Influenza - United States, 1988-89.
Problem/Condition: CDC monitors the emergence and spread of new influenza virus variants and the impact of influenza on morbidity and mortality annually from October through May.
Reporting Period Covered: This report covers United States influenza surveillance conducted from October 1988 through May 1989.
Description of System: Weekly reports from the vital statistics offices of 121 cities provided an index of influenza's impact on mortality; 58 WHO collaborating laboratories reported weekly identification of influenza viruses; weekly morbidity reports were received both from the state and territorial epidemiologists and from 153 sentinel family practice physicians. Nonsystematic reports of outbreaks and unusual illnesses were received throughout the year.
Results: During the 1988-89 influenza season, influenza A (H1N1) and B viruses were identified in the United States with essentially equal frequency overall, although both regional and temporal patterns of predominance shifted over the course of the season. Throughout the season increases in the indices of influenza morbidity in regions where influenza A (H1N1) predominated were similar to increases in regions where influenza B predominated. Only 7% of identified viruses were influenza A (H3N2), but isolations of this subtype increased as the season waned and it subsequently predominated during the 1989-90 season. During the 1988-89 season outbreaks in nursing homes were reported in association with influenza B and A (H3N2), but not influenza A (H1N1).
Interpretation: The alternating temporal and geographic predominance of influenza strains A (H1N1) and B during the 1988-89 season emphasizes the importance of continual attention to regional viral strain surveillance, since amantadine is effective only for treatment and prophylaxis of influenza A.
Actions Taken: Weekly interium analyses of surveillance data produced through the season allow physicians and public health officials to make informed choices regarding appropriate use of amantadine. CDC's annual surveillance allows the observed viral variants to be assessed as candiates for inclusion as components in vaccines used in subsequent influenza seasons.
Influenza virus infections remain an important cause of morbidity and mortality in the United States [1,2]. During major epidemics, hospitalizations among the elderly and persons with chronic health problems may increase two- to five-fold compared with nonepidemic periods ; mortality from pneumonia and influenza, as well as from cardiopulmonary and other chronic diseases that can be exacerbated by influenza infection, may also increase substantially .
The most effective available preventive measure is annual immunoprophylaxis with inactivated (killed-virus) trivalent vaccine, which contains a virus from each of the three major categories of influenza viruses that have co-circulated in human populations since 1977 (B, A [H1N1], and A [H3N2]). However, long-term control of influenza by vaccination is complicated by the propensity of the virus for antigenic variation . Antigenic drift, which occurs in viruses in all three groups, necessitates annual reformulation of the vaccine [6-8].
Influenza-associated mortality and morbidity can also be reduced by the appropriate use of chemoprophylaxis or therapy with an influenza-specific antiviral drug (i.e., amantadine or rimantadine) . Only amantadine is currently licensed for use in the United States. The efficacy of both drugs is limited to influenza A .
CDC conducts national influenza surveillance every season from October through May to monitor influenza-associated morbidity and mortality, to detect the emergence and spread of new viral variants, and to assess the appropriateness of these variants as vaccine component candidates. This report summarizes final influenza surveillance data collected by CDC during the 1988-89 season in the United States. It updates preliminary 1988-89 data included in the previously published influenza surveillance summaries and interim reports [11-13] and allows comparison with surveillance trends during the seasons that both preceded and followed 1988-89.
Sources for influenza surveillance information during the 1988-89 season were similar to those used in both previous and subsequent years. In addition to receiving occasional telephone reports of outbreaks and unusual illnesses, CDC systematically collects data through the following four surveillance systems:
1. State and territorial epidemiologists' reports. Influenza activity, as assessed by the state and territorial epidemiologists, is reported to CDC on a weekly basis as widespread, regional, sporadic, or no activity.
2. Sentinel physician surveillance network. One hundred fifty-three physicians participated in a surveillance network in which the number of patients with influenza-like illness (ILI) per total number of patient visits by age group and the number of hospitalizations for ILI were reported weekly. In addition, a subgroup of 103 physicians collected nasopharyngeal specimens from selected patients and submitted the specimens to a contract laboratory for virus identification.
