le many officials and experts we contacted believe aspects of the break investigation went quickly and well, nearly all of them also eve there were lessons to be learned. These lessons may be especially vant for acts of bioterrorism, where the outbreak of cases may be much e rapid and law enforcement agencies may need to be involved to vent terrorists from releasing additional biological agents. The time lable for decision-making and response may be compressed from days weeks to a matter of hours. The lessons we identified related primarily ddressing possible needs in five areas:

ocal surveillance and response capabilities,

communication among public health agencies,

coordination between public health and animal health efforts, apabilities of laboratories, and

efforts to distinguish between natural and unnatural events.

West Nile outbreak provided a number of lessons about surveillance. learned that many aspects of the surveillance network worked well, eding the response to the outbreak. These positive lessons can serve as lels for other communities that may have less substantial surveillance vorks. However, while several of the lessons are positive, the outbreak exposed some weaknesses.

human outbreak of West Nile began with a few unusual cases. The ntial that one or two persons' medical conditions could be an cation of some larger concern, such as an emerging infectious disease, not be readily apparent to the health professionals involved. In many s, such events might not be noticed until a number of physicians have rted the cases and the local health department identifies a cluster, or a ber of victims seek care for similar conditions at the same location. t responses by the doctors and nurses who first see such victims are cularly crucial in alerting the public health community to the ibility of a wider problem.

e West Nile outbreak, several actions were particularly important in iding this early alert, as well as in providing valuable evidence for the stigation. Among these actions are the following:

ne physician who encountered the first human cases at the local
>spital in Queens reported the unusual cluster of illnesses to local

8

GAO/HEHS-00-180 West Nile Virus Outbreak

public health officials. Such occurrences could easily go unrep for example, the physician does not consider the circumstance unusual enough to report or does not recognize a rare disease. • Epidemiologists and staff at the New York City Department of took a number of actions that were essential to containing the They quickly investigated and recognized the potential signific the initial case reports. Their interviews with patients and fam identified common features in how the patients were exposed doors, suggesting that a mosquito-borne disease might be invol canvassed all New York City area hospitals to identify potentia and throughout their investigation, they remained in daily touc many local, state, and federal officials who had quickly becom involved. These staff members said previous planning for biot response in place at the city health department was key to the of the Department's response.'

11

• Autopsies were performed on the victims. The New York City Department of Health and Office of Chief Medical Examiner w together to ensure that autopsies were performed on any fatal encephalitis. Autopsies were performed on over 25 fatal cases initially suspected as having viral encephalitis, including all 4 f of West Nile encephalitis that occurred among city residents. A to one assessment of the response, information obtained from autopsies pointed to a flavivirus as the cause and helped guide subsequent laboratory testing.12 Autopsy rates nationally have decreasing, at a time when public health officials believe they increase to help detect infectious diseases. The decline has be influenced by such factors as costs and jurisdictional and auth uncertainties.

12

"Officials indicated that most critical was the collaborative relationship in plac the health department and the Office of Emergency Management due to the bio planning efforts. This collaboration helped facilitate the rapid mobilization of e control measures, establishment of the public hotline, and rapid mobilization of the holiday weekend to assist in canvassing local neighborhoods with education 12Wun-Ju Shieh, Jeannette Guarner, Marci Layton, Annie Fine, James Miller, Der Grant L. Campbell, John T. Roehrig, Duane J. Gubler, and Sherif R. Zaki, "The R Pathology in an Investigation of an Outbreak of West Nile Encephalitis in New! Emerging Infectious Diseases, Vol. 6, No. 4 (May-June 2000).

ile the West Nile outbreak was identified more quickly than otherwise
ht have been expected because an astute physician reported two
sual cases, it still provides evidence that the reporting system could be
proved. The virus might have been identified earlier-perhaps by a week
ording to an involved official—if case reporting had been better and if
d baseline data showing past trends of encephalitis and related diseases
been available. Similarly, a physician we interviewed who had treated
st Nile patients said clinicians often do not know whom to call when a
ster of patients with a disease of unknown origin is noticed. Wildlife and
officials also indicated that within their fields there is a need for better
ormation and guidance about whom to contact in the public health
munity when an outbreak is suspected.

ese problems have been noted in other instances besides the West Nile
break. For example, a November 1998 workshop on public health
tems and emerging infections sponsored by the Institute of Medicine-
organization chartered by the National Academy of Sciences to examine
lic health policy matters-reported that physicians are not sure when
where to report suspicious cases of infection. The workshop also
orted that physicians are unaware of the need to collect and forward
ical specimens for laboratory analysis and may not be educated
arding the criteria used to launch a public health investigation. Unlike
case in New York City, where the health department had been actively
municating with physicians, the workshop found that there is often a
of communication between public health agencies and community
sicians, 13

