Technical guidelines for the application of seasonal influenza vaccine in China (2014–2015)

Abstract

Influenza, caused by the influenza virus, is a respiratory infectious disease that can severely affect human health. Influenza viruses undergo frequent antigenic changes, thus could spread quickly. Influenza causes seasonal epidemics and outbreaks in public gatherings such as schools, kindergartens, and nursing homes. Certain populations are at risk for severe illness from influenza, including pregnant women, young children, the elderly, and people in any ages with certain chronic diseases.

To strengthen the technical guidance for control and prevention of influenza, and to promote influenza vaccination in China, the Chinese Center for Disease Control and Prevention organized a panel of experts to review the latest international studies on influenza vaccination, including the “WHO Position Paper on Influenza Vaccines – November 2012,” and to compile the “Technical Guidelines for the Application of Seasonal Influenza Vaccines in China (2014–2015).” The guidelines systematically review the published literatures and unpublished works from the latest influenza research in China, including its etiology, clinical characteristics, laboratory diagnosis, epidemiology, disease burden, types of influenza vaccines, immune response mechanisms, durability of immunity, immunogenicity, vaccine efficacy, effectiveness, safety, cost-effectiveness and cost-benefit. On the basis of existing scientific evidence, the guidelines provide recommendations for influenza vaccination in the influenza season of 2014–2015.

In China, influenza vaccine is not incliuded in the national immunization program, and recipients shall pay for the vaccine voluntarily out of pocket. Points of Vaccination clinics (POVs) should provide immunization services for all individuals aged 6 months and above who are willing to be vaccinated and have no contraindications. To decrease the risk of severe complications due to influenza infection, the guidelines recommend annual seasonal influenza vaccines to be administered to pregnant women, children aged 6–59 months, adults ≥60 y of age, persons with specific chronic diseases, healthcare workers, and family members and caregivers of infants <6 months of age.

The guidelines can be used by staff of the Centers for Disease Control and Prevention at all levels who engage in influenza control and prevention, POVs staff, healthcare workers in the departments of pediatrics, internal medicine, and infectious diseases in medical institutions, and the staff of maternal and child health institutions. The guidelines will be updated periodically as new evidence becomes available.

Preface

Influenza is a respiratory infectious disease that endangers China's public health. According to the estimation, by the World Health Organization (WHO), 3–5 million severe cases and 250,000–500,000 deaths in the global scope are caused by seasonal epidemics of influenza every year.¹ Influenza vaccination is the most effective way to prevent influenza. To promote the prevention of influenza in China, and to decrease the morbidity and mortality caused by influenza, the Chinese Center for Disease Control and Prevention (China CDC) organized a panel of experts to review the latest progress domestic and overseas and compiled Technical Guidelines for the Application of Seasonal Influenza Vaccines in China (2014–2015).

The guidelines update the corresponding documents compiled by China CDC in 2007, 2008, and 2009 with reference to “WHO Position Paper on Influenza Vaccines – November 2012,” based on the existing scientific evidence and China's operational situation. The evidence was mainly from relevant documents recently published domestically and internationally and recommendations on influenza vaccines by health authorities, such as WHO, the United States Advisory Committee on Immunization Practices (ACIP), etc.

This guidance can be used by staff members of the CDCs at all levels who work on influenza control and prevention, POVs staff members, healthcare workers (HCW) in the departments of pediatrics, internal medicine and infectious diseases at medical institutions and the staff members of maternal and child health institutions. The guidelines will be updated and optimized according to the research progress in domestic and overseas.

Etiologic Basics, Clinical Characteristics, and Laboratory Diagnosis of Influenza

Influenza is an acute respiratory infectious disease caused by the influenza virus. The burden of influenza differs in a different area, and population, and about 5–10% of adults and 20–30% of children suffer from seasonal influenza in the world annually.² New variant strains constantly emerge and is widely transmitted among the population, causing repeated epidemics and outbreaks because of high genetic variability and multiple hosts of the viruses.

Etiologic basics of influenza

Influenza virus, belonging to the family orthomyxoviridae, is a single-stranded, segmented negative-sense RNA virus. It is further classified as type A, B, and C, according to the virus nucleoprotein and matrix protein.³ There are 8 RNA segments for influenza A and B virus, encoding 10–11 proteins. Hemagglutinin (HA) and neuraminidase (NA) proteins are located on the surface of the virus, playing an important role in the process of viral replication and infection. Influenza A virus could be classified as various sub-types according to the HA and NA proteins and their genetic characteristics. Currently, there are 18 (H1–18) HA subtypes and 11 (N1–11) NA subtypes being identified.⁴

RNA polymerase makes errors every 10,000 replications of nucleotides since no RNA proofreading enzyme is participating in the replication of influenza virus,⁵ resulting in a higher mutation frequency than other viruses. Moreover, the segmental genome of influenza virus makes gene reassortment possible when one cell is infected by different types of virus, leading to more profound changes in the viral genome. Therefore, most newly replicated influenza viruses have mutations.⁶ There are 2 kinds of mutation mechanisms for influenza virus, i.e., antigenic drift and antigenic shift.⁷ Antigenic drift is the result of the accumulation of point mutation in HA gene and NA gene during replication, which causes the repeated epidemics of influenza, e.g., seasonal influenza.^8–11 Antigenic shift only occurs with influenza A virus, which could generate new sub-types. This situation is usually caused by reassortment of influenza virus prevalent in human and animal, or by significant mutations of influenza virus in the animals that sometimes lead to direct infection of human breaking through the species' barrier. If the new subtype of the influenza virus with antigenic shift is capable of transmitting to human beings, the influenza pandemic occurs because human lacks immunity.⁷ For example, the 2009 pandemic H1N1 A virus [A (H1N1) pdm09] derives from the reassortment of strain in poultry, swine, and human.¹²

All subtypes of influenza A virus can infect poultry, especially aquatic birds, and pigs, horses, seals, whales, mink, and other mammals. Humans can be infected by a few sub-types of influenza A virus; among which the most common sub-types are H1-H3 and N1, N2.^2,13 Humans are the natural host for influenza B virus although it is also found in seals and ferrets. Influenza B virus mutates less frequently and can cause seasonal epidemic. Influenza C virus infects humans, dogs, and pigs with a much more stable structure than type A and B, and the antibody against influenza C virus is positive in 80% of people at the age of 7–10 y Therefore influenza C virus only causes sporadic cases of upper respiratory tract infection, and lower respiratory tract infection by influenza C virus is almost only observed in children aged < 2, such as bronchopneumonia.¹⁴

Influenza is mainly transmitted through droplet of the respiratory tract secretions and direct contact.^2,3 The incubation period ranges from 1 to 4 d (2 d on average). The virus can be shed in 24 to 48 hours before the onset of clinical symptoms, and the amount of viruses shed increases significantly in 0.5–1 d after infection and reaches its peak within 24 hours after the onset.¹⁵ Usually, adults and older children continue to shed viruses for up to 6 d (3–8 d), but the duration of virus excretion differs according to the strains infected. Hospitalized adult patients can excrete for one week or longer after the onset.¹⁶ The amount of virus excreted in younger children is the same as adults, but the amount decreases slower, and the duration lasts longer.¹⁷ Long-term excretion in the infants and toddlers is much more frequent (1–3 weeks) compared with the adults. The elderly, people with immunocompromised conditions, such as HIV infection, can shed viruses for a longer time.^15,18

Clinical characteristics and laboratory diagnosis of influenza

Influenza illness is characterized by the abrupt start of fever (up to 39 to 40°C) with chills, shiver, headache, myalgia and joint pain, extreme fatigue, anorexia and other systematic symptoms with sore throat and cough, nasal congestion, running nose, retrosternal discomfort, facial blushing, and conjunctival mild hyperemia. Some patients present symptoms such as vomit and diarrhea, etc.³ The symptoms of mild influenza are often similar to that of common cold, but the fever and systematic symptoms for influenza are severer. Viral pneumonia, secondary bacterial pneumonia, acute respiratory distress syndrome, shock, diffuse intravascular coagulation, manifestations and multiple complications in extra-pulmonary systems, such as cardiovascular system, nervous system, etc., may develop in severe cases.^2,3

The diagnosis and treatment are mainly based on the symptoms of the influenza. Commonly used screening standard (Influenza-like Illness, ILI: temperature of 38°C or higher, with cough or sore throat) is of certain sensitivity but poor specificity.^19,20 The accuracy of the diagnosis of influenza could be improved with a comprehensive analysis of ILI and in combination with other information during the clinical preliminary diagnosis. For example, the probability to be confirmed as influenza is 3 times higher for those patients with 5 characteristics of contact history with ILI patients, cough, cough with sputum in the day of onset, nasal congestion and running nose, anorexia and fever of higher than 37.8°C in the epidemic season of influenza; while the probability is 14 times lower for those without cough or fever higher than 37.8°C in the non-epidemic season of influenza.²¹ In addition, the progress of symptoms after influenza infection has certain features. The symptoms occur usually one day after infection, with systematic symptoms (fever, myalgia, fatigue, and headaches) most obvious in 2–3 d and recover faster than nasal discomfort and respiratory symptoms, which can persist for 8–9 d The symptoms occur in only about 1/4–2/3 infected people, irrelevant to the virus types or subtypes.^18,22 Firstly, the ear discomfort is the most common symptom, 33–73% of patients reported ear pressure anomalies while 33–47% of adults complained of an earache. Secondly, nasal congestion, running nose, sore throat, sneezing, hoarseness, and other upper respiratory tract symptoms occur in 58.8% of patients. About 21% of patients experience cough, dyspnea, chest discomfort, and other lower respiratory symptoms. The symptoms above are also not relevant to the virus types. The fever of higher than 37.8°C is reported by 34.9% of influenza patients, with influenza A virus infection higher than influenza B virus. The symptoms persist longer in elderly and in people with immunodeficiency.^18,23

The pathogen diagnosis of influenza is based on laboratory tests, including virus isolation, detection of viral antigen, nucleic acid, and antibody. Virus isolation is the “gold standard” for laboratory testing, but it needs strict requirements on sampling and transferring of specimen and needs longer time to test. Detection of virus nucleic acid can be used for early diagnosis, and RT-PCR (preferable real-time RT-PCR) is used for detecting influenza virus nucleic acid in the respiratory specimens (swabs, nasal swabs, aspirate from nasopharynx or trachea and sputum). The detection of virus, nucleic acid is of best specificity and sensitivity and can quickly distinguish virus types and sub-types, and the results can be available within 4–6 hours. Virus antigen detection can also be used for early diagnosis, and fluoroimmunoassays can detect antigens in the specimens of respiratory tract, monoclonal antibody can distinguish between type A and B within hours. Other methods include colloidal gold test for which the result can be obtained in 10–30 minutes. A variety of rapid diagnostic tests has been developed in recent y with a specificity of 90–95% and sensitivity of 50–70% compared with virus isolation and RT-PCR. However, their sensitivity in children and type A influenza is higher than that of adults and type B influenza. The interpretation of rapid diagnosis text should be in combination with patients' epidemiological history and clinical symptoms of the patients as there is a high probability of false positive during a non-epidemic period and false negative during an epidemic period. RT-PCR or virus isolation should be used for further confirmation in both situations. Antibody detection can be used for the retrospective investigation but is of little aid in early diagnosis of cases.