3. World Health Organization (WHO) collaborating laboratories in the United States. During 1988-89, 58 WHO collaborating laboratories (the majority from state or local health departments, with some university or hospital laboratories also participating) reported weekly the total number of specimens received for respiratory virus testing, as well as the number and type of influenza viruses identified.
4. CDC 121 Cities Surveillance System. Each week, the vital statistics offices of 121 cities reported the total number of death certificates filed that week due to all causes and the number of those for which pneumonia was identified as either the immediate or underlying cause of death or for which influenza was mentioned in any position on the certificate. A seasonal baseline was calculated by using a robust regression procedure in which a periodic regression model is applied to observed percentages of deaths due to pneumonia or influenza since 1983. An "epidemic threshold" for each season is arbitrarily set at 1.645 standard deviations above the seasonal baseline. Reported data are graphed against this seasonal baseline and epidemic threshold, providing an index for the impact of influenza on mortality .
State and territorial epidemiologists
Although sporadic ILI was reported by five state epidemiologists during the first surveillance week of the 1988-89 influenza season (CDC surveillance week 40, October 2-8, 1988), a sustained[unkeyable] regional level of influenza activity was first reported by Nebraska during the week ending December 17 (week 50) (Figure 1). Sustained widespread activity was first reported by Oklahoma the week ending January 14 (week 2 of 1989), by which time 10 states--in all regions except the Pacific region--were reporting regional or widespread levels of influenza activity (Figure 2). The number of states reporting either regional or widespread activity increased rapidly to a sustained peak of 32-36 states per week (primarily in the West South Central and South Atlantic regions) during each of 4 weeks beginning January 29 and ending February 25 (weeks 5-8) and then declined rapidly. The last reports of sustained widespread activity were received from Nebraska and Virginia the week ending March 11 (week 10). For the week ending March 25 (week 12), no states reported widespread activity and only seven states continued to report regional activity, a number that declined progressively through the week ending April 22 (week 15). Reports of sporadic ILI were received from <20 states per week from April 23 through May 6 (weeks 16-18).
Nationally, [is less than or equal to] 2% of patient visits reported by sentinel physicians before October 16 (week 41) were attributed to ILI. From October 16 through December 11 (weeks 42-49), 3%-4% of reported patient visits were attributed to ILI. This proportion increased steadily to a plateau of 7%-8% from December 25, 1988, through January 21, 1989, peaked at 11% during the 2-week period ending February 11 (weeks 5-6), and then dropped steadily to 3% by the week ending April 1 (week 13), where it remained essentially stable through the end of the influenza surveillance season (week 20, ending May 20, 1989) (Figure 3). Nationally, [Greater than or equal to] 4% of sentinel physician office visits were attributable to ILI for the 19 weeks from November 27, 1988, through March 25, 1989 (weeks 47-12). In individual surveillance regions, the peak proportion of sentinel physician office visits attributable to ILI ranged from 9% to 22%. These regional peaks temporally clustered during the first 2 weeks in February, but individually occurred as early as January and as late as mid-March.
Overall, 3% of visits to sentinel physicians for ILI resulted in hospitalizations. The highest proportion of total office visits per week that resulted in hospitalizations for ILI (0.3% of all office visits) was reported from January 15 to February 4, 1989 (weeks 3-5). This 3-week period immediately preceded the peak of both reported influenza virus isolations by WHO laboratories (an average of 610 isolates per week during weeks 5-7) and influenza-associated mortality (7.3% of reported deaths during week 8). Patients >64 years old accounted for only 9% of ILI visits to sentinel physicians but 35% of ILI-associated hospitalizations.
Four hundred eighty-three specimens submitted by sentinel physicians for testing yielded 140 influenza B and 69 influenza A viruses.
