Final Research Plan
Eating Disorders in Adolescents and Adults: Screening
September 24, 2020
Recommendations made by the USPSTF are independent of the U.S. government. They should not be construed as an official position of the Agency for Healthcare Research and Quality or the U.S. Department of Health and Human Services.
- Does screening for eating disorders in adolescents and adults improve health outcomes, including for specific subgroups of interest?
- What is the accuracy of primary care–relevant screening tools for eating disorders in adolescents and adults, including for specific subgroups of interest?
- What are the harms of screening for eating disorders in adolescents and adults, including for specific subgroups of interest?
- How effective are interventions for improving health outcomes in screen-detected or previously untreated adolescents and adults with eating disorders, including for specific subgroups of interest?
- What are the harms of interventions for eating disorders, including for specific subgroups of interest?
The Research Approach identifies the study characteristics and criteria that the Evidence-based Practice Center will use to search for publications and to determine whether identified studies should be included or excluded from the Evidence Review. Criteria are overarching as well as specific to each of the key questions.
|Anorexia nervosa, bulimia nervosa, binge eating disorder, other specified feeding and eating disorder, and avoidant restrictive food intake disorder, based on DSM-5 criteria or other valid diagnostic criteria (e.g., DSM-IV); subthreshold condition definitions (e.g., meeting most but not all DSM criteria for the disorders above, as defined by study authors) are also eligible
|Other DSM-5 categories of eating disorders (e.g., pica, rumination disorder) or potentially unhealthy eating behaviors or syndromes not recognized by DSM-5 (e.g., orthorexia)
|KQs 1–3: Unselected or explicitly asymptomatic adolescents and adults (age ≥10 years) without signs or symptoms of an eating disorder, including populations selected for increased risk of an eating disorder (e.g., based on age, sex, race/ethnicity, gender identity, or mental health comorbidity) and populations selected based on high BMI
KQs 4, 5: Studies enrolling adolescents and adults (age ≥10 years) who screen positive for eating disorders in a primary care setting or are identified through other population-based screening; studies enrolling populations from specialty settings who have not been previously treated for eating disorders are also eligible
All KQs: A priori subgroups of interest include those defined by age, sex, race/ethnicity, gender identity, sexual orientation, mental health comorbidity, or recruitment setting
|Studies limited to participants undergoing evaluation for bariatric surgery; studies limited to individuals who are underweight (BMI <18.5 kg/m2 for adults or BMI <5th percentile on growth charts for age and sex for adolescents) or with other physical signs or symptoms of an eating disorder
|KQs 1–3: Screening questionnaires designed to detect eating disorders or risk of eating disorders that are feasible for use for screening in primary care (i.e., brief, easy to interpret)
|KQs 1–3: Serologic screening (e.g., using leptin or other biomarkers)
|KQs 4, 5: Individual, group, or family therapy (cognitive behavioral therapy or other forms of psychotherapy); pharmacotherapy with FDA-approved medications; interventions that combine therapy with pharmacotherapy, or those that include other components such as education or nutritional counseling (in addition to therapy and/or pharmacotherapy) are also eligible
|KQs 4, 5: Public awareness campaigns without specific interventions linked to screening; complementary and alternative therapies, or those considered to be adjunctive therapy (e.g., acupuncture, herbal supplements, massage, light therapy); interventions designed for primary prevention only (e.g., school-based interventions to improve body dissatisfaction and prevent onset of eating disorders)
|KQs 1, 3: Screened vs. nonscreened groups
KQ 2: Comparison with acceptable reference standard (structured or semistructured diagnostic interview or a nonbrief [>5 minutes] unstructured interview with mental health clinician)
KQs 4, 5: No treatment, attention control, wait-list control, or minimal intervention (e.g., brief education about eating disorders); placebo-controlled studies of pharmacotherapy
|KQs 4, 5: Head-to-head comparisons of two active interventions
|KQs 1, 4: Eating disorder remission or symptom reduction, general health-related quality of life or function, eating disorder–related quality of life or function, depression, anxiety, suicide, and mortality
KQ 2: Sensitivity, specificity, positive and negative predictive values, positive and negative likelihood ratios, and diagnostic odds ratios
KQ 3: Anxiety, labeling, stigma, psychosocial harms, and false-positive results
KQ 5: Any harms that result as an effect of interventions such as increased depression, increased anxiety, worsening of the eating disorder, or adverse effects from medications
|KQs 1, 4: Screening or referral rates, attitudes about screening; intermediate outcomes (e.g., weight change, frequency of menses, frequency of specific behaviors [e.g., change in frequency of binge eating episodes])
KQ 2: Psychometric development and internal (e.g., split sample) validation studies of new instruments; assessment of provider or participant attitudes toward the screening questionnaire
|All KQs: RCTs
KQ 2: Cross-sectional and cohort studies of screening test accuracy are also eligible
KQs 3, 5: Cohort studies with a concurrent control group are also eligible
|All other study designs, including case-series and case-control studies; systematic reviews, and others
|Studies rated good or fair quality
|Studies rated poor quality
|All KQs: Primary care clinics or other settings applicable to primary care, including school-based health centers and other community settings
KQs 2, 4, 5: Settings referable from primary care are also eligible
|Settings or institutions not applicable to primary care (e.g., school classrooms, bariatric surgery centers) or inpatient/residential settings
|Research conducted in the United States or in populations similar to U.S. populations with services and interventions applicable to U.S. practice (countries categorized as “very high” on the Human Development Index [as defined by the United Nations Development Programme])
|Research not relevant to the United States in countries categorized as less than “very high” on the Human Development Index
|Full text published in English
|Not English language
Abbreviations: BMI=body mass index; DSM-5=Diagnostic and Statistical Manual of Mental Disorders (5th edition); DSM-IV=Diagnostic and Statistical Manual of Mental Disorders (4th edition); FDA=U.S. Food and Drug Administration; KQ=key question; RCT=randomized, controlled trial.
The draft Research Plan was posted on the USPSTF website for public comment from June 25, 2020, to July 28, 2020. Several comments were received about population inclusion and exclusion criteria, as well as suggestions to add additional subgroups to those already listed, including “chronic disease comorbidity,” “sexual minorities,” “athlete status,” and others. In response, the USPSTF added two additional groups based on a rationale for how these factors may affect outcomes of screening and treatment: “sexual orientation” (in addition to “gender identify”) and “recruitment setting.” The phrasing of most comments about subgroups suggested that commenters assumed that populations not explicitly listed as a subgroup would be excluded from the review scope. To address this, the USPSTF removed the “part b” from each key question that restated the question along with the a priori subgroups of interest. Instead, the USPSTF added “including for specific subgroups of interest” at the end of each key question. The a priori subgroups of interest are listed in the Table. One group emphasized the importance of other prevention strategies, including studies that identify persons with risk factors for eating disorders (e.g., those with elevated weight/shape concerns) and offer interventions aimed at preventing the onset of eating disorders. In response, the USPSTF added text to clarify that interventions designed for primary prevention only are not eligible for inclusion. A few comments asked for clarification on whether the scope included interventions for eating disorders that combine therapy and pharmacotherapy or include components (e.g., nutritional counseling) in addition to therapy or pharmacotherapy. In response, the USPSTF recommended being inclusive of studies that offer multicomponent interventions, given the expected low volume that meets our population and comparison eligibility criteria.