- Email: [email protected]
Adult obesity management in primary care, 2008–2013
Accepted Manuscript Adult obesity management in primary care, 2008–2013
Stephanie L. Fitzpatrick, Victor J. Stevens PII: DOI: Reference:
S0091-7435(17)30074-9 doi: 10.1016/j.ypmed.2017.02.020 YPMED 4950
To appear in:
Preventive Medicine
Received date: Revised date: Accepted date:
28 July 2016 17 February 2017 18 February 2017
Please cite this article as: Stephanie L. Fitzpatrick, Victor J. Stevens , Adult obesity management in primary care, 2008–2013. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Ypmed(2017), doi: 10.1016/j.ypmed.2017.02.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Adult Obesity Management in Primary Care, 2008-2013 RUNNING HEAD: Obesity Management in Primary Care, 2008-2013
Stephanie L. Fitzpatrick, PhD,1 Victor J. Stevens, PhD2 Department of Preventive Medicine, Rush University Medical Center, Chicago, IL
2
Center for Health Research, Kaiser Permanente Northwest, Portland, OR
CR
IP
T
1
Corresponding Author: Stephanie L. Fitzpatrick, PhD, (Present Address) Center for Health
US
Research, Kaiser Permanente Northwest, 3800 N. Interstate Ave., Portland, OR 97227, Phone:
M
AN
503-335-6773, Email: [email protected]
PT
Abstract: 248/250
ED
The authors have no conflicts of interest to disclose.
Text pages: 21
Tables: 4
AC
References: 28
CE
Word Count: 2,950/3,500
Figures: 1
1
ACCEPTED MANUSCRIPT Abstract In the U.S., the occurrence of weight counseling in primary care for patients with obesity decreased by 10% between 1995-1996 and 2007-2008. There have been several national recommendations and policies to improve obesity management since 2008. The purpose of this
IP
weight management counseling in the U.S. from 2008 to 2013.
T
study was to examine the rates of body mass index (BMI) screening, obesity diagnosis, and
CR
The National Ambulatory Medical Care Survey visit-level data for adults 18 and over with a primary care visit during survey years 2008-2009, 2010-2011, and 2012-2013 was
US
included in the analyses using SAS v9.3. Study outcomes included percent of visits with: BMI
AN
screening; obesity diagnosis; and weight counseling. We compared survey years on these outcomes using 2008-2009 as the reference as well as examined patient and practice-level
M
predictors. Analyses were conducted from 2015 to early 2017.
ED
Of the total 55,608 adult primary care visits sampled, 14,143 visits (25%) were with patients with obesity. BMI screening significantly increased between 2008-2009 and 2012-2013
PT
from 54% to 73% (OR = 1.75, 95% CI 1.28-2.41); however, percent of visits with an obesity
CE
diagnosis remained low at less than 30%. Weight management counseling during visits
0.41-0.92).
AC
significantly declined from 33% to 21% between 2008-2009 and 2012-2013 (OR = 0.62, 95% CI
Despite emerging recommendations and policies, from 2008 to 2013, obesity management in primary care remained suboptimal. Identifying practical strategies to enforce policies and implement evidence-based behavioral treatment in primary care should be a high priority in healthcare reform. Key words: obesity, primary care, weight management counseling
2
ACCEPTED MANUSCRIPT More than one-third of adults in the U.S. have a body mass index [BMI] ≥ 30 kg/m2 and are therefore at substantially increased risk for diabetes and cardiovascular disease (CVD).1,2 Behavioral weight management treatment is an effective first-line treatment for obesity with an average initial weight loss of 8-10%, which is associated with a significant reduction in risk for
T
diabetes and improvement in CVD risk factors.3,4 However, in 2005-2006, two-thirds of U.S.
IP
patients with obesity were not offered or referred to weight management treatment during their
CR
primary care visit.5 In addition, the rate of weight management counseling in primary care significantly decreased by 10% (40% to 30%) between 1995-1996 and2007-2008.6
US
There have been several national recommendations and policies implemented since 2008
AN
to improve obesity management in primary care. The U.S. Preventive Services Task Force (USPSTF),7 and a joint statement by the American Heart Association, American College of
M
Cardiology, and The Obesity Society8 recommend that physicians screen for overweight and
ED
obesity in their practices and provide or refer patients with risk factors for cardiovascular disease to intensive behavioral counseling. In 2011, the Center for Medicare & Medicaid Services
PT
(CMS) passed a decision to reimburse primary care physicians for delivering intensive
CE
behavioral therapy to treat patients with obesity.9 The CMS reimbursement policy is limited to coverage for Medicare beneficiaries and only reimburses primary care practitioners. When
AC
delivering the intensive behavioral therapy for obesity, physicians are expected to follow the 5 A’s counseling framework (i.e., Assess, Advise, Agree, Assist, Arrange)9-11 with 10-15 minute visits (maximum of 22 visits). In addition, CMS implemented the Electronic Health Record (EHR) Meaningful Use Incentive Program, where physicians receive financial incentives when they implement and use the EHR to document quality improvement measures.12 Physicians are incentivized to document
3
ACCEPTED MANUSCRIPT in the EHR BMI and a follow-up treatment plan to provide or refer the patient with BMI ≥ 25 to weight management treatment. The purpose of this study was to examine rates of patient BMI screening, obesity diagnosis, and provision of treatment for obesity by primary care physicians in the U.S. from
T
2008 to 2013 as well as examine the patient and practice characteristics associated with these
IP
outcomes.
CR
Methods
The National Ambulatory Medical Care Survey (NAMCS) is an annual survey conducted
US
by the National Center for Health Statistics that is used to characterize the utilization and
AN
provision of ambulatory care in the U.S.13 Currently, there is survey data publicly available from 1973 to 2013. A multi-stage probability sampling design is used, which consist of sampling
M
from primary sampling units (e.g., counties), physicians within the primary sampling units, and
ED
patient visits within practices. Using the medical chart (paper and/or electronic), physicians, office staff, or Census Bureau representatives complete physician and patient record survey
PT
forms regarding: outpatient practice characteristics, physician demographics, and visit-level data
CE
including patient demographics, reasons for the visit, diagnoses, and treatment. In terms of obesity-related data, the patient record form allowed surveyors to write in the patient height and
AC
weight, a checkbox to indicate diagnosis of obesity, and a checkbox to indicate if health education was provided (e.g., diet/nutrition, exercise, or weight reduction). In 2008-2009 and 2010-2011, the survey was completed using a paper form. However, in 2012-2013, surveys were completed mostly by Census representatives using a computer form.14 Each visit was weighted in order to obtain national estimates.
4
ACCEPTED MANUSCRIPT Similar methods applied in Ma et al.5 and Kraschnewski et al.6 to examine previous NAMCS survey year data were used in this study. We used survey years 2008-2009, 2010-2011, and 2012-2013 visit-level data with adults 18 and over who had an office-based outpatient visit with a general, family, or internal medicine physician. Because community health center visits
T
were excluded in the 20122013 surveys, community health center visits in survey years 2008-
IP
2009 and 2010-2011 were removed from the analytic dataset.14 Using the criteria specified in
CR
Ma et al.,5 patients with diabetes or coronary artery disease were considered at high risk for obesity-related disease complications and mortality. Patients with any one of the following
US
conditions were classified as moderate risk: a) hypertension, b) hyperlipidemia, c) sleep apnea,
AN
or d) asthma. Patients with no cardiovascular disease risk factors besides obesity were considered to be at low risk for obesity-related complications and mortality. We compared survey years
M
using a multivariable logistic regression model on the following study outcomes using 2008-
ED
2009 as the reference time period: 1) percent of visits with height and/or weight measured; 2) among patients with BMI ≥ 30, percent of visits with obesity diagnosis; and 3) among patients
PT
with obesity, rate of any weight-related education (i.e., any combination of diet/nutrition,
CE
exercise, and/or weight reduction education selected on the patient record form). In the model we adjusted for patient and practice-level variables that were found to be associated with these
AC
obesity-related metrics in previous studies.5,6,15 These variables included sex, age, race/ethnicity, insurance type, level of risks for obesity-related diseases or mortality, if the patient had been seen before, use of electronic medical records in the practice, and region of the U.S. in which the practice was located. Given that predictors of BMI screening, obesity diagnosis, and any weightrelated education were similar across years (data not shown), we combined all survey years to increase sample size in order to examine predictors of the outcomes for 2008-2013 (as one data
5
ACCEPTED MANUSCRIPT point) using the same patient and practice-level characteristics previously mentioned. Analyses were conducted from 2015 to early 2017 using PROC SURVEYFREQ AND SURVEYLOGISTIC in SAS v9.3. Two-sided P values less than .05 were considered significant. Results
T
There were 13,075 adult primary care visits sampled from 2008-2009, 10,951 from 2010-
IP
2011, and 31,582 visits sampled from 2012-2013. Of the total 55,608 adult primary care visits
CR
sampled, 14,143 visits (25%) were with patients with obesity. Table 1 presents the weighted proportions for visit, patient, and practice characteristics by survey year.
