UNDERSTANDING THE DISPROPORTIONALLY HIGH INCIDENCE OF CARDIAC
ARRESTS IN THE ALLEGHENY COUNTY POPULATION UNDER THE AGE OF 65
by
Thomas Uray
MD, University of Vienna, Austria, 2001
Submitted to the Graduate Faculty of
the Multidisciplinary Master of Public Health Program
Graduate School of Public Health in partial fulfillment
of the requirements for the degree of
Master of Public Health
University of Pittsburgh
2013
i
UNIVERSITY OF PITTSBURGH
GRADUATE SCHOOL OF PUBLIC HEALTH
This essay is submitted
by
Thomas Uray
on
August 7th, 2013
and approved by
Essay Advisor:
David Finegold, MD
Director, Multidisciplinary MPH Program
Professor of Pediatrics
Department of Human Genetics
Graduate School of Public Health
University of Pittsburgh
Essay Reader:
Cameron Dezfulian, MD
Assistant Professor
Department of Critical Care Medicine
Vascular Medicine Institute
School of Medicine
University of Pittsburgh
________________________________
________________________________
ii
Copyright © by Thomas Uray
2013
iii
David Finegold, MD
UNDERSTANDING THE DISPROPORTIONALLY HIGH INCIDENCE OF CARDIAC
ARRESTS IN THE ALLEGHENY COUNTY POPULATION UNDER THE AGE OF 65
Thomas Uray, MPH
University of Pittsburgh, 2013
ABSTRACT
Introduction- Every year approximately 300,000 out of hospital cardiac arrests (OHCA)
occur in the US. Across seven large North American urban populations, the City of Pittsburgh
had the highest crude rate in cardiac arrests. This discrepancy was most pronounced in young
patients 18 to 64 years of age with the lowest quartile median household income. The aim of this
thesis was to test the hypothesis that people in Pittsburgh who suffer a cardiac arrest live in
poorer neighborhoods compared to the Pittsburgh/Allegheny/National average. In addition, I
wanted to test the additional hypothesis that these cardiac arrest victims have more preexisting
comorbid conditions and engage in poor health behaviors (e.g. illicit drug use).
Methods- A retrospective analysis of all treated in- and out-of hospital cardiac arrest
patients at an inner-city hospital during a 16-month period was analyzed. Patients with all causes
of cardiac arrest between the age of 18 and 64 years were included. Next to existing data about
neurological outcome and basic demographic data, data about residence, insurance status,
employment, marital status, history of drug abuse and urine toxicology screening was analyzed.
Patient’s residences were geocoded and depicted in a map of Allegheny, which also contained
census tracts color coded by poverty status.
iv
Results- Between March 2011 and July 2012, 95 patients between the ages of 18 and 64
(median age of 53) were included in this analysis. Among those, 69 were living in Allegheny
County at time of arrest. 19 patients (20.2%) had a history of active drug abuse, and forty
patients (42.1%) had no insurance or Medicaid or Medicare listed as support. 25 (36.2%) were
living in a neighborhood with 20-30% poverty at the time of the arrest.
Conclusion- Cardiac arrest patients between the age of 18 and 64 treated at UPMC
Mercy have a higher incidence of being unemployed, uninsured, living alone, being obese and
using drugs than the rest of the population in Allegheny County. This major public health
problem needs to be addressed by local authorities by designing special prevention programs
targeted to this subpopulation.
