Physicians certified by the American Board of Obesity Medicine provide evidence-based care

Summary

Our objective was to determine the clinical services offered by American Board of Obesity Medicine (ABOM) Diplomates and whether guideline concordant services varied by clinical practice attributes. We conducted a cross-sectional analysis of the 2019 ABOM Diplomate survey (response rate 19.2%). Respondents (n = 494) self-reported services offered: nutrition, exercise, mental health, minimally invasive bariatric procedures, perioperative bariatric surgical care and FDA-approved anti-obesity medications. We graded concordance of services offered with three evidence-based obesity guidelines, and then conducted bivariate analyses comparing concordance by practice attributes. Most responding ABOM Diplomates offered nutrition (90.1%), exercise (67.8%) and mental health (76.7%). Few offered minimally invasive procedures (24.3%), and most provided perioperative surgical care (63.0%). Most (83.4%) prescribed FDA-approved medications—typically both short- and long-term agents (70.9%). Few Diplomates had low concordance with the American Heart Association/American College of Cardiology/The Obesity Society (AHA/ACC/TOS) guidelines (24.7%). Those who managed more obesity-related conditions and endorsed AHA/ACC/TOS guideline use had higher concordance with these recommendations. No differences in guideline concordance existed by population, clinical effort or location. We found similar findings regarding concordance with ) American Association of Clinical Endocrinologists/American College of Endocrinology and Obesity Medicine Association guidelines. In conclusion, most responding ABOM Diplomates offer evidence-based obesity medicine services. Clinicians may therefore have increased confidence in patient receipt of evidence-based care when referring to an ABOM Diplomate.

1 |. INTRODUCTION

The prevalence of obesity in adults has increased in the United States to 42.4%.¹ Obesity has been associated with an increased risk of death and chronic conditions including cardiovascular disease, type 2 diabetes mellitus and certain cancers.² In 2013, the American Medical Association recognized obesity as a complex, chronic disease that requires medical attention.³ However, research has found that the rate of obesity counselling among practicing physicians remains low.^4,5

While multiple factors likely contribute to the low rates, physicians receive little education or training to appropriately address obesity in the clinical setting. Competencies taught in medical school and internal medicine residency have identified obesity as major knowledge gaps.^6–9 Practicing primary care physicians have repeatedly identified insufficient training about obesity, as well as low confidence and self-efficacy with respect to weight management skills.^10–12 Despite these limitations, prior research has found that patients want their physician involved in their weight-loss plan,¹³ and patients have greater weight loss success in this scenario.¹⁴ This evidence indicates a role for physicians in the care of patients with obesity, but also highlights the need for education and training on obesity.

Certification from the American Board of Obesity Medicine (ABOM) began to address this gap in education and training in obesity in 2011. The organization established requirements regarding the pursuit of additional obesity education for physicians, and testing procedures to ensure that certified physicians have achieved competency in the science of and evidence-based treatments for obesity.¹⁵ To date, little information about the 2656 physicians who are ABOM Diplomates as of 2018 has been described.¹⁶ A recent publication reported their demographics, medical specialty (eg, internal medicine, family medicine), years in practice and practice region.^15,16 Another study reported on the geographic distribution of ABOM Diplomates,¹⁷ which found that all United States had an ABOM-certified adult medicine physician, although disparities existed in Diplomate availability relative to state-level obesity prevalence. This study also identified fewer paediatric medicine ABOM Diplomates were available. To date, little is known about the clinical practice habits and services offered by ABOM Diplomates.

In this study, our first aim was to describe the clinical obesity services provided by current ABOM-certified physicians. We hypothesized that ABOM Diplomates would offer a range of clinical obesity services. We explored the frequency of use of medications not approved by the Food and Drug Administration (FDA) for obesity treatment—off-label use—and physicians’ reasons for this use. We also examined whether differences in services offered differed by whether physicians treated mostly children or adults. Our second aim was to compare whether evidence-based guideline concordant care differed by clinical practice characteristics. We hypothesized that many Diplomates would offer services concordant with evidence-based weight management guidelines, and guideline-concordant care might differ by management of population treated, comorbid conditions, clinical effort, practice location, health system affiliation, reimbursement model and self-reported guideline usage.

