Launching a National Surveillance System After an Earthquake—Haiti, 2010 FREE

MMWR. 2010;59:933-938

2 figures, 1 table omitted

On January 12, 2010, Haiti experienced a magnitude-7.0 earthquake; Haitian government officials estimated that 230,000 persons died and 300,000 were injured. At the time, Haiti had no system capable of providing timely surveillance on a wide range of health conditions. Within 2 weeks, Haiti's Ministry of Public Health and Population (MSPP), the Pan-American Health Organization (PAHO), CDC, and other national and international agencies launched the National Sentinel Site Surveillance (NSSS) System. The objectives were to monitor disease trends, detect outbreaks, and characterize the affected population to target relief efforts. Fifty-one hospital and clinic surveillance sites affiliated with the U.S. President's Emergency Plan for AIDS Relief (PEPFAR) were selected to report daily counts by e-mail or telephone for 25 specified reportable conditions. During January 25–April 24, 2010, a total of 42,361 persons had a reportable condition; of these, 54.5% were female, and 32.6% were aged <5 years. Nationally, the three most frequently reported specified conditions were acute respiratory infection (ARI) (16.3%), suspected malaria (10.3%), and fever of unknown cause (10.0%). Injuries accounted for 12.0% of reported conditions. No epidemics or disease clusters were detected. The number of reports decreased over time. NSSS is ongoing and currently transitioning into becoming a long-term national surveillance system for Haiti. NSSS data could assist decision makers in allocation of resources and identifying effective public health interventions. However, data reporting and quality could be improved by additional surveillance education for health-care providers, laboratory confirmation of cases of disease, and Internet-based weekly reporting.

Before the January 12 earthquake, Haiti's national surveillance system focused on the following six immediately notifiable diseases: acute hemorrhagic fever syndrome, suspected meningococcal meningitis, suspected diphtheria, suspected acute flaccid paralysis, suspected measles, and bite by animal suspected of having rabies. Expansion of Haiti's national surveillance capabilities to monitor diseases and conditions of concern after the earthquake was a public health priority. Haiti is divided administratively into 10 departments; surveillance sites were spread across all departments, with additional sites sampled in Port-au-Prince. The 51 NSSS sites were selected from 99 PEPFAR sites that provided general care, based on their proximity to the earthquake epicenter, size, geographic representativeness, and capacity to submit data electronically after the earthquake. Selecting PEPFAR sites for NSSS enabled rapid establishment of post-earthquake surveillance despite the destruction of most governmental buildings, schools, homes, hospitals, and transportation and communication infrastructure in the West Department, which includes the capital Port-au-Prince, and much of the South-East Department. NSSS provided MSPP and its public health partners with daily information, including patient demographics and condition trends.

A standardized reporting form used by MSPP and PAHO during their summer 2008 response to Hurricane Gustav in the Caribbean was amended for NSSS to include a total of 25 conditions (including symptoms, suspected infectious diseases, acute injuries, and chronic conditions).* PEPFAR surveillance staff members at each site were instructed to report daily counts of the 25 conditions (as well as other, not specified conditions) with only one condition per new patient, and the total number of new patients examined each day for any condition. Patients were considered new if they had not been examined previously at the site for that condition. No explicit instructions were provided regarding which condition to report if a patient presented with more than one condition. Each patient was classified on the surveillance form by sex, age group (<5 years, ≥5 years, or unknown age), and morbidity and mortality status. Because the MSPP office was destroyed during the earthquake, for temporary data management the surveillance forms were submitted electronically (or if necessary, by telephone) to the CDC-Haiti office and then to the CDC Emergency Operations Center in Atlanta. A CDC epidemiology team entered data from the forms into a database and conducted data analyses. Cumulative daily surveillance reports were e-mailed from CDC-Atlanta to MSPP for immediate review, approval, and dissemination to public health partners working in Haiti. Frequencies of reported conditions were categorized as either from the two departments nearest the earthquake epicenter (West and South-East) or from the eight departments further away from the epicenter (North-West, North, North-East, Artibonite, Center, Grand Anse, Nippes, and South).

During January 25–April 24, 2010, a total of 48 of the 51 selected sites reported at least once to NSSS, with an average of 18 sites reporting each weekday (most sites did not report on weekends). The number of sites reporting decreased over time (both those nearest and further away from the epicenter), with an average of 23 sites reporting each weekday during January 25–March 14, 2010, and an average of 11 sites reporting during March 15–April 24, 2010.

