The impact of obesity on office-based physician visits

advertisement
The impact of obesity on office-based physician visits
William S. Pearson, Ph.D., M.H.A.1
Kavitha Bhat-Schelbert, M.D., M.S.2
Earl S. Ford, M.D., M.P.H.1
Ali H. Mokdad, Ph.D., M.P.H.3
1. Division of Adult and Community Health, Centers for Disease Control and Prevention, Atlanta
2. Department of Family Medicine, University of Pittsburgh Medical Center, Pittsburgh
3. Institute for Health Metrics, University of Washington, Seattle
Approximately two-thirds of American adults are overweight
and one-third of those are obese (Ogden, et al, JAMA 2006).
Background

The direct medical cost of overweight and obesity in the U.S. was estimated
to be $96.2 billion in 2002 dollars (Finkelstein, 2002).

Obesity accounts for 6-10% of medical expenditures in the U.S. (Thompson,
2001).

Between 1987 and 2001, the increase in obesity accounted for a 12%
increase in overall healthcare spending (Thorpe, 2004).

As obesity increases, visits to physicians for the care of co-morbid chronic
diseases such as diabetes increases (Pearson, 2007).

Raebel and colleagues estimated that for each unit increase in BMI, costs
related to inpatient and outpatient care increase by 2.3% (Raebel, 2004).
Background


Relatively little information is known about the
impact of obesity on the patient visit in terms of
resource utilization.
This study examined the amount of time spent
with the provider during the visit and the number
of prescription medications mentioned at each
visit.
Methods

2006 National Ambulatory Medical Care Survey (NAMCS).

Survey of non-federally-employed, office-based physician visits,
including community health centers, conducted yearly by
NCHS.

Survey uses a multi-stage probability design.
 First stage = Geographic Primary Sampling Units
 Second stage = Physician practices within PSU
 Third stage = Patient visits within the practices
Methods

Variables collected for this analysis:
age, race, sex of the patient
 obesity status
 payment source
 major reason for the visit
 total number of chronic conditions
 number of medications listed at the visit
 amount of time spent with the provider during the visit

Methods

Definition of variables
 Age categorized into four groups (18-34, 35-49, 50-64,
65+)
 Race classified as White, Black, Other
 Payment source was private pay, Medicare, Medicaid,
or other sources such as workers comp. or charity care,
combined with unknown sources
 Reason for visit: new problem, routine visit for a chronic
problem, visit for a flare-up of a chronic problem, visit
for pre or post surgery review, visit for preventive care
Methods

Fourteen co-morbid conditions: arthritis, asthma,
cancer, cerebrovascular disease, CHF, chronic renal
failure, COPD, depression, diabetes, hyperlipidemia,
hypertension, ischemic heart disease, obesity,
osteoporosis

Time spent with the provider in minutes

Total count of mentioned prescription medications
Methods

Analyses

24,392 visits were stratified by BMI of the patient.
 BMI
> 30 kg/m2 and BMI < 30 kg/m2.
Estimates of visit characteristics were made.
 Average time spent with the provider and average
number of medications mentioned at each visit were
calculated for each BMI category.
 Time spent with the provider was rounded to the closest
five minute interval among those with BMI < 30 kg/m2.
 Number of medications among those with BMI < 30
kg/m2 was rounded down to the closest whole number.