2 |. METHODS

2.1 |. Study design and sample

We analysed cross-sectional data from the 2019 ABOM Diplomate survey. The survey was designed by the ABOM, and the study team added questions to and refined the survey instrument prior to the distribution by ABOM. Appendix S1 contains the survey instrument. The ABOM notified all Diplomates about the survey via email, and physicians were offered an incentive to participate (entry into a lottery to receive a year of annual dues for either the Obesity Medicine Association or The Obesity Society). All Diplomates received an email from the Chairman of the ABOM to expect a survey from ABOM regarding their professional activities, along with three reminder emails to complete the survey. The ABOM provided a deidentified dataset to the study team. Information on demographics and location of the respondents were not available in the data. The response rate for the survey was 19.2% (n = 509), which is similar to other online surveys of physicians.¹⁸ For our analysis, we excluded Diplomates who did not practice clinically (n = 11) or did not report whether they provided clinical care (n = 4) (analytic sample = 494). The Johns Hopkins University School of Medicine Institutional Review Board exempted this study as not human subjects research.

2.2 |. Clinical obesity services

Our primary dependent variables were self-reported clinical obesity services, which included nutrition, meal replacements, physical activity, behavioural services, mental health services, minimally invasive procedures/devices, perioperative bariatric surgical care, anti-obesity medications approved by the Food and Drug Administration (FDA), off-label medications and supplements/other products. Table 1 provides details on the survey questions and responses used to create these variables. Respondents were also asked to identify reasons for any off-label medication use (Q15 in Appendix S1).

TABLE 1.

Overview of variables representing clinical obesity services

Variable	Responses for relevant survey questions
Nutrition	Dichotomized as “nutrition services” if any of the following were checked (vs “no nutrition services”): • Dietician/nutritionist works in the practice (Q5) • Nutritional assessment (Q11) • Nutrition and diet counselling (Q11) • Physician supervised meal replacement program (Q11) • Specific calorie-deficit recommended (Q12) • Yes to selling meal replacement products (Q25)
Meal replacements	Dichotomized as “meal replacements” if any of the following were checked (vs “no meal replacements”): • Physician supervised meal replacement program (Q11) • Yes to selling meal replacement products (Q25)
Physical activity	Dichotomized as “physical activity services” if any of the following were checked (vs “no physical activity services”): • Exercise specialist works in the practice (Q5) • Fitness evaluation (Q11) • Exercise counselling (Q11)
Behavioural services	Dichotomized as “behavioural services” if any of the following were checked (vs “no behavioural services”): • Psychologist works in the practice (Q5) • Behaviour counselling (Q11)
Mental health services	Dichotomized as “mental health services” if any of the following were checked (vs “no mental health services”): • Psychologist works in the practice (Q5) • Psychological evaluation (Q11) • Binge eating disorder monitored/treated in clinic (Q18) • Night eating syndrome monitored/treated in clinic (Q18)
Minimally invasive procedures or devices	Dichotomized as “minimally invasive procedures” if any of the following were checked (vs “no minimally invasive procedures”): • Minimally invasive and/or endoscopic procedures and devices is checked (Q11) • Complications from minimally invasive and/or endoscopic procedures and devices (Q17)
Perioperative bariatric surgical care	Dichotomized as “bariatric surgical care” if any of the following were checked (vs “no bariatric surgical care”): • Bariatric surgeon works in the practice (Q5) • Pre-op clearance for bariatric surgery (Q11) • Bariatric surgery (Q11) • Postoperative bariatric surgery care (Q11) • Complications from bariatric surgery (Q17)
FDA-approved anti-obesity medications	Categorized as: • “Prescribes medications” if any FDA-approved medications were checked (Q13) • “Does not prescribe” if “I do not prescribe medications” was checked (Q13) • “Not disclosed” if no answer selected (Q13) Categorized also as: • “Short-term medications only” if only benzphetamine, diethylpropion, phendimetrazine or phentermine checked • “Long-term medications only” if only orlistat, lorcaserin, phentermine/topiramate, naltrexone/bupropion, or liraglutide checked • “Short- & long-term medications” if at least one short- and one long-term medication checked • “None/Not disclosed” if “I do not prescribe medications” was checked or no answer selected
Off-label medication use	Dichotomized as “off-label medications used” if any of the following were checked (vs ‘no off-label medications used): • Bupropion, naltrexone, topiramate, metformin, GLP-1 agonists, pramlintide, lisdexamfetamine, SSRIs, SNRIs, zonisamide, diuretics, hCG, vitamin B12, other (Q14) Dichotomozied also as: • “Components of FDA-approved medications” if bupropion, naltrexone, or topiramate checked • “Diabetes medications” if metformin, GLP-1 agonists, pramlintide, or other specified SGLT2 inhibitors or DPP-IV inhibitors checked • “Psychiatric medications” if lisdexamfetamine, SSRIs, SNRIs, or zonisamide checked • “Other” if diuretics, hCG, Vitamin B12, or other (non-diabetic medication) checked
Supplements/other products	Dichotomized as “supplements used” if any of the following were checked (vs “no supplements used”): • hCG, vitamin B12, or diuretics used off-label for weight loss (Q14) • Yes to supplements/nutraceuticals used (Q16)

