have grown steadily and now represent 17.9% of the gross domestic product in 2016,² nearly half of which was accounted for by obesity-related conditions ($1.42 trillion/y).³ Dietary interventions focused on single nutrients (eg, vitamin D, omega-3 fatty acids) have not adequately reduced incident CVD.^4–7 Conversely, dietary patterns low in fruits, vegetables, and whole grains and high in red and processed meat, added sugars, sodium, and total energy are the leading determinants of CVD, Alzheimer disease, some cancers, and diabetes mellitus risks.^1,2 As a result, strategies that promote holistically healthier dietary patterns to reduce chronic disease risk are of contemporary importance.⁸

Despite the importance of diet quality, most clinicians and other members of the healthcare team do not currently assess or counsel patients about their food and beverage intake during routine clinical care^9,10 because of clinical constraints. Clinical practice barriers include lack of training and knowledge, lack of time, insufficient integration of nutrition services into many healthcare settings, lack of reimbursement, and competing demands during the visit.^10–13 Inadequate nutrition training among clinicians and other members of the healthcare team represents a structural contributor to diet-related disease.¹³ However, these barriers can be overcome. Recent evidence suggests that even brief training can increase physician self-efficacy in dietary counseling.¹⁴ In addition, a brief diet screener tool, particularly one that provides clinical decision support (CDS) and can be tracked through electronic health records (EHRs), is likely to reduce some of these barriers by providing an efficient and valid assessment of diet quality and identifying actionable modifications for improvement appropriate for practitioners without adequate training to intervene. Such assessment is a necessary first step to support dietary changes, which are ideally part of a more comprehensive plan.

Once diet is assessed, CDS can help clinicians and other members of the healthcare team engage patients in diet education and evidence-based behavior counseling. For instance, a systematic review conducted by the US Preventive Services Task Force found that in adults not at high risk for CVD, behavioral counseling for diet and physical activity resulted in modest benefits in blood pressure, low-density lipoprotein and total cholesterol levels, and body weight, resulting in a US Preventive Services Task Force C recommendation for a positive but small benefit.^15,16

The usability and ultimately use of rapid diet screener tools in clinical settings will depend on their ability to deliver the right data to the right people at the right time.¹⁷ The term right data implies that CDS tools provide not only accurate information but also, in the case of dietary screener data, actionable steps for improvement. In the context of a short clinical encounter, dietary data must be presented in a clear, easily digestible format to prompt shared decision-making between the right people: a patient and the clinician. Attention should be paid to the workflow of the encounter to ensure that EHR-based diet screeners and related CDS tools occur at the right time during the clinical encounter without interrupting the course of usual care.¹⁸

The objectives of this American Heart Association (AHA) scientific statement are to provide a rationale for adopting a rapid diet assessment tool in clinical settings, to discuss the emerging role of clinicians and other members of the healthcare team in diet screening and counseling, and to provide an evidence-based assessment of tools currently available and feasible for integration into the EHR. The statement specifically focuses on the rationale and feasibility for rapid diet assessment in clinical settings among adults 20 to 75 years of age, desired theory- and practice-based criteria for a rapid diet screener tool and review of existing tools, implementation and challenges of a diet screener tool, and considerations for integrating a rapid diet screener tool into the EHR. The target audience for this statement is clinicians and other members of the healthcare team who focus on preventing and treating chronic diseases in diverse clinical practice settings. The statement is not intended to comprehensively review tools needed for secondary prevention/therapeutic diets or specialized clinical, nonadult populations, nor does it focus on the applications of dietary screening data for population health monitoring or quality assessment improvement across practice settings.

Rationale for Clinician-Delivered Diet Assessment and Counseling

The massive health impacts of poor diet quality in the United States and globally¹ and the potential for large reductions in healthcare costs and enhanced quality of life from population-wide improvements in diet quality¹⁹ provide a strong rationale to increase the delivery of diet assessment, education, and counseling by clinicians and other members of the healthcare team in diverse healthcare settings. Moreover, the consistent evidence that low nutrition knowledge negatively affects diet quality and clinician-delivered diet counseling improves diet behaviors and cardiometabolic risk factors^15,16,20,21 supports adoption of routine integration of diet assessment and counseling into diverse healthcare settings.