v
TABLE OF CONTENTS
1.0
INTRODUCTION .........................................................................................................1
2.0
METHODS ....................................................................................................................3
3.0
RESULTS ......................................................................................................................5
3.1
DEMOGRAPHICS ...............................................................................................5
3.2
COMORBID CONDITIONS AND HEALTH BEHAVIORS ..........................5
3.3
SOCIAL BACKGROUND ...................................................................................6
3.4
NEIGHBORHOOD POVERTY..........................................................................7
3.5
MEDICAL CHARACTERISTICS .....................................................................7
3.6
OUTCOMES .........................................................................................................8
4.0
DISCUSSSION ..............................................................................................................9
4.1
LIMITATIONS ...................................................................................................11
4.2
CONCLUSION ...................................................................................................12
APPENDIX: TABLES AND FIGURES.................................................................................. 13
BIBLIOGRAPHY ........................................................................................................................19
vi
LIST OF TABLES
Table 1. Race and social characteristics; good versus bad outcome .............................................13
vii
LIST OF FIGURES
Figure 1. Insurance status ............................................................................................................. 14
Figure 2. Employment status ........................................................................................................ 14
Figure 3. Etiology overall ............................................................................................................. 15
Figure 4. In-hospital arrests (n=55) .............................................................................................. 15
Figure 5. Out-of hospital arrests (n=40) ....................................................................................... 16
Figure 6. Mapping of residences ................................................................................................... 17
Figure 7. Percentage of poverty in area of residence (n=69) ........................................................ 18
viii
1.0
INTRODUCTION
Ischemic heart disease is the leading cause of death in the United States.1 Every year,
approximately 300,000 out of hospital cardiac arrests (OHCA) occur in the US with an overall
survival rate of around eight percent.2,3 Previous cardiac arrest studies have shown that incidence
and survival rates vary by race. For instance, black people in urban Chicago were more likely to
suffer an OHCA across all age groups and had an overall 3-fold lower survival rate compared to
white people.4 These findings are consistent with data that link the ‘typical’ cardiac arrest victim
to an adverse socioeconomic background, i.e. lower average household income, lower
educational level, and poor neighborhood residence.5-8
Indeed, the annual incidence of cardiac arrests was approximately 2-fold higher in poor
vs. wealthy neighborhoods across seven large North American urban populations.7,8
Interestingly, Pittsburgh (Pennsylvania, USA) had the highest crude cardiac arrest incidence rate
among these seven populations (159.3 per 100,000 in Pittsburgh vs. 71.5 per 100,000 overall
incidence, respectively).8 This discrepancy was most pronounced in young patients 18 to 64
years of age with the lowest quartile median household income.8
The aim of this study was to understand factors contributing to the disproportionally high
cardiac arrest incidence rate among Pittsburghers 18-64 years of age. Previous data from
Allegheny County showed a 73% higher coronary heart disease mortality in black men compared
to white men.9
1
We tested the hypotheses that patients 18-64 years of age who reside in Allegheny county
and suffered a cardiac arrest are more likely to be black, are more likely live in a poor
neighborhood, are more likely to have preexisting co-morbid conditions (e.g. hypertension) or
unhealthy behaviors (e.g. smoking) that predispose them to cardiovascular disease and
potentially a cardiac arrest. We also examined whether cardiac arrest victims are less likely to
have health insurance compared to the overall Allegheny population. To put our findings into
perspective, we compared the results form Allegheny County with the national average.
2
2.0
METHODS
We conducted a retrospective study and analyzed cardiac arrest patients 18-64 years of
age that were transferred to University of Pittsburgh Medical Center (UPMC) Mercy Hospital
(Pittsburgh, PA) between March 1, 2011 and July 30, 2012.
All in- and out-of hospital cardiac arrests were identified through a preexisting cardiac
arrest database. Clinical and demographic data was obtained by cross-linking the cardiac arrest
database with individual electronic medical records.
We abstracted the following information from patient charts: Full address and zip code,
insurance status, race, marital status, religious status, first spoken language, history of drug
abuse, known past history of illicit drug abuse, urine drug screen at admission, smoking status,
preexisting hypertension, diabetes, hyperlipidemia, cardiovascular arterial disease (CAD), stroke,
and cause of arrest. We calculated body mass index (BMI) to classify overweight and obesity.10
Last, based on abstracted preexisting co-morbid conditions, we calculated the Charlson comorbidity index (CCI) for every patient.11
The cardiac arrest database contained information regarding initial cardiac arrest rhythm,
specific cardiac arrest treatments (i.e. mild therapeutic hypothermia), neurologic outcome
(cerebral performance category [CPC]12), date of withdrawal of care and time of death where
applicable.
3
A cerebral performance score of 1 (good function) or 2 (moderate disability) on a 5category scale was considered a good outcome; the remaining categories were 3 (severe
disability), 4 (a vegetative state), and 5 (death). Patients with good recovery or moderate
disability typically have sufficient cerebral function to live independently and work at least parttime.12
Patient addresses were geocoded and depicted in a map of Allegheny County using
commercially available geographic information system software (ArcGIS 10.1, Esri, Redlands,
CA, USA). Similarly, poverty status (defined as proportion of people living below the poverty
threshold) at the census tract level was obtained from 2011 American Community Survey (ACS)
data, color coded into 4 groups, and depicted in the same Allegheny County Map.