Abbreviations: DPP, dipeptidyl peptidase; FDA, Food and Drug Administration; GLP, glucagon like peptide; hCG, human chorionic gonadotropin, SGLT, sodium glucose transport protein; SNRI, serotonin norepinephrine reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor.

2.3 |. Concordance of clinical services with guidelines

We created variables to represent concordance of services offered with the following obesity care guidelines: (a) American Heart Association/American College of Cardiology/The Obesity Society (AHA/ACC/TOS) Adult Weight Management Guidelines,¹⁹ (b) American Association of Clinical Endocrinologists/American College of Endocrinology (AACE/ACE) Comprehensive Clinical Practice Guidelines for the Medical Care of Patients with Obesity²⁰ and (c) Obesity Medicine Association (OMA) Obesity Algorithm.²¹ One study team member (K. A. G.) proposed a strategy to align survey measures with each element of the guidelines, and consensus on the approach was reached during discussion with other physician team members (E. P. W. and F. C. S.). Supplemental Tables S1–S3 provide details on the measures used to assess concordance with each set of guidelines.

We used five key elements to assess guideline concordance with the AHA/ACC/TOS recommendations¹⁹: high intensity of services (defined as ≥2 visits per month) (offered), nutrition (offered), physical activity (offered), behavioural services (offered) and supplement use (not offered), which is based on a previously published strategy.²² We graded services as “high” if all five criteria were met; “moderate” if the practice met at least three of the five criteria; and “low” if the practice met less than three of the criteria.

We used six key elements to assess guideline concordance with the AACE/ACE guidelines²⁰: medical evaluation (performed), meal plan (offered), physical activity (offered), behavioural services (offered), long-term FDA-approved anti-obesity medications (offered) and bariatric surgical services (available). We graded services as “high” if all six criteria are met; “moderate” if the practice met at least four of the six criteria; and “low” if the practice met less than four of the criteria. Ideally, we would have also included an assessment of expertise of the interdisciplinary team or physician; however, the survey did not capture this information.

We used seven key elements to assess guideline concordance with the OMA recommendations²¹: medical evaluation (performed), nutrition (offered), physical activity (offered), behavioural services (offered), any FDA-approved anti-obesity medications (offered), supplement use (not offered) and bariatric procedures (available). We graded services as “high” if all seven criteria are met; “moderate” if the practice met at least four of the seven criteria; and “low” if the practice met less than four of the criteria. Ideally, we would have also included an assessment of the use of motivational interviewing; however, this was not available in the survey.

2.4 |. Clinical practice attributes

Our independent variables were self-reported clinical practice characteristics, which included self-reported patient panel attributes, clinic characteristics and clinical practice behaviours.