For instance, national data have shown that US adults 20 to 65 years of age have suboptimal diet quality, achieving 60% to 65% of ideal scores.²² Although diet quality is positively correlated with socioeconomic indicators, low nutrition knowledge was associated with low diet quality regardless of education and income.⁸ Similarly, in the Italian Moi-sani cohort, higher nutrition knowledge was associated with better adherence to the Mediterranean diet and a lower prevalence of obesity.²³ These data underscore the importance of increasing nutrition knowledge to improve diet quality.²²

In addition, increased delivery of diet assessment and counseling by clinicians and other members of the healthcare team aligns with the transformation of healthcare delivery in the United States to value-based accountable care and population health management. The national prevention strategy of the 2010 Affordable Care Act called for these efforts to extend to entire practice-based populations; to be focused on the behavioral determinants of health, including healthy eating; and to be better integrated with public health sector efforts.²⁴

The evidence that clinician-delivered diet assessment, education, and counseling can improve diet-related behaviors and cardiometabolic risk factors has been summarized in recent reviews.^25–27 Even brief diet counseling led to a significant, albeit modest, improvement in diet quality,²⁵ as well as body weight, total cholesterol, low-density lipoprotein cholesterol and fasting glucose concentrations, systolic and diastolic blood pressures, and incident diabetes mellitus.^26,27 Limited evidence demonstrated that intensive counseling involving registered dietitians increased dietary quality by increasing the intake of fruits, vegetables, and nuts and decreasing the intake of commercial baked goods.²⁸ As a result, the Centers for Disease Control,²⁹ US Preventive Services Task Force,²⁷ AHA/American College of Cardiology Foundation guidelines for the secondary prevention of coronary artery disease,³⁰ and American College of Cardiology/AHA guideline for the primary prevention of CVD recommend dietary counseling by clinicians and other members of the healthcare team.³¹ Recent evidence that diet quality may be suboptimal even after a diagnosis of coronary disease also supports outpatient diet counseling in primary and secondary prevention settings.^32,33 Last, limited evidence demonstrates the utility of dietary screeners in measuring dietary changes associated with behavioral interventions.³⁴ In summary, the efficacy of dietary counseling for improving diet quality and cardiometabolic risk factors both in the population at large and among those with established disease provides additional support for incorporating evidence-based dietary screening and counseling of patients into clinical practice.²⁵

Feasibility of Clinician-Delivered Diet Assessment and Counseling

Population-wide dietary assessment and counseling by clinicians and other members of the healthcare team have become more feasible in the past decade as the reach of health systems has expanded. The percentage of US adults <65 years of age who have CVD risk factors and are covered by healthcare insurance has increased significantly since passage of the 2010 Affordable Care Act.³⁵ The percentage of adults who have contact with clinicians and other members of the healthcare team annually is 85%; the average is 3 to 4 times yearly.³⁶ Primary care clinicians, cardiologists, and endocrinologists treat the majority of adults with hypertension, hyperlipidemia, and diabetes mellitus, and evidence-based dietary interventions are included in the management guidelines for these conditions.^37–39

Also in support of the feasibility and efficacy of dietary assessment and counseling by clinicians and other members of the healthcare team are data that show that most patients view clinicians as credible and trusted sources of dietary advice.^40,41 Many clinicians have indicated interest in discussing diet with their patients and view this activity as within the scope of their practice.^13,40,42 Low perceived knowledge, limited self-efficacy, and lack of time were cited as barriers.⁴³ Because the EHR provides CDS, documentation, and referral processes, it is more feasible to incorporate diet assessments into existing workflows with less clinical burden than previously. Even as medical nutrition training and competencies improve, a rapid diet screener can help standardize diet assessment in clinical practice and facilitate tracking, benchmarking, and the improvement of diet quality.