All data was analyzed with a commercially available statistical software package (Stata
SE Version 12.0, StataCorp, College Station, Texas, USA). Continuous data were compared
using Wilcoxon rank sum tests and are presented as means with standard deviations or median
with inter-quartile ranges (IQR) if not normally distributed. Categorical data between groups as
compared using two-sided chi square tests. A p-level smaller than 0.05 was considered
statistically significant.
4
3.0
RESULTS
DEMOGRAPHICS
During the period of March 1st 2011 and July 30th 2012, 197 patients after in- and out-of
hospital cardiac arrests due to different causes were treated at UPMC Mercy hospital. Of these,
we excluded 102 (52.8%) because they did not meet the pre-specified age criteria, and 95
(48.2%) patients were included in the final analysis cohort.
Among the 95 patients, 40 (42.1%) had an out of hospital cardiac arrest (OHCA). Among
those, 19 (47.5%) were male, 32 (80%) were white and 8 (20%) were black. Median age was 48
years (IQR: 39 to 57.5).
Of the remaining 55 patients (57.9%) who had an in-hospital cardiac arrest (IHCA), 33
(60%) were male, 43 (78.2%) were white and 12 (21.8%) were black. Median age was 54 years
(IQR: 41 to 59).
COMORBID CONDITIONS AND HEALTH BEHAVIORS
The median age-adjusted Charlson comorbidity index (CCI), calculated in 85 patients
(89.5%), was 2 (IQR: 1 to 3). Arterial Hypertension was documented in 53 cases (56.4%),
hyperlipidemia in 29 patients (30.9%), and diabetes in 30 patients (31%).
5
Preexisting coronary artery disease (CAD) was documented in 23 patients (24.5%), six
patients (6.4%) had a past medical history of cerebrovascular accident (CVA). The median body
mass index (BMI) for all 95 patients was 30kg/m2 (IQR: 25 to 38), and approximately a third
(29/95) of the patients were documented smokers. These numbers did not show any statistical
significant difference between OHCA and IHCA patients.
SOCIAL BACKGROUND
All 95 patients (100%) spoke English as their first language. Seventy-one patients
(74.7%) were either single or divorced, and 29 (30.5%) had either no religious denomination
listed or their denomination was unknown. These numbers did not differ between patients with
IHCA or OHCA patients (74.5% vs. 75% single/divorce and 31% vs. 30% without or unknown
religious denomination, respectively, both P<0.05)
Of all 95 patients, 10 (10.5%) did not have a next of kin (NOK) listed in their patient data
file. As expected, the proportion of patients without a NOK was slightly higher among OHCA
patients, although this different was not statistically significant (15% vs. 7.3%, P=0.34).
Forty patients (42.1%) had no insurance or Medicaid or Medicare listed as support. Of
these, 20 (21.1% out of 95) had no insurance (Figure 1). At time of cardiac arrest or prior to
hospital admission in case of in hospital cardiac arrests, 69 patients (72.6%) were unemployed or
disabled (Figure 2).
6
NEIGHBORHOOD POVERTY
Sixty-nine of the 95 patients (72.6%) had their residence in Allegheny County. Among
those 69 patients, 18 (26.1%) lived in neighborhoods with a poverty level (i.e. proportion of
people living below the poverty level) below 10%, 18 patients (26.1%) in neighborhoods with a
poverty level of 10-20%, 25 patients (36.2%) in neighborhoods with 20-30% poverty level, and 8
patients (11.6%) in areas with >30% poverty level (Figure 6 and 7).
MEDICAL CHARACTERISTICS
A history of active drug abuse was reported in 19 patients (20.2%), and in 20 cases
(23.2%) an illicit drug was identified on urine drug screen (UDS), and ethanol was detected in
three instances. Among the 21 patients with a positive UDS, opiates were most frequently used
in 9 instances (42.9%). The proportion of patients with a positive UDS was higher in OHCA vs.
IHCA patients, but did not reach statistical significance (28.2% vs. 14.6%, P=0.10).
Overall, drug overdose occurred in 8 patients (8.4%) and was the third most common
cause of arrest in both OHCA and IHCA patients. The two most common causes in OHCA were
cardiac and respiratory (34 [35.8%] cases each), in IHCA a respiratory cause of arrest was the
single most common cause followed by cardiac cause (Figures 3,4,5).
