Content uploaded by Raoult Ratard
Author content
All content in this area was uploaded by Raoult Ratard on May 02, 2018
Content may be subject to copyright.
Hurricane Katrina Deaths, Louisiana, 2005
Joan Brunkard, PhD, Gonza Namulanda, MS, and Raoult Ratard, MD
ABSTRACT
Objective: Hurricane Katrina struck the US Gulf Coast on August 29, 2005, causing unprecedented
damage to numerous communities in Louisiana and Mississippi. Our objectives were to verify,
document, and characterize Katrina-related mortality in Louisiana and help identify strategies to reduce
mortality in future disasters.
Methods: We assessed Hurricane Katrina mortality data sources received in 2007, including Louisiana
and out-of-state death certificates for deaths occurring from August 27 to October 31, 2005, and the
Disaster Mortuary Operational Response Team’s confirmed victims’ database. We calculated age-,
race-, and sex-specific mortality rates for Orleans, St Bernard, and Jefferson Parishes, where 95% of
Katrina victims resided and conducted stratified analyses by parish of residence to compare differ-
ences between observed proportions of victim demographic characteristics and expected values based
on 2000 US Census data, using Pearson chi square and Fisher exact tests.
Results: We identified 971 Katrina-related deaths in Louisiana and 15 deaths among Katrina evacuees in
other states. Drowning (40%), injury and trauma (25%), and heart conditions (11%) were the major causes
of death among Louisiana victims. Forty-nine percent of victims were people 75 years old and older.
Fifty-three percent of victims were men; 51% were black; and 42% were white. In Orleans Parish, the
mortality rate among blacks was 1.7 to 4 times higher than that among whites for all people 18 years old
and older. People 75 years old and older were significantly more likely to be storm victims (P⬍.0001).
Conclusions: Hurricane Katrina was the deadliest hurricane to strike the US Gulf Coast since 1928. Drowning
was the major cause of death and people 75 years old and older were the most affected population cohort.
Future disaster preparedness efforts must focus on evacuating and caring for vulnerable populations,
including those in hospitals, long-term care facilities, and personal residences. Improving mortality reporting
timeliness will enable response teams to provide appropriate interventions to these populations and to
prepare and implement preventive measures before the next disaster. (Disaster Med Public Health
Preparedness. 2008;:1–1)
Key Words: Hurricane Katrina, Louisiana, mortality, drowning, flooding, disaster preparedness
Hurricane Katrina struck the US Gulf Coast
on August 29, 2005 as a category 3 hurricane
on the Saffir-Simpson scale, causing unprec-
edented damage to numerous Louisiana and Missis-
sippi communities.
1
During the hours and days after
Hurricane Katrina, breaches in the levee infrastructure
resulted in flooding throughout approximately 80% of
New Orleans. A massive rescue and recovery effort
was launched by local, state, and federal governments
and nongovernmental organizations. Before Hurri-
cane Katrina, the deadliest hurricane to make landfall
in the United States during the previous 50 years was
Hurricane Audrey (1957), with an estimated 416
deaths.
2
Hurricane Andrew (1995), the last category
5 hurricane to strike the United States, caused 26
deaths.
2
Although several preliminary estimates exist
of the deaths attributable to Hurricane Katrina,
1,3,4
no prior report has systematically reviewed all of the
available death databases to accurately document
Hurricane Katrina mortality in Louisiana.
Our objectives were to verify and document the num-
ber of deaths from Hurricane Katrina among people
in Louisiana at the time of the storm and to charac-
terize the storm’s mortality burden by victim demo-
graphics, geographic location, timeline, and cause of
death. This report is the first to combine multiple death
databases to assess the number of storm-related deaths
among Louisiana residents and people who were in
Louisiana at the time of the storm and to provide
information regarding the causes of death. The find-
ings in this report will aid public health and emer-
gency preparedness efforts and may help reduce the
mortality burden in future natural disasters.
METHODS
Data Sources
Data sources included the Hurricane Katrina Disaster
Mortuary Operational Response Team (DMORT) da-
tabase and death certificates collected through Louisi-
ana vital statistics and out-of-state coroners’ offices.