WHO collaborating laboratories
During the 1988-89 influenza season, the 58 WHO collaborating laboratories tested 29,537 specimens for respiratory viruses and identified 5,125 influenza viruses;2,530 (49%) were influenza B and 2,595 (51%) were influenza A.
Of all influenza A isolates subtyped this season, 86% were influenza A(H1N1) viruses (Figure 4).
Reported influenza virus identifications peaked during a 5-week period from January 29 through March 3, 1989 (weeks 5-9), when 59% of all isolations were reported (range in individual regions, 41%-75%) (Figure 5).
The pattern of influenza virus isolations varied both from region to region and temporally within regions. During the early part of the season, influenza A viruses predominated among isolates reported from the New England, Middle Atlantic, East North Central, and Pacific regions. Influenza A isolations continued to predominate throughout the season in the New England region. From early February through early March, the marked predominance of influenza A viruses in each of the remaining three regions shifted to a 1- to 5-week period of co-dominance, during which time essentially equal proportions of influenza A and B viruses were isolated (February 12 through 18 in the East North Central region, February 26 through March 11 in the Middle Atlantic region, and January 29 through February 18 in the Pacific region). After this co-dominant period, a clear predominance of influenza B isolations emerged in all three regions (Figure 6). Influenza B isolations remained predominant throughout the rest of the season in the Pacific region. Influenza A viruses reemerged in essentially equal proportions during the last 3 weeks that isolations were reported from the Middle Atlantic region (week 16, beginning April 16, through week 18, ending May 6). Only one isolate of influenza B was reported from the East North Central region afterweek 16, when influenza A virus again accounted for the majority of the relatively few influenza isolates (Table 1).
In contrast, influenza B virus clearly predominated the influenza isolates early in the season through late January in the South Atlantic, East South Central, West South Central, and Mountain regions. In the East South Central region, influenza B viruses continued to predominate throughout the season. In the remaining three regions, a 1-to 6-week period of essentially equal proportions of influenza A and B isolation began some time after January (February 12 through 18 in the West South Central region, February 5 through March 4 in the Mountain region, and January 29 through March 11 in the South Atlantic region), then shifted into a predominance of influenza A isolations for the rest of the surveillance season (Figure 6, Table 1).
Influenza A and B viruses co-circulated in the West North Central region during the early part of the surveillance season. Beginning in mid-February and continuing throughout the season, influenza B isolations clearly predominated (Figure 6, Table 1).
Nationally, 74% of influenza A viruses reported were subtyped. Within each region except the New England, Pacific, and Middle Atlantic, 91%-98% of all subtyped influenza A viruses were H1N1. In these other three regions, influenza A(H1N1) viruses accounted for 83% (95/115, New England), 81% (105/130, Pacific), and 59% (256/432, Middle Atlantic) of subtyped isolates.
Within the Middle Atlantic region, 89% of all influenza A isolates were subtyped; 70% of the 157 influenza A isolates subtyped before February 11 (week 6) were influenza A(H1N1) and 30% were influenza A(H3N2). However, influenza A(H1N1) and A(H3N2) viruses were isolated with almost equal frequency (53% and 47%, respectively) from February 12 through May 20 (weeks 7-20).
Through all sources, 43 states identified and reported influenza B viruses; 35 reported influenza A(H1N1) viruses; 23 reported influenza A(H3N2) viruses; and 38 states, including nine that reported neither influenza A(H1N1) nor (H3N2), reported influenza A viruses that were not subtyped.
Antigenic drift and changes in vaccine composition, 1987-89.
During the preceding 1987-88 season, influenza A(H3N2) had been the predominant strain. Many 1987-88 season H3N2 isolates resembled two antigenically distinguishable strains first identified in China in 1987 (A/Sichuan/2/87 and A/Shanghai/11/87),(1*) both of which were associated with outbreak activity affecting all age [TABULAR DATA OMITTED] groups . Antigenic drift from the 1987-88 vaccine strain A/Leningrad/360/86 (H3N2) can be inferred from the reactions of each of these variants with ferret antisera (Table 2). On the basis of these and other findings, A/Sichuan/2/87 was chosen for inclusion in the 1988-89 vaccine  (Table 3).
