Changes in Influenza Activities Impacted by NPI Based on 4-Year Surveillance in China: Epidemic Patterns and Trends

Background Since the Non-pharmaceutical Intervention (NPI) by COVID-19 emerged, influenza activity has been somewhat altered. Objectives The aim of this study was to explore changes in influenza activities in the context of COVID-19 based on the sentinel hospitals/units in Guangdong, southern China. Methods The surveillance data in influenza-like illness (ILI) were collected from 21 cities in Guangdong between September 2017 and August 2021, while 43 hospitals/units were selected to analyze the predominant types of influenza, population characteristics, and seasonal features by three methods (the concentration ratio, the seasonal index, and the circulation distribution), based on a descriptive epidemiological approach. Results During the four consecutive influenza seasons, a total of 157345 ILIs were tested, of which 9.05% were positive for influenza virus (n = 14238), with the highest positive rates for both IAV (13.20%) and IBV (5.41%) in the 2018–2019 season. After the emergence of COVID-19, influenza cases decreased near to zero from March 2020 till March 2021, and the dominant type of influenza virus changed from IAV to IBV. The highest positive rate of influenza existed in the age-group of 5 ~  < 15 years in each season for IAV (P < 0.001), which was consistent with that for IBV (P < 0.001). The highest annual positive rates for IBV emerged in eastern Guangdong, while the highest annual positive rates of IAV in different seasons existed in different regions. Furthermore, compared with the epidemic period (ranged from December to June) during 2017–2019, the period ended three months early (March 2020) in 2019–2020, and started by five months behind (April 2021) during 2020–2021. Conclusion The highest positive rates in 5 ~  < 15 age-group suggested the susceptible in this age-group mostly had infected with infected B/Victoria. Influenced by the emergence of COVID-19 and NPI responses, the epidemic patterns and trends of influenza activities have changed in Guangdong, 2017–2021. Supplementary Information The online version contains supplementary material available at 10.1007/s44197-023-00134-z.


Introduction
Influenza (flu) is an acute respiratory infectious disease due to the influenza virus, which often causes a series of local outbreaks and seasonal epidemics, then resulting in socioeconomic burden [1]. Influenza-like illness (ILI), including influenza infection, is an important public health concern, but the timing and peak intensity vary considerably from season to season and regionally [2]. Since the emergence of Coronavirus Disease 2019  at the end of 2019, it has had a great impact not only on human health but also on the spread of influenza worldwide by the Non-pharmaceutical Intervention (NPI). In South Korea, the proportion of ILI in general outpatient clinics decreased to 49.8‰ in the 2019-2020 influenza season, in contrast, during 2016-2019 influenza seasons the rates of ILIs had maintained at the level of 71.9‰ to 86.2‰ [3]. Regarding China, in Hubei, during the emergency response to COVID-19 (Feb ~ Mar, 2020), the positive rates of influenza A virus (IAV) and influenza B virus (IBV) in ILIs were 4.17% and 0.29%, respectively, which were significantly lower than those in 2015-2019 (P IAV < 0.001, P IBV < 0.001) [4]. Additionally, in Zhenjiang of Zhejiang, the weekly average number of influenza cases between the 5th and 18th weeks in 2020 decreased by 97.5% compared with the first 4 weeks [5].
Continuous and systematic influenza surveillance could keep track of epidemic trends, predominant strains, antigenic variations and sensitivity of antiviral drugs. A previous study [6] found that the results of two methods (the concentration ratio and the circular distribution) were highly correlated (r = 0.905, P < 0.001), of which the peak of influenza virus activities appeared from October to March of following year during 2009-2018 in Qinghai. To learn the characteristics of influenza activities in Guangdong, we performed an epidemiological analysis by three methods (the concentration ratio, the seasonal index, and the circulation distribution), and explored the patterns of influenza epidemic in the context of COVID-19 based on the sentinel surveillance from 2017 till 2021.

Epidemiological Definition
Epidemiological definitions included, (i) ILI: a case had body temperature ≥ 38 ℃, accompanied with either cough or sore throat, but a lack of molecular detection; (ii) Influenza case: an ILI tested positive for nucleic acid of influenza virus; (iii) Positive rate (PR): PR was influenza-virus-positive rate in specimen of ILIs, in both IAV and IBV.

Region Classification
Guangdong is located in the southern China, which is classified into four regions based on geographical and cultural features. The four regions include the Rearl River Delta, the Eastern Guangdong, the Western Guangdong and the Northern Guangdong. Specifically, the pearl river delta region has nine cities, covering Guangzhou, Shenzhen, Huizhou, Foshan, Dongguan, Zhaoqing, Jiangmen, Zhongshan and Zhuhai; The Estern Guangdong region has four cities, covering Chaozhou, Jieyang, Shantou and Shanwei; The Western Guangdong region has three cities, including Maoming, Yangjiang and Zhanjiang; The North Guangdong region has five cities, covering Heyuan, Meizhou, Qingyuan, Shaoguan and Yunfu.

