Georgia epidemiology report, Vol. 12, no. 11 (Nov. 1996)

Georgia Epidemiology

Report

The Georgia Epidemiology Report is a publication of the Epidemiology Section of the Epidemiology and Prevention Branch, Division of Public Health, Department of Human Resources

November 1996
http://www.ph.dhr.state.ga.us Division Of Public Health
Patrick J. Meehan, M.D. - Director Epidemiology and Prevention Branch State Epidemiologist
Kathleen E. Toomey, M.D., M.P.H.- Director Epidemiology Section Paul A. Blake, M.D., M.P.H.-Chief
Notifiable Diseases
Jeffrey D. Berschling, M.P.H.; Karen R. Horvat, M.P.H. ; Amri B. Johnson, M.P.H. ; Jane E. Koehler, D.V.M, M.P.H.; Preeti Pathela, M.P.H.; Sabrina Walton, M.S.P.H.
Chronic Disease
Nancy E. Stroup, Ph.D.-Program Manager Patricia M. Fox, M.P.H.; David M. Homa, Ph.D., M.P.H.; Edward E. Pledger, M.P.A.
Tuberculosis
Naomi Bock, M.D., M.S.
HIV/AIDS/Sexually Transmitted Diseases
Kim Cook, M.D., M.S.P.H.-Program Manager Stephanie Bock, M.P.H.; Mary Lynn Gaffield, M.P.H.; Awal D. Khan, Ph.D., M.A.; Andrew Margolis, M.P.H.
Office of Perinatal Epidemiology
Roger W. Rochat, M.D. - Program Manager Mary D. Brantley, M.P.H.; Raymond E. Gangarosa, M.D., M.P.H.; Rebekah Hudgins, M.P.H.; Mary P. Mathis, Ph.D., M.P.H.; Florina Serbanescu, M.D.
Preventive Medicine Residents
Hussain R. Yusuf, M.B.B.S., M.P.H.; E. Anne Peterson, M.D.
EIS Officer
Michael S. Friedman, M.D.
Georgia Epidemiology Report Editorial Board
Editorial Executive Committee Paul A. Blake, M.D., M.P.H.- Editor Kathleen E. Toomey, M.D., M.P.H. Mary D. Brantley, M.P.H. Jeffrey D. Berschling, M.P.H.
Mailing List Edward E. Pledger, M.P.A.

Volume 12 Number 11
After the Floods: Responding to Potential Public Health Threats

W hen tropical storm Alberto hit the southeastern United States on July 5-6, 1994, the heavy rains that accompanied the storm caused record flooding in many parts of south and central Georgia. In some areas, more than 2 feet of rain fell in 24 hours, causing the Flint and other rivers to rise more than 20 feet above flood stage. Twenty-eight people drowned in flash floods that hit the area. Fifty-five of Georgia's counties were declared federal disaster areas with thirteen additional counties declared state disaster areas.

Figure 1. Flood Disaster areas in Georgia by Health District
Source: Georgia Emergency Management Agency (GEMA)

In the devastation that occurs during a flood, the health of the affected population can be jeopardized. In addition to obvious environmental havoc, floods often lead to the movement of large numbers of people fleeing the rising waters and to the disruption of normal public services. As a result, people may be exposed to physical, biological, and chemical hazards. In previous floods, increased numbers of injuries have often been reported that occur during the flood itself as well as during the clean up period. Electrocutions and asphyxiation from carbon monoxide have been reported when people returning to their homes came in contact with downed electrical lines and damaged fuel lines. Animals affected by a flood become more aggressive and unpredictable and increased numbers of animal and insect bites have been observed9,11.

In addition to injuries, other health effects ranging from skin rashes and allergic reactions to hyperthermia and hypothermia can be related either directly or indirectly to a flood. The stress of coping with a disaster can also influence the psychological health of the affected population9.

Contrary to popular belief, large outbreaks of communicable diseases are rarely observed after natural disasters in industrialized countries1,7,9. Although fears of cholera or typhoid epidemics are unfounded, the potential for increased incidence of illnesses such as diarrheal diseases that are associated with contaminated water must be considered4,9,10. The occurrence of vector borne diseases may be affected by the increased number of mosquito breeding sites created after a flood5,8,9. Increased numbers of respiratory infections, particularly among displaced persons housed in emergency shelters, have also been documented following flood disasters6,9,11.