Epidemiology

Different types and subtypes of influenza viruses co-circulate globally, and its activity intensity and dominant strains could differ across different seasons, different periods, or different areas. Seasonality varies with geographic and climatic areas, which is relevant to latitude, meteorology, and other factors. The annual epidemic of influenza results in serious health and economic burden. The risk of severe disease and death after influenza is higher in pregnant women, infants, elderly, patients with chronic underlying diseases, and other high-risk groups.

Matching of circulating strains and vaccine strains

The Global Influenza Surveillance Network (GISN) was established by WHO in 1952, and its name was changed to the Global Influenza Surveillance and Response System (GISRS) since 2011. Currently, there are 6 WHO Collaborating Centers (CCs), 4 Essential Regulatory Laboratories (ERLs), and 141 national influenza centers located in 111 WHO member states within GISRS. GISRS monitors the variation of influenza virus through surveillance network around the world and makes recommendations for the laboratory diagnosis and assessment of vaccine strains, antiviral drugs, and risk. Since 2009, the national influenza surveillance network has covered all cities in China, including 554 sentinel hospitals and 408 influenza surveillance network laboratories.

A(H1N1)pdm09, A(H3N2), Yamagata and Victoria lineage of type B have co-circulated in human since 2009. The active intensity and dominant strain vary with areas in the same period.²⁶ WHO recommends viruses for inclusion in influenza vaccines for use in the northern and southern hemisphere influenza seasons in February and September each year, according to the results of the global influenza surveillance.²⁷ WHO will consider changing vaccine strain(s) recommended in the next season if more and more influenza viruses are low reactive to the vaccine strains in a season.²⁸ Serum (antibody) drawn from ferrets immunized using influenza vaccine strain is used in Hemagglutinin Inhibition (HI) assay for isolated and vaccine strain. The isolated strains are named as low reactors if the HI titers are more than 8 times compared to the vaccine strain.^29,30 Mismatch between vaccine strains and epidemic strains is said to occur when most strains in a season are low reactors.³¹ The antigenic drift mainly causes the mismatch for influenza A viruses, and influenza B viruses are primarily caused by 2 reasons: antigenic drift, or lineage in the trivalent vaccine is different from circulating lineage.³¹ For example, the mismatch for type A influenza vaccine strains occurred in season of 2003/2004 and 2007/2008^28,32; and the mismatch for type B influenza vaccine strains occurred in 6 seasons from 2000 to 2011.³¹

Epidemiological characteristics and seasonality of influenza in China

The seasonal epidemics of influenza in winter and spring have been widely studied in temperate zone, but the seasonality and its driving factors in the tropical and subtropical regions have always been the problems bothering international scientific community.^33-35 More and more studies suggest that the seasonality of influenza in tropical regions, particularly in asia, vary substantially with periodic epidemic for half a year or a year and epidemics interspersed in year round.^34-37

A study of the influenza seasonality patterns in different regions of China by Yu et al. suggested that, the annual periodicity of influenza A epidemics increased with latitude.³⁸ The epidemic in winter with single annual peak in January and February each year in the north with latitude ≥atitude and February each year in the north with tics influenza inlatitude <27°N; and double peaks in January or February and between June and August annually in the mid-latitude regions with latitude atitude an<33°N. Epidemics peaked in January–February in Northern China (latitude ≥33°N) and April–June in southernmost regions (latitude <27°N). Provinces at intermediate latitudes (latitude within 27.4°N–31.3°N) experienced dominant semi-annual influenza A periodicity with peaks in January–-February and June–August. In contrast, influenza B activity predominated in colder months throughout most of China. The findings provide evidence for determining the optimal timing for region-based influenza vaccination in China in the future. The study also found that climate and latitude were major factors impacting on the seasonality of influenza. Low temperature is the predictor for influenza in the winter and annual periodic intensity in the northern China while the influenza activity in the spring in the southern part is related to the rainfall.

Disease burden

Health burden and indirect burden

It is estimated that 5–10% of adults and 20–30% of children are suffering from seasonal epidemic of influenza around the world each year,² leading to 30–50 million severe cases and 250,000–500,000 deaths.¹ The influenza-related hospitalizations and deaths are mainly observed in people with high-risk factors, such as pregnant women, infants, and young children, elderly, and patients with chronic underlying conditions. It is also easy to the HCWs to get infected, thus transmitting the virus to the patients with high-risk factors due to their professional characteristics.

All-age group

Studies show that considerable, substantial morbidity, hospitalization, and mortality are caused by seasonal epidemic of influenza in China every year. The population-based hospitalization surveillance for severe acute respiratory infection (SARI) in Jingzhou, Hubei Province, showed that the laboratory-confirmed influenza (LCI) associated SARI hospitalization was 115/100,000 and 142/100,000 in 2010/2011 and 2011/2012, respectively, with severe cases in the children < 5 y old.³⁹ From 1996 to 2000, the excess hospitalization was due to acute respiratory infection (ARI), pneumonia, cerebrovascular disease and ischemic heart disease related with influenza was 60.6 (95% CI: 52.8–67.2)/100,000, 29.3 (95% CI: 25.8–32.6)/100,000, 3.5 (95% CI: 1.4–5.8)/100,000, and 4.2 (95% CI: 1.2–7.0)/100,000, respectively in Hong Kong Special Administrative Region (Hong Kong SAR); the highest excess hospitalization rates were observed in children < 15 y old and elderly ≥ 75 y old.⁴⁰

A study on influenza-related excess mortality from 2003 to 2008 conducted in 8 cities of China estimated that the average annual excess mortality of influenza-associated respiratory and circulatory diseases was 12.4/100,000 and 8.8/100,000 in the northern and the southern cities, respectively, with more than 86% in the elderly ≥ 65 y old.⁴¹ Another influenza mortality burden study using the national representative mortality surveillance data, found that annual epidemics of seasonal infeluzna in China was associtead with 5.0, 11.1 and 13.8 excess respiratory, respiratory and circulatory andall-cause deaths per 100,000 popualtion during 2004–2009; 85% of influenza-associated excess mortality occurred in elderly.⁴² These results of influenza-associated excess mortality in China were similar to those in the developed countries of Europe and America, and analogous to the tropical and subtropical regions.

Pregnant women

Influenza can bring severe harms the pregnant women. The complications of the respiratory system, cardiovascular system, and other systems are more likely to occur in pregnant women infected with influenza virus, because of their immunological and physiological changes. During influenza seasons from 1974 to 1993 in the United States (US), the hospitalization rate for acute cardiopulmonary disease in pregnant women was 3–4 times of that of the postpartum women and not-pregnant women in childbearing age, and the highest rate was observed in the third trimester.⁴³ Another US study found that the hospitalization rate for pregnant women who were suffering from respiratory diseases was higher during influenza epidemic season than non-epidemic seasons from 1998 to 2002 (3.4 v.s. 1.8/1000 hospitalizations).⁴⁴ Compared with pregnant women without underlying diseases, the risk of hospitalization for influenza infection in the pregnant women with a history of asthma increased by 10 times (OR = 10.6).⁴⁵ A study in Canada suggested that pregnant women without underlying diseases were more likely to be hospitalized due to respiratory diseases in epidemic seasons compared with one year before becoming pregnant with an OR of 1.7 (95% CI: 1.0–2.8), 2.1 (95% CI: 1.3–3.3), and 1.3 (95% CI: 3.6–7.3), respectively. If the pregnant women suffer from more than one kind of underlying diseases, the risk of hospitalization was even higher with an OR of 7.9 (95% CI: 5.0–12.5) in the third trimesters.⁴⁶ A study in US during 1998–2005 reported that the influenza-related mortality in the third trimesters of pregnant women was highest with a rate of 3.1/100,000 live births.⁴⁷

Several studies showed that the risk of being admitted to the hospital, ICU, and death from influenza infection in the pregnant women during an epidemic season increased significantly.^48-52 The study on hospitalized cases for A(H1N1)pdm09 in 2009 found that pregnant women accounted for 51% in serious cases (ICU admission or death) and 31% of non-serious cases of hospitalization in women of childbearing age in China, although the pregnant women accounted for only 3% of women of child-bearing age. Among fatal cases of A(H1N1)pdm09 in China, 20% were pregnant women, in which only 7% are with chronic underlying diseases. Compared with non-pregnant healthy women of childbearing age, the risk of developing serious diseases in pregnant women was 3.3 times higher (95% CI: 2.7–4.0), and the risk of serious diseases in the second trimesters (OR = 6.1) and third trimesters (OR = 7.6) was further increased.⁵⁰

Limited studies showed that pregnant women infected with influenza could affect fetal and neonatal, such as stillbirth, neonatal death, preterm delivery, and low birth weight, etc.^53,54 A recent meta-analysis showed that influenza infection in the first trimester significantly increases the risks for a variety of congenital diseases such as multiple congenital malformations (OR = 2.0), neural tube defects (OR = 3.3), congenital heart disease (OR = 1.6), cleft lip (OR = 3.1), etc.⁵³ A US study reported that bipolar disorder in infants increased with influenza infection during pregnancy (OR = 3.8).⁵⁵

Children

Incidence is the highest in young children and decreases slightly with increasing age.⁵⁶ A study of serum antibodies found that the vast majority of children (93%) are infected with influenza virus at least once from birth to 6 y old.⁵⁷ In some highly epidemic seasons, the annual attack rate in children could reach as high as 50%.^58,59 The antibody titer tends to be higher, and the duration of virus is shedding longer in children. Therefore, children are important drivers of influenza transmission, especially to family members and schoolmates.^60,61

Influenza is an important cause of medical visits in children. It is estimated that about 10–15% of children who need to see a doctor is due to influenza infection each year.^62,63 The risk of serious complications after influenza infection in children < 5 y old is higher, with hospitalization rate of influenza-related diseases up to 921/100,000 person-year. In general, the admission rate is highest in infants <2 y old. For example, the admission rate for infants <1 y old due to influenza in the Hong Kong SAR is up to 2785/100,000 person-year.^64-66 The population-based study in Jingzhou suggested that influenza-associated SARI hospitalizations were mainly occurred in children <5 y old, and the hospitalization rate for this age group was up to 2021–2349/100,000 person-year, with the highest rate in 6 to 11 months old (3605–3805/100,000 person-year).³⁹ Another study in Hong Kong SAR showed that influenza hospitalization rate in infants of 2–6 month was highest and reached up to 1762/100,000 person-year.⁶⁷

Influenza can cause large numbers of absence from school in children and absence from work in their parents. A US study found that for every 100 students, during one season, there are 28 infections, 63 d of absence from school, 20 d of absence from work for parents and 22 families with secondary cases.⁶⁸ From 2003 to 2006, there are 662–1046 d of absence from school and 214–336 d of absence from work in parents per 10,000 people each year, caused by influenza A and B in Hong Kong SAR.⁶⁹ In Chinese children aged <5 y, 41% of children infected by influenza need to see a doctor more than twice on average.⁷⁰

In China, a series of studies on etiology of acute respiratory infection showed that the positive rate for influenza in ILI or ARI cases is 3.2% to 25.8% in outpatient and emergent children respectively,^70-75 and 1.4%–35.4% in inpatient children.^76-81

The risk of death in children with underlying diseases is significantly higher than healthy children. However, influenza-associated excess mortality in children is lower than those in the elderly and adults. A Hong Kong SAR study estimated that the A(H1N1) and A(H3N2) associated excess mortality in children of 0–4 y old was 0.91/100,000 and 0.18/100,000 person-year from 1998 to 2009.⁸²

Elderly

Influenza infection is an important cause of death for the elderly. About 2.5–8.1% of deaths in the elderly in the United Kingdom (UK) from 1999 to 2010 were caused by influenza.⁸³ The excess mortality due to influenza in Guangzhou, Hong Kong SAR, and other areas of China showed that influenza-associated excess respiratory and circulatory deaths, and all-cause deaths in the elderly were 64–147/100,000 and 75–186/100,000, respectively,^41,82,84,85 similar to neighboring Singapore,^85,86 and western developed countries such as Portugal⁸⁷ and US⁸⁸ (Table 1).