In 27 Patients (28.4%) a shockable rhythm (i.e. ventricular fibrillation or pulseless
ventricular tachycardia) was documented as first rhythm, 55 patients had a non-shockable
rhythm (i.e. asystole or pulseless electrical activity) and 13 patients (13.7%) had no documented
7
initial rhythm. OHCA patients were more likely to have a shockable rhythm compared to IHCA
patients (32.5 vs. 25.5%), but this difference was not statistically significant.
Mild therapeutic hypothermia, a post-cardiac arrest treatment recommended by the
American Heart Association,13 was applied in 45 cases (47.4%). Patients who suffered an out-ofhospital arrest were more twice as likely to be treated with hypothermia compared to patients
who suffered an in-hospital arrest (67.5% vs. 32.7%, P<0.001).
OUTCOMES
Thirty patients (31.6%) survived to hospital discharge, with no difference between
OHCA and IHCA patients (27.5 vs. 34.6%, p=0.47). A favorable neurologic outcome (=CPC
categories 1 or 2) at discharge was reported in 17 (17.9%) of patients, a good outcome (CPC 1-2)
at discharge was reported in 17 patients (17.9%), without any difference between OHCA and
IHCA patients (20.0% vs. 16.4%, P=0.30).
8
4.0
DISCUSSSION
This A previous study in cardiac arrest patients reported a disproportionally high
incidence of cardiac arrests in adult Pittsburghers ages 18 to 64.3 To understand factors
contributing to this high cardiac arrest incidence rate, we analyzed data from cardiac arrest
patients who were treated at an inner city hospital (UPMC Mercy, Pittsburgh, PA).
Our data show that a high proportion of cardiac arrest patients have pre-existing comorbidities, engage in unhealthy behaviors (i.e. more likely to be overweight and obese, more
likely to use illicit drugs), and have an adverse social background compared to the Allegheny
County or the whole US. For instance, they are more likely to live in poor neighborhoods, are
very likely to be unemployed or disabled and uninsured at the time of their arrest. Besides that,
our cohort was more likely to be single or divorced, which by itself is associated with adverse
health outcomes and reduced life expectancy.14
Per definition of the Centers for Disease Control and prevention (CDC), a body mass
index (BMI) between 25 and 29.9 is considered overweight, and a BMI of >30 is considered
obese in adults. In our study, 22.1% of patients were overweight and 44.2% were obese. This is
significantly higher than average adult obesity prevalence in Allegheny County, Pittsburgh and
the US (26.3%, 27.3% and 35.7%, respectively).19,20 Obesity is most prevalent in non-Hispanic
blacks15 and is associated with many adverse health outcomes, including cardiovascular disease
and sudden cardiac death.16-18
9
For example, obesity was identified as independent risk factor for sudden cardiac arrest in
middle-aged people aged 35 to 60 in an epidemiologic surveillance study in Portland, Oregon.19
In addition, hypertension and diabetes, which are known cardiovascular risk factors, were very
prevalent in our cohort.
A significant proportion of our patients either had a past history of illicit drug use or a
positive urine drug screen on admission. The most common substance group identified was
opiates. Currently, it is estimated that approximately 2% of the US adult population (~4.3
million) regularly take opioids.20 Regular opioid use is associated with lower education level,
female gender and non-Hispanic white race.20 The societal costs for prescription opioid abuse in
the US are estimated at 55.7 billion US dollars in 2007, over 90% comprising workplace and
health care costs.21
The incidence of sudden cardiac arrests is higher in areas of low socioeconomic status.7,8
In Allegheny County the number of people below poverty level is estimated at 12.4% for the
general population and 11.8% for the population between 18 and 64, which is lower than
national numbers (15.9% and 14.8%, respectively). In contrast, on average, our patients lived in
neighborhoods where 18.7% of the total population and 17.3% of 18 to 64 year olds live below
the poverty level. These numbers are more similar to the poverty levels of the City of Pittsburgh,
where 22.2% of the general population and 21.7% of 18-64 year olds are estimated to live below
the poverty level.
Similarly, the vast majority of our patients were unemployed at the time of their cardiac
arrest, which is in contrast to county or national estimates. Seventy-three percent of our patients
were unemployed or on disability compared to an unemployment civilian labor force rate of
7.2% in Allegheny County, 9.0% in Pittsburgh, and 9.1% in the whole US.
10
These findings are also consistent with a disproportionally high proportion of uninsured
patients. In our cohort 21.1% did not have any medical insurance at the time of their arrest vs. a
non-insurance rate of 11.7% in the 18-64 year old Allegheny County and 12.2% in the Pittsburgh
population.