Because of fears that the health of people living in flood-affected communities might be at risk, the Georgia Division of Public Health established a short-term surveillance system from July 6 to August 21, 1994, to monitor the occurrence of 24 types of illnesses and injuries that could have been affected by the flood. Similar systems were used during the floods that occurred in the Midwest in 19933. The purpose of the system was to detect disease outbreaks in flood affected

Epidemiology Section, Epidemiology & Prevention Branch, Two Peachtree St., N.W., Atlanta, GA 30303-3186

Phone: (404) 657-2588

FAX: (404) 657-2586

areas so that interventions could be implemented rapidly, as well as to estimate the additional morbidity that had been caused by the flood.
Methods
In the four health districts affected by the flood (Albany, Columbus, Dublin, and Macon), all health care facilities and clinics except the offices of private physicians were asked to participate in the system. Data collection was coordinated by district public health staff from approximately 130 facilities, including emergency shelters. Information was collected on incident cases of 24 conditions including broad categories of injuries, communicable diseases and other conditions associated with flood clean up. These data were sent by FAX daily to the district health offices, entered into a computer, and transmitted electronically, using CDC PC/WONDER, to the office of the State Epidemiologist, Georgia Division of Public Health in Atlanta. Data from flood disaster counties were compared to non-disaster counties as a control to estimate the burden of morbidity associated with the flood. Changes over time in the absolute number of cases of a particular outcome, as well as in the proportion of total morbidity due to that outcome, were used to monitor health problems. Differences in proportions were compared using chi-square tests.

ing types of facilities reporting in disaster and non-disaster counties. Milder cases of psychiatric morbidity (e.g., mild depression, anxiety) may have been more likely to be detected in the emergency shelters that were established in Albany than they would have been in the more typical setting of a hospital or clinic. However, even when data collected in the emergency shelters are excluded from the analysis, the excess of psychosocial conditions in the disaster counties remains statistically significant.
Figure 2. Diarrheal Illness in Flood Affected and Unaffected Counties by Week

Surveillance Results
During the 7 weeks that the surveillance system was active, no major disease outbreaks requiring public health intervention were recorded in any of the four districts. However, when federally declared disaster counties were compared with non-disaster counties, we observed that four outcomes accounted for a significantly greater proportion of total morbidity in flood-affected counties than in counties not affected by the floods (p <0.05). These outcomes were diarrheal illness, skin rash, heat-related illnesses and psychosocial symptoms.
Diarrheal illness We found that diarrheal diseases were significantly more com-
mon as a cause of morbidity in counties that were disaster areas than in counties that were not (Figure 2). This difference was most noticeable 3 to 4 weeks after the initial floods when 5% of people seeking treatment in the flood-affected areas did so for diarrheal disease compared with 2.3% in non-disaster counties. By that time, municipal water supplies were functioning normally, and although many private wells still remained contaminated by flood water, residents had been warned to boil water before use. By the seventh week of the surveillance, the proportion of people seeking treatment who did so for diarrheal disease was the same (approx. 2.2%) for both disaster and non-disaster counties, a finding which suggests that the problem causing excess diarrhea in the flooded counties--possibly contact with contaminated water during clean-up--had been corrected. The greatest contrast between disaster and non-disaster counties occurred among people younger than 15 years of age (Figure 3). We found that the proportion of morbidity attributed to diarrhea among children younger than 15 years of age was 1.6 times higher in disaster counties than in non-disaster counties over the surveillance period.
Psychosocial symptoms From the beginning of the surveillance period (i.e., from the first
day of the flood's impact) a statistically significant increase in the proportion of people being treated for psychosocial symptoms was observed in federally declared disaster counties compared to the proportion observed in non-disaster counties. During the 4 weeks after the flood, 4-5% of all visits to health care facilities in disaster counties were attributed to psychosocial conditions as compared with only 2-3% of visits in non-disaster counties. The contrast was most dramatic in the Albany district where psychosocial symptoms accounted for 6-8% of health care visits in disaster counties and only 0-2% of visits in nondisaster counties. This apparent excess may be due in part to the differ-