Desired Theory-Based Features of a Rapid Diet Screener Tool

All self-reported dietary tools are subject to some measurement error and bias; therefore, diet assessment tools must undergo extensive reliability and validity testing. In this advisory, we a priori established theoretical and practice-based criteria for an optimal diet screener tool (Table 1) applicable for assessing diet quality in the US adult population 20 to 75 years of age.^45,46 In reviewing the literature, we emphasized diet screener tools developed or used in clinical practice within the past 10 years that evaluated the total diet rather than single food groups (eg, fruits and vegetables) or nutrients. Tools that evaluated total diet were emphasized to ensure that the tools evaluated reflected up-to-date dietary guidance and captured foods to limit and encourage in recognition that dietary patterns are complex and individual components may have synergistic effects. A detailed summary of brief diet screener tools is provided in Supplemental Tables 1 and 2. For each identified criterion, we allocated 0, 0.5, or 1 point if the dietary screening tool met those criteria and summed them to generate a total score; screeners ranged from meeting 2.5 to 9 of 10 established criteria.

Of the 15 tools reviewed for this statement, the 3 that met the greatest number of theoretical and practice-based validity criteria were the Mediterranean Diet Adherence Screener (MEDAS)^47–51 and its variations,⁵² the modified, shortened Rapid Eating Assessment for Participants (REAP),^53–55 and the modified version of the previously validated Starting the Conversation tool (Table 2).^43,56* Although the Starting the Conversation tool met 0.5 fewer criteria than the PrimeScreen tool, we chose to discuss it given its frequency of use in clinical practice.

Table 2. Rapid Diet Screener Tool Options for Clinical Settings

Briefly, the original 14-item MEDAS⁵¹ was developed and validated in the PREDIMED study (Prevención con Dieta Mediterránea) conducted in Spain from 2003 to 2009. In 2011, the screener was compared with a full-length, 135-item food frequency questionnaire in >7000 older adults at high risk for CVD. There was moderate agreement between the MEDAS and food frequency questionnaire results (r=0.52, P<0.001). Higher MEDAS scores were associated with higher high-density lipoprotein cholesterol (P<0.0001) and lower body mass index, waist circumference, triglyceride concentrations, triglyceride/high-density lipoprotein ratio, fasting glucose concentrations, total cholesterol/high-density lipoprotein ratio, and 10-year coronary artery disease risk.⁵¹ The 14-item screener has subsequently been validated in smaller samples in Germany⁴⁸ and England.⁴⁷ More recently, the English version of the MEDAS was validated in a clinical sample of UK adults 55 to 80 years of age who were at high risk for CVD, although no significant associations between the MEDAS score and cardiometabolic risk factors were observed.⁴⁹

To date, 2 studies have applied the MEDAS to US populations. The first study used forward and backward translation from Spanish to English to incorporate the MEDAS into an EHR within the EPIC system in a US neurology clinical practice.⁵⁰ A corresponding electronic flow sheet provided CDS (ie, auto-scoring and auto-interpretation) to allow responses to be reviewed with patients. It was recommended that patients with scores <10 be referred to a registered dietitian nutritionist. Notably, the system automatically incorporated results from the MEDAS into the progress note, capturing data relevant for diet quality improvement and practice-based research. Although the English-translated tool was not specifically validated, lower MEDAS scores were associated with higher body mass index and higher prevalence of hypertension. The second study⁵² from the same research group developed the Mediterranean Eating Pattern for Americans, which added 2 questions and altered target intakes for some foods. The Mediterranean Eating Pattern for Americans was validated against a 156-item food frequency questionnaire, and although the concordance between the Mediterranean Eating Pattern for Americans and food frequency questionnaire was less than in the original validation study in Spain,⁵¹ higher scores remained positively associated with higher diet quality and were marginally, but not significantly, associated with lower body mass index.

Overall, the MEDAS and Mediterranean Eating Pattern for Americans screeners met most of the theoretical and clinical validity criteria (Table 1) for a rapid diet assessment screening tool for use in clinical settings. They assess total dietary quality, demonstrate criterion and predictive validity, are brief, and can be administered across diverse populations remotely or in person with a computer, mobile application, or paper form. Either can be used to provide CDS by numerous clinicians and other members of the healthcare team.