Table 1.

Comparison of influenza-associated excess mortality in the elderly aged ≥65 y in different countries and areas

			Influenza-associated excess mortality (/100,000)
First author and reference	Country/Region	Study period	All causes	Respiratory and circulatory diseases	Pneumonia and influenza
Yu 42	Whole China	2004–2009	145.5	117.8	—
Feng 41	3 Northern cities, China	2003–2008	150.8	106.0	3.1
Feng 41	5 Southern cities, China	2003–2008	75.4	64.3	3.6
Wang 84	Guangzhou, China	2010–2012	185.6	146.9	30.4
Yang 85	Guangzhou, China	2004–2006	111.3	104.1	—
Yang 85	Hong Kong SAR	2004–2006	103.7	78.7	—
Wu 82	Hong Kong SAR	1998–2009	89.7	—	—
Chow 86	Singapore	1996–2003	167.8	155.4	46.9
Yang 85	Singapore	2004–2006	80.5	46.0	—
Nunes 87	Portugal	1980–2004	155.8	—	16.9
Thompson 88	United States	1976–1999	132.5	98.3	22.1

The studies in both mainland China and Hong Kong SAR have shown that influenza-associated excess mortality in the elderly ≥65 y is much higher than group aged 0–64 y, with 84–95% of excess deaths in the elderly.^{41,42,82,84,85} Similarly, about 90% of influenza-associated excess mortality in US from 1976 to 2007 was in elderly aged ≥65 y.⁸⁹ The death risk increases with older age. Influenza-assoicated mortality was 11.3 times higher in elderly than all-age group in Singapore; and influenza-associated all-cause mortality was 16 times greater in elderly ≥85 y than in 65–69 y old in US.⁸⁸

Influenza can also lead to higher burden of hospitalization in the elderly. The population-based study in Jingzhou found that the admission rate of SARI confirmed as influenza was 89–141/100,000 in the elderly ≥65 y.³⁹ A study in Hong Kong SAR found that influenza-associated excess hospitalization rates per 100,000 population in elderly of 65–74 y and ≥75 y during 1996–2000 were: 84 and 266 for ARI, 59 and 176 for pneumonia and influenza, 24 and 54 for diabetes, 16 and 55 for cerebrovascular diseases, 10 and 56 for ischemic heart disease.⁴⁰ A US study showed that, in the patients with a diagnosis of influenza and pneumonia, influenza-associated excess hospitalization rate in elderly ≥65 y with chronic diseases was 476–636/100,000, which was 6–8 times of young people with chronic diseases; the excess hospitalization rate was 150–172/100,000 for those elderly without chronic disease, which was 6 to 7 times of the healthy young people.⁹⁰

In addition, the investigations of influenza outbreaks in nursing homes and other institutions in Belgium, Australia, France, and other countries, suggest that the outbreaks was more likely to occur in collective institutions, where the elderly were all under the risk of influenza infection throughout the year.^91-95

Persons with chronic underlying diseases

Compared with age-matched healthy adults, infelunza patients with chronic underlying diseases are more likely to suffer from serious diseases or die, and influenza-associated hospitalization and excess mortality are significantly higher than those without underlying disease. The high-risk underlying diseases include: respiratory disease (asthma, chronic bronchitis and emphysema, and other lung diseases), heart disease (atherosclerotic heart disease, cardiomyopathy, chronic congestive heart failure, and congenital heart disease), neurodevelopmental disorders (cerebral palsy, muscle malnutrition, and cognitive impairment), metabolic disease (diabetes), immune dysfunction (HIV/AIDS, chemotherapy, organ transplant patients using immunosuppressants, and long-term use of corticosteroids), chronic renal deficiency on dialysis treatment, chronic liver disease (liver cirrhosis), morbid obesity, blood system diseases (sickle cell anemia and Mediterranean anemia) and adolescents with long-term use of aspirin (risk of Reye's syndrome).

From 1996 to 2000, for the elderly aged ≥65 y in US, influenza-associated hospitalization rates for patients with or without chronic underlying diseases were 560/100,000 and 190/100,000 respectively.⁹⁶ From 2001 to 2007, influenza-associated hospitalization rate in British patients with chronic diseases was 71/100,000, 2.6 times of healthy control group; for influenza-associated deaths each year, the patients with chronic diseases accounted for 72%.⁹⁷ A study in South Korea showed that the risk of serious complications for patients with diabetes after influenza infection is 3.63 times of those without diabetes (95% CI: 1.15–11.51).⁹⁸ From 1999 to 2005, influenza-related hospitalization rate for the cancer patients in Taiwan was 70.1/100,000, which was significantly higher than that of the non-cancer groups (44.4/100,000, P < 0.0001).⁹⁹

Healthcare workers (HCW)

The risk of exposure to influenza virus is higher in HCW than general people due to their increased opportunities of contacting with the patients infected with influenza in routine clinical practices. A meta-analysis of 29 studies from 1957 to 2009 around the world showed that the LCI morbidity in HCW without influenza vaccination was 18.7% (95% CI: 15.8–22.1%) each season, 3.4 (95% CI: 1.2–5.7) times of healthy adults.¹⁰⁰ The risk of infection with influenza virus was even higher when influenza outbreak occurs in the hospital, with an attack rate of as high as 11–59% in HCW who are caring the influenza patients.¹⁰¹ The risk of influenza infection differs in HCW at different department, for which the HCW in clinical departments was obviously higher than HCW from the other departments,¹⁰² and the attack rate in nurses was higher than that of other HCW.¹⁰³

The burden of influenza has been investigated among HCW in China. A prospective randomized clinical trial suggested that the LCI occurred in 1.3%, 1.0%, and 1.3% in the HCW who are not wearing masks, wearing surgical masks, and N95 respirator, respectively.¹⁰⁴ A retrospective survey from October 2004 to May 2005 in Taiwan showed that the attack rate of ILI in the HCW without influenza vaccination was 15.0%.¹⁰⁵ In addition, influenza could also cause loss of the labor force in HCW; a retrospective cohort study in Hong Kong SAR showed that the average absent days from work due to ILI in HCW without influenza vaccination was 1.75 d.¹⁰⁶

Influenza infection in HCW increases the risk of nosocomial infection and community transmission. A study reported that 35% of HCW infected with influenza virus were asymptomatic,¹⁰⁷ more than 75% of the HCW continued to work after the influenza-like symptoms.^101,108 Therefore, the infection risk of HCW,¹⁰⁹ outpatients, inpatients and their families¹⁰¹ will be increased if effective protective measures are not taken.

Economic burden due to influenza

Studies on the economic burden by influenza in China are mostly conducted in the economically developed areas, and only one unpublished study covered 29 provinces across mainland China (except provinces of Tibet and Liaoning).^{63,69,70,110–112} Studies suggested that the direct medical cost of influenza outpatients was 156–398 RMB/person, and indirect cost was 198–244 RMB/person.^63,69,70,110 The economic burden of hospitalized influenza cases was obviously higher than that of outpatient cases. The survey in 29 provinces showed that the direct medical cost was 4783 RMB/person (IQR: 2949–7286), direct non-medical cost was 1307 RMB/person (IQR: 706–2193), and indirect cost was 626 RMB/person (IQR: 322–1053). The direct medical cost of hospitalized influenza patients aged ≥60 y in Shanghai was 10934 ± 12409 RMB/person, and the medical cost of patients with chronic underlying diseases (15072 ± 16654 RMB/person) was significantly higher than those without underlying diseases (7624 – 5997 RMB/person, P < 0.05).¹¹⁰ The direct medical cost of hospitalized influenza patients aged <5 y in Suzhou was 3931 RMB/person (IQR: 3024–5216); the direct medical cost of patients admitted to ICU (median of 8335 RMB/person, IQR: 7075–9986) was higher than the cases treated in the common ward (median of 3875 RMB/person, IQR: 3018–5078, P < 0.05).¹¹¹ A survey on the medical cost in hospitalized influenza cases in 3 cities (Changsha, Chengdu and Jinan) suggested that the median direct medical cost per person in children aged <15, patients aged 16–64 and elderly >65 was 1541, 5696 and 15094 RMB, respectively, and the direct medical cost in high-risk groups was 6 times of low-risk groups.¹¹²

It is estimated that the numbers of influenza outpatient in Zhuhai was 5568 in 2008 and 26275 in 2009, and the direct economic burden was 1,370,000 and 6,160,000 RMB, respectively.⁶³ From 2003 to 2005, the direct and indirect economic burden of hospitalization in children <18 y old with influenza in Hong Kong SAR was 2,880,000 and 3,650,000 USD.⁶⁹

Influenza Vaccines

History and current situation of influenza vaccine

Evolution of components of influenza vaccine

The research and development of influenza vaccine are evolved with the mature in technology of isolation and culture of influenza virus as well as with the changes in epidemic strains.¹¹³ Influenza A virus was first isolated from ferrets in 1933 and virus B was in 1940. In 1945, the bivalent whole-virus inactivated influenza vaccine containing virus A(H1N1) and B began to be widely used in US. In 1958, the new bivalent influenza vaccine containing A(H2N2) and B were developed with A(H2N2) replacing A(H1N1). In 1968, A(H3N2) virus led to the pandemic and replaced A(H2N2) in seasonal influenza, and then in 1970, the strains included in bivalent influenza vaccine was replaced by type A(H3N2) and B. In 1978, A(H1N1) virus appeared again and co-circulated with A(H3N2) and virus B. Subsequently, trivalent influenza vaccine [A(H1N1) + A(H3N2) + B] was developed in order to improve the protection against the 2 types A virus co-circulating at the same time. After the emergency of Victoria and Yamagata lineage of influenza B in 1987 and 1988, 2 lineages were co-circulated, but influenza vaccines was still trivalent, thus influenza B strain in vaccine was chosen between B(Victoria) and B(Yamagata) strains according to the results of global surveillance of influenza. Since 2012, trivalent vaccine strain of [A(H1N1)pdm09 + A(H3N2) + B(Victoria) or B(Yamagata)] was recommended by WHO, while another type B strain was also recommended to be included in quadrivalent influenza vaccine.

