Featured post

#Analysis of recent #scientific #information on #avian #influenza A(#H7N9) virus - 10 February 2017 (@WHO, edited)

  Title : #Analysis of recent #scientific #information on #avian #influenza A(#H7N9) virus - 10 February 2017. Subject : Avian Influenza, ...

16 Dec 2016

#Review of #global #influenza #activity, October 2015–October 2016 (@WHO, edited)


Title: #Review of #global #influenza #activity, October 2015–October 2016.

Subject: Seasonal Human Influenza, activity surveillance report.

Source: World Health Organization (WHO), Weekly Epidemiological Record, full PDF file (LINK). Extract, edited.

Code: [     ]


Weekly epidemiological record / 16 DECEMBER 2016, 91th YEAR Nos. 51/52, 2016, 91, 601–624


Review of global influenza activity, October 2015–October 2016


This report summarizes the 2015–2016 influenza season in the temperate regions of the northern hemisphere, the 2016 influenza season in the temperate regions of the southern hemisphere, and influenza activity in tropical and subtropical regions. Where available, transmission characteristics, illness and mortality data are presented by geographic region and/or country. The report covers influenza activity data collected from October 2015 until the end of October 2016. The data presented have been derived primarily from reports published by national ministries of health or other official bodies, or reported to WHO through FluNet1 or FluID.2 Distributions of virus subtypes by transmission zone are shown in Map 1. For the purposes of this report, the peak of influenza activity is the month/week with the highest percentage of respiratory specimens testing positive for influenza viruses; influenza activity is based on the positivity rate for influenza. In some cases, the absolute counts of influenza detections are used if no denominator data on total samples processed are available.


Review of the 2015–2016 influenza season  in the northern hemisphere

Human influenza transmission usually occurs in the winter months in the northern hemisphere temperate region but the exact timing and duration of the influenza season varies by country and by year.

The  influenza season can begin as early as October, but typically peaks around January or February.


North America

Timing of the season and circulating influenza viruses

In North American countries [Canada and the United States of America (USA)] (Figure 1), the 2015–2016 influenza season was characterized by later than usual (midMarch) peaks in influenza activity in both countries with a predominance of influenza A(H1N1)pdm09 viruses (81% of influenza A viruses subtyped in the USA3 and 90% in Canada4). A small proportion of influenza B virus detections occurred throughout the season in the USA, while in Canada the proportion increased towards the end of the season.


Morbidity and mortality

In Canada, the peak of influenza positivity reached approximately 35%, similar to the previous season. The proportion of patient visits for influenza-like illness (ILI) was just over 75 per 1000 patient visits at the peak of the influenza season, slightly less than in the previous season, with children aged <20 years accounting for the majority of ILI consultations.

Cumulative paediatric (aged <20 years) influenza-associated hospitalizations were higher during the 2015–2016 season compared with the prior 4 seasons according to provincial and territorial reporting.4

During the 2015–2016 season, cumulative hospitalization and rates of admission to intensive care units (ICUs) for influenza infections were higher in adults aged 45–64 years, compared with the previous season, and lower in adults aged >65 years, with adults in the 45–64 year age group accounting for the largest proportion of cumulative influenza-associated ICU admissions (46%), according to sentinel hospital surveillance.

Absolute number of deaths due to influenza, as reported by a subset of areas, was lower than that reported in the previous 3 seasons.

In the USA, the peak in influenza positivity (29% in March) was similar to that of the 2014–2015 season. The proportion of patient visits due to ILI reached only 3.6% at the peak, lower than in the previous season. Pneumonia and influenza (P & I) mortality was also much lower this season compared with the previous 3 seasons. The cumulative influenza-hospitalization rate (84.8 per 100 000) was substantially lower in elderly adults (aged ≥65 years) compared with the previous season (322.8 per 100 000). The majority of influenza-related hospitalizations were due to influenza A(H1N1)pdm09 virus infections.3



Timing of the season and circulating influenza viruses

In Europe, influenza surveillance is coordinated by the WHO Regional Office for Europe and the European Centre for Disease Control (ECDC) and data jointly published in weekly updates (Flu News Europe)5 between weeks 40 and 20. The timing of influenza detections during the season was similar to that of previous years (Figure 1). Influenza activity started to increase in the last few weeks of 2015 (similar to the pattern observed in the 2011–2012,6 2013–2014 and 2014–2015 seasons), peaked in early February and decreased throughout March and April. The increase was earliest in some eastern European countries. The timing of the peaks in 2016 varied between countries, but was earlier in countries within the European region that fall into the western Asia transmission zone7 where activity peaked most often in January; in western Europe activity peaked later, most often in March. However, the percentage of specimens testing positive for influenza was still above the threshold in many countries in April.

