OBJECTIVES:

Nearly one-quarter of individuals diagnosed with avoidant/restrictive food intake disorder (ARFID) require medical admission. There have been efforts to characterize ARFID in outpatient and intensive day treatment settings; however, authors of few studies have examined this presentation in the inpatient pediatric hospital setting. In this study, we aim to further characterize patients presenting to the hospital with concerns for ARFID.

METHODS:

This study involved a retrospective chart review of medically admitted patients with ARFID seen by the psychiatry consultation service at a tertiary care New England pediatric hospital from 2015 to 2016.

RESULTS:

The typical hospitalized patient with ARFID was a 12.9-year-old, white girl with previous history of outpatient mental health treatment, anxiety disorder, and gastrointestinal-related diagnoses admitted to adolescent medicine or pediatric hospitalist services with >1 year of feeding difficulties often triggered by a precipitating event. Despite >80% of subjects receiving evaluations as outpatients for feeding-related concerns, including 60.5% seeing their primary care provider, <20% were diagnosed with ARFID before hospitalization. The average length of admission was 8 days. All imaging, scopes, and swallow studies conducted during the admission were nonrevealing. Almost half of patients required enteral tube feeds, and 63.2% required psychiatric medications during the admission. Only 31.6% of patients had the ARFID diagnosis documented in their discharge notes.

CONCLUSIONS:

Consistent identification of ARFID remains variable, underrecognized by community providers, and underdocumented by hospital providers. Accurate recognition of ARFID and additional study into contributory factors and treatment approaches may help improve effective health care use and treatment outcomes.

Eating and feeding disorders are a common diagnostic presentation leading to medical hospitalization and psychiatric consultation and/or liaison referral.1  Included in this diagnostic category is avoidant/restrictive food intake disorder (ARFID) (see Table 1), a relatively new diagnosis in the current Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition2  that encompasses feeding issues that were previously diagnosed as a feeding disorder of infancy or early childhood or an eating disorder (ED) not otherwise specified. Prevalence data on ARFID in the general population are limited; recent studies suggest that 5% to 14% of patients presenting to inpatient ED programs and 22.5% of patients presenting to ED day treatment present with ARFID.3  Although there are well-defined treatment approaches for EDs, including the use of empirically supported clinical pathways for the management of anorexia nervosa (AN), there are few studies in which the empirical management of ARFID is explored, particularly in the inpatient medical setting.4  Despite the introduction of ARFID as a diagnosis and the recognized need for continued characterization of this clinical presentation, the majority of empirical efforts to date have been focused on outpatient or day treatment settings.5,6  It is important to note, however, that nearly one-quarter of individuals with ARFID presenting to an adolescent ED program at a pediatric tertiary care hospital are medically admitted after an initial evaluation.7  Therefore, it appears important to understand the presentation of ARFID in the inpatient pediatric setting, which may represent a different cross section of patients with ARFID.

TABLE 1

Diagnostic Criteria for ARFID

Criteria
An eating or feeding disturbance (eg, apparent lack of interest in food, avoidance based on the sensory characteristics of food, concern about aversive consequences of eating) as manifested by persistent failure to meet appropriate nutritional and/or energy needs associated with ≥1 of the following: significant wt loss (or failure to achieve expected wt gain or faltering growth in children), significant nutritional deficiency, dependence on enteral feeding or oral nutritional supplements, and marked interference with psychosocial functioning 
The disturbance is not better explained by lack of available food or by an associated culturally sanctioned practice. 
The eating disturbance does not occur exclusively during the course of AN or bulimia nervosa, and there is no evidence of a disturbance in the way in which one’s body wt or shape is experienced. 
The eating disturbance is not attributable to a concurrent medical condition or not better explained by another mental disorder. When the eating disturbance occurs in the context of another condition or disorder, the severity of the eating disturbance exceeds that routinely associated with the condition or disorder and warrants additional clinical attention. 
Criteria
An eating or feeding disturbance (eg, apparent lack of interest in food, avoidance based on the sensory characteristics of food, concern about aversive consequences of eating) as manifested by persistent failure to meet appropriate nutritional and/or energy needs associated with ≥1 of the following: significant wt loss (or failure to achieve expected wt gain or faltering growth in children), significant nutritional deficiency, dependence on enteral feeding or oral nutritional supplements, and marked interference with psychosocial functioning 
The disturbance is not better explained by lack of available food or by an associated culturally sanctioned practice. 
The eating disturbance does not occur exclusively during the course of AN or bulimia nervosa, and there is no evidence of a disturbance in the way in which one’s body wt or shape is experienced. 
The eating disturbance is not attributable to a concurrent medical condition or not better explained by another mental disorder. When the eating disturbance occurs in the context of another condition or disorder, the severity of the eating disturbance exceeds that routinely associated with the condition or disorder and warrants additional clinical attention. 