A second dietary screener tool that has been used frequently is the REAP and shorter REAP-S. The first version of the REAP was developed⁵³ with the intent of providing a tool to rapidly assess and discuss diet quality with patients, particularly in lower-income populations. The 27- to 31-item REAP questionnaire has been validated with the Healthy Eating Index⁵⁸ (r=0.49, P=0.0007), a diet-quality index used to evaluate concordance with US dietary guidance, and has been demonstrated to have good test-retest reliability (r=0.86). However, because the original tool emphasized limiting dietary fat, only updated modified versions reflect contemporary nutritional guidance.⁵⁴ A recent report that compared the unmodified REAP tool with a modified version found that older white adults with higher REAP scores had lower odds of executive dysfunction and that the results were similar between the 2 versions.⁵⁷ The shortened version (16-item REAP-S), which had less focus on dietary fat, was recently compared with the Healthy Eating Index–2010 in a group of healthy adult omnivores and vegetarians.⁵⁴ The REAP-S scores were moderately correlated with the Healthy Eating Index–2010 and with the nutrient density of the diet. Taken together, the REAP tools have some of the established criteria deemed important for an optimal brief diet assessment tool for use in diverse clinical settings, including a physician key to guide dietary advice.

Last, the modified Starting the Conversation tool⁴³ assesses dietary intake and provides CDS with only 8 questions. Although the tool was not validated against a complete dietary assessment method, it was sensitive to dietary changes over 2 years and able to be used by a range of clinicians and other members of the healthcare team.⁵⁶ It also has been used among patients with a diagnosis of heart failure, and among this clinical sample, better dietary habits moderated the association between heart failure and cognitive impairment.⁵⁹

Desired Practice-Based Features of a Diet Screener Tool

The adoption of evidence-based validated diet screener tools is necessary and desirable in clinical practice, yet their use must also be balanced against practical constraints: time and workflow. Thus, as it pertains to clinical utility, this section discusses existing tools ordered from the least to most time-consuming and includes both evidence-based data and expert opinion (summarized in Figure 1 and Table 2).

• Download figure
• Download PowerPoint

Almost 16 years ago, there was a call to action for clinicians and other members of the healthcare team to play a key role in combatting the twin epidemics of excess body weight and physical inactivity.⁶⁰ Yet, annual physician visits generally remain an overlooked opportunity to screen patients with regard to diet and physical activity. The challenges in assessing diet in clinical settings are practical and relate primarily to time available to assess diet quality, to provide actionable steps to patients, and to enter the information into the EHR.

To start a conversation about lifestyle related behaviors, the simplest tool available includes only 2 questions (Table 2). These questions can be asked by the clinician, medical staff, web-based educational site, or a community resource person.⁴⁴ Because of the brevity of the tool, it may be advantageous for clinicians to administer the questions, given the high level of respect for their advice and their ability to motivate patients to change behaviors.^60–65 Similar to other screening questions administered in primary care settings, patients who score higher than a certain threshold (indicating greater dietary risk) would be prompted for additional counseling or a more detailed dietary assessment.⁴⁴

Many clinicians have developed their own tools to start a discussion about diet quality and CVD prevention. Some of these tools share elements with validated rapid diet screener tools (Table 2).^{44,46,49–51,56,66–68} For example, clinicians have asked patients to rate their diet from 1 to 10 (poor to optimal) and then asked patients to indicate why they gave the score they did and to further indicate what food changes would be needed to be an 8 or 9 of 10. Some patients quickly identify eating out and having too much salt or carbohydrates. If the score is ≤6, the patient is referred for more formal counseling. In 2004, during his AHA presidential address, Dr Robert Eckel endorsed the use of a 3-minute diet and lifestyle patient interview¹² that was easy to implement and provided actionable goals for the patient to work on for a follow-up visit. Since that time, EHRs have become ubiquitous, making it easier to capture information elicited from a short interview and have it available for future discussions. Such discussion may encourage patients to reduce their intake of added sugar and substitute fish or chicken for red and processed meat.^14,69,70 Although many of these tools reflect prudent dietary advice, greater adoption of validated tools in clinical settings can assist in standardizing evidence-based approaches going forward.¹²Table 2 summarizes some of the available tools, emphasizing the degree to which they have been validated and their approximate length. Although many of the longer, validated tools provide greater ability to monitor changes in a patient’s diet over time and to correlate them with diet-related cardiometabolic risk markers, the tool that clinicians and other members of the healthcare team select will be determined largely by time constraints in their existing workflows.