Concentration Ratio
The concentration ratio (short for CR) [7] is a algorithm that comprehensively measures the tendency of influenza concentration by calculating the monthly distribution of influenza cases in each influenza season and labeling the seasonality. A series of computational formulas included, ; where f i is the number of influenza cases in the i th month, N is the cumulative number of influenza cases throughout the whole influenza season, and R describes the dispersion. The degree of concentration was divided into six stages by CR values, criteria were as following, 0.9 ≤ CR ≤ 1 for very high concentration, 0.7 ≤ CR < 0.9 for high concentration, 0.5 ≤ CR < 0.7 for moderate concentration, 0.3 ≤ CR < 0.5 for low concentration, 0 < CR < 0.3 for minimal concentration and CR = 0 for evenly distributed.

Seasonal Index
The seasonal index (short for SI) [8] is a statistical method, in which the seasonal pattern of influenza is shown as the ratio of influenza cases per month to the monthly average number of influenza cases in each influenza season. The calculation formula is SI = A/B × 100%, where A is the number of influenza cases per month and B is the monthly average number of influenza cases throughout the whole influenza season. A month when its SI value is greater than 100% is considered an epidemic period of influenza, otherwise it is non-seasonal one.

Circular Distribution
The Circular Distribution (short for CD) [6] is a method for processing periodic circular data, in which an influenza season was treated as a circle, the onset time was converted into an angle and the periodic data were transformed into linear data by trigonometric functional transformation. A set of equations included as following, where f i is the number of influenza cases in the i th month, γ describes the discrete trend, a is the mean angle and s is the standard deviation for a.
The Rayleigh test (Z test) was used to test the presence of the mean angle. The formula is Z = N 2 , where N is the number of influenza cases in the whole influenza season; if Z > Z 0.05 = 2.996 , the mean angle has statistical significance. The period of influenza epidemic ( a ± s ) was predicted as N > 100.

Statistical Analysis
Excel 2019 was used to deal with original data and draw figures. Data were analyzed by using SPSS v.24 (SPSS Inc., Chicago, IL). The chi-square ( X 2 ) test was performed for categorical variables with significance level at two-tailed P < 0.05. The statistical analyses on data include the indicators related to above three methods (CR, SI and CD).

General Characteristics
The number of ILI specimens collected for PCR testing during the four influenza seasons accounted for 23.91% (37,623), 24.15% (37,997), 26.67% (41,969) and 25.27% (39,756), respectively. The ratio of male-to-female was 1.30:1 (88 864/68 481) and the median age was 9 year-old (Interquartile range 3-34 year-old). Through PCR testing, a total of 14 238 influenza cases were sorted out, with 9.05% of positive rate. The annual positive rates showed an upward trend from 11.59% in the 2017-2018 influenza season to 18.58% in the 2018-2019 season, and then decreased to 5.51% and 1.28% in the 2019-2020 and the 2020-2021 seasons, respectively. These presented a tendency in positive rates decreasing during 2019-2021 seasons compared to the 2017-2019 seasons, even closing to zero from March 2020 to March 2021, on the back of remaining stable in the number of samples tested (Fig. 1).

Gender Distribution
The positive rate of influenza virus in female was 9.29% (6361/68481), which was higher than that in male (8.86%, 6361/88864) ( X 2 =8038.89, P < 0.001). In addition, compared the annual positive rates of influenza viruses by gender, there were no statistically difference in both male and female in the annual positive rates for both IAV and IBV for each season. The maximum positive rates for both IAV and IBV in boy and girl existed in the 2018-2019 season and the 2017-2018 season, respectively (Table S1).

Age-group Distribution
School children in the 5 ~ < 15 age-group shared the highest positive rates for influenza virus (17.36%), followed by ) and 25 ~ < 60 age-group (9.45%). By comparing the annual positive rates by agegroups, the highest positive rate of influenza virus existed in the 5 ~ < 15 age-group, ranging from 3.41% to 27.68%; in contrast, the lowest positive rates emerged in both 0 ~ < 5 age-group and ≥ 60 age-group, covering 0.33%-11.32% and varying 0.29%-11.13%, respectively. In addition, the annual positive rates of both IAV and IBV existed in different age-groups, the highest positive rate was in the agegroup of 5 ~ < 15 age-group in each influenza season for IAV (P < 0.001), which was consistent with that for IBV (P < 0.001). The maximum annual positive rates for both IAV and IBV appeared in 5 ~ < 15 age-group existed in the 2018-2019 season (18.45%) and the 2017-2018 season (18.17%), respectively (Table S2).  (Table S3).

Concentration Ratio
Based on the monthly number of influenza cases from 2017 to 2021 in Guangdong, the concentration of influenza activities in each influenza season was analyzed, shown in Table 3.