Adapted from American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Washington, DC: American Psychiatric Association; 2004:334

The authors of one study examined patients admitted to the medical setting with ARFID compared with patients presenting with AN,8  but no studies have yet examined the health care use of these patients and care provided in the assessment and management of ARFID in the pediatric hospital setting. Additionally, literature suggests that primary care providers (PCPs) are largely unfamiliar with the ARFID diagnosis.9  Without proper diagnosis and referral to appropriate treatments, patients experience decreased quality of life and increased health care costs.10 

Our aims in this study are to describe the demographic and psychosocial characteristics of patients admitted to a stand-alone pediatric hospital for medical complications associated with ARFID, to describe medical and psychiatric characteristics and psychiatric treatment dispositions, and to describe the health care use of these patients.

All evaluations completed by a psychiatry consultation service (PCS) at a stand-alone children’s hospital in New England were reviewed to identify medically hospitalized patients diagnosed with ARFID by PCS clinicians between January 2015 and October 2016. PCS is a multidisciplinary service that consists of child psychiatry attending physicians and fellows, attending pediatric psychologists, psychology postdoctoral fellows, psychology interns, and licensed social workers. The service provides consultation and liaison psychiatry services to patients with a wide range of general psychiatric and pediatric behavioral health needs admitted to inpatient medical and/or surgical units in an urban, tertiary care, pediatric hospital with ∼400 licensed beds. Thirty-eight patients were diagnosed with ARFID, representing 2.04% of evaluations completed by PCS during this period. For patients with multiple admissions meeting study inclusion criteria, only the first admission was sampled.

This study was approved by the hospital’s institutional review board. Information was collected by the research team (which included PCS psychiatry and psychology faculty and postdoctoral trainees) using Research Electronic Data Capture11  via chart review of the electronic medical record, including admission notes, laboratory tests, vitals, PCS initial evaluation, and progress notes. Interrater reliability across all variables was 94.5%. PCS clinicians used a standardized psychiatric evaluation template, which structured the interview process and documentation. Certain variables were not routinely assessed as part of the evaluations (eg, breastfeeding history, early attachment, temperament) and therefore were not included in analyses. For all other variables, if a data point was not specifically stated in the chart review, such as whether someone had history of depression, the response was coded as “no.”

Diagnostic categories before and during admission, medications prescribed, and medical workup were grouped by category for data analysis. Descriptive statistics and frequencies were evaluated by using IBM’s SPSS statistical software platform (IBM SPSS Statistics, IBM Corporation).12 

The mean age for the sample was 12.87 years (SD: 4.86), with a range of 5 to 26 years. The sex of more than two-thirds of the sample was female, and a majority of patients were identified as white, non-Hispanic. Please see Table 2 for additional demographic information.

TABLE 2

Demographic Characteristics

Characteristicsn (%)
Sex  
 Female 26 (68.4) 
 Male 12 (31.6) 
Race or ethnicity  
 White (non-Hispanic) 28 (73.7) 
 Hispanic 3 (7.9) 
 African American 2 (5.3) 
 South Asian (India, Pakistan, Bangladesh, or Sri Lanka) 2 (5.3) 
 Multicultural 2 (5.3) 
 Middle Eastern 1 (2.6) 
Primary insurance  
 Commercial or private 27 (71.1) 
 State or federal 10 (26.3) 
 Tricare 1 (2.6) 
Median household income (based on zip code), $ (thousands)  
 40–60 9 (23.7) 
 60–80 10 (26.3) 
 80–100 10 (26.3) 
 100–120 7 (18.4) 
 120–140 2 (5.3) 
Admitting service  
 Adolescent medicine 15 (39.5) 
 General pediatrics 14 (36.8) 
 Gastroenterology 8 (21.1) 
 Pulmonology 1 (2.6) 
Grade in school  
 Preschool 1 (2.6) 
 Kindergarten 3 (7.9) 
 Elementary level (1–5) 8 (21.0) 
 Middle school (6–8) 12 (31.6) 
 High school (9–12) 8 (21.0) 
 College and/or GED degree 5 (13.1) 
 Unknown 1 (2.6) 
Learning problems 14 (36.8) 
Academic intervention plans  
 Individualized education plan 13 (34.2) 
 504 plan 3 (7.9) 
Characteristicsn (%)
Sex  
 Female 26 (68.4) 
 Male 12 (31.6) 
Race or ethnicity  
 White (non-Hispanic) 28 (73.7) 
 Hispanic 3 (7.9) 
 African American 2 (5.3) 
 South Asian (India, Pakistan, Bangladesh, or Sri Lanka) 2 (5.3) 
 Multicultural 2 (5.3) 
 Middle Eastern 1 (2.6) 
Primary insurance  
 Commercial or private 27 (71.1) 
 State or federal 10 (26.3) 
 Tricare 1 (2.6) 
Median household income (based on zip code), $ (thousands)  
 40–60 9 (23.7) 
 60–80 10 (26.3) 
 80–100 10 (26.3) 
 100–120 7 (18.4) 
 120–140 2 (5.3) 
Admitting service  
 Adolescent medicine 15 (39.5) 
 General pediatrics 14 (36.8) 
 Gastroenterology 8 (21.1) 
 Pulmonology 1 (2.6) 
Grade in school  
 Preschool 1 (2.6) 
 Kindergarten 3 (7.9) 
 Elementary level (1–5) 8 (21.0) 
 Middle school (6–8) 12 (31.6) 
 High school (9–12) 8 (21.0) 
 College and/or GED degree 5 (13.1) 
 Unknown 1 (2.6) 
Learning problems 14 (36.8) 
Academic intervention plans  
 Individualized education plan 13 (34.2) 
 504 plan 3 (7.9) 