US
BMI Screening
AN
The measurement of both height and weight significantly increased between 2008-2009 and 2012-2013 from 54% to 73% (OR = 1.75, 95% CI 1.28-2.41) while adjusting for patient and
M
practice-level characteristics. Independent predictors of both height and weight being measured
ED
included: a) Hispanic ethnicity compared to White non-Hispanic (OR = 1.29, 95% CI 1.061.56); b) moderate risk (OR = 1.12, 95% CI 1.01-1.24) and high risk for obesity-related disease
PT
complications and mortality (OR = 1.30, 95% CI 1.14-1.49) compared to low risk; c) Medicaid
CE
compared to private insurance (OR = 1.28 (1.02-1.59); and d) presence of all electronic medical records (OR = 1.58, 95% CI 1.25-2.00) or part paper and part electronic (OR = 1.75, 95% CI
AC
1.28-2.41) versus no electronic records (Table 2). It should be noted that there was a significant increase in the use of electronic medical records between 2008-2009 and 2012-2013 (OR = 2.48, 95% CI 1.66-3.71,). Patients with Medicare versus private insurance and established patients (seen before in the clinic) versus new patients were less likely to have both height and weight measured during the visit (Table 2). Obesity Diagnosis and Weight Management Counseling
6
ACCEPTED MANUSCRIPT Figure 1 presents the weighted percentage of visits with obesity diagnosis, diet/nutrition education, exercise education, weight reduction education, or any combination of weight-related education among visits where a patient had a BMI ≥ 30. Across the years there was a nonsignificant decline in diagnosis and each specific domain of health education (Figure 1). There
T
was a significant decline of 33% to 21% in any combination of weight-related education between
IP
2008-2009 and 2012-2013 (OR = 0.62, 95% CI 0.41-0.92). Women compared to men (OR =
CR
1.41, 95% CI 1.29-1.53), Black non-Hispanic compared to White non-Hispanic (OR 1.47 (1.211.79), moderate (OR = 1.91 95% CI 1.62-2.26) and high risk patients (OR = 4.76, 95% CI 3.97-
US
5.70) compared to low risk, patients age 18-44 (OR = 2.34, 95% CI 1.86-2.94) or 45-64 (OR =
AN
1.87, 95% CI 1.56-2.24) compared to patients 65 and older, patients with Medicaid compared to those with private insurance (OR = 1.23, 95% 1.04-1.46), as well as established patients
M
compared to new (OR = 1.30, 95% 1.12-1.52) were more likely to have a diagnosis of obesity
ED
documented in their medical record (Table 3).
Weight management counseling was more likely to occur for patients who, identified as
PT
Black non-Hispanic (OR = 1.47, 95% CI 1.06-2.03), Hispanic (OR = 1.39, 95% CI 1.08-1.79) or
CE
“Other” for race/ethnicity (OR = 1.62, 95% CI 1.08-2.43), moderate (OR = 1.30 (1.02-1.65) and high risk (OR = 1.69, 95% CI 1.28-2.22), and had an obesity diagnosis documented in their
AC
medical record (OR = 3.37, 95% CI 2.88-3.95) (Table 4). Visits where weight management counseling occurred were, on weighted average, 21.95 (95% CI 21.04-22.86) minutes long. Obesity Management for Medicare Beneficiaries Only Because most of the policies to improve obesity management in primary care have been commissioned by CMS, we examined visits among Medicare beneficiaries only. There was a total of 16,620 primary care visits sampled with Medicare beneficiaries from 2008-2013 and
7
ACCEPTED MANUSCRIPT 3,863 (23%) of these were with patients with obesity. Percent of visits with both height and weight measured increased from 49% to 72% between 2008-2009 and 2012-2013 (OR = 1.57, 95% CI 1.05-2.34). Among visits with patients with Medicare and BMI ≥ 30, there was a nonsignificant decline in percent of visits with an obesity diagnosis (31% to 24%) and a non-
T
significant decline in percent of visits with any combination of weight-related education (28% to
IP
20%) between 2008-2009 and 2012-2013.
CR
Discussion
NAMCS data between 2008-2009 and 2012-2013 indicated a significant increase in BMI
US
screening (i.e., measure of both height and weight); however, the percent of visits with a
AN
documented obesity diagnosis declined by 5 percentage points (30% to 25%). Despite emerging national recommendations and healthcare policies, provision or referral to weight management
M
counseling significantly declined from 33% to 21% between 2008-2009 and 2012-2013.
ED
Medicare beneficiaries had similar increases BMI screening and decreases in obesity diagnosis and weight management counseling compared to the total population.
PT
There was a significant increase in practices with electronic medical records, which was a
CE
predictor of both height and weight being measured. Most electronic medical records allow easy input of vital signs and automatic calculation of BMI. Other practice characteristics that
AC
predicted measurement of both height and weight included Hispanic ethnicity, Medicaid insurance, and moderate and high risks for obesity-related diseases and mortality. Established patients or patients with Medicare were less likely to have both height and weight measured. This may be due to only weight and not height being measured during an encounter because the height for established and adult patients is assumed to be the same at each visit. However, in most electronic health records, an updated BMI cannot be calculated without both a weight and
8
ACCEPTED MANUSCRIPT height entered at each visit even if the height has not changed. It should be noted that weight was measured at 87% or more visits across the survey years, so the percent of visits with BMI screening would be higher if height did not need to be entered to update BMI. There is a need for more built in sophisticated clinical decision support tools within the electronic medical record
T
that can automatically extract previous vital sign data such as last recorded height and include it
IP
in calculations of BMI once a new weight has been entered during an encounter. This
CR
improvement may decrease burden on physicians or medical staff and perhaps increase screening and diagnosis.
US
Across the survey years, 70% or more of patients with obesity did not have a documented
AN
obesity diagnosis. Although electronic medical records facilitate the input of vital signs, documentation of an obesity diagnosis requires the provider to take additional steps. To
M
diagnose obesity, physicians would need to add obesity to the patient’s problem list. However,
ED
there is no real incentive for physicians to diagnose obesity since it is still not considered a billable medical condition by most insurance companies, with the exception of Medicare.