Figure 3. Diarrheal Illness in Flood Affected and Unaffected Counties by Age
The proportion of health care visits attributed to psychosocial conditions is much higher among adults (>15 yrs) than among children regardless of whether an area was flood affected or not (Figure 4). However, when disaster counties were compared with non-disaster counties, the proportion of morbidity among persons younger than 15 years of age that was attributed to psychosocial symptoms for the entire surveillance period was 3.5 times higher in the disaster counties than in non-disaster counties. In contrast, among people 15 to 54 years of age, the proportion of medical visits attributed to psychosocial morbidity was 40% higher in disaster counties than in non-disaster counties, although this difference leveled off by weeks 6-7.
* proportions are significantly different, p<0.05 ** proportions are significantly different by chi square test, p<0.01 *** proportions are significantly different by chi square test, p<0.001

- 2 -

These findings suggest that although children are less prone than adults to be treated for psychiatric illness in non-disaster situations, they are vulnerable to episodes of emotional symptoms during the stress of a disaster and its aftermath. Although the long-term effects of the flood on the psychological health of the most severely affected families are not known, the findings of the surveillance system suggest that most people recovered rapidly from any acute psychiatric symptoms brought on by the stress of the flood disaster.
Figure 4. Psychosocial Morbidity in Flood Affected and Unaffected Counties by Age

case load in flooded counties (0.4%) than non-flooded counties (0.2%). The difference between flooded and non-flooded counties was greatest during weeks 3-4 of surveillance and may, like the excess of skin rash and diarrhea, be related to clean-up activities that were going on during that period.
This surveillance system was a useful tool for monitoring the occurrence of potential post-flood public health threats but was laborintensive. Modification of the system, by selecting sentinel sites within the affected areas rather than surveying all providers, would likely have produced equivalent results with less intensive use of personnel resources. Although we found significant differences between flood-affected and unaffected counties regarding the proportion of morbidity attributable to diarrheal illness, skin rash, heat-related illness and psychosocial conditions, no large disease outbreaks occurred during the surveillance period. This information was useful for assuring the public and decision-makers that major epidemics were not occurring, enabling public health officials to respond more effectively to the flood disaster using objective data, and ensuring that resources were not wasted on unnecessary disease-control measures (e.g. mosquito spraying or mass vaccination campaigns).

Editorial Note
This model for active surveillance was modified July-August 1996 to monitor emergency room trends during the 1996 Olympics. (See June 1996 issue).

Skin rash The counties declared federal disaster areas also reported an
increase in the proportion of visits attributed to skin rash (Figure 5). This increase was at its highest during weeks 3 and 4 of surveillance and was likely associated with activities involved in cleaning up after the flood (e.g., using caustic cleaning agents, coming in contact with chemical irritants or biological agents in flood water). By week 6, the proportion of morbidity attributed to skin rash in flood-affected counties, although still greater than the proportion in less-affected counties, was no longer statistically different. Unlike the situation seen with diarrheal illness and psychiatric morbidity where increased illness as a result of the floods seemed to be mainly a problem among children, the proportion of morbidity in disaster counties attributed to skin rash was significantly greater than in non-flood counties for all age groups.
Figure 5. Skin Rash in Flood Affected and Unaffected Counties by Week