GED, general education development.

Regarding early development, 2 patients in the sample had a history of prematurity, and 5 had a history of admission to a NICU. A majority of patients (71.1%) reported a history of feeding difficulties before the episode leading to admission. Fourteen patients (36.8%) reported a history of developmental delays, with 23.7% having received birth-to-three services, and approximately one-third of the patients had a history of sensory concerns. Nine patients reported a history of trauma: medical trauma (n = 5), sexual abuse (n = 1), physical abuse (n = 2), and witnessing domestic violence (n = 3). Patients were mostly in middle school, and more than one-third received academic support (Table 2). More than one-quarter of patients reported a history of peer victimization.

Characterization of ARFID-related symptoms often varied among medical teams involved, as did terminology used to refer to these concerns. Only 31.6% of patients had a documented diagnosis of ARFID in their discharge notes, with the majority of patients receiving discharge diagnoses of food refusal, weight loss, restrictive eating, vomiting, and malnutrition (Table 3).

TABLE 3

Illness Course

Factors Related to Illness Coursen (%)
Medical diagnosis at discharge  
 Food refusal 14 (36.8) 
 Wt loss 14 (36.8) 
 ARFID 12 (31.6) 
 Malnutrition 8 (21.1) 
 Vomiting 6 (15.8) 
 Abdominal pain 5 (13.2) 
 Anorexia and/or restrictive eating 4 (10.5) 
 Gastroenteritis 4 (10.5) 
 Nausea 4 (10.5) 
 Constipation 3 (7.9) 
 Allergies 2 (5.3) 
 Cardiologic related 7 (18.4) 
 Other GI related 3 (7.9) 
 Infectious related 3 (7.9) 
 Pulmonary related 3 (7.9) 
 Dermatologic related 1 (2.6) 
Duration of feeding problems  
 <3, mo 14 (36.8) 
 4–6, mo 8 (21.0) 
 7–12, mo 6 (15.8) 
 >12, mo 9 (23.7) 
 Unknown 1 (2.6) 
Precipitating events  
 Infection and/or viral illness 10 (26.3) 
 Vomiting 7 (18.4) 
 Choking episode 2 (5.3) 
 Other 14 (36.8) 
Somatic symptom concerns reported  
 Pain (eg, muscle, stomachache, headache) 17 (44.7) 
 Vomiting or swallowing difficulties 9 (23.7) 
 Nausea 9 (21.1) 
 Constipation 6 (15.8) 
 Diarrhea 1 (2.6) 
 Other 6 (15.8) 
Other services consulted during admission  
 Nutrition 35 (92.1) 
 Gastroenterology 14 (36.8) 
 Feeding team 3 (7.9) 
 Pain service 2 (5.3) 
 Endocrinology 1 (2.6) 
 Otolaryngology 1 (2.6) 
 Neurology 1 (2.6) 
 Adolescent medicine (not admitting service) 1 (2.6) 
Co-occurring medical symptoms  
 Electrolyte abnormalities 28 (73.7) 
 Orthostatic instability 26 (68.4) 
 Bradycardia 20 (52.6) 
 Anemia 3 (7.9) 
Medical interventions during admission  
 Antacids 20 (52.6) 
 Laxatives 18 (47.4) 
 Antiemetics 16 (42.1) 
 Pain medication 11 (28.9) 
 Cyproheptadine 8 (21.1) 
 Other promotility agents 3 (7.9) 
 Antiinflammatories 2 (5.3) 
 Antihistamines 2 (5.3) 
 Antiflatulents 2 (5.3) 
 Dronabinol 1 (2.6) 
 Probiotics 1 (2.6) 
Factors Related to Illness Coursen (%)
Medical diagnosis at discharge  
 Food refusal 14 (36.8) 
 Wt loss 14 (36.8) 
 ARFID 12 (31.6) 
 Malnutrition 8 (21.1) 
 Vomiting 6 (15.8) 
 Abdominal pain 5 (13.2) 
 Anorexia and/or restrictive eating 4 (10.5) 
 Gastroenteritis 4 (10.5) 
 Nausea 4 (10.5) 
 Constipation 3 (7.9) 
 Allergies 2 (5.3) 
 Cardiologic related 7 (18.4) 
 Other GI related 3 (7.9) 
 Infectious related 3 (7.9) 
 Pulmonary related 3 (7.9) 
 Dermatologic related 1 (2.6) 
Duration of feeding problems  
 <3, mo 14 (36.8) 
 4–6, mo 8 (21.0) 
 7–12, mo 6 (15.8) 
 >12, mo 9 (23.7) 
 Unknown 1 (2.6) 
Precipitating events  
 Infection and/or viral illness 10 (26.3) 
 Vomiting 7 (18.4) 
 Choking episode 2 (5.3) 
 Other 14 (36.8) 
Somatic symptom concerns reported  
 Pain (eg, muscle, stomachache, headache) 17 (44.7) 
 Vomiting or swallowing difficulties 9 (23.7) 
 Nausea 9 (21.1) 
 Constipation 6 (15.8) 
 Diarrhea 1 (2.6) 
 Other 6 (15.8) 
Other services consulted during admission  
 Nutrition 35 (92.1) 
 Gastroenterology 14 (36.8) 
 Feeding team 3 (7.9) 
 Pain service 2 (5.3) 
 Endocrinology 1 (2.6) 
 Otolaryngology 1 (2.6) 
 Neurology 1 (2.6) 
 Adolescent medicine (not admitting service) 1 (2.6) 
Co-occurring medical symptoms  
 Electrolyte abnormalities 28 (73.7) 
 Orthostatic instability 26 (68.4) 
 Bradycardia 20 (52.6) 
 Anemia 3 (7.9) 
Medical interventions during admission  
 Antacids 20 (52.6) 
 Laxatives 18 (47.4) 
 Antiemetics 16 (42.1) 
 Pain medication 11 (28.9) 
 Cyproheptadine 8 (21.1) 
 Other promotility agents 3 (7.9) 
 Antiinflammatories 2 (5.3) 
 Antihistamines 2 (5.3) 
 Antiflatulents 2 (5.3) 
 Dronabinol 1 (2.6) 
 Probiotics 1 (2.6) 

