Epidemiologic Notes and Reports Hurricanes and Hospital Emergency-Room Visits -- Mississippi, Rhode Island, Connecticut
In September 1985, Hurricanes Elena and Gloria struck the Gulf of Mexico and Atlantic coasts of the United States, respectively, causing injuries, fatalities, and property damage. To determine the impact of the storms, health departments in three states -- Mississippi, Rhode Island, and Connecticut -- examined records of hospital emergency-room (ER) visits during and after the hurricanes.
Mississippi. On August 30, 1985, the National Weather Service issued a hurricane warning for portions of the Gulf Coast, including Mississippi. As Hurricane Elena approached Mississippi, coastal municipal and county govern- ments, through coordination with civil defense and emergency management agencies, evacuated low-lying areas. The eye of the hurricane passed over Gulfport on Monday morning, September 2. Sustained winds were recorded that day at 90 m.p.h., with gusts to 100 m.p.h.
The day after Hurricane Elena struck, the Office of Epidemiology, Mississippi State Department of Health, set up a hospital ER surveillance network to: (1) establish daily contact between the health department and the six hospitals serving the coast, so no reporting delay would occur if any unusual illness/injury patterns emerged, and (2) characterize the types of adverse health effects seen in the aftermath of the storm. ER personnel were asked to tally 24-hour totals for: (1) visits to the emergency room; (2) injuries, including lacerations and muscular strains; (3) gastroenteritis (because of loss of water pressure in local water systems); and (4) other conditions judged by the ER staff to be storm-related. In addition, ER personnel reported storm-related deaths of which they were aware.
Total visits peaked September 3 (551 visits), then, after approximately 1 week, decreased steadily to a level that represented the average daily baseline for the hospitals (331 visits) (Figure 1). The frequency of total injuries roughly paralleled that of total visits, and gastroenteritis never emerged as a problem during the 3 weeks of follow-up. At least three fatalities were thought to be storm-related: two in separate motor-vehicle crashes and one due to electrocution.
On September 26, 1985, the National Weather Service issued a hurricane warning for the Atlantic coast of the United States from North Carolina to Massachusetts. On Friday afternoon, September 27, Hurricane Gloria struck the New England coast. The storm weakened rapidly in its final approach and caused less damage than predicted. State health department investigators in both Rhode Island and Connecticut studied the effects of the storm on ER visits.
Rhode Island. On the evening of September 26, the governor of Rhode Island declared a state of emergency and implemented the state's disaster plan. Residents of low-lying coastal areas were evacuated to shelters. On September 27, schools and most businesses remained closed, and all nonemer- gency automobile travel was banned from 2 p.m. to 6 p.m. The peak wind velocity, 91 m.p.h., was recorded near 2 p.m. that day. Because the storm arrived at low tide and brought very little rain to the state, no significant flooding occurred. However, the wind downed trees and power lines, leaving more than half of Rhode Island residents without electricity.
To help assess the impact of the storm and the effectiveness of emergency measures, the Rhode Island Department of Health reviewed ER logs for September 20-October 6 at four coastal hospitals that serve approximately 325,000 persons. Medical records of persons identified in the logs for September 27-29 at three of these hospitals and death certificates for the entire state were also reviewed. Five storm-related fatalities were identified. Two of these persons sustained fatal injuries outdoors during the storm -- one, from a falling tree; another, in a boating incident. Two died from injuries related to the lack of electricity -- one fell in an unlighted area; one pedestrian was killed on a road with a nonfunctioning traffic light. One person suffered a cardiac arrest and fell from his roof while removing debris.
The number of ER visits dropped during the afternoon of the hurricane, then increased markedly over the next 2 days (Figure 2). Overall, 191 more patients were seen September 27-29 than the previous weekend of September 20-22, a 16% increase (Table 1). While the greatest relative increase in the rate of hospital admissions occurred the day of the hurricane, the greatest increase in ER visits was seen the following day. The age and sex distributions of patients were similar for the two periods.
Records of all ER visits to three of the four hospitals on September 27-29 were reviewed. Of 1,029 patients seen, 484 (47%) had sustained injuries. Among injured patients, the most common diagnoses were laceration (22%), abrasion or contusion (20%), sprain (14%), and fracture (12%). Four percent of the injured patients were admitted to the hospital. Over half the records reviewed included insufficient information to determine whether the visit was related to the storm or its aftermath. Eighty-nine (9%) records described visits clearly related to the storm; among them, 73 were injuries. Twenty-six (36%) of the 73 injured patients had lacerations, and another 11 (15%) had fractures. Nine storm-related chain-saw injuries were identified; four additional chain-saw injuries were reported without specifying circum- stances of injury. Compared with the 113 non-storm-related injuries, more of the storm-related injuries occurred among males (71%, compared with 60%), and among persons 40-49 years old (23%, compared with 6%).