Methods

Analyses
 Chi-square analyses were conducted comparing BMI
categories for proportions of visits that lasted > 20 minutes
and those that had > 2 medications mentioned.
 Logistic regression models were used to further test
differences between BMI categories, controlling for:
 1. age, sex and race of patient.
 2. age, sex, race, payment source, major reason for visit,
and total number of co-morbid chronic conditions.
 All tests conducted at α = 0.05.
 All analyses conducted in SUDAAN to account for
complex sampling design.
Results
Visit Characteristics
N = 24,239
BMI < 30 kg/m2
n=20,058
%, (% S.E.)
BMI > 30 kg/m2
n=4,181
%, (% S.E.)
Age in years (%)
18–34
35–49
50–64
65+
19.2 (0.9)
22.4 (0.6)
25.0 (0.6)
33.5 (1.1)
14.8 (1.0)
28.6 (1.1)
32.4 (1.0)
24.2 (1.3)
Sex (%)
Female
Male
61.5 (0.8)
38.5 (0.8)
61.3 (1.4)
38.7 (1.4)
Race (%)
White
Black
Other
86.2 (1.1)
8.2 (0.8)
5.6 (0.9)
83.1 (1.7)
13.5 (1.7)
3.5 (0.6)
Results
Visit Characteristics
N = 24,239
BMI < 30 kg/m2
n=20,058
BMI > 30 kg/m2
n=4,181
Payment Source (%)
Private Pay
Medicare
Medicaid
Other/Unknown*
48.6 (1.3)
27.6 (1.2)
8.8 (0.7)
15.0 (1.2)
53.2 (1.9)
21.7 (1.3)
10.6 (1.1)
14.5 (1.5)
Major Reason for Visit (%)
New problem
Chronic problem, routine
Chronic problem, flare-up
Pre/post surgery
Preventive Care
30.8 (0.9)
33.7 (1.4)
8.3 (0.5)
8.2 (0.6)
17.2 (1.1)
33.4 (1.5)
34.4 (1.7)
10.4 (0.8)
5.4 (0.7)
15.0 (1.3)
Number of co-morbid Chronic Conditions (%)
0
41.2 (1.4)
1–2
42.0 (1.1)
3+
16.8 (0.9)
25.5 (1.5)
46.6 (1.4)
27.9 (1.8)
Results
Visit Characteristics
N = 24,239
BMI < 30 kg/m2 BMI > 30 kg/m2
n=20,058
n=4,181
Average amount of time spent
with provider in minutes
20.8 (0.3)
22.8 (0.8)
Average number of meds
prescribed at visit
2.2 (0.1)
3.1 (0.1)
Results
Chi-square comparisons
BMI < 30 kg/m2 BMI > 30 kg/m2
n=20,058
n=4,181
%, (% SE)
%, (% SE)
p*
% of visits where time spent
with provider was > 20 minutes
26.9 (1.17)
31.5 (1.76)
.005
% of visits where number of
medications prescribed was > 2
32.2 (1.30)
49.2 (2.15)
<.001
Results
Amount of time spent with provider during visit > 20 minutes.
Model 1*
BMI > 30 kg/m2 (n = 4,181)
BMI < 30 kg/m2 (n = 20,058)
OR
95% CI
1.21
Ref.
1.05–1.41
Ref.
1.17
Ref.
1.02–1.36
Ref.
* controlling for age, sex & race
Model 2**
BMI > 30 kg/m2 (n = 4,181)
BMI < 30 kg/m2 (n = 20,058)
** controlling for age, sex, race, payment source, number of chronic conditions, &
major reason for visit
Results
Number of prescribed medications during visit > 2.
Model 1*
BMI > 30 kg/m2
BMI < 30 kg/m2
OR
95% CI
2.21
Ref.
1.89–2.58
Ref.
1.83
Ref.
1.57–2.14
Ref.
* controlling for age, sex & race
Model 2**
BMI > 30 kg/m2
BMI < 30 kg/m2
** controlling for age, sex, race, payment source, number of chronic conditions, &
major reason for visit
Discussion

Findings from this study:

First nationally-representative study to document impact
on provider time and numbers of medications.

Findings are significant after controlling for number of comorbid conditions and primary reason for the visit.
Discussion

Findings from other studies:

Andreyeva, 2004 – Health and Retirement Study – 33% increase in
cost with BMI > 30 kg/m2

Bertakis, 2005 – Prospective study of 509 participants – obesity was
associated with increased primary care visits, but not lengthier
visits.

Molenaar, 2008 – Framingham Heart Study – obese persons were
using more prescription drugs & more likely to receive medications
for hypertension and hyperlipidemia.

von Lengerke, 2005; Saez, 2006; van Dijk, 2006 – population based
studies in Europe – positive association between obesity and
increased usage of primary care.
Discussion

Limitations of this study:
 1. Exact dollar amounts not determined.
 2. Sampling unit is the visit, not the person.
 Possibility of multiple visits by one person.
 3. Data only on in-person visits.
Discussion

Implications of this study:

These findings demonstrate how obesity impacts the US
ambulatory healthcare system.

Aging of the country and increased prevalence of chronic
disease will place stronger demand on burdened medical
resources.

Obesity will only exacerbate this problem.

Obesity is a modifiable health issue.
Discussion

Future directions:



1. Demonstrate impact on ancillary services.
 Nursing services
 Physical therapy
 Dietary services
2. Demonstrate impact in other care settings
 Surgical services
 Long-term care
3. Focus on time not directly spent with patient
Download