The total percentage may be >100% when >1 factor is applied.

Most patients were admitted primarily for feeding difficulties (94.7%). One patient was admitted with scurvy and another for fever and abscess. Seven patients (18.4%) were admitted with a preexisting diagnosis of ARFID. The duration of feeding problems ranged from 2 days to 17.83 years, with a mean of 1.75 years (SD: 3.57 years; Table 3).

Approximately two-thirds of patients reported a variety of precipitating events contributing to their ARFID presentation (Table 3). The precipitating events categorized as “other” were wide ranging, including fears of nausea or choking (with or without witnessed event); medical insults, such as pancreatitis, surgical intervention, or diarrhea; and school education about meat processing. History of food allergies was noted in 10.5% of patients. Patients in the sample presented with a variety of somatic symptoms (Table 3).

Before the current admission, a majority of patients (81.6%) had other outpatient medical visits and/or tests, 28.9% had previous medical admissions, and 44.7% had visited other medical institutions, all for the same presenting problem. Of the patients who had previous outpatient care, 60.5% were seen for feeding concerns by their PCP and 37% had previously worked with a nutritionist. Past medical history was significant for various related illnesses (Table 4). Several patients had a history of gastrointestinal (GI) procedures (36.8%) or abdominal imaging (18.4%), with fewer patients having a history of gastrostomy tube placement (7.9%), nasojejunal or nasogastric tube placement (5.3%), or GI surgery (5.3%).

TABLE 4

Past Medical Diagnoses

Medical Diagnosesn (%)
Other GI related 23 (60.5) 
Allergies 13 (34.2) 
Pulmonary related 11 (28.9) 
Musculoskeletal or orthopedic related 5 (13.2) 
Cardiologic related 4 (10.5) 
Neurologic related 4 (10.5) 
Dermatologic related 3 (7.9) 
Genitourinary related 3 (7.9) 
Endocrine related 2 (5.3) 
Hematology and oncology related 2 (5.3) 
Infectious related 1 (2.6) 
Renal related 1 (2.6) 
Rheumatologic related 1 (2.6) 
No past medical diagnosis 5 (13.2) 
Medical Diagnosesn (%)
Other GI related 23 (60.5) 
Allergies 13 (34.2) 
Pulmonary related 11 (28.9) 
Musculoskeletal or orthopedic related 5 (13.2) 
Cardiologic related 4 (10.5) 
Neurologic related 4 (10.5) 
Dermatologic related 3 (7.9) 
Genitourinary related 3 (7.9) 
Endocrine related 2 (5.3) 
Hematology and oncology related 2 (5.3) 
Infectious related 1 (2.6) 
Renal related 1 (2.6) 
Rheumatologic related 1 (2.6) 
No past medical diagnosis 5 (13.2) 

The total percentage may be >100% because subjects may have received >1 past medical diagnosis.