Research, development and production history of influenza vaccine

In 1936, the success of chicken embryo culture of influenza virus made the mass production of human vaccine possible. The whole virus inactivated influenza vaccine made from chicken embryo was approved in US for the first time in 1941.¹¹⁴ In 1968, the split influenza vaccine was developed, which was prepared by pyrolysis thus became the routinely used due to its good immunogenicity and great reduction of the reactogenicity.^115–119 The subunit vaccine was produced by removing the internal antigen (i.e., nuclear capsid and matrix proteins) using additional purification steps based on the split vaccine.

In order to increase the immunogenicity of the subunit vaccine, manufacturers also developed influenza vaccine with adjuvants. Adjuvant is the material that could produce stronger immune response than the antigen alone when being inoculated with antigen. Therefore, adjuvants could save the antigen amounts and then increase the production of vaccine, which brought a potential benefit in the influenza pandemic. Currently, commercialized adjuvant containing influenza vaccines include MF59 oil-in-water emulsion vaccines, virus particles (virosome) vaccine, and monophosphoryl lipid (MPL) adjuvant pandemic influenza vaccine. However, higher standards of safety of adjuvants were proposed since the possible relevance between adjuvant A(H1N1)pdm09 vaccine and narcolepsy was reported in Finland and Sweden.

To further improve the immunogenicity of influenza vaccine in high-risk groups, intradermal vaccine, high dose vaccine, nasal spray live attenuated vaccine, etc. have been developed and marketed in recent years.

Source of influenza vaccine strains and production of vaccine

As early as 1947, 2 y after the commercialization of influenza vaccine, it was realized that the change of antigenicity of influenza virus HA (i.e., antigenic drift) would significantly reduce vaccine effectiveness,^120,121 the components of the vaccine were therefore needed to be updated annually to effectively protect against the new strains in the northern and southern hemispheres.^122,123 Influenza vaccine strains derive from the isolated strains in WHO GISRS; the original isolates became reference strains after passage and then were distributed to the manufacturers for preparation of seed batch. Usually, the growth of original wild strains in chicken embryo was poor, it was required to prepare a number of alternatives strains with similar antigenicity, and to analyze their growth characteristics and optimal culture conditions (e.g., time and temperature) for the preparation of high yields of reassortment strains.

The production of influenza vaccine by chicken embryo culture began to be used in US since 1945, which has been regarded as the most mature and safe production method recognized in the world after decades of production experiences. The traditional production procedures were roughly divided into 6 steps, including culture, harvesting, purification, disruption, inactivation, and final filtration of virus; then monovalent influenza vaccine is obtained, and these steps could be different with different technology used by manufacturers. National laws and regulations have all specific requirements on the chicken embryos eggs used in the production of influenza vaccine, e.g., the chicken embryo for passages of virus seed and preparation should come from specific pathogen free (SPF) chickens and the chicken embryo for vaccine production should come from healthy chicken raised in enclosed houses; the live chicken embryo of 9–11 days, without deformation and with clear blood vessels should be chosen. The antibiotics added into the inactivated influenza vaccine are not active ingredients and are reduced to traceable or undetectable levels in the final products. Most single dose of influenza vaccine no longer contains thimerosal, which was called as preservative-free vaccine.

In recent years, a lot of studies have been done to use of mammalian cells such as Madin-Darby Canine Kidney (MDCK cells) or African green monkey kidney cell (Vero cell) as medium for culturing inactivated influenza virus. The antigenic variability of influenza virus using cell culture was superior to using chicken embryo culture; in addition, shorter time is needed in cell culture is a unique advantage during influenza pandemics.

Influenza vaccine marketed internationally

Currently marketed influenza vaccines internationally are trivalent or quadrivalent Influenza Inactivated Vaccine (IIV) and Live Attenuated Influenza Vaccine (LAIV). The components of trivalent influenza vaccine include A(H3N2), A(H1N1), and one lineage of type B virus, while the components of quadrivalent vaccine include A(H3N2), A(H1N1), B(Victoria) and B(Yamagata).² There are 3 kinds of vaccines for the IIV, i.e., whole-virus vaccine, split virus vaccine, and subunit vaccine.² Standard dosage is 0.5 ml/dose by intramuscular or subcutaneous route (for 3 y old or above) with 15 μg of HA for each component and 0.25 ml/dose (for infants and young children aged 6 to 35 months) containing 7.5 μg of HA for each component. Currently, there is no inactivated influenza vaccine approved for use in infants less than 6 months of age. In most countries, the whole-virus vaccine has been substituted by safer split virus vaccine and subunit vaccine.²

In recent years, the influenza inactivated vaccine by intradermal route has been developed and launched for adults aged 18 to 59 y containing 9 μg of HA for each component, with the immunogenicity as intramuscular inoculation of 15 μg HA; and 15 μg of HA for elderly HAity as intramu with a better immunogenicity better than intramuscular inoculation of 15 μg HA, similar to adjuvant vaccine.^124-127 In 2010, high dose of IIV containing 60 μg of HA for each component was marketed in US for the elderly with a significantly higher effectiveness than the standard doses.^128-130 Currently, Europe has approved 3 kinds of subunit vaccines with adjuvants, 2 of which are oil-in-water emulsion MF59 and AS03 of MPL series respectively, and the third one is liposome-like particle adjuvants.

LAIV is developed by cold adapting the attenuated live virus that couldn't cause influenza and administered via nasal spray. In 2012, a quadrivalent LAIV was approved in US.² Attenuated live vaccines could induce mucous immune responses, simulating natural infection. Rapid attenuation of the mutated virus is crucial to attenuated vaccines production. Cold adaption technology is reliably and effectively used for influenza vaccine attenuation. Viruses are cultured at gradually changing temperature from 36°C to 25°C; to obtain cold-adapted strains [Cold-adapted prototype of influenza virus A/Ann Arbor/6/60(H2N2 and B/Ann Arbor/1/66], which are characterized by normal replication at the temperature between 25–33°C with the same titers as cultured in a normal temperature incubator. However, replication is limited at 38–39°C. Cold adapted strains are not capable of replicate effectively in the body at 37°C, thus is not pathogenic to human, although its limited replication could still induce an immune response. Table 2 provides details of influenza vaccines marketed in US.

Table 2.

Influenza vaccines marketed in the United States (2014–2015)

Product name	Manufacturer	Specification	Indications	Route
Standard dose of quadrivalent inactivated influenza vaccines.
Fluarix quadrivalent	GlaxoSmithKline	(Single dose) 0.5 ml/syringe	≥3 y	intramuscular
FluLava quadrivalent	ID company (GlaxoSmithKline)	(Single dose) 0.5 ml/syringe	≥3 y	intramuscular
		(Multiple doses) 5.0 ml/vial	≥3 y	intramuscular
Fluzone quadrivalent	Sanofi Pasteur	(Single dose) 0.25 ml/syringe	6–35 months	Intramuscular
		(Single dose) 0.5 ml/syringe	≥3 y	Intramuscular
		(single dose) 5.0 ml/vial	≥3 y	Intramuscular
		(Multiple doses) 5.0 ml/vial	≥5 months	Intramuscular
Standard dose of trivalent inactivated influenza vaccines
Afluria	CSL company	(Single dose) 0.5 ml/syringe (Needle-Free)	≥9 y	Intramuscular
		(Multiple doses) 5.0 ml/vial	≥9 y	Intramuscular
Fluarix	GlaxoSmithKline	(Single dose) 0.5 ml/syringe	≥3 y	Intramuscular
FluLaval	ID company (GlaxoSmithKline)	(Single dose) 0.5 ml/syringe	≥3 y	Intramuscular
		(Multiple doses) 5.0 ml/vial	≥3 y	Intramuscular
Fluvirin	Novartis	0.5 ml/syringe	≥4 y	Intramuscular
		5 ml/vial	≥4 y	Intramuscular
Fluzone	Sanofi Pasteur	(Single dose) 0.5 ml/syringe	≥3 y	Intramuscular
		(Multiple doses) 5.0 ml/vial	≥l months	Intramuscular
Fluzone Intradermal	Sanofi Pasteur	0.1 ml/Intradermal Needle	18 months –64 y	intracutaneous
Standard dose of trivalent inactivated influenza vaccines cultured from cells
Flucelvax	Novartis	0.5 ml/syringe	≥l8 y	Intramuscular
High doses of trivalent inactivated influenza vaccines
Fluzone High-dose	Sanofi Pasteur	0.5 ml/syringe	≥65 y	Intramuscular
Recombinant trivalent influenza vaccines
FluBlok	Protein Sciences	0.5 ml/vial	18–49 y	Intramuscular
Quadrivalent attenuated influenza vaccines
FluMist quadrivalent	AstraZeneca branch, Medimmune	0.2 ml/nasal spray	2–49 y	Nasal spray

Data source: Official websites of United States Food and Drug Administration.

Vaccines in the domestic market

Currently, marketed influenza vaccines in China are all trivalent inactivated influenza vaccine (TIV). There are 16 manufacturers providing seasonal influenza vaccines in China, which are mainly split vaccines and subunit vaccines. Table 3 presents details of manufacturers and product information.

Table 3.

Types and specifications of influenza vaccines marketed in China (by September 2014)

Manufacturers	Vaccine Type	Specification
Tasly skinner biological technology (Tianjin) co., LTD	subunit	0.5 ml
Lanzhou institute of biological products co. LTD	split	0.5 ml, 0.25 ml
	wholevirus	1.0 ml
Beijing Tiantan Biological products co., LTD	wholevirus	1.0 ml, 0.5 ml
	split	0.5 ml
Changchun Changsheng biotechnology co., LTD	wholevirus	0.5 ml
Jiangsu Xiansheng WeiKe biological pharmaceutical co., LTD	split	0.5 ml, 0.25 ml
Zhejiang Tianyuan Biological pharmaceutical co., LTD	split	0.5 ml, 0.25 ml
Dalian Hissen biological pharmaceutical co., LTD	split	0.5 ml
Shanghai institute of biological products co., LTD	split	0.5 ml, 0.25 ml
Changchun institute of biological products co., LTD	split	0.5 ml, 0.25 ml
Hualan biological product co., LTD	split	0.5 ml, 0.25 ml
Dalian Aleph biological pharmaceutical co., LTD	split	0.5 ml, 0.25 ml
Beijing Sinovac biotech co., LTD	split	0.5 ml, 0.25 ml
Abbott Trading (Shanghai) co., LTD	subunit	0.5 ml
Shenzhen Sanofi Pasteur biological products co., LTD	split	0.5 ml, 0.25 ml
Glaxosmithkline (China) investment co., LTD	split	0.5 ml, 0.25 ml
The Swiss Crucell co., LTD	subunit (virus particle)	0.5 ml

Data source: State Food and Drug Administration of China.