Of the 42,361 new patients with reportable conditions, 23,081 (54.5%) were female, and 13,798 (32.6%) were aged <5 years. Nationally, the three most frequently reported specified conditions were ARI, 6,910 (16.3%); suspected malaria, 4,366 (10.3%); and fever of unknown cause, 4,240 (10.0%). Injuries accounted for, 5,065 (12.0%) of reported conditions. Among patients aged <5 years, the three most frequently reported specified conditions nationally were ARI, 3,895 (28.2%); acute watery diarrhea, 2,560 (18.6%); and fever of unknown cause, 1,565 (11.3%).

The two departments nearest the epicenter accounted for 53.6% (22,717) of the reported conditions. The three most frequently reported specified conditions in the departments nearest the epicenter were ARI, 4,027 (17.7%); suspected malaria, 2,437 (10.7%); and fever of unknown cause, 2,238 (9.9%). Injuries accounted for 2,084 (9.2%) of the reported conditions. In the eight departments further from the epicenter, among the 19,644 conditions reported, the three most frequently reported were ARI, 2,858 (14.5%); watery diarrhea, 2,059 (10.5%); and fever of unknown cause, 2002 (10.2%). Injuries accounted for 2,977 (15.2%) of the reported conditions.

R Magloire, MD, Ministry of Public Health and Population, Haiti. K Mung, MD, Pan American Health Organization. S Harris, MD, Y Bernard, MD, R Jean-Louis, MD, H Niclas, Global AIDS Program Office, CDC-Haiti. P Bloland, DVM, Incident Response Coordination Team, Public Health Br, US Dept of Health and Human Svcs. J Tappero, MD, ST Cookson, MD, KM Tomashek, MD, C Martin, MSPH, E Mintz, MD, KA Lindblade, PhD, E Barzilay, MD, RC Neurath, MS, Emergency Operations Center; SJ Vagi, PhD, WR Archer, PhD, EK Sauber-Schatz, PhD, EIS officers, CDC.

NSSS was instituted to monitor disease trends, detect outbreaks, and characterize the affected population to target post-earthquake relief efforts. NSSS surveillance data and laboratory reports were used to respond to rumors and concerns of disease clusters and outbreaks by providing evidence that no unexpected or abnormal increases in disease had been detected. Although not unexpected in postdisaster settings, underreporting, unclear case definitions, and limited laboratory capacity compromised the data quality and completeness expected of an effective surveillance system.¹ Control of NSSS was transferred from CDC-Haiti and CDC in Atlanta to MSPP on April 25; however, collaboration continues between these agencies and PAHO. As Haiti moves from the postearthquake emergency response phase into the recovery phase, efforts to increase reporting, add surveillance sites, improve data quality, and meet long-term surveillance needs by amending the list of reportable conditions are ongoing. These efforts will help ensure that Haiti's MSPP has a sustainable national surveillance system that will better identify unmet health needs in order to set priorities for the allocation of resources for effective interventions for improving public health in Haiti.

Although NSSS was implemented rapidly, the system could not describe the immediate effects of the earthquake. For example, most persons with earthquake-associated injuries were treated or transported immediately after the earthquake, before NSSS began operation, 13 days later. In addition, before establishment of NSSS and continuing during the first few weeks after the earthquake, hundreds of thousands of persons migrated out of the area nearest the epicenter. By January 31, an estimated 570,000 persons had migrated out of Port-au-Prince alone.² The migration might explain why both the number and proportion of injuries were higher in those departments further from the epicenter. In addition, as internally displaced persons (IDPs) camps³ arose in the departments nearest the epicenter, safe water provisions and the availability of health care increased, which might have affected disease trends.

NSSS did not detect any unexpected disease clusters or outbreaks during the reporting period. A few suspected clusters of diarrhea, measles, hemorrhagic fever, and typhoid were reported directly to MSPP. However, using NSSS data and in consultation with the Haiti National Laboratory and surveillance site staff members, investigators determined that no unexpected or abnormal increases in disease had occurred. NSSS is limited in surveillance capacity because of incomplete reporting and patients seeking care at non-NSSS sites. Disease surveillance in Haiti could be improved by investigating unreported cases identified through laboratory data, increasing the capacity of the Haiti National Laboratory to perform diagnostic testing, and informing health-care providers, including those outside of NSSS surveillance sites, of the need to report immediately notifiable cases to MSPP immediately. Sensitivity and specificity of the surveillance system should be evaluated; plans are being developed to evaluate NSSS systematically.