Seventy per cent of typed influenza detections were influenza A viruses. Influenza A(H1N1)pdm09 viruses predominated (91%) in both sentinel and non-sentinel influenza detections, in contrast to the 2014–2015 season, during which influenza A(H3N2) virus accounted for the majority (77%) of subtyped influenza A virus detections. Of the influenza B viruses characterized to a lineage, 91% were of the Victoria lineage, while most (98%) in the 2014–2015 season were of the Yamagata lineage. Most countries in the region reported that detections of influenza A virus outnumbered those of influenza B virus; 14 out of 52 countries (27%) reported influenza B to be the predominant virus but there was no apparent geographical clustering of these countries.


Morbidity and mortality

Although the intensity of influenza activity in the 2015– 2016 season varied by subregion and country, with overall influenza positivity peaking at 53%, it was similar to that of the 2014–2015 season (55%).8 Many countries (16) reported periods of high intensity, though slightly fewer than in the 2014–2015 season (20); 3 countries reported very high influenza activity during the season (Finland in late January, and the Russian Federation and Ukraine in late January–early February), similar to that reported by 2 countries in 2014–2015. Countries in the eastern part of the region typically started reporting medium activity and returned to low activity earlier in the season compared with those in the western countries, indicating an east to west spread in the 2015–2016 season. Eight countries [Finland, France, Ireland, Romania, Slovakia, Spain, Sweden and the United Kingdom (UK)] reported data on laboratory-confirmed hospitalized cases. Influenza A(H1N1)pdm09 was the dominant subtype detected in hospitalized cases in both ICUs (98% of the subtyped A viruses) and other wards (99%). Most hospitalized laboratory-confirmed cases of influenza reported were among people aged 15–64 years, particularly adults aged 30–64 years. Seven countries reported data on hospitalized influenza-associated deaths with the highest numbers reported by France (362) and Spain (160); 52% (392/760) of deaths were in patients aged ≥65 years, and 43% (322/760) in the 30–64 year age group. Most of the fatal cases (86%) were infected with influenza A viruses and 14% were infected with influenza B viruses. Of the 397 influenza A viruses that were subtyped from these fatal cases, 97% (387/397) were A(H1N1)pdm09 and 3% (10/397) were influenza A(H3N2).

According to pooled analyses from the European system for monitoring excess mortality for public health action (based on data from 18 reporting countries: Belgium, Denmark, England, Estonia, Finland, Greece, France, Hungary, Ireland, Italy, Netherlands, Norway, Portugal, Scotland, Spain, Sweden, Switzerland and Wales), there was an overall excess all-cause mortality for the 2015– 2016 season at a level similar to that in the 2010–2011 and 2011–2012 seasons, i.e. unlike the very high mortality observed in the 2014–2015 season. Using a model that included influenza activity to directly estimate the effect of influenza, controlled for the effect of extreme ambient temperatures, there was an estimated excess influenza-attributable mortality for all ages at the same low levels as observed in the 2011–2012 and 2012–2013 seasons, again unlike the high excess mortality that was observed in the 2014–2015 season, particularly in the elderly aged ≥65 years. This pattern is consistent with a season in which influenza A(H1N1)pdm09 and influenza B viruses are predominant. However, the estimates varied between countries (EuroMOMO unpublished data).