Immune responses and immunity duration after TIV vaccination

Influenza vaccination based on the US Food and Drug Administration and the European Medicines Agency should induce^131,132: (1) Hemagglutination inhibition (HI) antibody ≥1:40; (2) seroconversion, which means HI antibody increase from <1:10 before vaccination, to ≥1:40 after vaccination or geometric mean titers (GMT) of HI antibody increase by over four times compared with ≥1:40e before vaccination.

Influenza vaccines produce both humoral and cellular immune responses. For the humoral immune response, influenza vaccination mainly induces antibodies against major surface glycoprotein such as HA and NA. The level of serum antibodies induced by inactivated vaccines in human is associated with age and the antibody level before vaccination. In peripheral blood, the amount of cells producing influenza-specific antibodies peak one week after vaccination, while serum antibody level in a healthy population with a history of influenza infection or vaccination peak 2–4 weeks after vaccination. However, in the elderly or population who never have been exposed to influenza, it might take 4 weeks or longer to achieve peak levels.^133,134 CD4 + T and CD8 + T lymphocytes also play important roles in influenza immunity. Compared with specific antibody responses, cellular immune cells could identify more conserved sites on virus surface / internal units as well as demonstrate better cross-reactivity for different virus subtypes.¹³⁵

Inactivated vaccines could induce rapid systematic and local immune responses in healthy youngsters.¹³⁴ Two weeks following vaccination, 90% of individuals produce HI antibodies of 1:40 or higher while the second injection induce little increase in antibody titer. Serum antibody peak occurs at 4–6 weeks after vaccination. Immunity gained by influenza virus infection or influenza vaccination decays gradually with time, the degree of decaying is associated with factors such as age, physical condition, and vaccine antigenicity. Clinical trials have documented that the protective effects of inactivated influenza vaccines against antigen similar strains could last for 6–8 months.¹¹⁶ After one year of vaccination, serum antibody level is significantly decreased; however, protective effects against a few strains might last longer. To match with the consistently mutated influenza virus, WHO recommendations update one or more vaccine strains in most seasons, although it is possible that the 3 vaccine strains are completely the same as the previous season. To guarantee the maximum protection in vaccinated population, most vaccinees are recommended to receive vaccines prior to flu season of the year even if the influenza vaccine components are exactly the same as the previous season, given that antibody titer in most vaccinated population had significantly reduced.

Immunogenicity, efficacy, and effectiveness of TIV

Vaccine efficacy measures how well a vaccine works in controlled clinical trials, whereas effectiveness relates to how well it works when used in routine immunization programmes. Efficacy and effectiveness of influenza vaccines can be affected by multiple factors, including vaccinee's age and immunity, matching the degree of vaccine strains and circulating strains, study design, measured indicators, and so on. There are many endpoints to evaluate efficacy and effectiveness of influenza vaccine, including LCI, medically attended acute respiratory illness (MAARI) or ILI, pneumonia and influenza associated hospitalization or death, seroconversion against circulating strains, etc. Efficacy or effectiveness using specific indicators such as LCI yield higher estimates than using non-specific indicators such as MAARI.¹³⁶ Observational studies are more likely to be influenced by bias when comparing with nonspecific outcomes compared with LCI. In addition, assessment of vaccine efficacy against LCI might be influenced by the sensitivity of the diagnostic methods. A simulation study conducted in 2012 showed that with every 1% decrease in diagnostic sensitivity of influenza infection, vaccine effectiveness would decrease approximately 4%.¹³⁷

Healthy adults

Inactivated influenza vaccines demonstrate good immunogenicity in healthy adults. Meta-analyses of randomized controlled trials (RCT) in Chinese showed that for split vaccines, no significant difference in seroconversion rates was observed between domestic vaccines and imported vaccines.^138,139 A systematic review of RCTs on influenza vaccination conducted in 2012 estimated that TIV was 59% effective against confirmed influenza among healthy adults aged 18–65 y (95% CI: 51–67%).¹⁴⁰ Another systematic review of RCTs on influenza vaccination among healthy adults conducted in 2014 had updated studies included to May 2013, showing that TIV vaccination among healthy adults aged 16–65 y was 60% effective against confirmed influenza (95% CI: 53–66%) and 16% effective against ILI (95% CI: 5–25%). When vaccine strains and circulating strains were matched, TIV vaccination could reduce 42% (95% CI: 9–63%) of ILI clinical visits.¹⁴¹

In China, there is a lack of RCT studies on influenza vaccine efficacy and effectiveness; instead, there are large numbers of cohort studies. A systematic review of domestic literature published between March 1998 and May 2008¹⁴² included 2 RCTs and 11 cohort studies, among which one RCT found that vaccine efficacy against ILI among adults aged 20–50 y was 73%,¹⁴³ and meta-analysis of cohort studies estimated that vaccine effectiveness against ILI among adults was 30% (95% CI: 17–41%). Because observational studies are subject to biases, meta-analysis of RCT with LCI as outcome can be used as the best evidence to show vaccine efficacy.

Pregnant women

Influenza vaccination during pregnancy can protect not only the pregnant women, but also their newborns from influenza through maternal antibodies¹⁴⁴. A randomized controlled clinical trial in Bangladesh between 2004 and 2005 enrolled 340 pregnant women in the third trimester, of whom 172 pregnant women received TIV, 168 received 23-valent pneumococcal polysaccharide vaccines as control. Follow-up was conducted during the whole pregnancy and 6 months after delivery. Results showed that compared with the control group, TIV vaccination could reduce 36% of febrile respiratory diseases in pregnant women and 29% of febrile respiratory diseases in infants, TIV vaccinated group had a 63% reduction in LCI infection within 6 months after birth.¹⁴⁵ A randomized controlled clinical trial among 194 pregnant women with HIV infection and 2,116 pregnant women without HIV infection in South Africa from 2011 to 2012 found that, influenza vaccination showed good immunogenicity in both HIV group and non-HIV group; both groups of pregnant women and their babies showed higher seroconversion rates than the placebo group one month after TIV vaccination. TIV vaccination provided certain degree of protection against influenza infection for both groups of pregnant women as well as babies without HIV exposure, with vaccine efficacy of 50.4% (95% CI: 14.5–71.2%) and 48.8% (95% CI: 11.6–70.4%) for pregnant women and babies without HIV infection, respectively; while vaccine efficacy in pregnant women with HIV infection was 57.7% (95% CI: 0.2–82.1%).¹⁴⁶ A prospective observational cohort study in US suggested that influenza vaccination during pregnancy reduced LCI infection and the risk of ILI hospitalization in infants by 41% and 39% respectively. Compared with unvaccinated group, infants of 2–3 months old in vaccinated pregnant women group had significantly higher titer of influenza antibody.¹⁴⁷ However, a retrospective study in US between 1997 and 2002 showed no significant decrease of ILI incidences among vaccinated pregnant women and their babies,¹⁴⁸ and a retrospective cohort study from 1995 to 2001 also found that no significant reduction in clinical visits due to the respiratory system disease in infants by vaccinated pregnant women.¹⁴⁹

Children

Children over 6 months will get protection against influenza virus infection when receiving influenza vaccination according to the recommended immunization schedule. Children under the age of 8 y old when vaccinated for the first time have better protective effects after receiving 2 doses than one dose. For example, antibody titer against A(H1N1), A(H3N2), and influenza B virus among children aged 5–8 y old were significantly higher after the second dose than the first dose.¹⁵⁰

The efficacy and effectiveness of influenza vaccination in children vary with season, influenced by study design, influenza activity, etc. Studies with confirmed influenza based on virus culture or RT-PCR were scarce. A controlled clinical trial with meningococcal or Japanese Encephalitis vaccine as control in Europe between 2007–2008 and 2008–2009 found that non-adjuvanted influenza vaccine and MF59-adjuvanted influenza vaccine demonstrated an efficacy of 43% (95% CI: 15–61%) and 86% (95% CI: 74–93%) against confirmed influenza among 6–71 month-old children respectively.¹⁵¹

A placebo controlled clinical trial found that influenza vaccination was 56% effective against confirmed influenza among children aged 3–9 y and 100% effective among those aged 10–18 y¹⁵² A randomized controlled trial conducted between 1985 and 1990 suggested that influenza vaccination among children aged 1–15 y could reduce 91% of A(H1N1) and 77% of A(H3N2) cases.¹⁵³ A randomized double-blinded placebo-controlled study in US included 786 children aged 6–24 months and estimated vaccine efficacy during 1999–2000 was 66% (95% CI: 34–82%) against virus culture-confirmed influenza, while during low activity season of 2000–2001 no significant decrease in influenza incidence were observed.¹⁵⁴ In studies with serology confirmed influenza as outcomes, adults might show “antibody ceiling” effects due to previous natural infection or passive immunity, which means they are less likely to have 4 times increase in antibody titer relative to that of the unvaccinated, resulting in overestimation of vaccine efficacy. However, it remains unclear whether such effects exist in children.

A 2012 meta-analysis of influenza vaccine effectiveness showed that in 6 studies covering 8 seasons, influenza vaccination showed moderately protective effects among children aged 6–59 months in 3 seasons.¹⁴⁰ A Chinese case-control study on influenza vaccine effectiveness between 2008 and 2013 reported that influenza vaccination showed moderately effectiveness against RT-PCR confirmed influenza among children aged 6–59 months. Vaccine effectiveness during 2010–2011 and 2011–2012 was 73.2% and 52.9% respectively (Table 4).^155,156 In 2012–2013, effectiveness against A(H1N1)pdm09 subtype among children aged 8–83 months was 67% (95% CI: 58–74%).

Table 4.

Influenza vaccine effectiveness in children in southern China, 2010–2012

	Vaccine effectiveness (%, 95% CI)
Age (months)/doses	2010–2011	2011–2012
6–35 months
2 doses	74.4 (44.6, 88.2)	68.8 (56.0, 77.9)
1 dose	69.3 (30.1, 86.5)	21.8 (−6.1, 42.4)
≥1 dose	72.3 (48.5, 85.1)	49.5 (35.3, 60.6)
36–59 months
≥1 dose	82.6 (−6.0, 97.1)	58.2 (38.7, 71.4)
6–59 months
≥1 dose	73.2 (52.2, 85.0)	52.9 (42.1, 61.7)
Virus type
Influenza A	73.5 (45.5, 87.1)	46.5 (24.4, 62.2)
Influenza B	98.6 (−2724.1, 100)	66.8 (46.3, 79.5)

Effectiveness of influenza vaccination among older children might be superior to younger children.^151,157 Meta-analysis in 2012 showed that vaccine effectiveness among children aged 6–23 y were 40% (95% CI:6–61%) and 60% (95% CI:30–78%) among 24–59 months.¹⁴⁰ A Chinese study also showed in 2011–2012, influenza vaccination was 58.2% and 49.5% effective for children aged 36–59 months and 6–35 months respectively (Table 4).¹⁵⁵

Two doses of vaccine among children aged 6–35 months showed better effectiveness than one dose. Vaccine effectiveness of 2 doses was 61.0% (95% CI: 44.1–72.8%) and 73.4% (95% CI: 54.7–84.3%) in 2008–2009 and 2009–2010 respectively, however, no significant effectiveness was found for one dose.¹⁵⁶ Similar results were reported in 2010–2012 (Table 4).¹⁵⁵ Therefore, to obtain maximum protection, younger children are recommended to receive 2 doses of vaccines.

