Influenza activity was low during October 3-November 6, 1999; influenza virus isolates were reported from 30 states, and four long-term-care facility outbreaks were reported from three states. The predominant viruses isolated were influenza type A(H3N2) viruses. This report summarizes influenza activity in the United States during October 3-November 6, 1999. It also summarizes U.S. influenza surveillance methodology, including the four primary sources of surveillance data, a modification to pneumonia and influenza (P&I) mortality reporting, and discusses detection and control of institutional influenza outbreaks.
Each week from October through May, volunteer physicians in 47 states and the District of Columbia report the number of patient visits and the number of those visits for influenza-like illness (ILI). ILI is defined as cough or sore throat and a temperature of ≥100 F (37.8 C). Baseline levels of total patient visits for ILI range from 0 to 3%. Levels >3% usually correlate with increased influenza activity.
Each week during October-May, state and territorial epidemiologists report statewide estimates of influenza activity to CDC. Activity levels are defined as: (1) no activity, (2) sporadic—sporadically occurring ILI or culture-confirmed influenza with no outbreaks detected, (3) regional—outbreaks of ILI or culture-confirmed influenza in counties with a combined population of <50% of the state's population, and (4) widespread—outbreaks of ILI or culture-confirmed influenza in counties with a combined population of ≥50% of the state's population.
Each week throughout the year, the vital statistics offices for 122 U.S. cities report the total number of death certificates received and the number of death certificates on which influenza or pneumonia is listed on Part I (immediate, intermediate, or underlying cause of death) or Part II (contributing cause of death). These data are used to calculate a P&I mortality curve. A periodic regression model incorporating a robust regression procedure is used to estimate a seasonal baseline for P&I deaths. An increase of 1.645 standard deviations above the seasonal baseline for P&I deaths is considered the epidemic threshold.
Each week from October through May, approximately 115 WHO and NREVSS collaborating laboratories in the United States report the total number of specimens received for respiratory virus testing and the number testing positive for influenza A(H1N1), A(H3N2), A (not subtyped) and influenza B. A subset of isolates are submitted for complete antigenic characterization to CDC.
From October 3 through November 6, 1999, 1% of patient visits to sentinel physicians were for ILI. Among the nine surveillance regions, patient visits for ILI ranged from 0 to 3% during the week ending November 6, except in the West South Central region, which reported 5% of patient visits for ILI. For the week ending November 6, state and territorial epidemiologists in New York, Indiana, and Puerto Rico reported regional activity, and 35 states reported sporadic activity. No state reported widespread activity. A long-term-care facility outbreak was identified in New York on September 30, in New York City on October 14, in California on October 17, and in Illinois on November 3. During the week ending November 6, 621 (7.4%) of 8414 total deaths in 122 U.S. cities were attributed to P&I; this proportion was above the epidemic threshold of 6.5%. The proportion of P&I deaths has remained above the threshold for 7 consecutive weeks.
From October 3 through November 6, WHO collaborating laboratories and NREVSS laboratories in the United States reported 117 influenza A and four influenza type B laboratory-confirmed infections out of 5198 specimens submitted for respiratory virus tests. All 49 subtyped influenza A viruses were H3N2 viruses. Three influenza B viruses were isolated from persons returning to Tennessee from a trip to Ireland. Another influenza B virus was confirmed by CDC in addition to those reported by WHO and NREVSS laboratories. All 51 U.S. influenza A(H3N2) isolates collected from September 6 through November 6 and antigenically or genetically characterized at CDC were influenza A/Sydney/5/97-like (H3N2) viruses, and all four influenza B isolates were characterized as B/Yamanashi/166/98-like viruses. Both of these strains are contained in the 1999-2000 influenza vaccine.
C Waters, P Smith, MD, State Epidemiologist, New York State Dept of Health. R Taylor, DVM, W Reimels, A Craig, MD, W Moore, MD, State Epidemiologist, Tennessee Dept of Health. R Murray, DrPH, DJ Vugia, MD, Acting State Epidemiologist, California Dept of Health Svcs. CE Jennings, SL Bornstein, MD, Illinois Dept of Public Health. Participating state and territorial epidemiologists and state public health laboratory directors. World Health Organization collaborating laboratories. Sentinel Physicians Influenza Surveillance System. National Respiratory and Enteric Virus Surveillance System Laboratories. Surveillance Systems Br, Div of Public Health Surveillance and Informatics, Epidemiology Program Office; Mortality Statistics Br, Div of Vital Statistics, National Center for Health Statistics; Respiratory and Enterovirus Br and Influenza Br and WHO Collaborating Center for Reference and Research on Influenza, Div of Viral and Rickettsial Diseases, National Center for Infectious Diseases; and an EIS Officer, CDC.