During the 1988-89 season, influenza A/Shanghai/11/87-like strains predominated among the influenza A(H3N2) isolates antigenically characterized by the WHO reference laboratory at CDC, although influenza A(H3N2) isolates were a minor component of that season. This strain replaced A/Sichuan/2/87 as the A(H3N2) vaccine component for the 1989-90 season  (Table 3).
Toward the end of the previous season (1987-88), isolations of both influenza B and A(H1N1) had increased. Antigenic variants of influenza B virus had circulated during the 1987-88 season; most outbreak-associated variants had resembled B/Victoria/2/87 , a new variant first identified in 1987 that was incorporated into the influenza vaccine for the 1988-89 season (Table 3).
During the 1988-89 season, influenza B viruses isolated outside Asia and characterized at CDC were predominantly B/Victoria/2/87-like. However, an antigenic variant, B/Yamagata/16/88, was first identified in Japan in the spring of 1988. Viruses isolated inside Asia were either B/Victoria/2/87-like or B/Yamagata/16/88-like, and the latter was chosen for inclusion in the 1989-90 vaccine  (Table 3). The antigenic relationships between B/Victoria/2/87, B/Yamagata/16/88, and B/Ann Arbor/1/86 (which had been the B component of the 1987-88 vaccine) can be inferred from the ferret antisera data (Table 2). [TABULAR DATA OMITTED]
Most influenza A(H1N1) viruses characterized during both 1988-89 and the two preceding seasons (1986-87 and 1987-88) were A/Taiwan/1/86-like. A/Taiwan/1/86 (H1N1) had been a stable trivalent vaccine component for two seasons and was retained for the 1989-90 vaccine [10,16,17] (Table 3).
CDC 121 cities surveillance system
The proportion of all deaths reported through the 121 Cities Surveillance System that were attributed to pneumonia and influenza exceeded the epidemic threshold for 11 weeks, from January 21 through April 8, 1989 (weeks 4-14), and peaked at 7.3% of deaths reported the week ending February 25 (week 8) (Figure 7).
From mid-November through the end of January (weeks 45-5), the South Atlantic, West South Central, East North Central, and Pacific regions reported influenza B outbreaks in middle or elementary schools, with high absentee rates and occasional school closures. Influenza B was the most frequently isolated virus in only two (South Atlantic and West South Central) of these four regions during this period. Outbreaks of influenza A (not subtyped) were reported from schools in the Pacific, New England, and South Atlantic regions. An outbreak of influenza A(H3N2) was reported from a nursing home in the Middle Atlantic region, which had the highest proportion (41%) of influenza A isolates subtyped as H3N2. An outbreak of influenza B in a nursing home was reported from the West South Central region. No reports of influenza A (H1N1) outbreaks in nursing homes were received, despite the relative predominance of this influenza virus among circulating subtypes.
Analysis of influenza surveillance data reveals distinct patterns of interplay of circulating influenza subtypes that vary among seasons. The co-dominance of two influenza types (influenzas B and A[H1N1]) during the 1988-89 season was very unusual. During most seasons one influenza type predominates, but others circulate in substantial numbers. Infrequently, such as during the 1989-90 season, circulating influenza strains are almost entirely (> 98%) one subtype.
The rare co-dominance of influenza B and influenza A (H1N1) viruses during the 1988-89 season and the shifting pattern of regional predominance over time make its mixed nature clearer than in other recent mixed influenza seasons such as 1991-92 . The variable spatial and temporal predominance of influenza A and B viruses during the 1988-89 season demonstrates the importance for influenza surveillance of continued virus culturing throughout the season. This alternating predominance also highlights the importance for practicing clinicians of continued attention to local influenza surveillance patterns, since amantadine is effective only against influenza A.