Poverty is strongly associated with race, and some researchers argue that racial health
disparities are entirely explained by differences in socioeconomic status.22-24 In our study 20.1%
of patients were black, which is significantly higher than the proportion of black people living in
Allegheny County (13.4%) or the US (13.1%).
LIMITATIONS
Major limitations of this study include its retrospective study design and the data
collection from a single hospital within a fairly small sample size, which limits generalizability
and any statistical inference. In order to correct this shortcoming, we are currently conduction a
follow-up investigation and will include several hundred cardiac arrest patients that were treated
at UPMC Presbyterian University Hospital for sudden cardiac arrest. We will analyze
approximately 420 patients in total, which should increase the external validity of our data.
In addition, our data do not include any autopsy reports. Overall, the non-forensic
autopsy rate has been sharply declining in the US over the past decades.25 We identified the
cause of arrest by reviewing death summaries within the medical record. Last, one must exert
caution when interpreting poverty data. The fact that a person lives in a “poor” neighborhood
does not necessarily mean that person is poor.
11
CONCLUSION
Together, our data suggests that patients who were treated at UPMC Mercy hospital after
a sudden cardiac arrest have more preexisting comorbid conditions, poor health behaviors and an
adverse socioeconomic background when compared to the average person in Allegheny County
or the US. The combination of being uninsured, unemployed and single in the setting of
preexisting obesity and poor health behaviors (smoking and illicit drug use) seems to be the ideal
recipe for disaster (i.e. cardiac arrest). Based on our findings, local public health institutions need
to invest in specially targeted prevention programs aimed at this population group. As a desirable
result, young patients at risk could be identified in advance and sufficiently treated by local
health providers in order to prevent the potentially devastating consequences of a sudden cardiac
arrest.
12
APPENDIX: TABLE AND FIGURES
Table 1. Race and social characteristics; good versus bad outcome
n=95
Good outcome
(n=20)
Bad outcome
(n=75)
p value
Black
3 (15%)
17 (22.7%)
0.55
No insurance
4 (20%)
16 (21.3%)
0.74
Medicare
3 (15%)
8 (10.7%)
0.38
Medicaid
3 (15%)
8 (10.7%)
0.38
Insurance
10 (50%)
43 (57.3%)
0.62
Employment
5 (25%)
13 (17.3%)
0.52
Unemployment
14 (70%)
51 (68%)
1.0
Disabled
1 (5%)
3 (4%)
1.0
Unknown status
0 (0%)
8 (10.7%)
0.2
13
Figure 1. Insurance status
Figure 2. Employment status
14
ca
re rd
sp iac
in irat
to
o
xi ry
ca
t
c ion
an he ere
b
ap m
r
hy o r al
la rha
ct
ic gic
sh
oc
tr k
au
m
se a
ps
un
i
kn s
ow
n
ca
re rdia
sp
c
ira
in
to tory
xi
ca
ti
ce on
r
eb
an he
ap mo ral
hy
rr
h
la
ct agi
c
ic
sh
oc
k
tr
au
m
a
se
ps
is
Etiology of cardiac arrest (n=95)
40
30
numbers 20
10
0
cause
Figure 3. Etiology overall
30
numbers
20
10
0
cause
Figure 4. In-hospital arrests (n=55)
15
ca
re rd
sp iac
in irat
to
o
xi ry
ca
t
c ion
an he ere
b
ap m
r
hy o r al
la rha
ct
ic gic
sh
oc
tr k
au
m
se a
p
un si
kn s
ow
n
15
10
numbers
5
0
cause
Figure 5.
Out-of hospital arrests (n=40)
16
Figure 6. Mapping of residences
17
30
20
number of patients
10
0%
>4
31
-4
0%
21
-3
0%
11
-2
0%
010
%
0
% poverty
Figure 7. Percentage of poverty in area of residence (n=69)
18
BIBLIOGRAPHY
1.
Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics--2013 update: a report
from the American Heart Association. Circulation. 2013;127:e6-e245.
2.
McNally B, Robb R, Mehta M, et al. Out-of-hospital cardiac arrest surveillance --- Cardiac Arrest
registry to Enhance Survival (CARES), United States, October 1, 2005--December 31,
2010. Morbidity and mortality weekly report. Surveillance summaries. 2011;60:1-19.
3.