Patient panel attributes included patient age groups and comorbidities managed. Participants reported the percentage of patients seen for obesity management who were children (2–10 years), adolescents (12–17 years) and adults (18+ years) (Q3 in Appendix S1). We identified physicians who saw only adults, a combination of adults and children/adolescents, or only children/adolescents. We also dichotomized this variable to identify physicians who saw majority (>50%) children/adolescents vs not. Respondents identified the comorbid conditions that they monitored/treated in patients with obesity (Q18). We summed the total number of comorbidities into a total count for each physician.

Clinic characteristics included number of clinical sessions, practice location, health system affiliation and reimbursement. Participants reported the average number of sessions per week spent in clinic where one session was defined as 4 hours (ie, morning or afternoon) (Q1). We categorized responses as one to two sessions per week, three to five sessions per week or six or more sessions per week. Respondents identified the type of community that they practiced in (Q4)—urban, suburban, rural or other—and reported whether their clinic was affiliated with a health system (Q6). Participants reported the reimbursement model their clinic used (Q21)—fee-for-service, insurance, both or other.

Clinical practice behaviours included self-reported use of several weight management guidelines including: (a) AHA/ACC/TOS adult weight management guidelines,¹⁹ (b) AACE/ACE Comprehensive Clinical Practice Guidelines for the Medical Care of Patients with Obesity,²⁰ (c) OMA Obesity Algorithm²¹ and (d) Pharmacologic Management of Obesity: An Endocrine Society Clinical Practice Guideline²³(Q20). Respondents also had the option to specify other guidelines that they used, and some physicians noted use of guidelines by the American Academy of Paediatrics. We examined responses individually, and created a dichotomous variables if any guideline use was endorsed vs not.

2.5 |. Statistical analysis

Our first aim was a descriptive analysis, and we calculated proportions using Chi square tests for the overall sample. Among those who reported offering certain services, we described the frequency of specific elements that comprised this area (eg, dietician onsite among those who offer nutritional services in clinic). We also examined these outcomes stratified by patient population treated-physicians who predominantly treated children/adolescents vs those who predominantly treated adults.

For our second aim, we conducted bivariate analyses using Chi square tests or ANOVA, as appropriate, to examine whether guideline concordance of services offered differed by clinical practice attributes. Among those who received low scores for concordance for each of the guidelines, we characterized the elements that contributed to their downgraded ratings.

We did not conduct multivariable logistic regressions, as the data did not contain key physician attributes (eg, years in practice) shown to be associated with differences in obesity practice habits among physicians.¹²

3 |. RESULTS

Among the 494 responding Diplomates in our analysis, few focused on treating obesity among majority of children/adolescents (8.7%). Table 2 describes the clinical practice attributes in our sample.

TABLE 2.

Clinical practice characteristics of American Board of Obesity Medicine (ABOM) Diplomates

	ABOM Diplomates (n = 494)
Patient panel attributes
Patient age groups
Adults only	52.3%
Adults and children/adolescents	44.0%
Children/adolescents only	3.7%
Children/adolescents comprise majority (>50%) of patients	8.7%
Comorbid conditions managed
Diabetes mellitus	89.5%
Hypertension	86.4%
Hyperlipidemia	86.2%
Non-alcoholic fatty liver disease	81.6%
Obstructive sleep apnea	78.1%
Gastroesophageal reflux disease	75.3%
Polycystic ovarian syndrome	71.9%
Mood disorders	71.5%
Binge eating disorder	66.8%
Constipation	64.4%
Chronic kidney disease	59.7%
Night eating syndrome	55.5%
Stress urinary incontinence	38.7%
Infertility	33.2%
Mean number of chronic conditions managed (SD)	10.2 (4.6)
Clinic characteristics
Mean number of half-day clinic sessions per week (SD)	6.7 (SD 2.8)
Clinical sessions groups
1–2 sessions per week	10.3%
3–5 sessions per week	21.7%
≥6 sessions per week	68.0%
Practice location
Urban	37.5%
Suburban	41.5%
Rural	11.3%
Multiple	9.7%
Health system affiliation	61.3%
Reimbursement model
Fee-for-service only	11.3%
Insurance and fee-for-service	33.6%
Insurance only	42.3%
Not disclosed/other	12.8%
Interdisciplinary team modela	57.3%
High intensity program offeredb	18.8%
Clinical practice behaviours
Guideline usage
AHA/ACC/TOS	30.8%
AACE/ACE	33.4%
OMA	65.6%
Endocrine Societyc	31.6%
Any guideline usedd	82.6%

Abbreviations: AACE/ACE, American Association of Clinical Endocrinologists/American College of Endocrinology; ABOM, American Board of Obesity Medicine; AHA/ACC/TOS, American Heart Association/American College of Cardiology/The Obesity Society; OMA, Obesity Medicine Association.