RESEARCH
Disaster Medicine and Public Health Preparedness 1
Published on August 28, 2008
Copyright 2008 by American Medical Association.
DMORT is a federal response team that provides assistance
with mortuary activities during disaster situations. Out-of-
state death certificates of Louisiana residents during the pe-
riod of August 27 to October 1, 2005, and others that state
coroners deemed worth consideration for potential associa-
tion with Hurricane Katrina were forwarded to the Louisiana
coroner’s office from coroners’ offices in 26 states and the
District of Columbia.
Katrina Classification
The primary classification of Katrina-related deaths assigned
by parish (county) coroners on death certificates was Inter-
national Classification of Diseases-10 code X37, victim of cat-
aclysmic storm. The majority of deaths had multiple hierar-
chical cause-of-death classifications; however, if trauma,
injury, or drowning was listed as a contributing cause of
death, these victims were categorized as drowning or injury
victims in our database. A systematic review of all of the
records in the DMORT database and of Louisiana death
certificates yielded a final database of confirmed victims. A
number of quality-assurance cross-checks were conducted,
including ensuring no duplication of records across data
sources and date of death occurring within a plausible time-
line (August 27–October 31, 2005) and within the geo-
graphic location for Katrina-related deaths (ie, death occur-
ring in or victim evacuated from the affected southeastern
Louisiana parishes). All deaths occurring before August 29,
2005 and after October 1, 2005 were reviewed to verify that
Hurricane Katrina was listed as a contributing cause of death,
and all of the deaths occurring after September 23, 2005 were
reviewed to determine whether they were associated with
Hurricane Rita, which struck southwest Louisiana on Sep-
tember 24, 2005. The final case definition for Katrina-related
deaths in this analysis included all of the deaths in the
DMORT database that were determined to be Katrina re-
lated, all Louisiana death certificates with victim of cataclys-
mic storm listed as the primary or a contributing cause of
death, and out-of-state death certificates for Louisiana resi-
dents that were classified as related to Hurricane Katrina.
Data from Louisiana vital statistics were collected at the local
levels through parish coroners’ offices and then sent to the
state level. Data from all 3 sources (DMORT, vital statistics,
and out-of-state death certificates) were entered into an
Access (Microsoft Corporation, Seattle, WA) database. De-
scriptive and inferential statistics were calculated using Stata
version 9.1 (StataCorp LP, College Station, TX). We calcu-
lated age-, race-, and sex-specific mortality rates (per 10,000
population) for victims in Orleans, St Bernard, and Jefferson
Parishes, where 95% of Katrina’s victims resided. We con-
ducted stratified analyses by parish of residence to compare
differences between observed proportions of victim demo-
graphic characteristics (sex, race/ethnicity, and age) and ex-
pected values based on 2000 US Census data, using Pearson
chi square and Fisher exact tests. We further stratified by race
within each age category, where sufficient observations ex-
isted, to determine whether there was an age-specific effect of
race among victims. We limited our comparative analyses to
Orleans, St Bernard, and Jefferson Parishes.
Mapping Using Geographic Information Systems
To identify high-mortality geographic areas, Louisiana death
records with addresses (n ⫽687) were geocoded using Envi-
ronmental Systems Research Institute’s ArcGIS version 9.2
(Redlands, CA). Most records were geocoded using the En-
vironmental Systems Research Institute street file reference
geospatial layer, which was used to match the address for
location of death to the street file address. Records with only
a nursing facility name were matched to the address of the
nursing facility in the Homeland Security Infrastructure Pro-
gram Public Health, Nursing Home geospatial layer.
Ethical Review
The study protocol was reviewed by the human subjects
coordinator at the Office of Workforce and Career Develop-
ment, Centers for Disease Control and Prevention, and de-
termined to be a public health response that did not require
further human subjects review.