In the future, if nutrition researchers and clinicians use consensus group methods⁷¹ when developing new rapid diet screener tools, it can help ensure that tool development reflects theoretical and practice-based expertise. Adding any time to existing clinician workflows is challenging, and partnerships between clinicians and other members of the healthcare team and nutrition providers and researchers are essential for consistent implementation of a rapid diet screener that measures and reflects current dietary guidance.

Implementation and Challenges of a Diet Screener Tool

Broad integration of diet screener tool into clinical settings represents a major change to current standards of practice and thus requires evidence of effectiveness to encourage adoption. The use of a brief diet screener tool has been documented to validly measure patients’ dietary quality in ambulatory settings.^46,53,56 As noted, there is growing recognition that behavioral lifestyle counseling can modestly benefit cardiometabolic health^15,16 and that more intensive counseling can lead to further improvement.²⁸ Although several rapid validated diet screener tools are now available (Supplemental Tables 1 and 2), few studies have evaluated the effectiveness of these tools in promoting improved diet quality.

A study evaluating a 27-question REAP tool surveyed medical students and primary care clinicians to determine the feasibility of implementation.⁵³ Overall, the clinicians rated the tool moderately high for ease of use, practicality, usefulness, and length. The 8-question Starting the Conversation tool was evaluated among adults with type 2 diabetes mellitus participating in a self-management intervention. The results indicated that the tool was feasible for use in a primary care setting and was sensitive to treatment-related changes in dietary intake.⁵⁶ Although this tool, like the REAP, does not reflect current guidance on dietary fat, an adapted version of the Starting the Conversation tool has been developed that reflects contemporary dietary recommendations.⁴³ The 14-item MEDAS questionnaire for adherence to the Mediterranean diet was evaluated within the context of a structured clinical documentation support tool built into the EHR.⁵⁰ This study demonstrated the feasibility of using the EHR to capture and document in clinic notes the level of adherence and to provide immediate and specific dietary feedback with a numeric score to discuss with patients.

In an era when deficiencies in medical nutrition education and training persist, diet screener tools with CDS, including guidance on when to refer to another clinician, can bridge gaps in knowledge and competencies. Taken together, evidence suggests that some clinicians and other members of the healthcare team may be able to implement rapid diet screener tools into their clinical practice and workflows. Tools of a length similar to the MEDAS, when incorporated into the EHR, have been shown to be feasible in clinical workflows in some settings,⁵⁰ and knowledge about how best to implement these tools to promote behavior change and to improve health is progressing. Ultimately, information gleaned from a diet screener tool should enable clinicians and other members of the healthcare team to apply the 5As framework for dietary behavior change (Table 3), which operationalizes a clinical process for guiding patients through the process of improving their diet. Assessment is an integral first step to providing dietary advice and securing additional support to adopt dietary changes.⁷² It should also eventually facilitate practice-level and system-level improvements in diet quality through tracking of diet-quality data in the clinical information system, reporting of diet outcomes, and engaging patients in efforts that will ultimately result in improved management of dietary risk and CVD at a population level. Given the lessons learned from implementing depression screening into clinical settings,⁷³ it is prudent to consider advocating for integration of dietary screening into the EHR and making it a standard billable practice to incentivize adoption by clinicians. In addition, if accountable care organizations add routine dietary screening as a core quality measure, it may more readily be adopted as a standard-of-care measure in primary care settings.

Considerations for Integrating a Rapid Diet Screener Tool Into the EHR

Clinicians, from primary care to specialist clinicians, often cite difficulty with completing clinical tasks, including documentation of the healthcare visit, during the allotted time frame.⁷⁴ The EHR is the ideal platform to prompt clinicians and other members of the healthcare team to capture dietary data and deliver dietary advice to patients⁷⁵ because it not only allows secure storage of data but also enables access to these data when needed at the point of care and for documentation purposes. There are 2 primary mechanisms for collecting such data via the EHR: a rapid diet screener tool completed by the patient using an EHR portal⁷⁶ or manual entry of data by clinicians and other members of the healthcare team into a dietary tool EHR template (Table 4).⁷⁷ For either approach, structured data elements are preferred not only for their ease of completion (yes/no versus fill-in-the-blank responses) but also to avoid the complicated downstream interpretation of unstructured (ie, narrative) data elements.