999 assessment by the Institute found that disease surveillance systems
lace at local, state, and federal levels rely on systems of disease
orting from health providers that are notorious for their poor sensitivity,
of timeliness, and minimal coverage.14 Because an effective medical
onse to a bioterrorist event would depend in part on the ability of
vidual clinicians to identify, accurately diagnose, and effectively treat
ases (including many that may be uncommon), the Institute reported

Some Inadequac for Surveillance

Better Commun Needed Among Health Agencies

titute of Medicine, Public Health Systems and Emerging Infections: Assessing the bilities of the Public and Private Sectors, Workshop Summary (Washington D.C.: ɔnal Academy Press, November 2000), p. 5.

itute of Medicine and National Research Council, Chemical and Biological Terrorism:
arch and Development to Improve Civilian Medical Response (Washington D.C.:
›nal Academy Press, 1999), p. 66.

cies in Resources e Were Exposed

that education about the threat posed by bioterrorism and abou diagnosis and treatment of various agents deserves priority.

Although this outbreak was relatively small in terms of the num human cases, it taxed the resources of one of the nation's larges health departments. The strain on resources is particularly note because local health departments in the United States have initia all the investigations that led to the recognition of infectious dis outbreaks. At the time of the West Nile outbreak, the New York Department of Health had a unit of about eight people responsib surveillance and case investigations related to over 50 reportable diseases. Officials told us that having even this small number of staff available was critical to the quick response to the initial ou Once the outbreak was identified, these and other staff assigned from other agencies and departments worked long hours, seven week.

We reported in 1999 that surveillance for important emerging inf diseases is not comprehensive in all states, leaving gaps in the na surveillance network.15 Many state epidemiologists reported ina staffing for generating and using laboratory data—often conside reliable for case investigation purposes than physician-reported performing infectious disease surveillance. The Institute of Medi workshop reported that, in general, epidemiological investigatio surveillance efforts are challenged by a variety of factors. These changes in the health care system and the continuing use of pap disease-reporting systems in many locations, where surveillance consequently sporadic and inadequate.

16

unication Is ng Public

cies

Experts consider rapid and reliable communication among publi agencies to be essential to bioterrorism preparedness and coord Timely dissemination of information allows public health official decisions with the most current information available. During the outbreak, however, officials indicated that the lack of leadership initial stages of the outbreak and the lack of sufficient and secure

15 Emerging Infectious Diseases: Consensus on Needed Laboratory Capacity Co Strengthen Surveillance (GAO/HEHS-99-26, Feb. 1999), p. 2.

16 Public Health Systems and Emerging Infections, p. 4.

communication among the large number of agencies involved vented them from sharing information efficiently.

y officials interviewed pointed to the lack of clear reporting guidelines ne source of confusion. Knowing who was in charge or could act as an ncy spokesperson, and which agency was responsible for what, would e allowed each agency to operate more effectively. Some officials gested that each agency should have one "point person" overseeing rations and the flow of information.

ing the outbreak, local, state, and federal officials held daily conference
s coordinated by the City or State Department of Health, or CDC.
ing these calls, officials received up-to-date information on such topics
he human and animal surveillance systems, test results from each
oratory, and schedules for mosquito spraying. While these calls were
sidered necessary to ensure that all parties heard the same information,
y sometimes involved over 100 people and lasted 2 hours or more.17 As a
lt, key officials had less time to investigate the outbreak in the
oratory and in the field. Additionally, veterinary health officials were
cerned because they were not always included in these calls.

17

le a secure electronic communication network was in place at the time
e initial outbreak, not all involved agencies and officials were using it
e time. For example, because CDC's laboratory was not linked to the
York State network, the New York State Department of Health had to
is an intermediary in sharing CDC's laboratory test results with local
th departments. CDC and the New York State Department of Health
ratory databases were not linked to the database in New York City, and
ratory results consequently had to be manually entered there.
sicians, local health departments, and laboratory officials indicated that
ng the outbreak, it was sometimes difficult to determine the status of
ents' samples and of the laboratory results. During and since the
reak, however, officials indicated that the use and utility of the
'ork have improved for West Nile surveillance and information sharing.
g the network, the state has put together an interactive surveillance
em for mosquito, bird, and human disease reports. Since the fall of
, access to the network has been provided to more health officials,
ding animal health agencies, for tracking West Nile in animals and

Links Between
Animal Health
Becoming Mor

wvolved official indicated that these problems were improved with the implementation andard format and agenda for each call.

12

GAO/HEHS-00-180 West Nile Virus Outbreak