Please see Table 5 for additional information about past and current psychiatric diagnoses. More than 65% of patients presented with some form of anxiety disorder. The majority of patients (71.1%) had a history of outpatient psychotherapy, almost 20% of patients had a history of day treatment (ie, partial hospitalization or intensive outpatient level of care), 10.5% had received inpatient psychiatric care, and 1 patient had a history of residential care. Approximately one-third of patients had previous treatment with psychiatric medication: antidepressant (16%), benzodiazepines (13%), antipsychotics (13%), and stimulants (10%).

TABLE 5

Past and Current Psychiatric Diagnoses

DiagnosesPast, n (%)Diagnosed by PCS, n (%)
Eating and feeding disorders   
 ARFID 7 (18.4) 38 (100) 
Anxiety disorders   
 Generalized anxiety disorder 5 (13.2) 13 (34.2) 
 Unspecified or other specified 3 (7.9) 10 (26.4) 
 Social anxiety disorder 0 (0.0) 2 (5.3) 
 Specific phobia 0 (0.0) 2 (5.3) 
 Separation anxiety disorder 0 (0.0) 1 (2.6) 
 Selective mutism 0 (0.0) 1 (2.6) 
 Panic disorder 1 (2.6) 0 (0.0) 
Obsessive-compulsive and related disorders   
 Obsessive-compulsive disorder 1 (2.6) 4 (10.5) 
 Other specified 1 (2.6) 1 (2.6) 
Somatic symptom and related disorders   
 Psychological factors affecting other medical conditions 1 (2.6) 3 (7.9) 
 Somatic symptom disorder 0 (0.0) 1 (2.6) 
Depressive disorders   
 Major depressive disorder 1 (2.6) 5 (13.2) 
 Unspecified or other specified 0 (0.0) 3 (7.9) 
Trauma- and stressor-related disorders   
 Posttraumatic stress disorder 1 (2.6) 3 (7.9) 
 Adjustment disorder 0 (0.0) 3 (7.9) 
 Other specified 0 (0.0) 1 (2.6) 
Neurodevelopmental disorders   
 Attention-deficit/hyperactivity disorder 7 (18.4) 6 (15.8) 
 Autism spectrum disorder 0 (0.0) 2 (5.3) 
 Motor disorder 1 (2.6) 1 (2.6) 
 Specific learning disorder 1 (2.6) 0 (0.0) 
Neurocognitive disorder 1 (2.6) 1 (2.6) 
Substance use disorder 1 (2.6) 1 (2.6) 
Disruptive, impulse-control, and conduct disorders   
 Oppositional defiant disorder 3 (7.9) 0 (0.0) 
DiagnosesPast, n (%)Diagnosed by PCS, n (%)
Eating and feeding disorders   
 ARFID 7 (18.4) 38 (100) 
Anxiety disorders   
 Generalized anxiety disorder 5 (13.2) 13 (34.2) 
 Unspecified or other specified 3 (7.9) 10 (26.4) 
 Social anxiety disorder 0 (0.0) 2 (5.3) 
 Specific phobia 0 (0.0) 2 (5.3) 
 Separation anxiety disorder 0 (0.0) 1 (2.6) 
 Selective mutism 0 (0.0) 1 (2.6) 
 Panic disorder 1 (2.6) 0 (0.0) 
Obsessive-compulsive and related disorders   
 Obsessive-compulsive disorder 1 (2.6) 4 (10.5) 
 Other specified 1 (2.6) 1 (2.6) 
Somatic symptom and related disorders   
 Psychological factors affecting other medical conditions 1 (2.6) 3 (7.9) 
 Somatic symptom disorder 0 (0.0) 1 (2.6) 
Depressive disorders   
 Major depressive disorder 1 (2.6) 5 (13.2) 
 Unspecified or other specified 0 (0.0) 3 (7.9) 
Trauma- and stressor-related disorders   
 Posttraumatic stress disorder 1 (2.6) 3 (7.9) 
 Adjustment disorder 0 (0.0) 3 (7.9) 
 Other specified 0 (0.0) 1 (2.6) 
Neurodevelopmental disorders   
 Attention-deficit/hyperactivity disorder 7 (18.4) 6 (15.8) 
 Autism spectrum disorder 0 (0.0) 2 (5.3) 
 Motor disorder 1 (2.6) 1 (2.6) 
 Specific learning disorder 1 (2.6) 0 (0.0) 
Neurocognitive disorder 1 (2.6) 1 (2.6) 
Substance use disorder 1 (2.6) 1 (2.6) 
Disruptive, impulse-control, and conduct disorders   
 Oppositional defiant disorder 3 (7.9) 0 (0.0) 

The total percentage may not equal 100% because subjects may have received >1 psychiatric diagnosis.