References
1. Centers for Disease Control, "Flood Disasters and Immunization--California." MMWR;1983; 32:171-2, 178.
2. Centers for Disease Control, "Injuries and illnesses related to Hurricane Andrew--Louisiana, 1992." MMWR;1992; 42:13.
3. Centers for Disease Control, "Morbidity surveillance following the Midwest flood--Missouri, 1993." MMWR;1993; 42:797-8.
4. Centers for Disease Control, "Outbreak of diarrheal illness associated with a natural disaster--Utah." MMWR;1983; 32:662-4.
5. Centers for Disease Control, "Rapid Assessment of vectorborne diseases during the Midwest flood--United States, 1993." MMWR;1994; 43:481.
6. Centers for Disease Control, "Surveillance of shelters after Hurricane Hugo-Puerto Rico." MMWR;1990; 39:41.
7. De Ville de Goyet C, Zeballos JL: "Communicable diseases and epidemiological surveillance after sudden natural disasters." In Baskett P, Weller R (eds): Medicine for Disasters. London:Wright, 1988, pp252-69.
8. Pan American Health Organization (1981). "Emergency Health Management after Natural Disasters". PAHO Office of Emergency Preparedness and Disaster Relief Coordination, Scientific Publication 407:1-76.
9. Seaman J. (1984) "Epidemiology of natural disasters. (Vol.5, Contributions to Epidemiology and Biostatistics series, ed. Kingberg, M.A.)
10. Toole MJ: "Communicable disease epidemiology following disasters." Ann Emerg Med. 1992;21:418-20.
11. Usher JH (1973) Philippine flood disaster." J Roy Naval Med Sevice 59:81-3.

Heat-related illness Although the number of medical visits attributed to heat-related
causes was small, it represented a significantly greater portion of the

This report was contributed by Annemarie Wasley, Division of Environmental Hazards and Health Effects, National Center for Environmental Health, CDC; and Kathleen E. Toomey, Epidemiology & Prevention Branch, GA Division of Public Health.
We would also like to acknowledge the work of the DIstrict Flood Team Leaders: Paul Morrison (Albany), Don McKinney (Macon), Tommy Craft (Dublin) and Darryl Harrelson (Columbus), who coordinated the data collection in their Districts.

- 3 -

The Georgia Epidemiology Report Epidemiology and Prevention Branch Two Peachtree St., NW Atlanta, GA 30303-3186

November 1996

Volume 12 Number 11

Reported Cases of Selected Notifiable Diseases in Georgia

Profile for August 1996

Selected Notifiable Diseases Campylobacteriosis Giardiasis

Total Reported for August 1996 68 88

Previous 3 Months Total

Ending in August

1996 1995 1994

249

350

402

218

160

108

Previous 12 Months Total

Ending in August

1996 1995 1994

818

1107

893

680

505

476

Meningococcal Disease Rubella Salmonellosis Shigellosis Viral Meningitis Tuberculosis Congenital Syphilis Early Syphilis Other Syphilis Cryptosporidiosis E. coli O157:H7 Legionnaires' Disease Lyme Disease Mumps Pertussis

9

31

12

13

153

82

91

0

0

0

7

0

0

7

156

444

494

528

1594

1530 1409

77

226

396

609

731

1840 1239

8

20

31

35

85

72

106

56

161

214

191

752

768

796

1

4

16

12

57

54

60

166

466

675

596

2200

2635 2922

87

222

351

231

989

1028

886

5

26

56

2

89

81

8

5

17

17

7

36

37

18

0

0

3

22

5

38

107

0

0

5

48

1

35

119

1

1

3

3

7

13

16

1

4

12

13

26

33

39

The cumulative numbers in the above table reflect the date the disease was first diagnosed rather than the date the report was received at the state office; and therefore are subject to change over time due to late reporting. The 3 month delay in the disease profile for a given month is designed to minimize any changes that may occur. This method of summarizing data is expected to provide a better overall measure of disease trends and patterns in Georgia.

Report Period

Total Cases Reported *

Last 12 Mos 11/95 to 10/96 5 Yrs Ago 11/90 to 10/91 Cumulative 01/80 to 10/96

2407 1451 16584

Percent Female
18.4 12.4 14.2

AIDS Profile Update

MSM

Risk Group Distribution (%) IDU MSM&IDU HS Blood Unknown

Race Distribution (%) White Black Other

44.8 17.1

4.3

58.5 20.0

5.3

52.5 19.0

6.0

16.2

1.5

16.1

34.4 62.8 2.8

9.2

2.0

5.1

46.4 52.2 1.4

10.8

2.1

9.6

41.2 56.8 1.9

MSM - Men having sex with men

IDU - Injection drug users

* Case totals are accumulated by date of report to the Epidemiology Section

HS - Heterosexual

- 4-