PT
Despite the overall low rate of diagnosed obesity, patients who were female, 18-64 years of age,
CE
Black non-Hispanic, Medicaid, established patient, or at moderate or high risk for obesity-related disease complications were more likely to have a diagnosis. Previous survey based studies have
AC
indicated that women as well as young and middle-aged adults are more likely to have an obesity diagnosis, perhaps because members of these populations are more likely to discuss concerns about their weight with their physician that prompted an obesity diagnosis at the end of the visit.15,16 Patients at high risk for obesity-related disease complications and mortality perhaps prompt physicians to regularly monitor their BMI and address progress with lifestyle changes and weight loss in order to prevent CVD events. Similarly, patients from racial/ethnic minority
9
ACCEPTED MANUSCRIPT backgrounds or with Medicaid may also be more likely to be at high risk for obesity-related diseases and more frequent users of care;16 thus maybe increasing the likelihood of receiving an obesity diagnosis and some health education. Given the findings in Kraschnewski et al., 6 the percent of primary care visits with weight
T
management counseling continued to significantly decline by 12 percentage points between
IP
2008-2009 and 2012-2013, despite national recommendation and policies established during this
CR
time period. As previously mentioned, there is no immediate financial incentive to address and manage obesity in the primary care office for adult patients 18-64 years old.17 Although,
US
Medicare reimburses intensive behavioral therapy for obesity for their beneficiaries,9 the
AN
occurrence of weight management counseling during a visit also decreased among this patient population. Most physicians are unaware of this reimbursement policy or how to effectively
M
implement it without disrupting current clinical workflow given the number of visits that must
ED
occur (i.e., 22 total). Additional physician specific barriers to providing weight management counseling include time constraints, discomfort with discussing weight issues, lack of training in
PT
weight management, and lack of knowledge regarding available treatments.18-21 Patients with an
CE
obesity diagnosis were three times more likely to receive weight management counseling than those without a diagnosis. Visits that did include weight management counseling were about 20
AC
minutes in duration, which is consistent with the expected length of an intensive behavioral therapy session using the CMS model (i.e., 10-15 minutes).9 However, based on 2005-2006 NAMCS data, only 8% (i.e., ~1.6 minutes) of the 20 minute visit was actually spent addressing obesity.22 Thus, even if weight management counseling is indicated in the medical chart that does not guarantee that a sufficient amount of time is spent providing high quality counseling. Limitations
10
ACCEPTED MANUSCRIPT There are several limitations in this study. First, the NAMCS data is a cross-sectional survey based on only one random week of clinic visits per physician over a year. Thus, diagnosis of obesity and weight management counseling could have occurred during another encounter not included in the survey sample. Second, the focus on primary care visits as the unit of analysis,
T
may have resulted in oversampling of sicker patients or frequent users of primary care. Third,
IP
since the CMS Electronic Medical Record Meaningful Use Incentive Program was not fully
CR
implemented until 2011,23 perhaps not enough time has passed to see the impact of financial incentives on obesity management in primary care. Fourth, there was a change in the sampling
US
design and methodology for the 2012-2013 NAMCS.14 Specifically, community health center
AN
visits were excluded, Census Bureau representatives were more likely to complete the survey forms than physicians and clinic staff, and the survey was completed using a computerized form
M
instead of a paper form. Increased use of the Census Bureau representatives and computerized
ED
forms may explain the increased number of visits sampled in 2012-2013. However, it is not clear if it was these changes or actual practice patterns that impacted the increase or decrease in
CE
Conclusions
PT
documentation of obesity-related metrics.14
Despite emerging national recommendations and policies since 2008, obesity
AC
management in primary care is still suboptimal. A recent study indicated that a majority of patients want to discuss weight loss with their physician.24 Thus, there is a need for primary care delivery redesign to facilitate rather than hinder physicians addressing obesity with their patients. Part of this redesign is increasing physician skills in starting the conversation to sensitively address obesity with a patient.25 It is not realistic to expect primary care physicians to deliver intensive behavioral weight loss counseling to all of their patients with obesity. Thus, the
11
ACCEPTED MANUSCRIPT process of ordering referrals and coverage of obesity management specialist (e.g., registered dietitians, psychologists) and community-based programs should be made easier in order to increase referral options for physicians and access to care for patients.26 In addition, the electronic medical record is emerging as a tool to not only facilitate BMI screening, but could
T
also be used to facilitate weight management counseling during an encounter.27,28 Given the
IP
obesity epidemic and increased risk for chronic diseases, identifying practical strategies to
CR
enforce policies and implement evidence-based treatment services in primary care should be a
AC
CE
PT
ED
M
AN
US
high priority in healthcare reform.
12
ACCEPTED MANUSCRIPT References 1.
Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. JAMA. 2012;307(5):491497. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in
Butryn ML, Webb V, Wadden TA. Behavioral treatment of obesity. Psychiatr Clin North
CR
3.
IP
the United States, 2011-2012. JAMA. 2014;311(8):806-814.
Am. 2011;34(4):841-859.
Wadden TA, Webb VL, Moran CH, Bailer BA. Lifestyle modification for obesity: new
US
4.
T
2.
AN
developments in diet, physical activity, and behavior therapy. Circulation. 2012;125(9):1157-1170.
Ma J, Xiao L, Stafford RS. Adult obesity and office-based quality of care in the United
M
5.
6.
ED
States. Obesity (Silver Spring, Md.). 2009;17(5):1077-1085. Kraschnewski JL, Sciamanna CN, Stuckey HL, et al. A silent response to the obesity
Moyer VA, Force USPST. Screening for and management of obesity in adults: U.S.
CE
7.
PT
epidemic: decline in US physician weight counseling. Med Care. 2013;51(2):186-192.
Preventive Services Task Force recommendation statement. Ann Intern Med.
8.
AC
2012;157(5):373-378.
Jensen MD, Ryan DH, Apovian CM, et al. 2013 AHA/ACC/TOS Guideline for the Management of Overweight and Obesity in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society. J Am Coll Cardiol. 2013.
9.
DHHS Centers for Medicare and Medicaid Services. Intensive Behavioral Therapy (IBT) for Obesity. Medicare Learning Network; 2012. 13
ACCEPTED MANUSCRIPT 10.
Vallis M, Piccinini-Vallis H, Sharma AM, Freedhoff Y. Clinical review: modified 5 As: minimal intervention for obesity counseling in primary care. Can Fam Physician. 2013;59(1):27-31.
11.
Alexander SC, Cox ME, Boling Turer CL, et al. Do the five A's work when physicians
T
counsel about weight loss? Fam Med. 2011;43(3):179-184. Center for M, Medicaid S. EHR Incentive Programs. Vol 2013.
13.
National Center for Health S. NAMCS/NHAMCS Ambulatory Health Care Data. Vol
CR
IP
12.
2016.
National Center for Health S. NAMCS/NHAMCS Questionnaires, Datasets, & Related
US
14.
15.
AN
Documentation. Vol 2016.
Ahmed NU, Delgado M, Saxena A. Trends and disparities in the prevalence of
16.
ED
Prev Med. 2016;89:70-75.
M
physicians' counseling on diet and nutrition among the U.S. adult population, 2000-2011.
Savageau JA, McLoughlin M, Ursan A, Bai Y, Collins M, Cashman SB. Characteristics
PT
of frequent attenders at a community health center. J Am Board Fam Med.
17.
CE
2006;19(3):265-275.
Kushner RF, Sarwer DB. Medical and behavioral evaluation of patients with obesity.
18.
AC
Psychiatr Clin North Am. 2011;34(4):797-812. Bleich SN, Bennett WL, Gudzune KA, Cooper LA. National survey of US primary care physicians' perspectives about causes of obesity and solutions to improve care. BMJ open. 2012;2(6):10.1136/bmjopen-2012-001871. Print 002012. 19.
Bleich SN, Bennett WL, Gudzune KA, Cooper LA. Impact of physician BMI on obesity care and beliefs. Obesity (Silver Spring, Md.). 2012;20(5):999-1005.
14
ACCEPTED MANUSCRIPT 20.
Kolasa KM, Rickett K. Barriers to providing nutrition counseling cited by physicians: a survey of primary care practitioners. Nutr Clin Pract. 2010;25(5):502-509.
21.
Kushner RF. Barriers to providing nutrition counseling by physicians: a survey of primary care practitioners. Prev Med. 1995;24(6):546-552. Tsai AG, Abbo ED, Ogden LG. The time burden of overweight and obesity in primary
IP
care. BMC Health Serv Res. 2011;11:191-6963-6911-6191.
T
22.
Center for Medicare and Medicaid Services. EHR Incentive Programs. Vol 2014.
24.
Sherson EA, Yakes Jimenez E, Katalanos N. A review of the use of the 5 A's model for
CR
23.
US
weight loss counselling: differences between physician practice and patient demand. Fam
Stop Obesity Alliance. Why Weight? A Guide to Discussing Obesity & Health With Your Patients. Vol 2014.
Fitzpatrick SL, Wischenka D, Appelhans BM, et al. An Evidence-based Guide for
ED
26.
M
25.
AN
Pract. 2014;31(4):389-398.
Obesity Treatment in Primary Care. Am J Med. 2016;129(1):115.e1-7. Bordowitz R, Morland K, Reich D. The use of an electronic medical record to improve
PT
27.
Tang JW, Kushner RF, Cameron KA, Hicks B, Cooper AJ, Baker DW. Electronic tools to assist with identification and counseling for overweight patients: a randomized
AC
28.
CE
documentation and treatment of obesity. Fam Med. 2007;39(4):274-279.
controlled trial. J Gen Intern Med. 2012;27(8):933-939.