Influenza vaccine effectiveness in children could partly decay over time. No interference of previous influenza vaccinations with current vaccination was observed. Vaccine effectiveness in 2010–2012 (same vaccine components) decreased from 68.9% (95% CI: 57.5–77.2%) at 1–3 months post-vaccination to 48.4% (95% CI:33.8–59.7%) at 4–6 months post-vaccination. Vaccination in both seasons showed similar effectiveness to vaccination only in the present season (55.9% and 56.8% respectively).¹⁵⁵

Children with chronic underlying diseases may have lower immunogenicity relative to that of healthy children,¹⁵⁸ however, another study suggested that children with asthma had similar immunogenicity to healthy children, even when corticosteroids therapy is required.¹⁵⁹

Students

Influenza is an important cause of school absence. Between November 2000 and June 2001, a study in China among 400 primary school students aged 7–12 y found that ILI incidence in vaccination group was 9.5% (19/200) during the 7 months observation period, significantly lower than that of control group (25.5%, 51/200).¹⁶⁰

Influenza vaccination among students could directly increase effectiveness and herd immunity, which could block the spreading of influenza among families and communities, contributing to indirect protective effects for the community population. A study of PCR confirmed influenza was conducted in 4,455 subjects in 4 schools with influenza vaccination organized and 4 schools that did not vaccinate during 2010–2011 in US, with PCR confirmed influenza as an outcome. The study found that the vaccination rates in vaccination school and controlled school were 32.7% and 2.7% respectively, with a vaccine efficacy of 67.0% (95% CI: 45.4–80.1%). Compared with the controlled school, the influenza incidence in vaccination school decreased by 30.8% (95% CI: 10.1–46.8%) and unvaccinated students experienced significantly higher school absenteeism than vaccinated students (4.3 v.s. 2.8 days/100 school days). Meanwhile, compared with controlled school, an indirect protective effect was observed among unvaccinated children when vaccine coverage achieved 50% in the vaccination school.¹⁶¹ Another study included 28 middle and primary schools in US showed that during influenza peak periods, ILI incidence (17%) in families with children with vaccination intervention school (coverage rate: 47%) was significantly lower than that of controlled school (26%). In addition, ILI incidence (8%) in adult family members in vaccination intervention school was significantly decreased compared with that of a controlled group (13%).¹⁶² A comparative study of 2 primary schools (coverage rate: 52%) and 2 neighboring schools (coverage rate: 28%, controlled group) showed that when absenteeism ≥1 day due to fever or cough was set as an indicator, school with high vaccination rates had a significantly lower absenteeism rate (26.5%) than that of controlled (38.9%).¹⁶³

The elderly

The elderly experience a declining immunogenicity induced by vaccination, which might be attributed to the immunosenescence.¹⁶⁴ A review of 31 studies on HAI antibody response among population aged might antly lower absenteshowed that during influenzawith seroconversion rates of 60%, 62%, and 58% of A(H1N1), A(H3N2), and B-type among young adults, the seroconversion rates in elderly was only 42%, 51%, and 35%. When serum protection (HAI antibody titer ≥1:40) was selected as an indicator, protection rates against three subtypes in the elderly were 69%, 74%, and 67% respectively, compared with 83%, 84%, and 78% in young adults.¹⁶⁵ Although HAI antibody titer y were 69%, 74 adult correlates with about 50% of clinical protection, it may be different in the elderly.

A randomized controlled study among community residents aged ≥60 y in US showed that during seasons with vaccine strains matching with circulating strains, influenza vaccination showed a protective efficacy of 58% (95% CI: 26–77%) against serology confirmed influenza, however, this study might overestimate vaccine efficacy due to “anti-ceiling” effects.¹⁶⁶ Vaccination in nursing homes showed a 20–40% of effectiveness against acute respiratory diseases among the elderly.^167,168 Nonetheless, influenza outbreak among the elderly in nursing homes showed that in the situation of antigen drift, vaccination showed no obvious effectiveness.^169,170

Influenza vaccination among the elderly aged ≥65 y could reduce influenza-related complications, hospitalization, and deaths. A meta-analysis of 20 cohort studies in 1995 found that influenza vaccination among the elderly could prevent 53% (95% CI:35–66%) of pneumonia, 50% (95% CI:28–65%) of hospitalization, and 68% (95% CI:56–76%) of deaths.¹⁷¹ A meta-analysis of 18 cohort studies among community-based elderly population in 2007 showed that influenza vaccination could reduce 27% of pneumonia hospitalization and 48% of deaths.¹⁷² However, these studies evaluated non-specific clinical outcomes instead of LCI, which might compromise the conclusion.

Persons with chronic diseases

HIV-infected subjects with low CD4+ cell counts showed poor immunogenicity after TIV and might had decreased vaccine effectiveness,^173-175 and a second dose of vaccine failed to increase immune responses.¹⁷⁴ Moderate immune responses were induced by one dose of TIV (with or without adjuvants) in adults or children with cancer or receiving organ transplants,^176,177 and a study found that a second dose significantly increased serum antibody protective level.¹⁷⁶ A study showed that HIV-infected adults and children might benefit from the influenza vaccine containing MF59 adjuvants.¹⁷⁸

Influenza vaccine effectiveness among different populations was described in 2011 systematic review,¹⁷⁹ demonstrating that influenza vaccination was effective in patients with certain chronic diseases.^180–190 There was the lack of high-quality evidence with regards to crucial outcomes such as pneumonia, hospitalization, and deaths. A 2006 systematic review¹⁸⁶ conducted meta-analysis for 2 RCTs^191,192 in COPD patients that showed an 81% (95% CI: 52–93%) efficacy against influenza, with an average reduction of 0.37 (95% CI: 0.11–0.64) COPD exacerbation. Chinese cohort studies also showed that TIV vaccination could reduce COPD- and chronic bronchitis- related acute infection and hospitalization.^193–196 A cohort study in Chengdu found that compared with the control group, TIV reduced duration of hospitalization due to acute COPD exacerbation by 3.3 d and 7.1 d at 3 months, 6 months after vaccination respectively.¹⁹³ A 2013 Cochrane systematic review¹⁸³ and the WHO evidence summary¹⁹⁷ identified only one RCT studied influenza vaccine effectiveness in patients with asthma,¹⁹⁸ which did not demonstrate that vaccination can decrease asthma incidence or influenza-related asthma duration and severity.

A 2013 meta-analysis of RCTs showed that influenza vaccination could decrease risk of cardiovascular events among cardiovascular patients (RR: 0.64, 95% CI: 0.48–0.86), with greater effectiveness among patients with acute coronary syndrome (RR: 0.45, 95% CI: 0.32–0.63).¹⁸⁰ No systematic review on influenza vaccination among patients with liver diseases, renal diseases or diabetes was found. Only one RCT on influenza vaccination among liver cirrhosis patients in South Korea was found with an efficacy of 76% (95% CI:18–93%) against LCI.¹⁹⁹

WHO summarized 2 RCTs in HIV-infected people finding that influenza vaccine showed a 75% (95% CI:9–96%) efficacy against confirmed influenza.^200–202 A systematic review of malignant tumors of the hematologic system included 2 RCTs. Meta-analysis showed an efficacy of 44% (95% CI:28–56%) against influenza virus infection, 61% (95% CI: 22–81%) against pneumonia and 83% (95% CI: 69–91%) against hospitalization respectively.¹⁹⁰

Healthcare workers

The majority of HCW are healthy adults, and thus the immunogenicity, efficacy, and effectiveness of influenza vaccines among HCW could refer to healthy adults, which should be optimal.

A systematic review on influenza vaccination among HCW in 2011 suggested that evidence of the reduction of LCI among HCW is still limited.²⁰³ There was only one high-quality RCT with LCI among HCWs as outcomes published in 1999, finding that vaccine efficacy against serologically confirmed influenza among HCW was 88% (95% CI: 59–96%).²⁰⁴ Two other RCTs and one cohort study reported other outcomes on vaccination among HCW, including ILI, duration of ILI symptoms, and absenteeism due to respiratory infections. However, there are some limitations in randomization, allocation concealment, blinding, ILI definitions, as well as a lack of labratory confirmation, biased the results.^205-207 Two relevant cohort studies in China reported that influenza vaccination among HCW could reduce absenteeism, incidence and clinical visits due to ILI and respiratory diseases, and decrease clinical visits due to cardiovascular and cerebrovascular diseases as well as diabetes in HCWs.^208,209 Nonetheless, neither study was blinded nor laboratory confirmed. One cohort study in Chaoyang and Xuanwu District in Beijing in 2004 estimated that within 3–6 months after influenza vaccination, the protection rates for ILI and respiratory diseases was 17.3%, with a reduction of 29.9% in associated clinical visits; the effectiveness for cerebrovascular diseases and diabetes was 67.1%, with a reduction of 75.5% in associated clinical visits.²⁰⁹ Another cohort study among HCW in Xuanwu District in Beijing in 2006 estimated that within 3 months after vaccination, the effectiveness against ILI was 39.5%. There was a significant difference regarding average disease duration of vaccination group and control for ILI patients (1.7 d in vaccination group and 3 d in the control group).²⁰⁸

HCW infected with influenza could contribute to influenza outbreak in the hospital that can directly infect patients, on the other hand, could affect healthcare system due to staff shortages. Therefore, influenza vaccination of HCW plays an important role in reducing influenza–related diseases among patients. Related studies were not found in China. A few RCTs and cohort studies were conducted abroad,^210-216 but with severe biases, such as lack of blinding, low statistic power, loss to follow-up not considered, and case selection bias, etc. Cochrane systematic review conducted a meta-analysis for relevant RCTs^210-213,217 and found that influenza vaccination of HCW was not significantly effective against serology confirmed influenza (OR: 0.86, 95% CI: 0.44–1.68), incidence of pneumonia (OR: 0.71, 95% CI: 0.29–1.71), and pneumonia death (OR: 0.87, 95% CI: 0.47–1.64) among elderly patients in long-term care institutions, but was significantly effective against ILI incidence (RR: 0.71, 95% CI: 0.58–0.88), clinical visits due to ILI (OR: 0.48, 95% CI: 0.33–0.69), and all-cause mortality (OR: 0.68, 95% CI: 0.55–0.84). In addition, some modeling studies demonstrated that in long-term healthcare institutions and general hospitals, there was a linear association between influenza vaccination among HCW and number of infected patients, without herd immunity threshold. When the percentage of HCW vaccinated increased from 0% to 100%, a maximum of 60% influenza infection among patients could be prevented.^218,219

Safety of TIV

TIV is considered safe. However, transient local reaction on injection sites are common (>1/100). Vaccinees without previous exposure to influenza vaccine antigens such as infants might experience fever, general malaise, or muscle pain.² Such adverse events rarely occur in adults.²²⁰