^aInterdisciplinary team defined as a clinic with other healthcare professionals (eg, registered dietician, psychologist, exercise specialist) delivering care along with a physician.

^bHigh intensity defined as offering ≥ 2 sessions per month.

^cPharmacologic Management of Obesity: An Endocrine Society Clinical Practice Guideline.

^dPhysicians who endorsed use of any guideline for weight management, which included other guidelines, such as American Academy of Paediatrics.

Diplomates commonly monitored and treated chronic conditions associated with obesity (mean 10.2 conditions). On average, these physicians’ clinical effort was high—mean number of clinic sessions per week was 6.7 (SD 2.8). Most clinics combined care from physicians with other healthcare professionals (57.3%). Few offered a high-intensity program (18.8%).

3.1 |. Clinical services provided by ABOM Diplomates

Nearly all responding Diplomates reported offering nutrition services in their practice (90.1%) (Table 3), which was typically accomplished with nutritional assessments (82.0%) and counselling (89.0%) (Supplemental Table S4). About half of respondents reported that their practice offered meal replacements (47.0%). Physical activity and behavioural services were included less often (67.8% and 66.2%, respectively)—and typically accomplished through exercise counselling (94.0%) and behavioural counselling (91.4%). Most Diplomates offered mental health services in their practice (76.7%) (Table 3), and monitoring/treating binge eating disorder was common (87.1%) (Supplemental Table S4).

TABLE 3.

Self-reported clinical obesity services offered by American Board of Obesity Medicine (ABOM) Diplomates

	ABOM Diplomates (n = 494)	Treats majority adults (n = 450)	Treats majority children (n = 43)	P-valuea
Clinical services
Nutrition	90.1%	90.4%	86.1%	.36
Meal replacements	47.0%	49.3%	23.3%	<.01
Physical activity	67.8%	66.9%	76.7%	.19
Behavioural services	66.2%	66.0%	69.8%	.62
Mental health services	76.7%	76.7%	79.1%	.72
Minimally invasive bariatric procedures/devices	24.3%	26.2%	4.7%	<.01
Perioperative bariatric surgical care	63.0%	65.1%	39.5%	<.01
Medications and supplements
FDA-approved anti-obesity medications				<.01
Prescribes	83.4%	86.2%	53.5%
Does not prescribe	7.5%	4.7%	37.2%
Not disclosed	9.1%	9.1%	9.3%
Types of FDA-approved medications used				<.01
Short-term medications only	5.5%	4.9%	11.6%
Short- and long-term medications	70.9%	75.3%	23.3%
Long-term medications only	7.1%	6.0%	18.6%
None/Not disclosed	16.6%	13.8%	46.5%
Off-label medications for weight loss				.02
Prescribes	76.9%	78.0%	67.4%
Does not prescribe	14.2%	12.7%	27.9%
Not disclosed	8.9%	9.3%	4.7%
Supplements/other products used				<.01
No	66.2%	64.0%	88.4%
Yes	25.1%	26.9%	7.0%
Not disclosed	8.7%	9.1%	4.7%

Abbreviations: ABOM, American Board of Obesity Medicine; FDA, Food and Drug Administration.

^aP-values calculated using Chi square tests comparing Diplomates who reported treating a majority adults as compared to those treating a majority children/adolescents.

Few responding Diplomates offered care related to minimally invasive procedures (24.3%) (Table 3), but perioperative bariatric surgical care occurred more commonly (63.0%). Most of these physicians provide pre- and post-operative bariatric care (71.1% and 69.6%, respectively) (Supplemental Table S4).