Nichol G, Thomas E, Callaway CW, et al. Regional variation in out-of-hospital cardiac arrest
incidence and outcome. JAMA : the journal of the American Medical Association.
2008;300:1423-1431.
4.
Becker LB, Han BH, Meyer PM, et al. Racial differences in the incidence of cardiac arrest and
subsequent survival. The CPR Chicago Project. The New England journal of medicine.
1993;329:600-606.
5.
Feero S, Hedges JR, Stevens P. Demographics of cardiac arrest: association with residence in a
low-income area. Academic emergency medicine : official journal of the Society for
Academic Emergency Medicine. 1995;2:11-16.
6.
Lerner EB, Fairbanks RJ, Shah MN. Identification of out-of-hospital cardiac arrest clusters using
a geographic information system. Academic emergency medicine : official journal of the
Society for Academic Emergency Medicine. 2005;12:81-84.
7.
Reinier K, Stecker EC, Vickers C, Gunson K, Jui J, Chugh SS. Incidence of sudden cardiac arrest
is higher in areas of low socioeconomic status: a prospective two year study in a large
United States community. Resuscitation. 2006;70:186-192.
8.
Reinier K, Thomas E, Andrusiek DL, et al. Socioeconomic status and incidence of sudden cardiac
arrest. CMAJ : Canadian Medical Association journal = journal de l'Association
medicale canadienne. 2011;183:1705-1712.
9.
Traven ND, Kuller LH, Ives DG, Rutan GH, Perper JA. Coronary heart disease mortality and
sudden death among the 35-44-year age group in Allegheny County, Pennsylvania.
Annals of epidemiology. 1996;6:130-136.
19
10.
Stavig GR, Leonard AR, Igra A, Felten P. Indices of relative body weight and ideal weight charts.
Journal of chronic diseases. 1984;37:255-262.
11.
Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic
comorbidity in longitudinal studies: development and validation. Journal of chronic
diseases. 1987;40:373-383.
12.
Safar P. Resuscitation after Brain Ischemia. In: Grenvik A, Safar P, eds. Brain Failure and
Resuscitation. New York: Churchill Livingstone; 1981:155-184.
13.
Peberdy MA, Callaway CW, Neumar RW, et al. Part 9: post-cardiac arrest care: 2010 American
Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency
Cardiovascular Care. Circulation. 2010;122:S768-786.
14.
Macintyre S. The effects of family position and status on health. Social science & medicine.
1992;35:453-464.
15.
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 : the journal of the American
Medical Association. 2012;307:491-497.
16.
Haslam DW, James WP. Obesity. Lancet. 2005;366:1197-1209.
17.
Tavora F, Zhang Y, Zhang M, et al. Cardiomegaly is a common arrhythmogenic substrate in adult
sudden cardiac deaths, and is associated with obesity. Pathology. 2012;44:187-191.
18.
Wilson PW, D'Agostino RB, Sullivan L, Parise H, Kannel WB. Overweight and obesity as
determinants of cardiovascular risk: the Framingham experience. Archives of internal
medicine. 2002;162:1867-1872.
19.
Noheria A, Teodorescu C, Uy-Evanado A, et al. Distinctive profile of sudden cardiac arrest in
middle-aged vs. older adults: A community-based study. International journal of
cardiology. 2013.
20.
Parsells Kelly J, Cook SF, Kaufman DW, Anderson T, Rosenberg L, Mitchell AA. Prevalence a
and characteristics of opioid use in the US adult population. Pain. 2008;138:507-513.
21.
Birnbaum HG, White AG, Schiller M, Waldman T, Cleveland JM, Roland CL. Societal costs of
prescription opioid abuse, dependence, and misuse in the United States. Pain medicine.
2011;12:657-667.
22.
Schwartz RS. Racial profiling in medical research. The New England journal of medicine.
2001;344:1392-1393.
20
23.
LaVeist TA. Disentangling race and socioeconomic status: a key to understanding health
inequalities. Journal of urban health : bulletin of the New York Academy of Medicine.
2005;82:iii26-34.
24.
Bleich SN, Thorpe RJ, Jr., Sharif-Harris H, Fesahazion R, Laveist TA. Social context explains
race disparities in obesity among women. Journal of epidemiology and community
health. 2010;64:465-469.
25.
Shojania KG, Burton EC. The vanishing nonforensic autopsy. The New England journal of
medicine. 2008;358:873-875.
21