The temporal and geographic changes in predominance of influenza B or A(H1N1) illustrated clearly by strain surveillance were not reflected in distinguishable changes in patterns of visits to physicians' offices or outbreak activity. Reports of outbreaks in schools associated with both influenza B and A were received in essentially equal numbers, though reports of influenza A-associated outbreaks tended to be among older age groups. Although 86% of all 1988-89 influenza A isolates subtyped were H1N1, outbreaks in nursing homes were reported only in association with influenza B, A (not subtyped), and A (H3N2).
The "herald wave" phenomenon, in which a relatively small cluster of influenza virus infections occurring during the latter half of one influenza season is predictive of the predominant virus for the following season, was first described by Glezen and colleagues [18,19]. Subtle herald waves both preceded and closed the 1988-89 influenza season. The co-dominant nature of the 1988-89 season had been heralded by a late season increase in isolation of both B and A(H1N1) influenza viruses during the 1987-88 season, during which influenza A(H3N2) had been predominant among circulating viruses . A herald wave of influenza A(H3N2) isolations that began late in the 1988-89 season presaged the predominance of this viral subtype during the subsequent 1989-90 season .
Although the influenza A (H3N2) component of the trivalent influenza vaccine had been stable for three seasons, from 1983 to 1986 [20-22], sufficient antigenic diversity was identified among circulating A (H3N2) viruses during each season from 1986 through 1991 (including 1988-89) to warrant an annual change in this vaccine component [11,14-16,23-25].
B/Victoria/2/87 was first identified in 1987, and B/Yamagata/16/88 was first recognized in 1988. The vaccine component was changed from B/Victoria/2/87 to B/Yamagata/16/88 between the 1988-89 and 1989-90 seasons . However, influenza B viruses closely related to both these reference strains continued to co-circulate through 1991 [11-15], although the identification of B/Victoria lineage viruses was increasingly infrequent after 1990 [12,13,16,24].
A/Taiwan/1/86 had initially been isolated during the spring of 1986, incorporated into a supplemental monovalent vaccine for the 1986-87 season , and subsequently into the trivalent vaccine for the 1987-88 season. In the absence of substantial identified antigenic drift among circulating influenza A (H1N1) viruses, A/Taiwan/1/86 remained the only constant vaccine component through five seasons (including 1988-89) until 1992-93 [11,14,15,23-25] (Table 3).
Analysis of influenza surveillance in the United States during the 1988-89 season revealed an unusual co-dominance of influenza B and Influenza A (H1N1) viruses. The spatial and temporal predominance of these virus types varied throughout the season, complicating clinical decision-making regarding the appropriate use of antiviral agents.
(1*) Widespread: outbreaks of influenza-like illness (ILI) or culture-confirmed influenza in counties having a combined population of [is greater than or equal to] 50% of the state's total population. Regional: Outbreaks of ILI or culture-confirmed influenza in counties having a combined population of <50% of the state's total population. Sporadic: sporadically cases of ILI or culture-confirmed influenza, with no outbreaks detected.
[unkeyable] Sustained reporting: same level reported for [is greater than or equal to]2 consecutive weeks.
(1*) The name given to an influenza reference virus indicates both when and where it was first isolated. For example, A/Sichuan/2/87 (H3N2) was the second influenza A (H3N2) virus isolated in Sichuan, China, in 1987 and A/Shanghai/11/87 (H3N2) was the eleventh influenza A (H3N2) virus isolated in Shanghai, China, in 1987.
[1.] Glezen WP, Couch RB. Interpandemic influenza in the Houston area, 1974-76. N Engl J Med 1978;298:587-92.
[2.] Glezen WP. Serious morbidity and mortality associated with influenza epidemics. Epidemiol Rev 1982;4:25-44.
[3.] Barker WH. Excess pneumonia- and influenza-associated hospitalizations during influenza epidemics in the United States, 1970-1978. Am J Public Health 1986;76:761-5.