On April 25, 2010, in an attempt to simplify data entry, increase reporting, and improve availability of data, NSSS began allowing weekly (instead of daily) reporting of daily counts of the 25 conditions and added the ability of NSSS sites to enter their data directly and electronically through the PEPFAR Internet-based system, known as the Monitoring, Evaluation, and Surveillance Interface (MESI).† Through MESI, NSSS began transitioning into a weekly, long-term national disease surveillance system for Haiti. In addition, MSPP, PAHO, and CDC have supported development of a complementary surveillance system, the Internally Displaced Persons Surveillance System, to better represent IDPs living in camps and served by nongovernmental organization (NGO) clinics.³

Many previously documented challenges of postdisaster public health surveillance were experienced in Haiti, including logistical constraints, absence of baseline information, unavailable denominator data, and underreporting of conditions.⁴ Despite these challenges, NSSS was a valuable element of the public health response, providing daily reports to public health partners in Haiti during an emergency response and serving as a tool to respond to rumors or concerns of increases in disease. Incorporation of NSSS into PEPFAR's MESI will improve the long-term sustainability of the system by streamlining data entry, improving data quality, providing data on a standardized platform, and complying with International Health Regulations of World Health Organization.⁵ Through planned improvement to NSSS and ongoing surveillance, MSPP will gain both routine and postdisaster baseline data on diseases to characterize trends that will help identify and support health priorities in Haiti.

This finding in this report are based, in part, on contributions by N Barthelemy, EJ Baptiste, PA Joseph, D Lafontant, J Boncy, Haiti Ministry of Public Health and Population; J Moya, P Tabard, Pan American Health Organization; J Grant, MS Massoudi, R Miramontes, D Moffett, M Riggs, H Williams, Public Health Br, Incident Response Coordination Team, US Dept of Health and Human Svcs; C El Bcheraoui, T Chiller, K Diallo, M Dillon, M Humphrys, G Gomez, J Pruckler, field team, CDC-Haiti; M Brown, N Celestin, K Jean-Charles, P Joseph, Global AIDS Program Office, CDC-Haiti; D Broz, J Burkholder, J Cortes, E Choudhary, A Espinosa Bode, M King, J Painter, S Papagari Sangareddy, H Patrick, Y Redwood, A Schnall, T Simon, C Thomas, Z Wang, A Wolkin, C Yen, F Yip, Emergency Operations Center, and M Wellman, Geospatial Research, Analysis, and Svcs Program, CDC.

Little was known about diseases and injuries in Haiti immediately after the January 12, 2010 earthquake.

What is added by this report?

Creation of the National Sentinel Site Surveillance (NSSS) System enabled reporting, during January 25–April 24, 2010, of 42,361 reportable conditions. No clusters or outbreaks of disease were detected; the three most frequently reported conditions were acute respiratory infections (16.3%), suspected malaria (10.3%), and fever of unknown cause (10.0%). Injuries accounted for 12.0% of conditions.

What are the implications for public health practice?

Continued improvements to NSSS, including Internet-based reporting, improved data quality, and a standardized platform, will allow for long-term sustainability of a system that will provide critical information for decision making, resource allocation, and compliance with the International Health Regulations of the World Health Organization.

* The 25 conditions were as follows: (infectious) fever of unknown cause, suspected malaria, suspected dengue fever, acute hemorrhagic fever syndrome, acute watery diarrhea, acute bloody diarrhea, suspected typhoid fever, acute respiratory infection, suspected measles (fever and rash), tuberculosis, and tetanus; (noninfectious) acute malnutrition, skin disorder, renal failure, pregnancy complications or third trimester without previous care, mental health or psychological health, and chronic diseases not accounted for in other conditions; (injury) trauma, fracture, cerebral concussion from head injury, laceration from weapon or dagger injury, burns, wounds (infected), crush injury syndrome, and amputation.

† Available at http://www.mesi.ht.

REFERENCES

5 Available.