RESULTS
We identified 971 Katrina-related deaths that occurred in
Louisiana and at least 15 deaths that occurred among Loui-
siana Katrina evacuees in other states, for a conservative
storm-related death total of 986 victims. The state coroner
was forwarded 446 out-of-state death certificates for Louisi-
ana residents. Of these, 15 were clearly related to Hurricane
Katrina, and 431 were classified as indeterminate because no
indication of hurricane association was listed on the death
certificate. Our total mortality estimate of 986 victims likely
represents a lower bound estimate for Katrina mortality in
Louisiana. Including all deaths classified as indeterminate
(DMORT, n ⫽23; and out-of-state, n ⫽431) yields an upper
bound estimate of 1440. We limited all subsequent analyses
to deaths we determined to be related to Hurricane Katrina
(n ⫽986).
Cause of Death
Excluding the 15 out-of-state deaths, we found that of the
971 people who died in Louisiana as a result of Hurricane
Katrina, data on cause of death were available for 800 people.
Three hundred eighty-seven victims drowned, and 246 peo-
ple sustained trauma or injuries severe enough to cause their
deaths. The mechanism of injury was unspecified for 226
trauma or injury deaths; specified injury-related causes of
death included heat exposure (n ⫽6), unintentional firearms
death (n ⫽4), homicide (n ⫽2), suicide (n ⫽4), gas poisoning
(n ⫽3), and electrocution (n ⫽1; Table 1).
Victim Demographic Characteristics
Among the 971 Hurricane Katrina victims who died in
Louisiana, 512 (53%) were men; 498 (51%) were black
(non-Hispanic/Latino); 403 (42%) were white (non-Hispanic/
Latino), and 18 (2%) were Hispanic/Latino (Table 2). Other
Hurricane Katrina Deaths, Louisiana, 2005
2Disaster Medicine and Public Health Preparedness VOL. /NO.
victim races included Asian (n ⫽4), American Indian (n ⫽
4), and other (n ⫽1). Data on race/ethnicity were missing
for 42 victims (4%), age was missing for 22 victims (2%), and
sex was missing for 4 victims (⬍1%).
The mean age of Katrina victims was 69.0 years (95% con-
fidence interval [CI], 67.8 –70.2), and their age range was 0 to
102 years. Approximately 50% of the people who died as a
result of Hurricane Katrina in Louisiana were 75 years old
and older. Fewer than 10% of victims were younger than 45
years old. In Orleans Parish, where the majority of victims
lived and died, all age categories of victims (except those
45–54 years old) were divergent from the overall parish age
distribution of the population. The most dramatic difference
occurred among people 75 years old and older (Pearson chi
square [
2
] 2400; degrees of freedom (df)⫽1; P⬍.0001;
Table 3). In Orleans Parish, men were more affected (
2
17.4,
df ⫽1, P⬍.001), and women were less affected (
2
18.7,
df ⫽1, P⬍.001) by storm mortality, relative to their
underlying population distributions. In both St Bernard and
Jefferson Parishes, the sex and racial distributions of Katrina
victims were not significantly different (P⬎.05) from census
population figures for those parishes, except for Hispanic/
Latino victims in St Bernard Parish (P⫽.03; Table 2).
In Orleans Parish, the mortality rate among blacks was 1.7 to
4 times higher than among whites for people 18 years old and
older, indicating that the effect of age on mortality con-
founded the effect of race. Chi square and Fisher exact tests
assessing differences in proportions of black and white vic-
tims within age groups found that blacks were significantly
more likely to be storm victims than whites in all age group
categories 30 years old and older in Orleans Parish (P⬍.05).
Race-specific mortality rates were also higher among blacks
55 to 64 years old in St Bernard Parish and 75 years old and
older in Jefferson Parish, but results from race- and age-
specific stratified analyses in these 2 parishes are limited by
small number of observations (n ⱕ5). Among the largest age
cohort of victims (people 75 years old and older), we further
stratified by both race and sex in Orleans and St Bernard
Parishes and found that black men 75 years old and older
were significantly overrepresented in Orleans Parish (P⬍
.001; Table 2).
Timeline
Of 799 deaths in vital statistics records in which date of death
was recorded, 650 (81%) occurred on August 29, 2005.
Seven deaths occurred in the 2 days preceding the storm and
4 deaths occurred after October 31, 2005 (Fig. 1). For the 171
victims who were listed only in DMORT and for whom a
date of death was not specified, date when victims were found
was available for 129 people. The majority of these people
(n ⫽82, 64%) were recovered during the second and third
weeks after the storm (September 5–19, 2005).