[4.] Lui KJ, Kendal AP. Impact of influenza epidemics on mortality in the United States from October 1972 to May 1985. Am J Public Health 1987;77:712-6.
[5.] Van Voris LP, Young JF, Bernstein JM, et al. Influenza viruses. In: Belshe RB, ed. Textbook of human virology. Littleton, MA: PSG Publishing Company, Inc., 1984, 267-97.
[6.] Striver HG, Graves P, Eickhoff TC, et al. Efficacy of 'Hong Kong' vaccines in preventing 'England' variant influenza A in 1972. N Engl J Med 1973;289:1267-71.
[7.] Meiklejohn G, Eickhoff C, Graves P, et al. Antigenic drift and efficacy of influenza virus vaccines, 1976-77. J Infect Dis 1978;138:618-24.
[8.] Noble GR. Epidemiological and clinical aspects of influenza. In: Beare AS, ed. Basic and applied influenza research. Boca Raton, Florida: CRC Press, 11-50.
[9.] ACIP. Prevention and control of influenza: recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR 1992;41(No. RR-9), 1-17.
[10.] Dolin R, Reichman RC, Madore HP, Maynard R, Linton PN, Webber-Jones J. A controlled trial of amantadine and rimantadine in the prophylaxis of influenza A infection. N Engl J Med 1982;307:580-4.
[11.] CDC. Update: influenza activity--worldwide, 1988-89. MMWR 1989;38:817-8.
[12.] CDC. Influenza--United States, 1989-90 and 1990-91 seasons. In: CDC Surveillance Summaries. MMWR 1992;41(No. SS-3):35-45.
[13.] CDC. Influenza surveillance--United States, 1991-92. In: CDC Surveillance Summaries. MMWR 1992;41(No. SS-5): 35-43.
[14.] CDC. Influenza--United States, 1987-88 season. MMWR 1988;37:497-503.
[15.] CDC. Update on influenza activity--United States and worldwide, with recommendations for influenza vaccine composition for the 1988-1989 season. MMWR 1988;37:241-4.
[16.] CDC: Update: influenza activity--United States and worldwide, and composition of the 1992-93 influenza vaccine. MMWR 1992;41:315-7;323.
[17.] ACIP. Monovalent influenza A (H1N1) vaccine, 1986-87. MMWR 1986;35:517.
[18.] Glezen WP, Couch RB, Taber LH, et al. Epidemiologic observations of influenza B virus infections in Houston, Texas, 1976-77. Am J Epidemiol 1980;111:13-22.
[19.] Glezen WP, Couch RB, Six HR. The influenza herald wave. Am J Epidemiol 1982;116:589-98.
[20.] ACIP. Recommendation: Influenza vaccine, 1983-84. MMWR 1983;32:333-7.
[21.] ACIP. Prevention of influenza. MMWR 1984;33:253-60;265-6.
[22.] ACIP. Prevention and control of influenza. MMWR 1985;34:261-8;273-5.
[23.] ACIP. Prevention and control of influenza. Recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR 1990;39:(RR-7).
[24.] CDC. Update: Influenza activity--United States and worldwide, and the composition of the 1991-92 influenza vaccine. MMWR 1991;40:231-3;239-40.
[25.] CDC. Influenza--United States--1986-87 Season. MMWR 1988; 37:466-70,475.
[26.] CDC. Antigenic variation of recent influenza A(H1N1) viruses. MMWR 1986;35:510-2.
|Printer friendly Cite/link Email Feedback|
|Author:||Chapman, Louisa E.; Tipple, Margaret A.; Folger, Suzanne Gaventa; Harmon, Maurice; Kendal, Alan P.;|
|Publication:||Morbidity and Mortality Weekly Report|
|Date:||Mar 19, 1993|
|Previous Article:||Surveillance for and comparison of birth defect prevalences in two geographic areas - United States, 1983-88.|
|Next Article:||Welcoming remarks.|