Patients were most frequently admitted to adolescent medicine and general pediatrics (Table 2). For most patients, the psychiatric consultation occurred on the second day of admission (range: first to eighth day of admission), and average length of stay was 8 days (SD: 6.25; range: 2–32 days). Patients averaged 5 PCS sessions (range: 1–32 sessions) during their medical admission. A number of other services were often consulted (Table 3). Patients presented with a number of common co-occurring medical concerns (Table 3). Almost half of patients (47.4%) required enteral feeds (ie, via nasogastric, nasojejunal, or gastrostomy tube) during their admission.

Regarding medical testing during admission, 16 patients (42.1%) received imaging procedures, 5 (13.2%) patients had GI-related scopes, 4 (10.5%) patients had swallow studies, and 3 (7.9%) patients had head imaging; all of these imaging procedures revealed normal physiologic functioning. Thirteen patients had abdominal imaging solely for evaluating placement of nasogastric or gastrostomy tubes.

In-hospital nutritionists calculated patient weight compared with that of peers (z score) as well as the ideal patient weight (percentage mean body weight [%MBW]). The mean z score of body mass was −1.66 on admission (SD: 1.61) and −1.45 on discharge (SD: 1.38). The average %MBW on admission was 85.9% and improved to 87.6% on discharge. At discharge, 19 patients had an increase in weight, 8 had weight that remained unchanged, and 11 patients actually had a decrease in weight over the course of their admission. Ten of these 11 patients had a slight decrease in weight that was not deemed to impact their medical stability for discharge. One patient had significant weight loss but was discharged from the hospital without a comprehensive multidisciplinary plan.

Medical treatments of ARFID and related symptoms are reviewed in Table 3. Twenty-four patients (63.2%) were prescribed psychiatric medications to help with the feeding process and comorbid psychiatric symptoms: benzodiazepines (55.3%), antidepressants (13.2%), and antipsychotics (13.2%). A behavioral feeding plan was implemented for 60.5% of patients. Relaxation strategies, such as guided imagery, were taught to 42.1% of patients. Ten patients required a 1:1 care assistant during their admission: 7 had this type of support during meals only, and 3 had this support at all times (day and night) because of other safety concerns. Most often, this type of support was implemented on the first day of admission (7 patients). Team meetings including family members, the primary medical team, and relevant consultative services were coordinated in 21.1% of cases, with some patients (10.5%) requiring multiple team meetings during their admission.

Twenty-three patients (60.5%) were discharged to routine outpatient care, with 4 patients (10.5%) discharged to preexisting outpatient services and 19 patients (50.0%) discharged with referrals to outpatient individual psychotherapy. Seven patients (23.7%) were discharged to ED outpatient programs, such as partial hospitalization or intensive outpatient programs, and 2 patients were referred to general psychiatry day treatment programs. Four patients (10.5%) were discharged to inpatient level of care (3 to ED facilities and 1 to a general psychiatric hospital). Of note, 2 patients (5.3%) were discharged from the hospital with guidance to follow-up in an outpatient medical clinic as needed, without psychological referrals.

ARFID was formally introduced as a diagnosis in 2013; however, consistent diagnosis and treatment of this disorder has not been immediate. In the current study, >80% of patients had received outpatient evaluations for the same problem, but only 18.4% had previously been diagnosed with ARFID. This appears consistent with previous studies revealing limited familiarity with and diagnosis of ARFID among pediatricians and pediatric subspecialists in Canada.9  This finding suggests a continued need for further medical education to PCPs and inpatient hospitalist and other subspecialty (eg, adolescent medicine, psychiatry, GI) providers regarding the diagnosis and treatment of ARFID. Despite the average length of illness duration being almost 2 years in our sample and ARFID existing as a diagnosis 2 years before the study period, the majority of these patients were not recognized as such in their communities. Even more relevant, <40% of our study sample had ARFID included in their medical discharge diagnoses, despite the fact that all the patients had received an ARFID diagnosis by PCS during their hospitalization. Although diagnoses such as food refusal and malnutrition characterize some aspects of the clinical presentation, they do not fully convey the clinical presentation to multidisciplinary hospital providers or to outpatient providers (eg, PCPs and nutritionists) who will continue to care for these patients in the community. Increasing familiarity in the community about ARFID would likely lead to more timely and accurate diagnosis, more opportunities for appropriate treatment, and decreased risk of hospital admission.