Connecticut. The eye of Hurricane Gloria passed directly over Connecticut at 2 p.m. The extent of storm-related damage was similar to that described in Rhode Island. Investigators from the State of Connecticut Department of Health Services and the Department of Epidemiology and Public Health, Yale University School of Medicine, focused their study on a single coastal community and the effects of the storm on emergency medical services.
The city chosen was one of the most storm-damaged in the state. Approxi- mately 1,000 of the city's 51,430 residents were evacuated from shoreline homes. Telephone service was largely uninterrupted, but more than 70% of electrical utility customers lost power, many for 4-5 days.
During September 27-29, medical care was available primarily through the ER at the local 149-bed hospital. Two private walk-in medical-care centers remained closed until electrical power was restored September 30.
Two sources of information on medical emergencies in the community were examined: records of hospital ER visits and telephone calls to the 911 emergency number. ER log entries for September 21-October 4 were reviewed to determine the number of trauma and nontrauma visits. ER-patient records for September 27-30 (the hurricane period) and October 1-4 (the posthurricane period) were abstracted, and discharge diagnoses were assigned to one of 31 diagnostic categories, 14 of which related to trauma. Telephone calls to the 911 emergency number were classified as medical or nonmedical by emergency personnel. The 911 telephone log and incident reports were reviewed, and medical emergencies were divided into seven injury and eight noninjury categories.
During the hurricane period, both the total number of ER visits and the number of ER visits for trauma were increased, compared to the pre- and posthurricane periods (Figure 3). The proportion of total visits resulting from trauma did not change significantly, nor did the percentage of total visits that led to hospital admission from the ER. Trauma-related visits peaked September 28 (64 visits), but the most common date of injury occurrence, when specified, was September 27 (65 injuries). Compared to the posthurricane period, the proportion of total ER visits during the hurricane period was significantly greater in two of the 31 diagnostic categories (corneal abrasion, odds ratio 3.9; bee stings, odds ratio 17.3) and signifi- cantly less in one (psychiatric, odds ratio 0.23) (for all three, p 0.05). No difference was found in the age or sex distribution of persons seeking emergency medical care during these two periods.
During the prehurricane period, the average daily number of calls to the 911 number for both medical and nonmedical assistance was seven (range 3-11 and 4-10, respectively). The number of calls peaked on the day of the hurricane: 25 medical calls (four trauma) and 49 nonmedical calls (mostly for tree damage and downed electrical wires). Nonmedical calls reached a second but lower peak on October 1 (27 calls), temporally associated with the resumption of electrical power to most residents.
Reported by FE Thompson, MD, Mississippi State Dept of Health; J Brondum, DVM, L House, R Keenlyside, MD, W Hollinshead, MD, Rhode Island Dept of Health; C Averbach, R Bisson, MD, R Kohn, B Mosellie, J Parker, A Phelps, K Smith, MD, Dept of Epidemiology and Public Health, Yale University School of Medicine, New Haven, A Brandt, MD, Milford Hospital, G Kraus, MD, Milford Health Dept, M Cartter, MD, J Hadler, MD, State Epidemiologist, Connecticut Dept of Health Svcs; Div of Field Svcs, Div of Surveillance and Epidemiologic Studies, Epidemiology Program Office, Office of the Director, CDC.
Editorial Note: Previous surveillance studies during severe storms have underscored the need to plan for an increase in hospital staffing needs during the aftermath of the storm and to expect a considerable number of injuries to occur during clean-up activities (1-3). In Mississippi, demands on ER services peaked immediately after the hurricane, and injuries comprised an important fraction of the visits. Despite the loss of water pressure, no outbreaks of gastroenteritis were identified. Overall morbidity and mortality were probably best minimized by effective area evacuation before the arrival of the storm. Public information should continue to emphasize safety during the clean-up period. In Connecticut, the number of calls to the emergency number actually increased the day electrical power was restored.
Three factors probably contributed to the increase in ER visits following the hurricanes. First, some patients who might have gone to ERs the day of the hurricanes delayed their visits until the following days. Second, as in the Connecticut community, lack of electricity kept some alternate sources of medical care, particularly walk-in medical-care centers, closed, while hospitals were able to rely on emergency generators to remain open. Finally, some hurricane-related injuries did occur: in Rhode Island, 47% (89/191) of the overall increase in visits may have been hurricane-related.
In general, injury surveillance is hampered by poor documentation of the circumstances of injury in ER records. In Mississippi, considerable variation in reporting occurred from hospital to hospital and, on occasion, within the same hospital. For example, some hospitals included backache as an injury, others counted it in the "other" category, and others did not count it at all. Efforts should be directed towards improving data recording systems to increase their utility in evaluating the response of public health services during natural disasters. Health departments should adopt guidelines for information to be recorded by selected ERs in disaster areas, emphasizing circumstances of injury.
Disclaimer All MMWR HTML documents published before January 1993 are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.**Questions or messages regarding errors in formatting should be addressed to firstname.lastname@example.org.
Page converted: 08/05/98
This page last reviewed 5/2/01