Three of four influenza surveillance systems indicated that influenza activity was low from October through early November in the United States; however, 30 states reported laboratory-confirmed cases of influenza, and four long-term-care-facility outbreaks were reported. The 122 cities mortality reporting system data indicated that P&I mortality was above epidemic thresholds for 7 consecutive weeks; however, these results must be viewed with caution because recent changes have been made to the reporting system.
In 1993, the WHO International Classification of Diseases, Ninth Revision (ICD-9) coding guidelines were updated to International Statistical Classification of Diseases and Related Public Health Problems, 10th Revision (ICD-10), and were implemented by CDC's National Center for Health Statistics (NCHS) in 1999.1 For ICD-10, the application of a coding rule was broadened such that when pneumonia is listed by a certifying physician on a death certificate as the underlying cause of death, nosologoists should give preference to coding the cause of death to an alternative condition that might have led to the pnuemonia. Preliminary results from an NCHS comparability study have shown that the ICD-10 coding rule change will result in a substantial decrease in the number of reported pneumonia-related deaths (CDC, unpublished data, 1999).
In response to ICD-10, CDC requested that the 122 cities report pneumonia deaths to the surveillance system if pneumonia is listed anywhere on the death certificate. This may partially account for the observed increase in reported P&I deaths above threshold levels; baseline and threshold levels of P&I mortality are estimated using the previous 5 years' mortality data. CDC continues to evaluate the impact of these changes in reporting criteria on P&I mortality estimates.
Influenza introduced into hospitals and long-term-care facilities by patients, visitors, or staff can cause nosocomial outbreaks that can occur year-round, but tend to occur during periods of increased influenza activity, usually December-March. Institutional outbreaks can result in high attack rates among staff and patients and increased patient mortality, particularly among elderly and other vulnerable populations, such as bone marrow transplant patients.2- 5 In a survey of Emerging Infections Network (EIN) physicians,* conducted during the spring of 1999, 344 (74%) of 462 reported diagnosing influenza in hospitalized patients, and 65 (14%) recognized one or more nosocomial influenza cases during the preceding influenza season. Despite the frequent diagnosis of influenza among hospitalized patients, only 163 (35%) of 458 of the EIN physicians reported that their hospitals had a written policy for the control of nosocomial influenza outbreaks.6
When influenza outbreaks occur in health-care institutions, early recognition and initiation of control measures are important because influenza can spread rapidly in these settings.2,7- 10 The use of rapid diagnostic tests to confirm an influenza outbreak can facilitate the immediate activation of control measures such as cohorting ill patients, initiating droplet precautions, and using antiviral medications for influenza prophylaxis and treatment. Four influenza antiviral medications are available. Amantadine and rimantadine are approved for both treatment and prophylaxis of influenza type A but not influenza type B. Zanamivir and oseltamivir are active against influenza A and B viruses and are approved for the treatment but not the prophylaxis of influenza.7,8,10† Although antiviral medications are an important adjunct for the prevention and control of influenza, they are not a substitute for vaccination. Vaccination is the primary means of preventing influenza and is recommended for persons at high risk for influenza-related complications and persons who may transmit influenza to those at high risk, including health-care workers.7
Influenza surveillance data collected by CDC are updated weekly during October-May and are available by telephone, (888) 232-3228, or fax, (888) 232-3299 and requesting document number 361100, or through CDC's National Center for Infectious Diseases, Division of Viral and Rickettsial Diseases, Influenza Branch World-Wide Web site, http://www.cdc.gov/ncidod/diseases/flu/weekly.htm.
Influenza Activity—United States, 1999-2000 Season. JAMA. 1999;282(24):2295-2296. doi:10.1001/jama.282.24.2295-JWR1222-2-1