Further complicating the diagnostic picture, patients with ARFID are more likely to experience comorbid medical conditions, including GI symptoms and food allergies.13  The majority of patients in our sample had comorbid GI conditions and GI-related somatic symptoms, although they were infrequently admitted to the GI service. The extent of medical evaluation and invasive procedures conducted in our sample was notable in highlighting the risk of high health care use in effort to rule in or rule out the feeding disorder, especially because findings were negative for this sample group. Consistent with previous literature,13  our patient sample had a long duration of illness before diagnosis. Although a triggering event such as infection or choking was often identified, close to one-third of the sample had no identified acute trigger. Patients with ARFID are often described as picky eaters in early childhood,13  with a tendency for chronic weight issues. Interestingly, the average %MBW for our sample was comparable with that of the broader ARFID population sampled by Fisher et al,13  suggesting that %MBW may not be the sole indicator of medical necessity for hospitalization of patients with ARFID. Other acute medical findings seen in the majority of our patients included electrolyte abnormalities, orthostatic instability, and bradycardia. Regardless, patients in this study were on average significantly underweight (>1.5 SDs below the mean body weight), suggesting a need for further consideration of how weight stability or refeeding occurs during medical hospitalizations.

To the best of our knowledge, this is the first study in which patients diagnosed with ARFID by a PCS within a pediatric inpatient setting are characterized. Similar to earlier efforts to characterize patients presenting with ARFID to an ED day treatment program,6  we found higher rates of developmental delays, learning disorders, and patients requiring academic support in our sample than usually seen with AN and bulimia nervosa, suggesting potentially unique vulnerabilities in patients with ARFID. When comparing our findings with those in studies of patients with ARFID in other settings,13,14  our sample had a similar younger mean age and higher male population (30%) than typically seen in patients with AN and bulimia nervosa. Generalized anxiety disorder was identified in approximately one-third of our sample, whereas major depressive disorder was diagnosed in 13.2% of patients, only slightly higher than rates seen in the broader ARFID population characterized by Fisher et al.13  Given previous research indicating that early involvement of psychiatry consultation in the pediatric hospital setting reduces length of stay and cost of admission for patients with comorbid medical and psychiatric conditions,15  a clinical pathway that implements this approach for patients with ARFID could improve patient and hospital outcomes. The majority of patients in our sample were discharged to outpatient levels of psychiatric and medical care, with half of the patients receiving referrals for outpatient individual psychotherapy. Unfortunately, given the retrospective nature of our study, data were not collected beyond discharge from the medical admission. As such, we were unable to gather information about patient outcomes after hospitalization, including whether patients and families followed-up with treatment referrals.

There are inherent limitations the authors recognize within the current study. The nature of retrospective research design limits data analysis, especially with missing data for some variables of interest in the broader field of ARFID research. There is also inherent sampling bias because all patients included in this study were referred to PCS and diagnosed with ARFID by the end of their admission, whereas patients hospitalized with ARFID presentations but not referred to PCS were not included. Patients in this study represent a relatively small sample size within 1 medical setting, potentially limiting generalizability to other institutions. Finally, our sample did not include patients with severe developmental disorders or genetic syndromes known to be associated with difficulty with eating and nutrition, hence limiting comparison to the literature on feeding struggles in these patient populations.

Future research is needed to continue to guide best practice in the management of ARFID in the acute pediatric care setting. The small volume of patients referred to PCS for ARFID concerns over almost 2 years in the current study suggests an area for future study. Because of the significant variability in the way medical providers documented their observations in the medical records and their minimal use of the ARFID diagnosis, an artifact of hospital medical versus psychiatric billing data, we were unable to identify cases of ARFID not referred to PCS. This phenomenon may reflect a lack of knowledge about and poor identification of this psychiatric diagnosis by providers in both outpatient and inpatient medical settings, highlighting the need to educate pediatric providers not only about ARFID but also about the multidisciplinary treatment approach, including psychiatry involvement. Additional studies are needed to examine referral patterns and prevalence of ARFID among multiple pediatric institutions. This study is a first step in guiding the future development of a clinical pathway for the assessment and management of patients presenting with ARFID to pediatric hospital settings. Standardization of practice in the evaluation of ARFID, reduction of unnecessary medical evaluations, and a better understanding of the implications of medical interventions, such as enteral feeds and reasonable discharge criteria (such as whether demonstrated weight gain before discharge leads to better outcomes), also warrant future study to support more efficient and effective health care use. Although the assessment of health care use is a potential area for reducing cost or invasive interventions in patients with ARFID, these procedures and tests may also be necessary to rule out medical conditions in a portion of patients. In future studies, researchers should also more rigorously evaluate the benefit and risk of psychotropic agents to aid in inpatient ARFID treatment. It will be helpful to better understand the relationship between biopsychosocial risk factors and subsequent development of ARFID. For example, we found several areas that were not consistently assessed during psychiatric evaluations of patients with ARFID in our sample, including temperament, early feeding history, and history of breastfeeding. Understanding the relationship between these early childhood factors and current ARFID etiology may guide more focused clinical interview and interventions. Furthermore, gathering consistent information on ARFID populations may, in turn, help to inform future research, specifically predictive, risk, and protective factors, ultimately enriching the understanding of this population and potentially contributing to more accurate diagnosis and stronger evidence-based treatment interventions. Finally, multidisciplinary provider education is needed to support awareness of ARFID across all levels of pediatric care.