Where Victims Lived and Died
Data on parish of residence and parish of death were available
for 934 and 854 people, respectively. The majority of hurri-
cane victims lived in Orleans Parish (73%), followed by St
Bernard (17%), Jefferson (5%), Plaquemines (1%), and St
Tammany Parishes (1%). Victims died primarily in Orleans
Parish (70%), St Bernard Parish (14%), Jefferson Parish
(4%), and East Baton Rouge Parish (3%). Data on both
parish of residence and death were available for 844 storm
victims; more than 80% of these victims died in their parish
of residence. Sixty-five percent of Hurricane Katrina victims
in Louisiana died of injury or drowning. The majority of these
deaths occurred in Eastern Orleans Parish, specifically the
lower ninth ward; in Lakeview and Gentilly, adjacent to Lake
Pontchartrain; and in St Bernard Parish (Figs. 2 and 3).
Drowning and injury-related deaths occurred predominantly
near levee infrastructure breaches.
Data related to place of death or where victims were found
were available for 877 people (Table 3). Deaths or places
where victims’ bodies were recovered occurred predomi-
nantly in residences (36%), hospitals (22%), and nursing
facilities (12%). More than 25% of victims (n ⫽262) were
found in other locations, including roadways, the New Or-
leans airport, the Convention Center, and the Superdome.
For the deaths recorded in the DMORT database only that
did not have date of death available, it is unclear whether the
person died at the location or the body was brought to the
location after death. The date the body was found was avail-
able for 129 of 171 people that appear only in the DMORT
database. Of these 129 people, 80 (62%) were recovered
before September 15, 2005.
At least 70 people who were classified as hospital inpatients
died during the period August 29, 2005 to September 2,
2005, in New Orleans hospitals and an additional 57 victims
were recovered from New Orleans hospitals and brought to
the DMORT facility from September 5 to 12, 2005. From
August 29 to September 3, 2005, at least 71 people died in
nursing facilities in Orleans, St Bernard, and Jefferson Par-
ishes, and an additional 7 bodies were recovered from nursing
TABLE 1
Cause of Death: Katrina-Related Mortality in Louisiana,
2005
Cause of Death No. %
Drowning 387 40
Injury and trauma
Mechanism specified 20 2
Mechanism unspecified 226 23
Illness
Heart disease 107 11
Other illness* 46 5
Unspecified, Katrina-related 185 19
Total 971 100
*Diabetes mellitus, chronic obstructive pulmonary disease, septicemia,
and cancer.
Hurricane Katrina Deaths, Louisiana, 2005
Disaster Medicine and Public Health Preparedness 3
TABLE 2
Demographic Data for Katrina-Related Deaths: Louisiana, 2005
All Victims (N ⴝ971)* Orleans Parish (n ⴝ682) St Bernard Parish (n ⴝ157) Jefferson Parish (n ⴝ42)
Characteristic No. (%) Rate† No. (%) Census, %‡ P§ Rate No. (%) Census, % PRate No. (%) Census, % P
Sex
Male 512 (53) 16 374 (55) 47 ⬍.001 20 64 (41) 48 .06 1 21 (50) 48 .80
Female 455 (47) 12 306 (45) 53 ⬍.001 27 93 (59) 52 .06 1 21 (50) 52 .80
Race/ethnicity兩兩
Black 498 (51) 14 459 (67) 67 .69 24 12 (8) 8 .98 1 10 (24) 23 .86
White 403 (42) 15 196 (29) 27 .20 25 141 (90) 84 .06 1 31 (74) 65 .25
Hispanic/Latino 18 (2) 8 14 (2) 3 .13 6 2 (1) 5 .03 0 1 (2) 7 .37
Age, years
⬍18¶ 20 (2) 1 9 (1) 27 ⬍.001 1 1 (1) 25 ⬍.001 0 0 (0) 25 ⬍.001