Adverse Events Following Immunization (AEFI) refers to reactions or events that could result in illness or injury during or after vaccination. Pilot surveillance system of AEFI in China was started in 2005, expanding to the whole country in 2008. From 2005 to 2012, 7095 adverse events associated with seasonal influenza vaccines (including split, whole virus, and subunit vaccines) were reported.^221-226 Since 2010, reported adverse events were categorized. Between 2010 and 2012, 5092 adverse events were collected, among which 98 were severe AEFIs, 14994 were non-severe AEFIs.^224-226

Children

In China, the majority of safety studies on influenza vaccination among children are RCTs before and after marketing. Studies found that there were no significant differences regarding safety between domestic and imported influenza vaccines, most adverse events were local reactions (e.g., erythema, swelling, scleroma, pain, burning around injection sites) and systematic reactions (e.g. fever, headache, dizziness, sleepiness, fatigue, muscle pain, general discomfort, nausea, vomiting, abdominal pain, diarrhea etc.). Most reactions were mild, and severe reactions were rare.^227-239 An Australian study found that febrile convulsion among young children was frequent especially after inactivated influenza vaccine, and the risks were even higher when the children received pneumococcal conjugate vaccines at the same time, whereas no increased risks were found in children aged >4 y.⁹⁶

Adults

Safety studies of influenza vaccination in adults in China are mostly RCT before and after marketing. Common reactions are local reactions such as erythema, swelling, redness, pain in injection sites and systematic reactions such as transient fever, headache, and muscle pain.^143,227-241 Analysis of placebo-controlled study in adults and vaccine adverse events reporting system in US showed that most common adverse events observed in adults with inactivated influenza vaccination are pain, redness, muscle pain in injection site, and headache.^242,243

Pregnant women and newborns

Abundant oversea data suggested that influenza vaccination among pregnant women is not harmful to the gestational outcome or fetus.^244-251 A prospective cohort study in Jiangsu province enrolled 122 pregnant women who received A(H1N1)pdm09 monovalent vaccines, with 104 unvaccinated pregnant women as control, showing that AEFI incidence in the vaccinated group was 3.3%. No significant difference regarding incidence of post-term pregnancy or spontaneous abortion was observed between 2 groups. Both groups delivered healthy newborns (Apgar score ≥7 ).²⁵² Given that the vaccination rates among pregnant women for seasonal influenza is extremely low in China, there are little systematic safety data on influenza vaccination in pregnant women and newborns

Persons with chronic diseases

A blinded randomized cross-over research among 1952 adults and children with asthma between September and November 2000 in US showed that within 2 weeks of TIV vaccination, no increase of asthma exacerbation in any age group was reported.²⁵³ Between October and December 1999, a prospective cohort study in US found that adverse events rates were similar between outpatients and inpatients with one or more chronic diseases in any age group of ≥65 or 18–64 y²⁵⁴ Limited domestic literatures showed that influenza vaccine group developed adverse events compared with the unvaccinated group, although the incidence of adverse events were low (about 5%–16%) and usually mild and temporary.^195,255,256

Immunocompromised population

Safety data in patients infected with HIV after vaccination is limited, and no evidence proved that influenza vaccination influences HIV infection or immunity.⁹⁶ Data with regards to another immunocompromised population is also limited. Small-scale study showed that vaccination among kidney, heart, or liver transplant recipients did not affect the functions of transplanted organs, and no rejection reaction was induced.^257-260 Safety data among this population is a lack in China.

Immediate anaphylaxis

Individuals might be allergic to components of influenza vaccine and experience an immediate anaphylaxis. Systematic allergic reactions include but not limited to systematic urticaria, stridor, edema of the oral cavity, tongue as well as throat, dyspnea, vomiting, blood pressure decreased, low consciousness level, and shock. There could also be a few mild symptoms such as redness of eyes or hoarseness.^261,262 Systematic allergies after influenza vaccination are rare.^263,264 Based on reports in Hong Kong, approximately 0.9 cases of immediate hypersensitivity among 1 million doses of vaccinations.²⁶⁵

Ocular and respiratory symptoms (ORS)

ORS is defined as acute and self-limited reactions induced by inactivated influenza vaccines, characterized by prominent ocular and respiratory symptoms, which was first reported in 2000–2001 season in Canada. The original definition of ORS was one or more of the following symptoms occurred within 2–24 hours after inactivated influenza vaccination: red eyes, cough, stridor, chest distress, dyspnea, sore throat, or facial edema, which could not be alleviated within 48 hours of onset.²⁶⁶ ORS was only found to be associated with one vaccine used in 2000–2001 season in Canada.²⁶⁷ After improving related manufacture technology, the incidence of ORS in Canada was significantly decreased,²⁶⁸ and with milder symptoms that could be recovered within 24 hours without treatment.²⁶⁹ However, a few placebo-randomized control studies showed that ORS after influenza vaccination was coincided and not associated with vaccination.⁹⁶

Guillain-Barré syndrome (GBS)

As reported, population with GBS history was more susceptible to GBS recurrence than those without GBS history.²⁷⁰ Therefore, the expected risk of GBS among population with GBS history after influenza vaccination should be higher than those without GBS history. With regards to the association between inactivated influenza vaccines and GBS incidence, results varied among different studies. Some studies suggested that inactivated influenza vaccination might contribute to increased GBS incidence,^271,272 while others demonstrated that risk of GBS was not significantly increased after influenza vaccination^273-278; and several studies even implied that influenza vaccination was associated with decreased GBS incidence.²⁷⁹ Based on a few studies that showed an association between influenza vaccination and GBS, expected risk of GBS was low after influenza vaccination, with approximately 1 case per 1 million vaccinated populations.^271,276 As a preventive measure, persons with a history of GBS within 6 weeks after influenza vaccination are not recommended for vaccination. These persons might use anti-influenza viral agents as alternatives.⁹⁶

No literatures on GBS after TIV vaccination were found in China. Between September 2009 and March 2010, 89,600,000 doses of A(H1N1)pdm09 monovalent vaccines were used in China, and 8,067 cases of adverse events were reported, among which 6,552 were defined as vaccine-related adverse events, and 1083 cases were rare or severe adverse events, of which 1,050 were allergic reaction. There were a total of 11 GBS cases in vaccine recipients, with an incidence of 0.1 per 1 million doses, much lower than the baseline incidence between 2008 and 2010 (1.9–2.7/1000, 000 in adults).²⁸⁰

Thiomersal

Thiomersal is a kind of antibacterial compounds containing mercury, which may be used in inactivated influenza vaccines to decrease the probability of bacterial growth. Although more and more evidence showed that vaccines containing thiomersal would not increase adverse events,^281-290 several public health institutions abroad suggested that as one of the strategies to reduce mercury exposure, measures should be taken to reduce or prohibit thiomersal in vaccines.^281,282 Current Chinese Pharmacopeia sets limits for thimerosal contents (≤100 μg/ml) for a single dose vaccine. Presently, imported influenza vaccines do not contain thimerosal, and half of domestically manufactured influenza vaccines contain thimerosal that complies with Chinese Pharmacopoeia.

Safety of co-administration with other vaccines

Inactivated vaccines do not interfere with immune responses to other inactivated or live attenuated vaccines. Current evidence on co-administering influenza vaccines and other attenuated live vaccines is limited. For the population aged ≥50 y, herpes zoster vaccines were administered with TIV at the same time or spaced with 4 week's interval showed similar safety and immunogenicity.⁹⁶ Co-administration of pneumococcal polysaccharide vaccine and TIV induces protective antibody responses without increase in incidence or severity of adverse events. DaPT vaccine could also be vaccinated among adults together with TIV.²⁹¹

Health economics evaluations of influenza vaccination

Influenza vaccination can reduce influenza associated with emergency visits, hospitalization, deaths, and treatment cost, thus resulting in economic benefits. In addition, the economic benefits of influenza vaccination are also demonstrated by reduced productivity loss due to absenteeism.

A most recent systematic review of 51 health economics studies²⁹² found that influenza vaccination was cost-saving (lower social costs for vaccinated population compared with unvaccinated population) in 22 studies (12 in children, 8 in the elderly, and 2 in pregnant women). Cost-effectiveness ratio in 13 studies was lower than $10,000 or cost-benefit ratio was close to 1 (the WHO recommended criteria is that intervention is considered highly cost-effective when cost-effectiveness ratio was lower than GDP per capita in the country; when the ratio of cost to effectiveness was 1–3 times of GDP per capita, intervention is considered cost-effective, and when cost-effective ratio were over 3 times GDP per capita, intervention is considered not cost-effective). Cost-effectiveness ratio in 13 studies ranged from $10,000 to $50,000 or cost benefits ratio lower than 6. Cost-effectiveness ratio in 3 studies was higher than $50,000. Most studies demonstrated that influenza vaccination among children could reduce costs or was considered highly cost-effective, and influenza vaccination in the elderly and pregnant women was cost-effective.

Among the 51 studies included in the above systematic review, most focused on high- or and medium-income countries and areas,²⁹² with few studies in China. A study in Hong Kong SAR showed that 90% of community-based elderly population aged ≥65 y received influenza vaccines resulted in a net social benefits of 3.78 billion HKD. Vaccination given to the elderly who visit clinics due to diseases other than cold or fever between October and December, would lead to a net social benefits of 3.01 billion HKD when vaccination rates were 62% and 76% among patient aged 65–74 y or ≥75 y, respectively.²⁹³ Another study in Hong Kong SAR showed that compared with unvaccinated group, an annual influenza vaccination for the elderly aged ≥65 y in long-term inhabitation in nursing homes was demonstrated to be more cost-effectiveness, improve health outcomes (quality adjusted life year increased 0.058 per person), and was cost saving (334 RMB cost reduction per person), with a cost-effective ratio of 6.39.²⁹⁴

A prospective population-based study in Taipei, Taiwan in 2000 and 2001 showed that among 226,997 local subjects aged ≥65 y, 35.6% were vaccinated with influenza vaccines, with 15% reduction in pneumonia and influenza hospitalization among the vaccinated population.²⁹⁵ Compared with unvaccinated population, vaccinated population had a 29% reduction in influenza–related deaths. The incremental cost-effectiveness ratio (difference of total costs between the vaccinated and the unvaccinated divided by the difference of life year or death numbers) was $309 for each life-year saved and $3,899 for each death avoided. Compared with the other health interventions in this group (such as cancer screening with an incremental cost-effectiveness of over $20,000), influenza vaccination was demonstrated to be more cost-effective. Another study in Taiwan in 2012 showed that an influenza vaccination rates of 57% in 34112 cancer patients aged 20–64 y living in Taiwan districts reduced 110 influenza-related deaths, 391 influenza-related hospitalization, 159 emergent visits, and 563 outpatient visits, resulting in a net social benefits of $ 22,300,000.²⁹⁶

Recommendations for Influenza Vaccination in the 2014–2015 Season

Currently, influenza vaccine is not incliuded in the national immunization program in China, and recipients shall pay for the vaccine voluntarily out of pocket. In some places (e.g., Zhuhai of Guangdong Province, Ningbo of Zhejiang Province, and Suzhou of Jiangsu Province, etc.), influenza vaccines have been covered by the medical or social insurance reimbursement system, and in some individual regions (Beijing, Karamay of Xinjiang Autonomous Region, etc.), free vaccination is given to certain people through the government fiscal subsidies. To raise the public awareness of the harm of influenza and the effectiveness of vaccine, and to gradually improve the vaccine coverage for the high-risk groups, the CDC at all levels should actively organize and carry out trainings to publicity, health education and personnel. For organizing and guiding vaccination, the choice of dosage, priority population, vaccination procedures, contraindications, inoculation timing, etc., should be emphasized. Vaccine leaflet is approved based on clinical trials and its update takes some time; therefore, when the product leaflet is different from this guideline regarding indicated population, doses, contraindications, etc., the leaflet should be referred.