15
ACCEPTED MANUSCRIPT Table 1. Patient, Practice, & Visit Characteristics of U.S. Adult Primary Care Visits by Survey Year: %†, (95% CI) 2008-2009 (n = 13,075) 2010-2011 (n = 10,951) 2012-2013 (n = 31,582) Age group, y 18-44 29.0 (27.9, 30.8) 30.0 (27.9, 32.2) 27.7 (26.4, 29.0) 45-64 38.1 (36.8, 39.5) 38.3 (36.8, 39.7) 38.3 (37.3, 39.3) 65 and up 32.9 (30.8, 35.0) 31.7 (29.3, 34.1) 34.0 (32.6, 35.4) Sex Female 59.8 (58.1, 61.4) 57.0 (55.1, 58.8) 57.3 (56.1, 58.5) Male 40.2 (38.6, 41.9) 43.0 (41.2, 44.9) 42.7 (41.5, 43.9) Race/Ethnicity White non-Hispanic 74.9 (71.7, 78.1) 73.4 (69.3, 77.4) 73.0 (70.9, 75.2) Black non-Hispanic 10.0 (7.7, 12.3) 12.7 (9.6, 15.9) 10.1 (8.9, 11.2) Hispanic 11.0 (8.7, 13.4) 9.2 (6.0, 12.5) 12.5 (10.7, 14.3) Other 4.1 (2.7, 5.5) 4.7 (3.0, 6.3) 4.4 (3.6, 5.2) 20.4 (18.4, 22.4) 21.4 (18.9, 23.9) 28.9 (28.2, 29.6) Body Mass Index ≥ 30 kg/m2 Risks for Obesity-Related Disease/Mortality Low 21.3 (19.6, 22.9) 20.5 (18.7, 22.4) 20.2 (19.0, 21.3) Moderate 55.4 (54.0, 56.9) 55.5 (53.9, 57.0) 57.2 (56.2, 58.3) High 23.3 (21.6, 25.0) 24.0 (21.9, 26.1) 22.6 (21.6, 23.6) Insurance Private 57.7 (54.9, 60.5) 53.8 (50.4, 57.1) 50.8 (48.9, 52.6) Medicare 29.2 (27.0, 31.5) 31.5 (29.0, 34.1) 33.9 (32.3, 35.4) Medicaid 5.8 (4.3, 7.4) 7.2 (5.9, 8.5) 7.9 (6.7, 9.0) Other 7.2 (5.5, 9.0) 7.5 (5.9, 9.2) 7.5 (6.2, 8.7) Electronic Medical Record System No 46.7 (40.5, 52.9) 37.8 (32.3, 43.3) 23.3 (20.3, 26.3) Yes, part paper part electronic 14.4 (9.8, 19.0) 9.3 (5.7, 12.9) 11.6 (9.0, 14.2) Yes, all electronic 38.9 (32.3, 45.4) 52.8 (47.4, 58.3) 65.1 (61.6, 68.6) Region Northeast 16.1 (11.1, 21.1) 19.8 (15.8, 23.8) 19.3 (17.7, 20.9) Midwest 25.8 (20.0, 31.6) 24.2 (20.0, 28.4) 19.8 (18.5, 21.0) South 37.9 (31.9, 43.8) 35.1 (30.0, 40.2) 36.2 (34.4, 38.0)
T P
I R
C S
U N
A
D E
M
T P
E C
C A
16
ACCEPTED MANUSCRIPT West 20.2 (15.6, 24.9) 20.9 (17.2, 24.6) 24.8 (22.9, 26.6) Height & Weight measured Neither 11.2 (9.0, 13.4) 10.3 (8.0, 12.5) 8.1 (6.8, 9.3) Either 34.6 (30.3, 38.9) 33.8 (29.2, 38.5) 18.7 (16.7, 20.8) Both 54.1 (50.0, 58.7) 55.9 (50.6, 61.2) 73.2 (70.8, 75.5) † All percentages are population percentages estimated from a weighted analysis taking into account the complex sampling stratification and clustering.
T P
I R
C S
U N
A
D E
M
T P
E C
C A
17
ACCEPTED MANUSCRIPT Table 2. Predictors of Body Mass Index Screening, 2008-2013 Adjusted OR (95% CI) 1.00 [Reference] 1.04 (0.83-1.31) 1.75 (1.28-2.41)
IP
T
1.09 (0.95-1.24) 1.02 (0.92-1.14) 1.00 [Reference]
CR
1.04 (0.97-1.11) 1.00 [Reference]
AC
CE
PT
ED
M
1.00 [Reference] 1.21 (0.93-1.58) 1.29 (1.06-1.56) 1.18 (0.88-1.58)
US
AN
Survey Years 2008-2009 2010-2011 2012-2013 Age group, y 18-44 45-64 65 and up Sex Female Male Race/Ethnicity White non-Hispanic Black non-Hispanic Hispanic Other Risks for Obesity-Related Disease/Mortality Low Moderate High Insurance Private Medicare Medicaid Other Use an Electronic Medical Record System No Yes, part paper part electronic Yes, all electronic Patient seen before? No Yes Region Northeast Midwest South West
1.00 [Reference] 1.12 (1.01-1.24) 1.30 (1.14-1.49) 1.00 [Reference] 0.82 (0.72-0.92) 1.28 (1.02-1.59) 0.83 (0.63-1.08) 1.00 [Reference] 1.75 (1.28-2.41) 1.58 (1.25-2.00) 1.00 [Reference] 0.83 (0.71-0.97) 0.92 (0.61-1.37) 1.00 (0.70-1.44) 1.21 (0.83-1.75) 1.00 [Reference]
18
ACCEPTED MANUSCRIPT Table 3. Predictors of Obesity Diagnosis, 2008-2013 Adjusted OR (95% CI) 1.00 [Reference] 1.08 (0.90-1.29) 1.11 (0.86-1.45)
IP
T
2.34 (1.86-2.94) 1.87 (1.56-2.24) 1.00 [Reference]
CR
1.41 (1.29-1.53) 1.00 [Reference]
AC
CE
PT
ED
M
1.00 [Reference] 1.47 (1.21-1.79) 1.14 (0.94-1.39) 0.55 (0.42-0.71)
US
AN
Survey Years 2008-2009 2010-2011 2012-2013 Age group, y 18-44 45-64 65 and up Sex Female Male Race/Ethnicity White non-Hispanic Black non-Hispanic Hispanic Other Risks for Obesity-Related Disease/Mortality Low Moderate High Insurance Private Medicare Medicaid Other Use an Electronic Medical Record System No Yes, part paper part electronic Yes, all electronic Patient seen before? No Yes Region Northeast Midwest South West
1.00 [Reference] 1.91 (1.62-2.26) 4.76 (3.97-5.70) 1.00 [Reference] 0.93 (0.77-1.11) 1.23 (1.04-1.46) 1.22 (0.85-1.74) 1.00 [Reference] 1.12 (0.92-1.38) 1.10 (0.94-1.29) 1.00 [Reference] 1.30 (1.12-1.52) 1.35 (1.00-1.82) 1.26 (0.98-1.62) 1.03 (0.80-1.33) 1.00 [Reference]
19
ACCEPTED MANUSCRIPT Table 4. Predictors of Any Weight-Related Education, 2008-2013 Adjusted OR (95% CI) 1.00 [Reference] 0.94 (0.69-1.28) 0.62 (0.41-0.92)
IP
T
0.99 (0.77-1.27) 1.09 (0.89-1.33) 1.00 [Reference]
CR
0.94 (0.82-1.07) 1.00 [Reference]
AC
CE
PT
ED
M
1.00 [Reference] 1.47 (1.06-2.03) 1.39 (1.08-1.79) 1.62 (1.08-2.43)
US
AN
Survey Years 2008-2009 2010-2011 2012-2013 Age group, y 18-44 45-64 65 and up Sex Female Male Race/Ethnicity White non-Hispanic Black non-Hispanic Hispanic Other Risks for Obesity-Related Disease/Mortality Low Moderate High Insurance Private Medicare Medicaid Other Obesity Diagnosis No Yes Use an Electronic Medical Record System No Yes, part paper part electronic Yes, all electronic Patient seen before? No Yes Region Northeast Midwest South West
1.00 [Reference] 1.30 (1.02-1.65) 1.69 (1.28-2.22) 1.00 [Reference] 0.78 (0.63-0.98) 0.97 (0.68-1.37) 1.24 (0.77-2.00) 1.00 [Reference] 3.37 (2.88-3.95) 1.00 [Reference] 1.00 (0.70-1.44) 1.00 (0.79-1.28) 1.00 [Reference] 1.04 (0.81-1.32) 0.82 (0.51-1.30) 0.69 (0.47-1.03) 0.60 (0.39-0.94) 1.00 [Reference]
20
ACCEPTED MANUSCRIPT
35%
b
30% 25% 20% 15%
T
10%
2008-2009 (n = 2693)
IP
5%
2010-2011 (n = 2426)
0%
AN
US
CR
2012-2013 (n = 9024)
AC
CE
PT
ED
M
Figure 1. Weighted percentage of patient visits with obesity diagnosis, diet/nutrition education, exercise education, weight reduction education, and any weight-related education by survey year among visits with patients with body mass index ≥ 30. a Adjusted for age, sex, race/ethnicity, insurance type, risks for obesity-related diseases or mortality, if the patient had been seen before, presence of electronic medical record, and practice region. b Percent of visits with any weight-related education in 2012-2013 was significantly lower than the percent in 2008-2009 (OR = 0.62, 95% CI 0.41-0.92).
21
ACCEPTED MANUSCRIPT Highlights More than one-third of adults in the U.S. meet criteria for obesity.
Screening for obesity significantly increased 2008 to 2013.
From 2008 to 2013, weight counseling in primary care significantly declined.
Racial/ethnic minorities and high risk for CVD more likely to receive counseling.
Less than 1/3 Medicare patients received weight counseling each year, 2008 to 2013.