Most responding Diplomates reported prescribing FDA-approved anti-obesity medications (83.4%) (Table 3), and most prescribed both short- and long-term medications (70.9%). Supplemental Table S5 displays the frequency of specific medications within these groups. Few physicians specifically stated that they did not prescribe these medications (7.5%). Off-label medication use was common (76.9%). Among Diplomates that prescribed off-label medications, 89.2% used components of FDA-approved anti-obesity medications, 91.1% medications to treat diabetes mellitus that are associated with weight loss, 47.4% psychiatric medications and 19.7% other medications. Supplemental Table S6 displays the frequency of specific medications among Diplomates who prescribed off-label medications. Among Diplomates who used off-label medications, the following reasons for their use were cited: lack of insurance coverage for FDA-approved medications (71.3%), medication cost (77.1%), contraindication for FDA-approved medications (33.7%), better outcomes (13.4%) and fewer side effects (10.3%). Few Diplomates reported using supplements (25.1%), and when used, vitamin B12 (53.2%) was the most common product reported (Supplemental Table S5).

We found several differences in clinical services by population treated when comparing Diplomates who treat majority adults vs majority children/adolescents (Table 3). Diplomates who treated majority children/adolescents were significantly less likely to use meal replacements, minimally invasive procedures/devices and perioperative bariatric care as compared to Diplomates who treat majority adults. These paediatric-focused Diplomates were also less likely to use anti-obesity medications, off-label medication, or supplements/ other products as compared to adult medicine Diplomates.

3.2 |. Concordance with AHA/ACC/TOS guidelines and differences by practice characteristics

Overall, 30.8% of responding Diplomates reported using the 2013 AHA/ACC/TOS guidelines (Table 2). Regarding services delivery concordance with these guidelines, 24.7% of Diplomates had low concordance, 65.6% moderate concordance and 9.7% had high concordance. Among Diplomates graded as low (n = 122), they were most frequently downgraded for lack of high-intensity services (96.7%), physical activity (86.1%) and behavioural services (90.2%). Lack of nutrition (36.9%) or use of supplements (29.5%) occurred less frequently.

We found a significant difference in AHA/ACC/TOS guideline concordance by number of chronic conditions treated, where responding Diplomates with moderate and high concordance treated a greater number of chronic conditions than those with low (mean 11.1, 12.0, and 7.2, respectively [P < .01]). Respondents who did not disclose their reimbursement model used were less likely to report moderate or high concordance with guidelines (60.3% had low AHA/ACC/TOS concordance vs 8.9% for fee-for-service, 22.0% for insurance and 19.9% for both [P < .01]). We also found a significant positive association between AHA/ACC/TOS guideline use with greater concordance with these guidelines (P < .01). There were no differences in AHA/ACC/TOS guideline concordance by population treated, number of clinical sessions per week, practice location or health system affiliation (data available upon request).

3.3 |. Concordance with AACE/ACE guidelines and differences by practice characteristics

Overall, 33.4% of responding Diplomates reported using the 2016 AACE/ACE guidelines (Table 2). We found high concordance with these guidelines—20.9% had low concordance, 44.9% moderate concordance and 34.2% high concordance. Among Diplomates graded as low (n = 103), lack of comprehensive medical examination (80.6%), physical activity (80.6%), behavioural services (74.8%), long-term anti-obesity medication use (73.8%) and available bariatric surgical care (73.7%). Lack of a meal plan (37.9%) occurred less often.

We found a significant difference in AACE/ACE guideline concordance by number of chronic conditions treated, where responding Diplomates with moderate and high concordance treated a greater number of chronic conditions than those with low (mean 10.9, 12.8 and 4.4, respectively (P < .01)). We also found a significant positive association between health system affiliation with greater concordance (ie, 77.8% of high AACE/ACE concordant respondents were affiliated with a health system vs 27.2% not affiliated [P < .01]). Respondents who did not disclose their reimbursement model used were less likely to report moderate or high concordance with guidelines (73.0% had low AACE/ACE concordance vs 16.1% for fee-for-service, 24.8% for insurance and 10.2% for both [P < .01]). Finally, we also found a significant positive association between AACE/ACE guideline use with greater concordance with these guidelines (P < .01). There were no differences in AACE/ACE guideline concordance by population treated, number of clinical sessions per week or practice location (data available upon request).