We appreciate the contributions of Dr Anna Mayo in data collection and preliminary analyses.

Dr Tsang conceptualized and designed the study, drafted the initial manuscript, and participated in data collection and analysis; Dr Hayes participated in data analysis, contributed to the initial manuscript draft, and reviewed and revised the manuscript; Drs Bujoreanu and Samsel participated in data collection and analysis and reviewed and revised the manuscript; Dr Ibeziako participated in study design and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted.

FUNDING: No external funding.

1
Carter
BD
,
Kronenberger
WG
,
Baker
J
, et al
.
Inpatient pediatric consultation-liaison: a case-controlled study
.
J Pediatr Psychol
.
2003
;
28
(
6
):
423
432
2
American Psychiatric Association
.
Diagnostic and Statistical Manual of Mental Disorders
. 5th ed.
Washington, DC
:
American Psychiatric Association
;
2004
:
334
338
3
Norris
ML
,
Spettigue
WJ
,
Katzman
DK
.
Update on eating disorders: current perspectives on avoidant/restrictive food intake disorder in children and youth
.
Neuropsychiatr Dis Treat
.
2016
;
12
:
213
218
4
Lock
J
.
An update on evidence-based psychosocial treatments for eating disorders in children and adolescents
.
J Clin Child Adolesc Psychol
.
2015
;
44
(
5
):
707
721
5
Norris
ML
,
Katzman
DK
.
Change is never easy, but it is possible: reflections on avoidant/restrictive food intake disorder two years after its introduction in the DSM-5
.
J Adolesc Health
.
2015
;
57
(
1
):
8
9
6
Nicely
TA
,
Lane-Loney
S
,
Masciulli
E
,
Hollenbeak
CS
,
Ornstein
RM
.
Prevalence and characteristics of avoidant/restrictive food intake disorder in a cohort of young patients in day treatment for eating disorders
.
J Eat Disord
.
2014
;
2
(
1
):
21
7
Norris
ML
,
Robinson
A
,
Obeid
N
,
Harrison
M
,
Spettigue
W
,
Henderson
K
.
Exploring avoidant/restrictive food intake disorder in eating disordered patients: a descriptive study
.
Int J Eat Disord
.
2014
;
47
(
5
):
495
499
8
Strandjord
SE
,
Sieke
EH
,
Richmond
M
,
Rome
ES
.
Avoidant/restrictive food intake disorder: illness and hospital course in patients hospitalized for nutritional insufficiency
.
J Adolesc Health
.
2015
;
57
(
6
):
673
678
9
Katzman
DK
,
Stevens
K
,
Norris
M
.
Redefining feeding and eating disorders: what is avoidant/restrictive food intake disorder?
Paediatr Child Health
.
2014
;
19
(
8
):
445
446
10
Williams
KE
,
Riegel
K
,
Gibbons
B
,
Field
DG
.
Intensive behavioral treatment for severe feeding problems: a cost-effective alternative to tube feeding?
J Dev Phys Disabil
.
2007
;
19
(
3
):
227
235
11
Obeid
JS
,
McGraw
CA
,
Minor
BL
, et al
.
Procurement of shared data instruments for Research Electronic Data Capture (REDCap)
.
J Biomed Inform
.
2013
;
46
(
2
):
259
265
12
IBM SPSS Statistics for Windows [computer program]. Version 24.0. Armonk, NY: IBM Corporation; 2016
13
Fisher
MM
,
Rosen
DS
,
Ornstein
RM
, et al
.
Characteristics of avoidant/restrictive food intake disorder in children and adolescents: a “new disorder” in DSM-5
.
J Adolesc Health
.
2014
;
55
(
1
):
49
52
14
Eddy
KT
,
Thomas
JJ
,
Hastings
E
, et al
.
Prevalence of DSM-5 avoidant/restrictive food intake disorder in a pediatric gastroenterology healthcare network
.
Int J Eat Disord
.
2015
;
48
(
5
):
464
470
15
Bujoreanu
S
,
White
MT
,
Gerber
B
,
Ibeziako
P
.
Effect of timing of psychiatry consultation on length of pediatric hospitalization and hospital charges
.
Hosp Pediatr
.
2015
;
5
(
5
):
269
275

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: Dr Ibeziako receives grant support from the National Heart, Lung, and Blood Institute. All authors have indicated they have no financial relationships relevant to this article to disclose.