Black 10 1 9 80 .22 5 1 11 .11
White 0 0 0 13 .62 0 0 80 .20
18–29 18 (2) 2 14 (2) 19 ⬍.001 0 0 (0) 16 ⬍.001 0 0 (0) 16 .001
Black 13 2 12 66 .16 0 0 9 —
White 5 1 2 26 .54 0 0 82 —
30–44 47 (5) 3 30 (4) 22 ⬍.001 6 9 (6) 23 ⬍.001 0 2 (5) 23 ⬍.01
Black 31 4 27 64 ⬍.01 9 1 7 .50
White 15 1 3 29 .03 6 8 85 1.00
45–54 119 (12) 14 91 (13) 13 .88 17 16 (10) 14 .18 1 5 (12) 14 .83
Black 78 18 72 63 ⬍.01 18 1 6 1.00
White 38 8 16 31 ⬍.01 19 15 87 .71
55–64 137 (14) 25 95 (14) 8 ⬍.001 39 23 (15) 9 ⬍.01 2 8 (19) 9 .05
Black 80 31 72 61 ⬍.01 180 5 5 ⬍.01
White 54 19 23 32 .08 33 17 88 .05
65–74 136 (14) 36 104 (15) 6 ⬍.001 38 20 (13) 8 .02 1 4 (10) 7 .36
Black 85 50 80 55 ⬍.001 114 2 3 .14
White 43 17 19 38 ⬍.001 34 16 90 .13
ⱖ75 472 (49) 121 338 (50) 6 ⬍.001 221 88 (56) 6 ⬍.001 10 23 (55) 5 ⬍.001
Black 198 (42) 150 186 (55) 44 ⬍.001 183 2 (2) 3 1.00 27 5 (22) 8 .03
Male 96 214 91 34 ⬍.001 476 2 39 .15 2
Female 102 117 95 66 ⬍.001 0 0 61 .15 3
White 247 (52) 94 133 (39) 51 ⬍.001 234 85 (97) 91 .12 8 17 (74) 86 .13
Male 90 120 56 33 .03 154 21 38 .01 7
Female 157 81 77 67 .03 282 64 62 .01 10
*Missing values: sex (n ⫽4), race (n ⫽42), age (n ⫽22).
†Mortality rate per 10,000 population.
‡2000 US Census figures for Orleans Parish (N ⫽484,674), St Bernard Parish (N ⫽67,229), and Jefferson Parish (N ⫽455,466).
§Pearson chi square or Fisher exact test (N ⬍10).
兩兩Non-Hispanic/Latino.
¶Includes 10 newborns; information on race available for only 1 of 10 infants.
Hurricane Katrina Deaths, Louisiana, 2005
4Disaster Medicine and Public Health Preparedness VOL. /NO.
facilities in these parishes during the weeks after Hurricane
Katrina.
Out-of-State Deaths
The state coroner’s office received 446 death certificates from
26 states and the District of Columbia for Louisiana residents
who died from August 27 to October 1, 2005 and other death
certificates that coroners’ offices deemed worthy of consider-
ation for potential association with Hurricane Katrina. The
majority of out-of-state deaths occurred in Texas (221, 50%),
followed by Alabama (47, 11%) and Mississippi (43, 10%).
Of the 446 Louisiana residents who died out of state, 53%
were female; 59% were white; and their mean age was 67.3
years (95% CI 65.3%– 69.3%). The majority of evacuees had
lived in Orleans Parish (40%), Jefferson Parish (22%), St Tam-
many Parish (5%), and St Bernard Parish (4%). Seventeen
out-of-state deaths (4%) occurred among Calcasieu Parish res-
idents, likely Hurricane Rita evacuees. Of the 446 deaths, only
15 people were classified as Katrina-related deaths by a state
coroner and thus included in our final mortality count.
DISCUSSION
This report provides a conservative estimate of deaths that
occurred in New Orleans and Louisiana associated with Hur-
ricane Katrina. We document at least 971 Katrina-related
deaths among people in Louisiana and 15 deaths among
Louisiana Katrina evacuees. Older adults were clearly the
most affected population cohort during Hurricane Katrina,
particularly people 75 years old and older, who made up 49%
of victims in Louisiana, whereas their age cohort represents
fewer than 6% of both the greater New Orleans and the
overall Louisiana population. Children and younger adults
were underrepresented among storm victims relative to their
proportional population size.