3.4 |. Concordance with OMA algorithm and differences by practice characteristics

Overall, 65.6% of responding Diplomates reported using the 2017–2018 OMA algorithm (Table 2). We also found high concordance with the algorithm, as 14.2% has low concordance, 59.1% had moderate concordance and 26.7% had high concordance. Among Diplomates graded as low (n = 70), lack of comprehensive medical examination (87.1%), physical activity (91.4%), behavioural services (85.7%), anti-obesity medication use (74.3%) and available bariatric procedural care (80.0%). Lack of nutrition (57.1%) and supplement use (18.6%) occurred less often.

We found a significant difference in concordance by number of chronic conditions treated, where responding Diplomates with moderate and high concordance treated a greater number of chronic conditions than those with low (mean 10.7, 12.4 and 3.7, respectively [P < .01]). We also found a significant positive association between health system affiliation with greater concordance with these guidelines (ie, 76.5% of high OMA concordant respondents were affiliated with a health system vs 23.5% not affiliated [P < .01]). Respondents who did not disclose their reimbursement model used were less likely to report moderate or high concordance with guidelines (60.3% had low OMA concordance vs 3.6% for fee-for-service, 8.6% for insurance and 7.2% for both [P < .01]). Finally, we also found a significant positive association between OMA algorithm use with greater concordance with these guidelines (P < .01). There were no differences in OMA algorithm concordance by population treated, number of clinical sessions per week or practice location (data available upon request).

4 |. DISCUSSION

This study is the first to describe the clinical obesity services offered by ABOM Diplomates. Overall, most responding Diplomates endorsed use of key elements in the treatment of obesity including nutrition, behavioural services, perioperative bariatric surgical care and FDA-approved anti-obesity medications. A prior report describing ABOM Diplomates found that 40.2% reported practicing obesity medicine for less than 2 years and had a background in internal medicine or family medicine (66.0%).¹⁵ Their clinical effort dedicated to obesity medicine varied—38.0% spent less than a quarter of their time to obesity medicine whereas 19.4% spent over 75% of their clinical practice in this area. Our study provides additional insight into the clinical practice characteristics of ABOM Diplomates. For example, these physicians manage obesity as well as multiple comorbid conditions associated with the disease—typically 10 related conditions. Few practices were located in rural areas, and they varied in reimbursement model used.

Most responding Diplomates treated adults with obesity relative to paediatric-focused obesity medicine physicians. A prior study of ABOM Diplomates also reported differences in the number of paediatric physicians relative to adult medicine physicians.¹⁷ We did find some differences in clinical services offered when comparing adult and paediatric ABOM Diplomates including lower use of meal replacements, perioperative bariatric surgical care, and anti-obesity medications among paediatric Diplomates. These differences may reflect the differences between adult and paediatric weight management guidelines.^19–21,24 Given that our sample size of paediatric physicians is small, the population of ABOM Diplomates should be reassessed, as this group grows, to understand whether the observed differences persist. Future analyses should also consider examining differences among paediatric and adolescent obesity medicine practices.

Most responding ABOM Diplomates reported prescribing FDA-approved anti-obesity medications, and most used both short- and long-term options. The use of off-label medications to treat obesity was also common, although use of FDA-approved medications was higher (76.9 vs 83.4%, respectively). A prior study of bariatric physicians reported that off-label medication use was more common than the long-term FDA-approved anti-obesity medications available at that time.²⁵ It is clinically lawful for physicians to prescribe a medication off-label once the medication is approved by the FDA,²⁶ and many off-label uses are accepted as standard of care. In a nationally representative sample of a general population, off-label medication use accounted for 21% of all prescriptions.²⁷ A framework exists to categorize off-label medication use as supported, suppositional, or investigational to help guide physicians’ appropriate off-label medication use.²⁸ Diplomates in our sample were predominantly using components of FDA-approved medications or anti-diabetic drugs with documented weight-loss benefits, which would be most consistent with supported off-label use (ie, moderate to high level of certainty in net health benefit). It appears the primary drivers of off-label use was financial - either a lack of insurance coverage for FDA-approved options or medication costs. While benefits coverage for outpatient visits for obesity counselling has improved, coverage of anti-obesity medications has lagged behind. For example, Medicare began covering intensive behavioural therapy for obesity as a benefit in 2011²⁹; however, anti-obesity medications are excluded from coverage. Policy changes, such as the Treat and Reduce Obesity Act,³⁰ propose addressing this issue, and passage of this legislation might help reduce the frequency of off-label prescribing among ABOM Diplomates.