Seasonal Index
The  Table 4. Compared with the 2017-2020 season, the influenza epidemic period was delayed in 2020-2021 (December vs. March 2021).

Circular Distribution
According to the number of influenza cases, the epidemic period for each influenza season was calculated separately by using the circular distribution method, as the epidemic periods shown in Table 5

Discussion
The overall positive rate of influenza in ILI was 9.05% (14 238/157 343) from 2017 till 2021 in this study, and the annual positive rates of influenza virus in ILIs showed a Similarly, the decreases of PRs also occurred in Canada during 2020-2021, with IAV and IBV being only 1.5‰ and 2.8‰ times of 2014-2020, respectively [9]. During the COVD-19 pandemic, to inhibit the spread of SARS-CoV-2 by NPI, the Guangdong health administration implemented three levels of public health emergency responses, in which the first phase started on 23 January 2020 and lasted until 23 February 2020, the second phase varying from 24 February to 8 May 2020, and the third phase covered from 9 May 2020 to 31 August 2021. Based on the substantial decline in influenza virus activity during 2019-2021, the following reasons have been suggested. Firstly, both influenza virus and SARS-CoV-2 belong to respiratory infectious disease pathogens and have the same mode of transmission, therefore the general precautions (including the wearing of masks and hand washing, etc.) for COVID-19 prevention could reduce the frequency of influenza virus infections [10]. Secondly, fewer reports of influenza cases during the COVID-19 epidemic were due to both that people altered their health-seeking behavior and influenza is a self-limiting illness. Thirdly, online teaching and other methods reduced mass gatherings, which could hold down the influenza outbreaks in school. Finally, the viral interference between SARS-CoV-2 and influenza virus might be one of reasons for low influenza virus circulation [11,12]. Previous studies in southern China showed that seasonal influenza epidemics mainly were attributed to IAV subtypes of A/H1N1pdm and A/H3N2 in addition to IBV (B/Victoria and B/Yamagata) [13,14]. Four types/subtypes alternately circulated during the four consecutive seasons in this study, similarly report occurred in Western Saudi Arabia [15]. Moreover, IAV dominating in Guangdong was line with a previous report in Iran [16]. However, after the emergence of COVID-19, the dominant type of influenza virus was changed from IAV into IBV, which is consistent with that of the overall situation in China [17]. Since then, a novel evolutionary branch (V1A.3a.2) evolved from B/Victoria gene dominated in southern China [18] of which the travel restrictions in NPI during the COVID-19 pandemic had affected the spread of other types of influenza strains in different regions [19].
The female than male in this study accounted for a higher positive rate (9.29%/ 8.86%), which were different from those in a previous study (12.4% vs 23.0%) [20]. School children in the 5 ~ < 15 age-group had the highest positive rates for influenza virus (17.36%), followed by the 15 ~ < 25 (12.39%), which was lower than the positive rate in Georgia in this age-group (48.2%) [21]. Similar to the results in this study, the patients infected with B/ Victoria in Italy were mainly aged 5 ~ < 15 age-group (51.7%) [22]. The highest positive rates in 5 ~ < 15 agegroup indicated the susceptible mostly had infected with infected B/Victoria, which is the dominant lineage and takes an absolute epidemic advantage in the 2020-2021 season, regardless of the NPI reducing the risk of SARS-COV-2 transmission during the pandemic. In addition, younger than 5 years old group had the largest proportion in ILI specimens (37.49%), while the positive rate in this group was only 5.39%, which may be influenced by other non-influenza viruses. Adam K et al. reported that in the < 5 years old, the positive rates of adenovirus, rhinovirus and respiratory syncytial virus were higher than that for influenza virus [23], which suggested that this population was susceptible to multiple respiratory pathogens. Meanwhile, personnel coordination and the quality of samples may also be one of the reasons for the low positive rate in this age-group of < 5 years old. Analysis on epidemic pattern of infectious diseases usually was adopted with different epidemiological methods, of which three methods including CR, SI and CD in this study [6][7][8]. The method CR represents a concentration trend based on algebraic cumulative calculation throughout whole season, while the method SI is a ratio based on the specific measured value to monthly average value, revealing the seasonal pattern, and the method CD uses circular periodic in trigonometric function to assess statistical distribution, projecting the epidemic period.
In southern China (including Guangdong), the epidemic periods of influenza in this study began in November across six months in the 2017-2019 influenza seasons, but the period in 2019-2020 ended three months earlier than that in 2018-2019, while the period in 2020-2021 was from April till August. The epidemic characteristic changes might relate with the prevention and control measures to COVID-19 (such as human mobility, social distance and personal hygiene), but it is still worth exploring the main causes for the changes of influenza epidemic in the context of the COVID-19 pandemic. In conclusion, the results in this study presented the changes in the epidemic patterns and trends of influenza activities impacted by NPI based on the sentinel surveillance.