The components of antigen and target age groups

The components of Northern hemisphere trivalent influenza vaccines in 2014–2015 season are A/California/7/2009(H1N1) like strains, A Texas/50/2012(H3N2) like strains, and B/Massachusetts/2/2012 (Yamagata lineage) like strains. The quadrivalent vaccine components recommended by WHO contain 2 lineages of type B viruses, which include the same strains as TIV plus B/Brisbane/60/2008 (Victoria lineage) like strain.

Currently, approved influenza vaccines in China are all TIV, including 2 formulations of 0.25 ml and 0.5 ml, respectively, which can be given to the population aged nes in ChinThe formulation of 0.25 ml, containing 7.5 μg of each hemagglutinin subtypes per dose, is used for 6–35 months infants; the formulation of 0.5 ml, containing 15 μg of each hemagglutinin subtypes per dose, is used for aged ≥36 months.

Recommended priority groups for influenza vaccination

The influenza vaccine is safe and effective. Points of vaccination (POVs) should offer influenza vaccination services for people who have willingness to vaccination and without contraindications. Considerable scientific evidence suggests that severity and clinical outcome of influenza infection vary. Therefore, we recommend the following individuals as the priority vaccination groups based on the experience of other countries, WHO position paper, and combined with China's local conditions.

Pregnant women

Pregnant women at any pregnancy stages should be offered influenza vaccine. The recommendation is based on the following evidence: the risk of pregnant women suffering from severe influenza is high; influenza vaccine is safe during any stage of pregnancy, the vaccine is effective in preventing pregnant women and the infants against influenza. Furthermore, the influenza-related disease burden in infants under 6 months of age, to whom current vaccines can't be given, is very high.

The family members and caregivers of infants <6 months of age

The current influenza vaccine cannot be given to infants under 6 months of age. They can get protection indirectly through vaccination of their mothers in pregnancy and by vaccinating reduce exposure of family members and caregivers.

Infants and children aged 6–23 months old

This group is more likely to experience serious diseases and influenza-related hospitalization, thus should be the priority for influenza vaccination. The effectiveness of the vaccine in this age group is highly dependent on the matching degree between vaccine strains and circulating strains.

Children aged 2–5 y

The burden of influenza is also high but lower than that of children <2 y of age. After influenza vaccination, the immune response in this age group is better than that of children <2 y of age

Elderly ≥60 y of age

Influenza-related death burden is highest in the elderly, so the elderly are an important target population for influenza vaccination. Vaccination is the most effective way to protect the elderly from influenza at present.

Persons with specific chronic underlying diseases

Persons with specific chronic underlying diseases, such as cardiovascular disease (except for simple hypertension), chronic respiratory disease, hepatic and renal dysfunction, blood system disease, nervous system disease, nervous-muscle function disorders, metabolic disease (diabetes), and patients with immune deficiency, are more likely to experience serious diseases when get infected. Thus, these persons should be given high priority for vaccination.

Healthcare workers

HCW is a priority group for influenza vaccination, not only to protect HCW themselves and maintain the regular medical service during influenza season, but also to reduce the spreading from HCW to vulnerable patients.

Dosage

Children aged 6 months to 8 y old

Studies demonstrated that children aged 6 months to 8 y who have never received influenza vaccination, require 2 doses of vaccine, with an interval of ≥4 weeks of the first vaccination to induce protection. Children who have received one or more influenza vaccines in a preceding year should receive one dose only.

Children >8 y old and adults

Only one dose is needed.

Vaccination timing

In general, protective antibodies develop within 2 to 4 weeks after influenza vaccination, and the antibody titer begins to decay after 6–8 months. The peak time and duration of annual influenza activity vary by regions in China. In order to ensure protection before the influenza season, influenza vaccination should be arranged as soon as the vaccines and the immunization services are available.

Pregnant women at any pregnancy stages should receive influenza vaccine. It is recommended that pregnant women vaccination should be provided as soon as possible.

Vaccination sites and routes

TIV should be given by intramuscular or deep subcutaneous route. Upper arm deltoid areas are preferred for adults and older children while the anterolateral thigh is preferred for infants and young children. Subcutaneous injection should be adopted for patients with thrombocytopenia or another hemorrhagic disease because of the risk of bleeding will be triggered by intramuscular injection.

Contraindications

People with documented allergy history to eggs or any ingredients in vaccines should not be vaccinated. Patients with mild-to-moderate acute diseases, with or without fever are recommended to vaccinate after symptoms vanished. GBS occurring within 6 weeks after previous influenza vaccination is not a contraindication to vaccination, but special attention should be paid. Specific contraindications to a product should refer to the product manual and the doctor's advice.

Co-administering influenza vaccine with other vaccines

Inactivated influenza vaccine and other inactivated or attenuated live vaccines can be vaccinated in different sites at the same time, or can be vaccinated at any time interval for which the immunogenicity and safety of influenza vaccine and other vaccines are not influenced.

Special vaccination policy in local districts of China

At present, free influenza vaccination policy to certain people has been implemented through the government fiscal subsidies in some regions. For example, in Beijing, free influenza vaccination service is provided for people ≥60 y old and primary and middle school students since 2007; in Karamay of Xinjiang Autonomous Region, free influenza vaccination service is provided for people ≥ 60 y old and children aged 3–7 y old since 2008; in 8 counties of Fushun, Jinzhou and Tieling city, Liaoning province, free influenza vaccination pilot programs are implemented for ≥65 y elderly. In some places, influenza vaccines have been included in the medical, social insurance system or new rural co-operative medical system, where people can get reimbursed, such as in Xi'an of Shaanxi province, Zhuhai of Guangdong Province, Ningbo of Zhejiang Province, and Suzhou of Jiangsu Province.

Implementation of influenza vaccination service

Requirements of vaccination sites

The persons who are willing to receive influenza vaccine should go to the public health department designated POVs for influenza vaccination. POVs should be eligible and released to the public. The procedures of the influenza vaccination in POVs are as follows:

Registry

The eligible age group for influenza vaccine and contraindications should be strictly remembered and screened. Children's vaccination certificates should be examined, and the possible adverse reactions and precautions after vaccination should be informed. The vaccinees or guardians should sign “the informed consent.”

Vaccination room

vaccination doctors safely provide vaccination.

Observation room

observe for 30 minutes after inoculation.

Precautions

1. Vaccination doctors should strictly follow the requirements of vaccinating.

2. The vaccine should be fully shaken before inoculation. The vaccine should not be used if the following situations occur: deposits occur which being undispersed, foreign materials, frozen, vials with cracks, or unclear labels.

3. The vaccine should be kept and transported at 2–8°C in the refrigerator or cold box, and frozen is strictly prohibited. Take the vaccine out from the refrigerator or cold box before inoculation. Ice packs should be replaced timely when they are melting.

4. Disinfect the injection sites with 75% ethanol from inside to outside by spiral before vaccination with disinfection diameter ≥5 cm, and immediately vaccinate after area is dry.

5. There should be obvious influenza vaccination identifiers at POVs, and inoculated rooms/tables.

6. 1:1000 adrenaline and other treatment medicine and equipment should be available at the POVs for emergency use, in case a rare serious allergic reaction occurs.

7. The syringes and other related wastes should be collected and destroyed promptly after vaccination in accordance with the regulations.

8. The vaccinees should be informed to stay for 30 minutes to observe any reactions after inoculation. The consulting phone number should be provided to the vaccinees, so the vaccinees may consult for any discomforts at any time.

Vaccination records

If the vaccinees are children, doctors should record the vaccine name, dose, batch number, manufacturer, inoculation time, POV, signature of the doctor on children's vaccination certificates after the vaccination. If the vaccinees are other people who pay for vaccination or receive vaccination for free, their information should be collected, reported, and recorded in accordance with the relevant regulations. And the information of vaccinees and summary data on influenza vaccination should be reported to “information management system for immunization program” within the required timeline.

Surveillance and evaluation of adverse reactions

To guarantee the safety of influenza vaccination, departments of disease control and POVs at all levels should monitor adequately adverse events following immunization (AEFI) of the influenza vaccine. Once adverse reactions or events occur, AEFI surveillance system should start to operate. AEFI surveillance report, investigation, diagnosis, and treatment, should be implemented in accordance with the requirements of “National AEFI Surveillance Program.” If POVs have learned that a suspected adverse reaction occurs after influenza vaccination, they should report the AEFI to the disease control department immediately. AEFI reporting process is listed as follows:

1. AEFI card should be submitted within 24 h after being aware of abnormal reactions.

2. All the AEFIs except for common reactions should be investigated within 48 h after being reported, and a case investigation form should be submitted within 3 d after the investigation.

3. The content reported must be clear including major clinical process, especially the time interval between vaccination and symptoms. Meanwhile, clinical-related data, such as the vaccination information should be collected.

4. Any death, severe disabilities, mass abnormal reaction suspected to be associated with the vaccination, and suspected unusual reaction with significant influence on society, should be reported within 2 h after being reported.

5. Please submit the completed report cards and case investigation forms to the local disease control department, and inform the administrative department of CDC via phone call. Once received AEFI report, the local CDC should immediately organize the AEFI expert panel to investigate and analyze, in order to give the preliminary conclusions as soon as possible and write the investigation report. When disagreements exist, the AEFI should be reported to the local municipal medical association to make a conclusion. When determining whether or not it is an abnormal reaction, one should be noted that the following items do not belong to abnormal reactions:

   1. The typical response caused by vaccine due to the inherent characteristics of vaccine;
   2. The harm to the vaccinee by unqualified vaccine;
   3. The harm to the vaccinee due to the violation of the vaccination specifications, standard procedures, the guidelines for vaccine, immunization programs by vaccination sites;
   4. The coincided disease occurred when vaccinee is during the incubation period or prodromal stage of individual disease when vaccinated;
   5. The vaccinee or his guardian does not provide information on subject's health conditions and contraindications which exist according to vaccine leaflet, etc., and vaccination trigger the acute onset of or exacerbate original disease;
   6. Psychogenic reactions due to psychological factors of the individual or groups.

When dealing with AEFI cases, vaccination sites should collect and sort all clinical data, and prepare related information of influenza vaccination for use by AEFI investigation and diagnosis expert panel at higher level. It is highlighted that clinical units and any individuals in vaccination sites are not qualified to make AEFI diagnosis.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.