A total of 178 people 75 years old and older died in their
homes, 115 of drowning. There are at least 2 possible expla-
nations for these findings. The first is that older people may
7
650
54
29 13 13 12 5534 4
0
100
200
300
400
500
600
700
Bef ore 8/29/ 05
8/29/2005
We ek 1
We ek 2
We ek 3
We ek 4
We ek 5
We ek 6
We ek 7
We ek 8
We ek 9
After10/30/2005
Date
Number of Deaths
FIGURE 1
Timeline of Hurricane Katrina-related deaths in Louisiana, vital statistics (N ⴝ799)
TABLE 3
Place of Death or Where Body Was Found (N ⴝ877)
Location No. %
Residence 317 36
Hospital
Inpatient 168 19
Emergency department 24 3
Dead on arrival 3 ⬍1
Nursing facility 103 12
Convention center 10 1
New Orleans International Airport 14 2
Superdome 7 1
Other location 231 26
Hurricane Katrina Deaths, Louisiana, 2005
Disaster Medicine and Public Health Preparedness 5
have been less likely to leave their homes during the storm
evacuation because of prior experience with false alarms
regarding the intensity and potential threat of a hurricane,
fear of their abandoned homes being looted, or not wanting
to be separated from medical or other routines. Another
possible explanation is that older people are more likely to
die of drowning and injury during a hurricane or flood and
that comorbidities contribute to their vulnerability to storm-
associated mortality. It is likely that both factors contributed
to the disproportionate representation of people 75 years old
and older among Katrina victims.
In the days following Hurricane Katrina, multiple large hos-
pitals in New Orleans flooded and were reportedly operating
without power or sanitation services in extreme heat
5
while
ambient temperatures in the greater New Orleans area were
ⱖ90°F (32°C) during the days and weeks following Hurri-
cane Katrina.
6
As a result of rapidly deteriorating conditions
in the New Orleans hospitals and extreme difficulties in
evacuating their existing patients, hospitals in the downtown
New Orleans area were reportedly not admitting new patients
in the days following the storm.
5
At least 70 hospital inpa-
tients died in New Orleans hospitals, and an additional 57
storm victims’ bodies were recovered from hospitals in the days
immediately following the storm, indicating that their storm-
related cause of death occurred in the hospital. The power
outages and dire conditions reported in certain New Orleans
hospitals after the storm underscore the importance of disaster
preparedness. Hospitals in hurricane-prone areas should ensure
that adequate generators are available in elevated locations that
are not prone to flooding. Water, food, and other supplies should
be available to maximize the likelihood that patients using
ventilators and other life-support interventions will survive if a
similar emergency situation occurs. More than 70 people died in
nursing facilities in Orleans, St Bernard, and Jefferson Parishes
on August 29, 2005 and in the days immediately following
Hurricane Katrina, indicating that more targeted emergency
and disaster preparedness planning, especially with respect to
evacuation capability, is needed for these types of institutions.
FIGURE 2
Location of Hurricane Katrina deaths, southeast Louisiana, 2005 (N ⴝ687)
Hurricane Katrina Deaths, Louisiana, 2005
6Disaster Medicine and Public Health Preparedness VOL. /NO.
The overall proportions of deaths among non-Hispanic
blacks and whites in the most affected parishes—Orleans, St
Bernard, and Jefferson—were remarkably consistent with
their pre-Katrina race/ethnicity distributions from the 2000
US Census (Table 2). However, stratified analyses evaluating
the effect of race within age groups revealed that the domi-
nant effect of age on overall storm mortality masked the
differential effect of race in most age groups in Orleans
Parish, where race-specific mortality rates were on average
2.5 times higher among blacks compared with whites. Older
black people in Orleans Parish, particularly men, were dis-
proportionately represented relative to their underlying pop-
ulation distribution. Of note, only 4 storm victims were
Asian, although Asians make up 2% of the Orleans Parish
population and 1% of the overall Louisiana population.
