DETROIT Varicella incidence decreased during 1996 in the United States but not necessarily because of the vaccine, according to new surveillance data presented at the 31st Annual National Immunization Conference recently held here.
To document changes in the epidemiology of varicella secondary to the introduction of varicella vaccine, a project for the five-year active surveillance of varicella in three different sites the United States was established by the Centers for Disease Control and Prevention (CDC) at the end of 1994.
The goal of the project is to determine the baseline incidence and epidemiology of varicella pre- and post-vaccine licensure in order to monitor the effect of the varicella vaccine program on disease occurrence and complications, said Jane Seward, MBBS, MPH, National Immunization Program (NIP).
Although the vaccine was licensed in March 1995 and received in the private sector as early as May, the public sector did not receive the vaccine until the second half of 1996 because of a delay in signing the federal contract.
"We did not anticipate the one-year delay in the signing of the federal contract so it took a little longer for the vaccine to penetrate into the community," she said.
The three communities participating in the surveillance are west Philadelphia, Antelope Valley, Calif., and Travis County, Texas. The study populations in the Antelope Valley and Philadelphia are approximately 300,000 each while the study population in Travis County is approximately 600,000.
Each site uses many sources of reporting including day care centers, public and private schools, universities, public and private health care providers, hospitals and families.
New cases are reported bimonthly and sentinels which fail to report are contacted by phone. Investigators contact every case to obtain clinical and disease tracking information and follow-up interviews are conducted with all susceptible household members for occurrence of secondary or tertiary cases. Complete information on age, gender, race/ethnicity, date of disease onset, disease severity, complications, duration of illness, use of medications and vaccination status is recorded.
Varicella has been a reportable disease in Texas since 1972 with an average of 20,000 cases per year statewide. However, Jan Pelosi, MPH, Immunization Division, Texas Department of Health (TDH), said that figure represents only about 3% of actual disease incidence.
Based on available data, a dramatic decline (49%) was observed from 1995 to 1996 in the number of reported cases in Travis County, Pelosi said. Approximately 4,000 cases were reported to the Austin-area surveillance project at the TDH in 1995, but only 3,100 were actually confirmed; 1,550 cases were confirmed in 1996.
The high disease incidence in 1995 may have reflected an outbreak year, but the significant decline in 1996 could be a combination of vaccine usage and the standard decline of cases after an outbreak year. "Estimated coverage rates of varicella vaccine among 2-year-olds in Travis County for 1995 and 1996 was less than 20%, so the likelihood of vaccine causing the decrease in disease is less likely than the cyclical variation of the disease," Seward said. "However, use of the vaccine varied in some sectors and was probably very high in the private sector compared with the public sector."
Most cases (78%) occurred in whites with fewer cases among blacks (18%), which is similar to 1995 figures. This is attributed to the small black and Hispanic population in Travis County.
However, the number of Hispanic cases of varicella increased from 1995 to 1996, which is still unexplained, Pelosi said.
The distribution of cases by age is slightly different in Texas than compared with the other test sites. Most cases in Travis County occurred in children younger than 5, possibly because of the large number of children enrolled in day care at an early age because of two-income families.
Data sources in Travis County include about 400 reporting sentinels: licensed day care centers, elementary and high schools, five acute care hospitals, primary care providers and 10 public health clinics.
Pelosi said varicella vaccine has been generally accepted based on the number of telephone calls and letters from parents, schools and day care facilities. However, opposition from physicians who don't want to administer the vaccine has also been noted.
The study population in west Philadelphia is different from that in Texas: the population is primarily low-income families with one-quarter of the population living in poverty, said Kimberly L. Goodnow, MPH, Philadelphia Department of Public Health, Infection Control Division.
Although the vaccine wasn't available until October 1996, the incidence of varicella declined almost 50% during that year. Part of the decline, Goodnow said, is most likely associated with of a severe snow storm in January 1996, which may have halted disease transmission because children were out of school for several days.
Forty breakthrough cases of varicella have been reported since usage of varicella vaccine began; only six cases (15%) required hospitalization because of an existing medical condition.
Most cases (44%) in Philadelphia occurred in 5- to 9-year-olds, followed by preschool-age children (30%); of those cases, most were among blacks (81%). Most other cases were non-Hispanic (64%).
Goodnow said various complications related to varicella were also reported; 8% of reported cases experienced fever, vomiting, diarrhea, otitis media, bacterial super-infections, pneumonia or encephalitis. The most severe complications were among those younger than 1 year of age and older than 15 years, Goodnow said.
This project consists of more than 200 sites including public and private schools, hospitals, public health and inner-city clinics, licensed day care centers and private physicians. All schools in Philadelphia were included in the study, even those outside the sample population.
Schools reported most of the cases, but an increase in cases reported from hospitals was observed in 1996.
A survey of private-sector physicians in the active surveillance sites revealed that 83% of 47 responded; only eight of the providers were administering varicella vaccine to their patients.
The Antelope Valley Surveillance Project is located primarily in a rural area with the majority of population clustered in three larger cities and about 25 smaller towns. Antelope Valley consists of primarily white middle-class families and the estimated population of blacks is about 6%, said Carol Peterson, MD, MPH, County of Los Angeles Department of Health Services.
Based on available data, 7,222 cases of varicella occurred among 2- to 18-year-olds in Antelope Valley during 1995, which is approximately half of the national rate. The age group with the highest incidence was the 5- to 9-year-olds with 1- to 4-year-olds having the second highest incidence rate. A significantly higher rate of incidence was observed in blacks.
The Antelope Valley surveillance project is taking a different approach to evaluating data and is using population-based surveillance, rather than a sample population. Surveillance sites include all public and private schools and day care facilities with an enrollment of 12 or more, all hospitals and clinics, primary care physicians and three large correctional facilities in the area. Also added was an informal surveillance unit for families.
Another objective for this project is to estimate the completeness of the surveillance data using a capture/recapture method. This was originally developed for the numeration of wildlife population and later used by the Census Bureau, Peterson said, but is becoming increasingly popular in epidemiology.
For capture-recapture analysis of surveillance data for children age 2 to 18 years, the project used two data sources: schools (including day care and preschools) and health care providers (including physicians, hospitals and clinics). Using these data, investigators estimated 7,222 cases of varicella should have occurred over the two-year period. However, 3,999 cases were missed.
Overall completeness of surveillance data from all data sources was 62%. Schools were the most efficient source of data for varicella and health care providers were the least efficient.
"Even an exhausting active surveillance will result in a substantial number of missed cases," she said. "While there needs to be a lot more work before we can really apply capture and recapture varicella surveillance, I think it may be a simple, cost-effective means to estimating missed cases."
While the CDC doesn't think these surveillance projects would be feasible for the whole country, we expect to learn a lot from the surveillance, said Seward.
"These sites are doing a marvelous job of active surveillance," she said.
For more information:
- Goodnow KL, Watson B, Lutz J, et al. Epidemiology of varicella in an inner-city population. Abstract 280. Presented at the 31st Annual National Immunization Conference. May 19-22. Detroit.
- Luckey P, Gonzalez O, Howell B, et al. Varicella active surveillance project 1995 and 1996; Travis County, Texas. Abstract 284. Presented at the 31st Annual National Immunization Conference. May 19-22. Detroit.
You can express your views on this article, or other relevant themes, in the Infectious Diseases in Children Specialty Forums.