AC
CE
PT
ED
M
AN
US
CR
IP
T
22
Stephanie L. Fitzpatrick, Victor J. Stevens PII: DOI: Reference:
S0091-7435(17)30074-9 doi: 10.1016/j.ypmed.2017.02.020 YPMED 4950
To appear in:
Preventive Medicine
Received date: Revised date: Accepted date:
28 July 2016 17 February 2017 18 February 2017
Please cite this article as: Stephanie L. Fitzpatrick, Victor J. Stevens , Adult obesity management in primary care, 2008–2013. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Ypmed(2017), doi: 10.1016/j.ypmed.2017.02.020
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Adult Obesity Management in Primary Care, 2008-2013 RUNNING HEAD: Obesity Management in Primary Care, 2008-2013
Stephanie L. Fitzpatrick, PhD,1 Victor J. Stevens, PhD2 Department of Preventive Medicine, Rush University Medical Center, Chicago, IL
2
Center for Health Research, Kaiser Permanente Northwest, Portland, OR
CR
IP
T
1
Corresponding Author: Stephanie L. Fitzpatrick, PhD, (Present Address) Center for Health
US
Research, Kaiser Permanente Northwest, 3800 N. Interstate Ave., Portland, OR 97227, Phone:
M
AN
503-335-6773, Email: [email protected]
PT
Abstract: 248/250
ED
The authors have no conflicts of interest to disclose.
Text pages: 21
Tables: 4
AC
References: 28
CE
Word Count: 2,950/3,500
Figures: 1
1
ACCEPTED MANUSCRIPT Abstract In the U.S., the occurrence of weight counseling in primary care for patients with obesity decreased by 10% between 1995-1996 and 2007-2008. There have been several national recommendations and policies to improve obesity management since 2008. The purpose of this
IP
weight management counseling in the U.S. from 2008 to 2013.
T
study was to examine the rates of body mass index (BMI) screening, obesity diagnosis, and
CR
The National Ambulatory Medical Care Survey visit-level data for adults 18 and over with a primary care visit during survey years 2008-2009, 2010-2011, and 2012-2013 was
US
included in the analyses using SAS v9.3. Study outcomes included percent of visits with: BMI
AN
screening; obesity diagnosis; and weight counseling. We compared survey years on these outcomes using 2008-2009 as the reference as well as examined patient and practice-level
M
predictors. Analyses were conducted from 2015 to early 2017.
ED
Of the total 55,608 adult primary care visits sampled, 14,143 visits (25%) were with patients with obesity. BMI screening significantly increased between 2008-2009 and 2012-2013
PT
from 54% to 73% (OR = 1.75, 95% CI 1.28-2.41); however, percent of visits with an obesity
CE
diagnosis remained low at less than 30%. Weight management counseling during visits
0.41-0.92).
AC
significantly declined from 33% to 21% between 2008-2009 and 2012-2013 (OR = 0.62, 95% CI
Despite emerging recommendations and policies, from 2008 to 2013, obesity management in primary care remained suboptimal. Identifying practical strategies to enforce policies and implement evidence-based behavioral treatment in primary care should be a high priority in healthcare reform. Key words: obesity, primary care, weight management counseling
2
ACCEPTED MANUSCRIPT More than one-third of adults in the U.S. have a body mass index [BMI] ≥ 30 kg/m2 and are therefore at substantially increased risk for diabetes and cardiovascular disease (CVD).1,2 Behavioral weight management treatment is an effective first-line treatment for obesity with an average initial weight loss of 8-10%, which is associated with a significant reduction in risk for
T
diabetes and improvement in CVD risk factors.3,4 However, in 2005-2006, two-thirds of U.S.
IP
patients with obesity were not offered or referred to weight management treatment during their
CR
primary care visit.5 In addition, the rate of weight management counseling in primary care significantly decreased by 10% (40% to 30%) between 1995-1996 and2007-2008.6
US
There have been several national recommendations and policies implemented since 2008
AN
to improve obesity management in primary care. The U.S. Preventive Services Task Force (USPSTF),7 and a joint statement by the American Heart Association, American College of
M
Cardiology, and The Obesity Society8 recommend that physicians screen for overweight and
ED
obesity in their practices and provide or refer patients with risk factors for cardiovascular disease to intensive behavioral counseling. In 2011, the Center for Medicare & Medicaid Services
PT
(CMS) passed a decision to reimburse primary care physicians for delivering intensive
CE
behavioral therapy to treat patients with obesity.9 The CMS reimbursement policy is limited to coverage for Medicare beneficiaries and only reimburses primary care practitioners. When
AC
delivering the intensive behavioral therapy for obesity, physicians are expected to follow the 5 A’s counseling framework (i.e., Assess, Advise, Agree, Assist, Arrange)9-11 with 10-15 minute visits (maximum of 22 visits). In addition, CMS implemented the Electronic Health Record (EHR) Meaningful Use Incentive Program, where physicians receive financial incentives when they implement and use the EHR to document quality improvement measures.12 Physicians are incentivized to document
3
ACCEPTED MANUSCRIPT in the EHR BMI and a follow-up treatment plan to provide or refer the patient with BMI ≥ 25 to weight management treatment. The purpose of this study was to examine rates of patient BMI screening, obesity diagnosis, and provision of treatment for obesity by primary care physicians in the U.S. from
T
2008 to 2013 as well as examine the patient and practice characteristics associated with these
IP
outcomes.
CR
Methods
The National Ambulatory Medical Care Survey (NAMCS) is an annual survey conducted
US
by the National Center for Health Statistics that is used to characterize the utilization and
AN
provision of ambulatory care in the U.S.13 Currently, there is survey data publicly available from 1973 to 2013. A multi-stage probability sampling design is used, which consist of sampling
M
from primary sampling units (e.g., counties), physicians within the primary sampling units, and
ED
patient visits within practices. Using the medical chart (paper and/or electronic), physicians, office staff, or Census Bureau representatives complete physician and patient record survey
PT
forms regarding: outpatient practice characteristics, physician demographics, and visit-level data
CE
including patient demographics, reasons for the visit, diagnoses, and treatment. In terms of obesity-related data, the patient record form allowed surveyors to write in the patient height and
AC
weight, a checkbox to indicate diagnosis of obesity, and a checkbox to indicate if health education was provided (e.g., diet/nutrition, exercise, or weight reduction). In 2008-2009 and 2010-2011, the survey was completed using a paper form. However, in 2012-2013, surveys were completed mostly by Census representatives using a computer form.14 Each visit was weighted in order to obtain national estimates.
4
ACCEPTED MANUSCRIPT Similar methods applied in Ma et al.5 and Kraschnewski et al.6 to examine previous NAMCS survey year data were used in this study. We used survey years 2008-2009, 2010-2011, and 2012-2013 visit-level data with adults 18 and over who had an office-based outpatient visit with a general, family, or internal medicine physician. Because community health center visits
T
were excluded in the 20122013 surveys, community health center visits in survey years 2008-
IP
2009 and 2010-2011 were removed from the analytic dataset.14 Using the criteria specified in
CR
Ma et al.,5 patients with diabetes or coronary artery disease were considered at high risk for obesity-related disease complications and mortality. Patients with any one of the following
US
conditions were classified as moderate risk: a) hypertension, b) hyperlipidemia, c) sleep apnea,
AN
or d) asthma. Patients with no cardiovascular disease risk factors besides obesity were considered to be at low risk for obesity-related complications and mortality. We compared survey years
M
using a multivariable logistic regression model on the following study outcomes using 2008-
ED
2009 as the reference time period: 1) percent of visits with height and/or weight measured; 2) among patients with BMI ≥ 30, percent of visits with obesity diagnosis; and 3) among patients
PT
with obesity, rate of any weight-related education (i.e., any combination of diet/nutrition,
CE
exercise, and/or weight reduction education selected on the patient record form). In the model we adjusted for patient and practice-level variables that were found to be associated with these
AC
obesity-related metrics in previous studies.5,6,15 These variables included sex, age, race/ethnicity, insurance type, level of risks for obesity-related diseases or mortality, if the patient had been seen before, use of electronic medical records in the practice, and region of the U.S. in which the practice was located. Given that predictors of BMI screening, obesity diagnosis, and any weightrelated education were similar across years (data not shown), we combined all survey years to increase sample size in order to examine predictors of the outcomes for 2008-2013 (as one data
5
ACCEPTED MANUSCRIPT point) using the same patient and practice-level characteristics previously mentioned. Analyses were conducted from 2015 to early 2017 using PROC SURVEYFREQ AND SURVEYLOGISTIC in SAS v9.3. Two-sided P values less than .05 were considered significant. Results
T
There were 13,075 adult primary care visits sampled from 2008-2009, 10,951 from 2010-
IP
2011, and 31,582 visits sampled from 2012-2013. Of the total 55,608 adult primary care visits
CR
sampled, 14,143 visits (25%) were with patients with obesity. Table 1 presents the weighted proportions for visit, patient, and practice characteristics by survey year.