7
Although Hispanic/Latino and Asian race/ethnic groups ap-
pear to have been less affected by storm mortality relative to
their proportional population size, victim numbers in these
groups are small, limiting statistical interpretation. Similarly,
small numbers of observations (n ⱕ5) in our stratified analyses
of race/ethnicity within age groups in St Bernard and Jeffer-
son Parishes limited our ability to assess potential associations
between demographic characteristics and mortality. The role
of poverty in Katrina mortality is unclear because we did not
have income data for victims, but socioeconomic status mer-
its future research to determine whether age- and race-asso-
ciated poverty may have increased the vulnerability of these
populations or limited their ability to evacuate.
The primary strength of this study is the use of multiple death
databases to systematically verify each death as related to
Hurricane Katrina using consistent and specific criteria. In-
corporating DMORT data allowed us to identify 171 victims
who were not classified as Katrina related in vital statistics/
death certificate data, including 17 victims whom coroners
were unable to identify by DNA matching or other methods.
In large-scale high-mortality disasters, DMORTs should be
FIGURE 3
Location of Hurricane Katrina deaths, Orleans Parish, 2005
Hurricane Katrina Deaths, Louisiana, 2005
Disaster Medicine and Public Health Preparedness 7
considered as an additional data source to better document
mortality.
Limitations
This study is subject to a number of limitations. First, the
disaster response aftermath of Hurricane Katrina may have
limited the ability to precisely document all deaths. Al-
though unlikely to be a large number, it is possible that some
people who died during the storm were never found or
documented. Second, classifying people who were evacuated
and later died out of state from Katrina-related causes is
inherently difficult, especially as regards older people who
had serious preexisting medical conditions. It is unknown
whether these people would have died had the storm oc-
curred or not. It is also unknown whether the storm exacer-
bated preexisting medical conditions enough to lead to
death. The majority of the 446 out-of-state deaths were from
chronic medical conditions, primarily heart disease, respira-
tory disease, and cancer. The clearest association with Kat-
rina in our review of out-of-state death certificates was 1
person whose body was found in the Gulf of Mexico off the
Florida coast. His death certificate specified drowning while
trying to save a family member who resided in the greater
New Orleans area as the cause of death.
Determining whether Hurricane Katrina contributed to out-
of-state deaths was particularly difficult in instances in which
patients died in hospitals during the days and weeks following
Katrina (n ⫽283), for suicides (n ⫽5), for accidents (n ⫽
25) and for cases pending further investigation or toxicology
reports (n ⫽15). The role of Hurricane Katrina in the
majority of the 446 out-of-state deaths will probably never be
clearly delineated because coroners in different states may
have used different criteria for classifying victims as storm
related. The 15 deaths with victim of cataclysmic storm or
other Katrina-related indications on their death certificates
almost certainly represent a lower bound estimate of out-of-
state deaths among Louisiana Katrina evacuees.
Hurricane Katrina death data from Louisiana vital statistics
first became available approximately 2 years after the storm.
Instituting an electronic reporting system for recording death
certificates would aid with timely use of data for both gov-
ernment and community preparedness and disaster response
efforts.
8
Alternately, if existing technological infrastructure is
unavailable, active mortality surveillance efforts should be
initiated immediately to document deaths in the early stages
of rescue and recovery responses, not only to aid with accu-
rate mortality reporting and victim identification but also to
mobilize needed resources (eg, body bags, mobile laboratories,
morgues) in a timely manner.
CONCLUSIONS
Despite the above limitations, this report provides the most
complete picture of Katrina-related mortality in Louisiana to
date. At least 986 people in Louisiana died as a result of
Hurricane Katrina, making it the deadliest hurricane to strike
the US Gulf Coast in more than 75 years. To prevent
hurricane-related mortality on this scale from occurring in
the future, disaster preparedness efforts must focus on evac-
uating and caring for older people, including those residing in
hospitals, long-term care facilities, and personal residences.
Adequate mortality reporting in disaster settings, particularly
when infrastructure has been damaged or destroyed, is vital to
ensuring timely collection of mortality data. Rapid identifi-
cation of vulnerable populations and risk factors during di-
sasters will enable response teams to provide appropriate
interventions to these populations and to prepare and imple-
ment preventive measures before the next disaster. Projec-
tions of a sustained or intensifying cycle of Atlantic Ocean
hurricane activity throughout the coming decades
9,10
and the
unprecedented landfall of 2 category 5 storms (Dean and
Felix) during the 2007 hurricane season underscore the crit-
ical need for all levels of government to be ready to evacuate
and care for vulnerable populations during future storms.