The feasibility of conducting wellness assessments via patient-facing EHR portals has been demonstrated.^76,78 A current limitation to collecting data from patients via EHR portals is their reach. Although nearly 9 in 10 office-based physicians have adopted some form of an EHR, 50% do not take advantage of more advanced features.⁷⁹ In addition, underserved populations, for example, racial minorities or those from rural areas, are less likely to have access to and to be regular users of patient portals.⁷⁸ Healthcare systems with opt-out strategies for patient portal access (ie, enrolling all patients by default and requiring them to actively unenroll) may have fewer disparities in use among underserved populations than those with opt-in strategies, which require potential users to register to use the system. Complex or multistep patient portal registration processes could also serve as a barrier to the use of EHR portals.

Manual entry of dietary data into an EHR data template by clinicians and other members of the healthcare team is preferable to using a separate platform in which to enter relevant data. Data entered on an external website are not stored in the EHR, and although that information is available for use to counsel the patient during the session, the external website does not provide a record of the patient-clinician dietary discussion. Conversely, dietary data via the EHR are available for use during a clinical encounter and are permanently recorded in the patient’s health record. To deliver the data from rapid dietary screeners to the point of care, the EHR must be programmed to make the data available.⁸⁰ Feedback from rapid diet screener tools can be presented to clinicians within the EHR in the form of a best-practice alert. A pop-up message can remind clinicians to perform a task. Best-practice alerts can also prompt referrals to dietitians and other resources within the healthcare system. In addition, key messages to improve diet quality can be added to after-visit summaries with the use of keyboard shortcuts. More sophisticated EHR-embedded web applications can be delivered to the point of care to give clinicians options for providing next steps to patients and to prompt clinicians to engage in shared decision-making for health behaviors.

Most common EHR platforms have extant functionality for displaying a web application within the EHR viewing screen so that clinicians and other members of the healthcare team do not have to navigate to an external website (Figure 2A). Such web applications can be programmed to automatically populate with EHR data and to be interactive to enhance shared decision-making at the point of care.⁸¹ Mobile health applications can be designed to mimic EHR-embedded applications and may serve to extend the impact of the intervention beyond the point of care (Figure 2B).⁸⁰

• Download figure
• Download PowerPoint

Historically, the use of myriad EHR platforms and versions of platforms has presented technical challenges to scalability that have been difficult to surmount.⁸² Specifically, the differences between EHR platforms and versions therein required programming web applications and other EHR-embedded solutions that were specific to each platform and version of the platform. Recently, standardized approaches have emerged for transmitting health data that will more seamlessly allow rapid diet screeners to be implemented in the EHR.^83,84 These approaches also enhance the scalability of such tools across institutions. Although technically it has become easier to disseminate EHR-based tools widely across platforms and versions of platforms, institutional approval processes for such tools need to be integrated into the EHR if they are not part of the scheduled EHR software update.

Once rapid diet screeners tools are implemented in the EHR, it is critical to quantify usability, workflow integration, and level of patient-clinician communication with the use of such tools to ensure their continued impact on health. Before EHR implementation, prototypes of rapid diet screeners can be tested by end users using a think-aloud protocol.⁸⁵ Assessing usability of the assessment before implementation can determine whether changes need to be made before implementation for use within certain clinics or among certain clinicians and other members of the healthcare team. Gathering these data ahead of time can improve the uptake of the application in the real world.⁸⁶

New Calls to Action and Current Gaps in Clinician-Delivered Diet Assessment, Education, and Counseling