US
BMI Screening
AN
The measurement of both height and weight significantly increased between 2008-2009 and 2012-2013 from 54% to 73% (OR = 1.75, 95% CI 1.28-2.41) while adjusting for patient and
M
practice-level characteristics. Independent predictors of both height and weight being measured
ED
included: a) Hispanic ethnicity compared to White non-Hispanic (OR = 1.29, 95% CI 1.061.56); b) moderate risk (OR = 1.12, 95% CI 1.01-1.24) and high risk for obesity-related disease
PT
complications and mortality (OR = 1.30, 95% CI 1.14-1.49) compared to low risk; c) Medicaid
CE
compared to private insurance (OR = 1.28 (1.02-1.59); and d) presence of all electronic medical records (OR = 1.58, 95% CI 1.25-2.00) or part paper and part electronic (OR = 1.75, 95% CI
AC
1.28-2.41) versus no electronic records (Table 2). It should be noted that there was a significant increase in the use of electronic medical records between 2008-2009 and 2012-2013 (OR = 2.48, 95% CI 1.66-3.71,). Patients with Medicare versus private insurance and established patients (seen before in the clinic) versus new patients were less likely to have both height and weight measured during the visit (Table 2). Obesity Diagnosis and Weight Management Counseling
6
ACCEPTED MANUSCRIPT Figure 1 presents the weighted percentage of visits with obesity diagnosis, diet/nutrition education, exercise education, weight reduction education, or any combination of weight-related education among visits where a patient had a BMI ≥ 30. Across the years there was a nonsignificant decline in diagnosis and each specific domain of health education (Figure 1). There
T
was a significant decline of 33% to 21% in any combination of weight-related education between
IP
2008-2009 and 2012-2013 (OR = 0.62, 95% CI 0.41-0.92). Women compared to men (OR =
CR
1.41, 95% CI 1.29-1.53), Black non-Hispanic compared to White non-Hispanic (OR 1.47 (1.211.79), moderate (OR = 1.91 95% CI 1.62-2.26) and high risk patients (OR = 4.76, 95% CI 3.97-
US
5.70) compared to low risk, patients age 18-44 (OR = 2.34, 95% CI 1.86-2.94) or 45-64 (OR =
AN
1.87, 95% CI 1.56-2.24) compared to patients 65 and older, patients with Medicaid compared to those with private insurance (OR = 1.23, 95% 1.04-1.46), as well as established patients
M
compared to new (OR = 1.30, 95% 1.12-1.52) were more likely to have a diagnosis of obesity
ED
documented in their medical record (Table 3).
Weight management counseling was more likely to occur for patients who, identified as
PT
Black non-Hispanic (OR = 1.47, 95% CI 1.06-2.03), Hispanic (OR = 1.39, 95% CI 1.08-1.79) or
CE
“Other” for race/ethnicity (OR = 1.62, 95% CI 1.08-2.43), moderate (OR = 1.30 (1.02-1.65) and high risk (OR = 1.69, 95% CI 1.28-2.22), and had an obesity diagnosis documented in their
AC
medical record (OR = 3.37, 95% CI 2.88-3.95) (Table 4). Visits where weight management counseling occurred were, on weighted average, 21.95 (95% CI 21.04-22.86) minutes long. Obesity Management for Medicare Beneficiaries Only Because most of the policies to improve obesity management in primary care have been commissioned by CMS, we examined visits among Medicare beneficiaries only. There was a total of 16,620 primary care visits sampled with Medicare beneficiaries from 2008-2013 and
7
ACCEPTED MANUSCRIPT 3,863 (23%) of these were with patients with obesity. Percent of visits with both height and weight measured increased from 49% to 72% between 2008-2009 and 2012-2013 (OR = 1.57, 95% CI 1.05-2.34). Among visits with patients with Medicare and BMI ≥ 30, there was a nonsignificant decline in percent of visits with an obesity diagnosis (31% to 24%) and a non-
T
significant decline in percent of visits with any combination of weight-related education (28% to
IP
20%) between 2008-2009 and 2012-2013.
CR
Discussion
NAMCS data between 2008-2009 and 2012-2013 indicated a significant increase in BMI
US
screening (i.e., measure of both height and weight); however, the percent of visits with a
AN
documented obesity diagnosis declined by 5 percentage points (30% to 25%). Despite emerging national recommendations and healthcare policies, provision or referral to weight management
M
counseling significantly declined from 33% to 21% between 2008-2009 and 2012-2013.
ED
Medicare beneficiaries had similar increases BMI screening and decreases in obesity diagnosis and weight management counseling compared to the total population.
PT
There was a significant increase in practices with electronic medical records, which was a
CE
predictor of both height and weight being measured. Most electronic medical records allow easy input of vital signs and automatic calculation of BMI. Other practice characteristics that
AC
predicted measurement of both height and weight included Hispanic ethnicity, Medicaid insurance, and moderate and high risks for obesity-related diseases and mortality. Established patients or patients with Medicare were less likely to have both height and weight measured. This may be due to only weight and not height being measured during an encounter because the height for established and adult patients is assumed to be the same at each visit. However, in most electronic health records, an updated BMI cannot be calculated without both a weight and
8
ACCEPTED MANUSCRIPT height entered at each visit even if the height has not changed. It should be noted that weight was measured at 87% or more visits across the survey years, so the percent of visits with BMI screening would be higher if height did not need to be entered to update BMI. There is a need for more built in sophisticated clinical decision support tools within the electronic medical record
T
that can automatically extract previous vital sign data such as last recorded height and include it
IP
in calculations of BMI once a new weight has been entered during an encounter. This
CR
improvement may decrease burden on physicians or medical staff and perhaps increase screening and diagnosis.
US
Across the survey years, 70% or more of patients with obesity did not have a documented
AN
obesity diagnosis. Although electronic medical records facilitate the input of vital signs, documentation of an obesity diagnosis requires the provider to take additional steps. To
M
diagnose obesity, physicians would need to add obesity to the patient’s problem list. However,
ED
there is no real incentive for physicians to diagnose obesity since it is still not considered a billable medical condition by most insurance companies, with the exception of Medicare.
PT
Despite the overall low rate of diagnosed obesity, patients who were female, 18-64 years of age,
CE
Black non-Hispanic, Medicaid, established patient, or at moderate or high risk for obesity-related disease complications were more likely to have a diagnosis. Previous survey based studies have
AC
indicated that women as well as young and middle-aged adults are more likely to have an obesity diagnosis, perhaps because members of these populations are more likely to discuss concerns about their weight with their physician that prompted an obesity diagnosis at the end of the visit.15,16 Patients at high risk for obesity-related disease complications and mortality perhaps prompt physicians to regularly monitor their BMI and address progress with lifestyle changes and weight loss in order to prevent CVD events. Similarly, patients from racial/ethnic minority
9
ACCEPTED MANUSCRIPT backgrounds or with Medicaid may also be more likely to be at high risk for obesity-related diseases and more frequent users of care;16 thus maybe increasing the likelihood of receiving an obesity diagnosis and some health education. Given the findings in Kraschnewski et al., 6 the percent of primary care visits with weight
T
management counseling continued to significantly decline by 12 percentage points between
IP
2008-2009 and 2012-2013, despite national recommendation and policies established during this
CR
time period. As previously mentioned, there is no immediate financial incentive to address and manage obesity in the primary care office for adult patients 18-64 years old.17 Although,
US
Medicare reimburses intensive behavioral therapy for obesity for their beneficiaries,9 the
AN
occurrence of weight management counseling during a visit also decreased among this patient population. Most physicians are unaware of this reimbursement policy or how to effectively
M
implement it without disrupting current clinical workflow given the number of visits that must
ED
occur (i.e., 22 total). Additional physician specific barriers to providing weight management counseling include time constraints, discomfort with discussing weight issues, lack of training in
PT
weight management, and lack of knowledge regarding available treatments.18-21 Patients with an
CE
obesity diagnosis were three times more likely to receive weight management counseling than those without a diagnosis. Visits that did include weight management counseling were about 20
AC
minutes in duration, which is consistent with the expected length of an intensive behavioral therapy session using the CMS model (i.e., 10-15 minutes).9 However, based on 2005-2006 NAMCS data, only 8% (i.e., ~1.6 minutes) of the 20 minute visit was actually spent addressing obesity.22 Thus, even if weight management counseling is indicated in the medical chart that does not guarantee that a sufficient amount of time is spent providing high quality counseling. Limitations
10
ACCEPTED MANUSCRIPT There are several limitations in this study. First, the NAMCS data is a cross-sectional survey based on only one random week of clinic visits per physician over a year. Thus, diagnosis of obesity and weight management counseling could have occurred during another encounter not included in the survey sample. Second, the focus on primary care visits as the unit of analysis,
T
may have resulted in oversampling of sicker patients or frequent users of primary care. Third,
IP
since the CMS Electronic Medical Record Meaningful Use Incentive Program was not fully
CR
implemented until 2011,23 perhaps not enough time has passed to see the impact of financial incentives on obesity management in primary care. Fourth, there was a change in the sampling
US
design and methodology for the 2012-2013 NAMCS.14 Specifically, community health center
AN
visits were excluded, Census Bureau representatives were more likely to complete the survey forms than physicians and clinic staff, and the survey was completed using a computerized form
M
instead of a paper form. Increased use of the Census Bureau representatives and computerized
ED
forms may explain the increased number of visits sampled in 2012-2013. However, it is not clear if it was these changes or actual practice patterns that impacted the increase or decrease in
CE
Conclusions
PT
documentation of obesity-related metrics.14
Despite emerging national recommendations and policies since 2008, obesity
AC
management in primary care is still suboptimal. A recent study indicated that a majority of patients want to discuss weight loss with their physician.24 Thus, there is a need for primary care delivery redesign to facilitate rather than hinder physicians addressing obesity with their patients. Part of this redesign is increasing physician skills in starting the conversation to sensitively address obesity with a patient.25 It is not realistic to expect primary care physicians to deliver intensive behavioral weight loss counseling to all of their patients with obesity. Thus, the
11
ACCEPTED MANUSCRIPT process of ordering referrals and coverage of obesity management specialist (e.g., registered dietitians, psychologists) and community-based programs should be made easier in order to increase referral options for physicians and access to care for patients.26 In addition, the electronic medical record is emerging as a tool to not only facilitate BMI screening, but could
T
also be used to facilitate weight management counseling during an encounter.27,28 Given the
IP
obesity epidemic and increased risk for chronic diseases, identifying practical strategies to
CR
enforce policies and implement evidence-based treatment services in primary care should be a
AC
CE
PT
ED
M
AN
US
high priority in healthcare reform.