About the Authors
Dr Brunkard and Ms Namulanda are with the Centers for Disease Control and
Prevention; Dr Ratard is with the Louisiana Office of Public Health.
Address correspondence and reprint requests to Dr Joan Brunkard, Centers for
Disease Control and Prevention, 4770 Buford Hwy NE, MS F-22, Atlanta, GA
30341 (jbrunkard@cdc.gov).
J.B. and R.R. designed the study and analyzed the data; J.B. wrote the report;
G.N. geocoded all of the death addresses and created the figures; and all of the
authors reviewed and revised the manuscript. J.B. had full access to the data and
takes responsibility for the integrity of the data and the accuracy of the data
analysis.
The findings and conclusions in this report are those of the authors and do not
necessarily represent the views of the Centers for Disease Control and Prevention.
Received for publication July 21, 2008; accepted August 15, 2008.
Authors’ Disclosures
The authors report no conflicts of interest.
Acknowledgments
The authors thank D. Bensyl, W.R. Daley, and D. Koo for useful comments
on the manuscript.
ISSN: 1935-7893 © 2008 by the American Medical Association and Lip-
pincott Williams & Wilkins.
DOI: 10.1097/DMP.0b013e31818aaf55
REFERENCES
1. Knabb RD, Rhome JR, Brown DP. National Hurricane Center. Tropical
cyclone report: Hurricane Katrina, August 23–30, 2005. Miami, FL:
National Oceanic and Atmospheric Administration, National Weather
Service, National Hurricane Center. http://www.nhc.noaa.gov/pdf/
TCR-AL122005_Katrina.pdf. Accessed October 2, 2007.
2. National Oceanic and Atmospheric Administration (NOAA). Hurri-
cane Research Division. Frequently asked questions. Miami, FL:
NOAA. http://www.aoml.noaa.gov/hrd/tcfaq/E12.html. Published June 1,
2007. Accessed October 2, 2007.
3. Louisiana Department of Health and Hospitals. Hurricane Katrina:
deceased reports; reports of missing and deceased. Baton Rouge: Loui-
Hurricane Katrina Deaths, Louisiana, 2005
8Disaster Medicine and Public Health Preparedness VOL. /NO.
siana Department of Health and Hospitals. http://www.dhh.louisiana.gov/
offices/page.asp?ID⫽192&Detail⫽5248. Published August 2, 2006. Ac-
cessed October 2, 2007.
4. The Earth Institute at Columbia University. Hurricane Katrina de-
ceased-victims list. http://www.katrinalist.columbia.edu/stats.php. Ac-
cessed October 2, 2007.
5. Brinkley D. The Great Deluge: Hurricane Katrina, New Orleans, and the
Mississippi Gulf Coast. New York: HarperCollins; 2006.
6. National Climatic Data Center, National Oceanic and Atmospheric
Administration. Global summary of the day. Asheville, NC: National
Climatic Data Center. http://www.ncdc.noaa.gov/oa/ncdc.html. Accessed
October 2, 2007.
7. US Census Bureau. Decennial census. Census 2000 summary file 1 (SF
1) 100-percent data, detailed tables. Washington, DC: US Census
Bureau. http://factfinder.census.gov. Accessed October 2, 2007.
8. Stephens KU, Grew D, Chin K, et al. Excess mortality in the aftermath
of Hurricane Katrina: A preliminary report. Disaster Med Public Health
Preparedness. 2007;1:15–20.
9. National Oceanic and Atmospheric Administration (NOAA), Cli-
mate Prediction Center. Atlantic hurricane season outlook update.
Camp Springs, MD: NOAA. http://www.cpc.noaa.gov/products/outlooks/
hurricane.shtml. Accessed October 2, 2007.
10. Trenberth KE, Shea DJ. Atlantic hurricanes and natural variability in
2005. Geophys Res Lett. 2006;33:L12704.
Hurricane Katrina Deaths, Louisiana, 2005
Disaster Medicine and Public Health Preparedness 9