Despite the evidence supporting the value of diet assessment, education, and counseling by clinicians, data have shown sizable gaps in the percentage of patients who report receiving clinician-delivered dietary counseling. According to data from National Health Interview Surveys, the prevalence of dietary counseling by physicians in 2000 was 24%, increasing to 33% in 2011.¹⁰ Disparities in diet counseling related to access to health care and differences based on ethnicity and sex were found. Men and the uninsured were significantly less likely to receive diet counseling; paradoxically, those with higher socioeconomic status were more likely to report more diet counseling despite generally having higher diet quality.¹⁰

Recent evidence suggests that nutrition education for clinicians, combined with the creation of a culture that values dietary counseling, increased clinician engagement.⁸⁷ However, survey data have consistently demonstrated that physicians cite insufficient nutrition training and knowledge as a major barrier to engaging patients in diet counseling, even during their peak learning years.^9,88,89 These data align with ongoing gaps in medical nutrition education and training,^90,91 underscoring an imminent need to better integrate nutrition education and competencies into medical school curricula.⁹² Recently, both the American College of Lifestyle Medicine⁶⁹ and the American Association of Medical Colleges have called for greater incorporation of behavioral and social sciences into medical school and training curricula and for competencies related to behavior counseling.⁹³ Similarly, the AHA and American College of Lifestyle Medicine have called for renewed efforts to improve medical nutrition education, training, and professional competencies across healthcare professionals.^11,69 Although the largest body of evidence emphasizes the need for nutrition training in medical schools and residencies, training for physician assistants, nurses, pharmacists, and others is also limited and important.^94,95 In the absence of such training, a valid rapid diet screener tool with CDS can help bridge the gap in conducting diet assessment and providing feedback in clinical settings.

Future Research Opportunities to Improve Existing Diet Screener Tools

Although some groups have altered available diet screener tools to conform to current dietary guidance,^43,57 thorough validation of these modified tools is needed, particularly in diverse populations. Included in these validation studies should be the predictive validity in relation to cardiometabolic risk factors in both primary and secondary prevention settings. Relevance for secondary prevention of chronic disease (eg, specialist visit for hyperlipidemia, diabetes mellitus, hypertension, and coronary heart disease)⁴⁶ may require more focused assessment of specific aspects of dietary intake, including eating frequency and timing. Of contemporary interest is the use of machine learning to develop and validate diet screener tools. One example of these efforts was conducted in a group of French-speaking adults in Quebec for whom a supervised learning approach was used to identify food patterns most predictive of health outcomes that included a maximum of 6 questions.⁹⁶ Subsequently, the algorithm classified individuals into mutually exclusive subgroups optimized to identify individuals at risk of having low diet quality. Although supervised learning within diet assessment may enhance tool development for specialized populations, the primary advantage of supervised learning versus static approaches is that it allows a highly specific assessment in only a few minutes. A summary of key areas to direct future research endeavors is provided in Table 5.

Summary

This AHA scientific statement provides compelling rationale for adopting a rapid diet screener tool for use in clinical settings. This call to action is made after an assessment of the feasibility for clinicians and other members of the healthcare team to screen diet quality and to provide feedback in clinical settings, an evaluation of existing rapid diet screener tools against optimal theory and practice-based features, and consideration of the necessity and practicality of integrating a rapid diet screener tool into the EHR.

It is evident that clinicians generally view dietary counseling as within their scope of practice, and clinician-delivered dietary counseling modestly improves diet quality and diet-related cardiometabolic risk factors. However, clinicians are constrained by lack of nutrition knowledge and time for patient counseling. Several existing rapid diet screener tools have been identified that meet theory- and practice-based validity criteria, some of which take <5 minutes to complete, have been integrated into EHRs, and provide immediate actionable dietary feedback. The process of integrating rapid diet screener tools into a health system is a complex but critical facet of improving primary and secondary prevention that ultimately involves concomitant consideration of the structural barriers (ie, time, reimbursement) that impede clinical care. This AHA scientific statement is designed to accelerate efforts to make diet quality assessment an integral part of office-based care delivery by encouraging critical conversations among clinicians, individuals with diet/lifestyle expertise, and specialists in information technology. In the future, providing regular diet assessment and recommendations based on validated clinical tools will help patients address the lifestyle changes they need for healthier lives and reduce the public health and economic burdens from CVD and other chronic diseases linked to poor diet quality.

Footnotes