12
ACCEPTED MANUSCRIPT References 1.
Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. JAMA. 2012;307(5):491497. Ogden CL, Carroll MD, Kit BK, Flegal KM. Prevalence of childhood and adult obesity in
Butryn ML, Webb V, Wadden TA. Behavioral treatment of obesity. Psychiatr Clin North
CR
3.
IP
the United States, 2011-2012. JAMA. 2014;311(8):806-814.
Am. 2011;34(4):841-859.
Wadden TA, Webb VL, Moran CH, Bailer BA. Lifestyle modification for obesity: new
US
4.
T
2.
AN
developments in diet, physical activity, and behavior therapy. Circulation. 2012;125(9):1157-1170.
Ma J, Xiao L, Stafford RS. Adult obesity and office-based quality of care in the United
M
5.
6.
ED
States. Obesity (Silver Spring, Md.). 2009;17(5):1077-1085. Kraschnewski JL, Sciamanna CN, Stuckey HL, et al. A silent response to the obesity
Moyer VA, Force USPST. Screening for and management of obesity in adults: U.S.
CE
7.
PT
epidemic: decline in US physician weight counseling. Med Care. 2013;51(2):186-192.
Preventive Services Task Force recommendation statement. Ann Intern Med.
8.
AC
2012;157(5):373-378.
Jensen MD, Ryan DH, Apovian CM, et al. 2013 AHA/ACC/TOS Guideline for the Management of Overweight and Obesity in Adults: A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and The Obesity Society. J Am Coll Cardiol. 2013.
9.
DHHS Centers for Medicare and Medicaid Services. Intensive Behavioral Therapy (IBT) for Obesity. Medicare Learning Network; 2012. 13
ACCEPTED MANUSCRIPT 10.
Vallis M, Piccinini-Vallis H, Sharma AM, Freedhoff Y. Clinical review: modified 5 As: minimal intervention for obesity counseling in primary care. Can Fam Physician. 2013;59(1):27-31.
11.
Alexander SC, Cox ME, Boling Turer CL, et al. Do the five A's work when physicians
T
counsel about weight loss? Fam Med. 2011;43(3):179-184. Center for M, Medicaid S. EHR Incentive Programs. Vol 2013.
13.
National Center for Health S. NAMCS/NHAMCS Ambulatory Health Care Data. Vol
CR
IP
12.
2016.
National Center for Health S. NAMCS/NHAMCS Questionnaires, Datasets, & Related
US
14.
15.
AN
Documentation. Vol 2016.
Ahmed NU, Delgado M, Saxena A. Trends and disparities in the prevalence of
16.
ED
Prev Med. 2016;89:70-75.
M
physicians' counseling on diet and nutrition among the U.S. adult population, 2000-2011.
Savageau JA, McLoughlin M, Ursan A, Bai Y, Collins M, Cashman SB. Characteristics
PT
of frequent attenders at a community health center. J Am Board Fam Med.
17.
CE
2006;19(3):265-275.
Kushner RF, Sarwer DB. Medical and behavioral evaluation of patients with obesity.
18.
AC
Psychiatr Clin North Am. 2011;34(4):797-812. Bleich SN, Bennett WL, Gudzune KA, Cooper LA. National survey of US primary care physicians' perspectives about causes of obesity and solutions to improve care. BMJ open. 2012;2(6):10.1136/bmjopen-2012-001871. Print 002012. 19.
Bleich SN, Bennett WL, Gudzune KA, Cooper LA. Impact of physician BMI on obesity care and beliefs. Obesity (Silver Spring, Md.). 2012;20(5):999-1005.
14
ACCEPTED MANUSCRIPT 20.
Kolasa KM, Rickett K. Barriers to providing nutrition counseling cited by physicians: a survey of primary care practitioners. Nutr Clin Pract. 2010;25(5):502-509.
21.
Kushner RF. Barriers to providing nutrition counseling by physicians: a survey of primary care practitioners. Prev Med. 1995;24(6):546-552. Tsai AG, Abbo ED, Ogden LG. The time burden of overweight and obesity in primary
IP
care. BMC Health Serv Res. 2011;11:191-6963-6911-6191.
T
22.
Center for Medicare and Medicaid Services. EHR Incentive Programs. Vol 2014.
24.
Sherson EA, Yakes Jimenez E, Katalanos N. A review of the use of the 5 A's model for
CR
23.
US
weight loss counselling: differences between physician practice and patient demand. Fam
Stop Obesity Alliance. Why Weight? A Guide to Discussing Obesity & Health With Your Patients. Vol 2014.
Fitzpatrick SL, Wischenka D, Appelhans BM, et al. An Evidence-based Guide for
ED
26.
M
25.
AN
Pract. 2014;31(4):389-398.
Obesity Treatment in Primary Care. Am J Med. 2016;129(1):115.e1-7. Bordowitz R, Morland K, Reich D. The use of an electronic medical record to improve
PT
27.
Tang JW, Kushner RF, Cameron KA, Hicks B, Cooper AJ, Baker DW. Electronic tools to assist with identification and counseling for overweight patients: a randomized
AC
28.
CE
documentation and treatment of obesity. Fam Med. 2007;39(4):274-279.
controlled trial. J Gen Intern Med. 2012;27(8):933-939.