Responding ABOM Diplomates endorsed using a variety of evidence-based guidelines to inform their practice, and the OMA Obesity Algorithm²¹ was the most commonly cited. In examining three evidence-based obesity management guidelines—AHA/ACC/TOS, AACE/ACE and OMA—we found that the majority of physicians had moderate or high concordance with services recommended in these guidelines. Guideline-recommended treatments have been associated with clinically significant weight loss in clinical trials¹⁹; however, other studies of physicians document pessimism regarding obesity treatment efficacy and lower rates of physician referral to obesity treatment programs.^10,31,32 Our findings may provide other clinicians increased confidence in patient receipt of evidence-based obesity services among ABOM Diplomates, which could improve the attitudes and referring practices of other clinicians.

In addition, we identified several factors associated with higher likelihood of guideline concordance. Greater concordance was associated with greater number of obesity-related chronic conditions managed and health system affiliation, which may reflect physicians who have a comprehensive approach to care and access to other resources. We found no significant difference between fee-for-service only, insurance-only, and both fee-for-service + insurance, although physicians who did not disclose their reimbursement model had lower guideline concordance. Finally, we identified a positive association between reported guideline use and guideline concordance with these services. Organizations, such as ABOM, might support education about and use of these evidence-based guidelines as a strategy to promote evidence-based obesity care.

This study has several limitations. Self-report of clinical services is subject to bias, and may not reflect actual services offered. Prior research has found that if physicians report delivering a service it is likely to have been done.³³ Verification of services was not possible given our deidentified data. Given the structure of some questions (ie, “check all that apply”), some respondents may have failed to indicate an offered service. We used answers from several questions to indicate services as a means to decrease the potential for misclassification. For example, physical activity services was identified as offered if any of the following were checked: exercise specialist works in the practice (Q5), fitness evaluation (Q11), or exercise counselling (Q11). The survey did not include questions about demographics of Diplomates; therefore, we cannot determine whether our sample is representative of ABOM Diplomates. The survey did not contain questions to distinguish academic, healthcare system and private practice physicians. We also elected not to conduct multivariable regression analyses, as we could not adjust for key physician attributes (eg, years in practice) that have been associated with differences in obesity practice habits.¹² The response rate to the survey was low, although is similar to other online surveys of physicians.¹⁸ We are unable to determine whether non-respondents may differ from responding ABOM Diplomates in regards to evidence-based services offered. The recruitment strategy and incentive offered may have selected for a more activated population of ABOM Diplomates. Finally, given the aforementioned limitations, we did not attempt to identify predictors of evidence-based care in this study.

5 |. CONCLUSION

We found that most responding ABOM Diplomates endorsed use of evidence-based services in the treatment of obesity including nutrition, physical activity, behavioural services, perioperative bariatric surgical care and FDA-approved anti-obesity medications. The majority of physicians offered services with moderate to high concordance with several clinical practice guidelines. Clinicians may therefore have increased confidence in patient receipt of evidence-based care when referring to an ABOM Diplomate.

Study Importance Questions.

What is already known about this subject?

• There is an increasing number of American Board of Obesity Medicine (ABOM) Diplomates

• Demographics of these physicians have been described, but little is known about their clinical practices

What are the new findings in your manuscript?

• Most ABOM Diplomates endorse using many obesity services including nutrition, exercise, behavioural services, perioperative bariatric surgical care, and FDA-approved anti-obesity medications

• The majority offered services with moderate to high concordance with three evidence-based clinical practice guidelines

How might your results change the direction of research or the focus of clinical practice?

• Clinicians may increase referrals to ABOM Diplomates, given the likelihood that their patients will receive evidence-based obesity care