15
ACCEPTED MANUSCRIPT Table 1. Patient, Practice, & Visit Characteristics of U.S. Adult Primary Care Visits by Survey Year: %†, (95% CI) 2008-2009 (n = 13,075) 2010-2011 (n = 10,951) 2012-2013 (n = 31,582) Age group, y 18-44 29.0 (27.9, 30.8) 30.0 (27.9, 32.2) 27.7 (26.4, 29.0) 45-64 38.1 (36.8, 39.5) 38.3 (36.8, 39.7) 38.3 (37.3, 39.3) 65 and up 32.9 (30.8, 35.0) 31.7 (29.3, 34.1) 34.0 (32.6, 35.4) Sex Female 59.8 (58.1, 61.4) 57.0 (55.1, 58.8) 57.3 (56.1, 58.5) Male 40.2 (38.6, 41.9) 43.0 (41.2, 44.9) 42.7 (41.5, 43.9) Race/Ethnicity White non-Hispanic 74.9 (71.7, 78.1) 73.4 (69.3, 77.4) 73.0 (70.9, 75.2) Black non-Hispanic 10.0 (7.7, 12.3) 12.7 (9.6, 15.9) 10.1 (8.9, 11.2) Hispanic 11.0 (8.7, 13.4) 9.2 (6.0, 12.5) 12.5 (10.7, 14.3) Other 4.1 (2.7, 5.5) 4.7 (3.0, 6.3) 4.4 (3.6, 5.2) 20.4 (18.4, 22.4) 21.4 (18.9, 23.9) 28.9 (28.2, 29.6) Body Mass Index ≥ 30 kg/m2 Risks for Obesity-Related Disease/Mortality Low 21.3 (19.6, 22.9) 20.5 (18.7, 22.4) 20.2 (19.0, 21.3) Moderate 55.4 (54.0, 56.9) 55.5 (53.9, 57.0) 57.2 (56.2, 58.3) High 23.3 (21.6, 25.0) 24.0 (21.9, 26.1) 22.6 (21.6, 23.6) Insurance Private 57.7 (54.9, 60.5) 53.8 (50.4, 57.1) 50.8 (48.9, 52.6) Medicare 29.2 (27.0, 31.5) 31.5 (29.0, 34.1) 33.9 (32.3, 35.4) Medicaid 5.8 (4.3, 7.4) 7.2 (5.9, 8.5) 7.9 (6.7, 9.0) Other 7.2 (5.5, 9.0) 7.5 (5.9, 9.2) 7.5 (6.2, 8.7) Electronic Medical Record System No 46.7 (40.5, 52.9) 37.8 (32.3, 43.3) 23.3 (20.3, 26.3) Yes, part paper part electronic 14.4 (9.8, 19.0) 9.3 (5.7, 12.9) 11.6 (9.0, 14.2) Yes, all electronic 38.9 (32.3, 45.4) 52.8 (47.4, 58.3) 65.1 (61.6, 68.6) Region Northeast 16.1 (11.1, 21.1) 19.8 (15.8, 23.8) 19.3 (17.7, 20.9) Midwest 25.8 (20.0, 31.6) 24.2 (20.0, 28.4) 19.8 (18.5, 21.0) South 37.9 (31.9, 43.8) 35.1 (30.0, 40.2) 36.2 (34.4, 38.0)
T P
I R
C S
U N
A
D E
M
T P
E C
C A
16
ACCEPTED MANUSCRIPT West 20.2 (15.6, 24.9) 20.9 (17.2, 24.6) 24.8 (22.9, 26.6) Height & Weight measured Neither 11.2 (9.0, 13.4) 10.3 (8.0, 12.5) 8.1 (6.8, 9.3) Either 34.6 (30.3, 38.9) 33.8 (29.2, 38.5) 18.7 (16.7, 20.8) Both 54.1 (50.0, 58.7) 55.9 (50.6, 61.2) 73.2 (70.8, 75.5) † All percentages are population percentages estimated from a weighted analysis taking into account the complex sampling stratification and clustering.
T P
I R
C S
U N
A
D E
M
T P
E C
C A
17
ACCEPTED MANUSCRIPT Table 2. Predictors of Body Mass Index Screening, 2008-2013 Adjusted OR (95% CI) 1.00 [Reference] 1.04 (0.83-1.31) 1.75 (1.28-2.41)
IP
T
1.09 (0.95-1.24) 1.02 (0.92-1.14) 1.00 [Reference]
CR
1.04 (0.97-1.11) 1.00 [Reference]
AC
CE
PT
ED
M
1.00 [Reference] 1.21 (0.93-1.58) 1.29 (1.06-1.56) 1.18 (0.88-1.58)
US
AN
Survey Years 2008-2009 2010-2011 2012-2013 Age group, y 18-44 45-64 65 and up Sex Female Male Race/Ethnicity White non-Hispanic Black non-Hispanic Hispanic Other Risks for Obesity-Related Disease/Mortality Low Moderate High Insurance Private Medicare Medicaid Other Use an Electronic Medical Record System No Yes, part paper part electronic Yes, all electronic Patient seen before? No Yes Region Northeast Midwest South West
1.00 [Reference] 1.12 (1.01-1.24) 1.30 (1.14-1.49) 1.00 [Reference] 0.82 (0.72-0.92) 1.28 (1.02-1.59) 0.83 (0.63-1.08) 1.00 [Reference] 1.75 (1.28-2.41) 1.58 (1.25-2.00) 1.00 [Reference] 0.83 (0.71-0.97) 0.92 (0.61-1.37) 1.00 (0.70-1.44) 1.21 (0.83-1.75) 1.00 [Reference]
18
ACCEPTED MANUSCRIPT Table 3. Predictors of Obesity Diagnosis, 2008-2013 Adjusted OR (95% CI) 1.00 [Reference] 1.08 (0.90-1.29) 1.11 (0.86-1.45)
IP
T
2.34 (1.86-2.94) 1.87 (1.56-2.24) 1.00 [Reference]
CR
1.41 (1.29-1.53) 1.00 [Reference]
AC
CE
PT
ED
M
1.00 [Reference] 1.47 (1.21-1.79) 1.14 (0.94-1.39) 0.55 (0.42-0.71)
US
AN
Survey Years 2008-2009 2010-2011 2012-2013 Age group, y 18-44 45-64 65 and up Sex Female Male Race/Ethnicity White non-Hispanic Black non-Hispanic Hispanic Other Risks for Obesity-Related Disease/Mortality Low Moderate High Insurance Private Medicare Medicaid Other Use an Electronic Medical Record System No Yes, part paper part electronic Yes, all electronic Patient seen before? No Yes Region Northeast Midwest South West
1.00 [Reference] 1.91 (1.62-2.26) 4.76 (3.97-5.70) 1.00 [Reference] 0.93 (0.77-1.11) 1.23 (1.04-1.46) 1.22 (0.85-1.74) 1.00 [Reference] 1.12 (0.92-1.38) 1.10 (0.94-1.29) 1.00 [Reference] 1.30 (1.12-1.52) 1.35 (1.00-1.82) 1.26 (0.98-1.62) 1.03 (0.80-1.33) 1.00 [Reference]
19
ACCEPTED MANUSCRIPT Table 4. Predictors of Any Weight-Related Education, 2008-2013 Adjusted OR (95% CI) 1.00 [Reference] 0.94 (0.69-1.28) 0.62 (0.41-0.92)
IP
T
0.99 (0.77-1.27) 1.09 (0.89-1.33) 1.00 [Reference]
CR
0.94 (0.82-1.07) 1.00 [Reference]
AC
CE
PT
ED
M
1.00 [Reference] 1.47 (1.06-2.03) 1.39 (1.08-1.79) 1.62 (1.08-2.43)
US
AN
Survey Years 2008-2009 2010-2011 2012-2013 Age group, y 18-44 45-64 65 and up Sex Female Male Race/Ethnicity White non-Hispanic Black non-Hispanic Hispanic Other Risks for Obesity-Related Disease/Mortality Low Moderate High Insurance Private Medicare Medicaid Other Obesity Diagnosis No Yes Use an Electronic Medical Record System No Yes, part paper part electronic Yes, all electronic Patient seen before? No Yes Region Northeast Midwest South West
1.00 [Reference] 1.30 (1.02-1.65) 1.69 (1.28-2.22) 1.00 [Reference] 0.78 (0.63-0.98) 0.97 (0.68-1.37) 1.24 (0.77-2.00) 1.00 [Reference] 3.37 (2.88-3.95) 1.00 [Reference] 1.00 (0.70-1.44) 1.00 (0.79-1.28) 1.00 [Reference] 1.04 (0.81-1.32) 0.82 (0.51-1.30) 0.69 (0.47-1.03) 0.60 (0.39-0.94) 1.00 [Reference]
20
ACCEPTED MANUSCRIPT
35%
b
30% 25% 20% 15%
T
10%
2008-2009 (n = 2693)
IP
5%
2010-2011 (n = 2426)
0%
AN
US
CR
2012-2013 (n = 9024)
AC
CE
PT
ED
M
Figure 1. Weighted percentage of patient visits with obesity diagnosis, diet/nutrition education, exercise education, weight reduction education, and any weight-related education by survey year among visits with patients with body mass index ≥ 30. a Adjusted for age, sex, race/ethnicity, insurance type, risks for obesity-related diseases or mortality, if the patient had been seen before, presence of electronic medical record, and practice region. b Percent of visits with any weight-related education in 2012-2013 was significantly lower than the percent in 2008-2009 (OR = 0.62, 95% CI 0.41-0.92).
21
ACCEPTED MANUSCRIPT Highlights More than one-third of adults in the U.S. meet criteria for obesity.
Screening for obesity significantly increased 2008 to 2013.
From 2008 to 2013, weight counseling in primary care significantly declined.
Racial/ethnic minorities and high risk for CVD more likely to receive counseling.
Less than 1/3 Medicare patients received weight counseling each year, 2008 to 2013.
AC
CE
PT
ED
M
AN
US
CR
IP
T
22