North Africa and the Middle East

Timing of the season and circulating influenza viruses

In North Africa and the Middle East, influenza activity began to increase in October 2015, peaked at the end of 2015, and returned to low levels by mid-May 2016. In most countries, the timing of the season generally corresponded to patterns seen in previous years. In Egypt, there was a peak in influenza activity at the beginning of 2016 which was almost exclusively due to influenza A(H1N1)pdm09 viruses followed by a smaller peak of activity with influenza B viruses between March and June. In Algeria and Tunisia, influenza A(H1N1) pdm09 co-circulated with influenza A(H3N2) with very few influenza B virus detections. In countries in the Middle East (Bahrain, the Islamic Republic of Iran, Jordan, Oman, Pakistan and Qatar), influenza A(H1N1) pdm09 predominated with a peak of activity at the end of October. Influenza B viruses co-circulated throughout the season in Oman. From May to October 2016, activity continued at low levels with influenza A(H3N2) viruses circulating in Egypt and influenza A (not subtyped) viruses circulating in Qatar. The proportions of circulating viruses differ from those of the 2014–2015 season in the region, when there was co-circulation of influenza A(H1N1)pdm09, influenza A(H3N2) and influenza B in approximately equal proportions. Additionally, Bahrain, the Islamic Republic of Iran and Qatar all reported an increased number of samples tested for influenza this season, compared with previous years.


Morbidity and mortality

Influenza activity peaked at similar levels to those of 2014–2015, except in Egypt and Qatar, where activity was higher in 2015–2016, and in Algeria, where activity was lower. In Pakistan, the absolute number of ILI cases was similar to that of past seasons, with the exception of 2014–2015 where the ILI count was lower, while the absolute number of severe acute respiratory infection (SARI) cases at the peak of activity was substantially greater than that measured in previous years. The peak in SARI cases occurred a few weeks following the peak of influenza activity.


Eastern Asia

Timing of the season and circulating influenza viruses

In eastern Asia, influenza activity generally started to increase a few weeks later than in previous seasons, especially in Japan and the Democratic People’s Republic of Korea. Increasing activity started at the beginning of December 2015 and reached a peak in February or March 2016, except in Mongolia, where activity peaked at the end of 2015. In most areas, the predominant circulating virus was influenza A(H1N1)pdm09 in the first half of the season, followed by predominantly influenza B viruses in the second half of the season. Compared with the 2014– 2015 season, little influenza A(H3N2) virus activity was reported, except in northern China, where A(H3N2) predominated at the beginning of the season. In northern China the proportions of influenza B detections, followed by influenza A(H1N1)pdm09 detections, increased later in the season. For most of the season in northern China, the predominant virus was of the influenza B/Victoria lineage, unlike during 2014–2015 where influenza detections were almost exclusively influenza B/Yamagata viruses at the end of the season. Of the influenza B viruses characterized, the majority were of the Victoria lineage in Mongolia and the Yamagata lineage in Japan.


Morbidity and mortality

In northern China, the proportion of visits due to ILI at national sentinel hospitals peaked at about 5.5% around mid-February, followed by a second, smaller peak in mid-March. The peak in the proportion of consultations due to ILI during 2015–2016 was higher than in the previous 3 seasons. In Mongolia, ILI rates from sentinel healthcare facilities followed historical patterns. The proportion of pneumonia cases among hospitalized patients was above the tolerance range at the beginning of January, as well as in February and March, but did not exceed levels of the previous year. The number of deaths among patients hospitalized with pneumonia, however, was higher compared with the 2014–2015 season. In Japan, the cumulative ILI count was increased compared with the previous season, with children (aged <15 years) accounting for nearly half of the influenza-positive cases detected this season. Elderly adults (aged >60 years) with ILI accounted for a small proportion of cases, smaller than in the previous season. Similarly, the number of hospitalizations for influenza was increased for those aged <15 years compared with the previous season and decreased for those aged >60 years. The weekly proportion of visits due to ILI at sentinel healthcare sites in the Democratic People’s Republic of Korea followed a trend seen in past influenza seasons, with an increase above baseline in mid-January, a peak in mid-February and a return to baseline levels by May. The peak in the weekly ILI rate was just over 50%, slightly greater than in 2014–2015, but lower than the peak during the 2013–2014 season.


Review of the 2016 influenza season in the southern hemisphere

Human influenza transmission usually peaks mid-year in the southern hemisphere temperate region, but the exact timing and duration of the influenza season varies by country and by year. The influenza season can begin as early as May, but typically peaks around August.


Southern cone of South America

Timing of the season and circulating influenza viruses

The timing of the 2016 influenza season varied for the countries in the southern cone of South America. Argentina and Uruguay experienced earlier influenza seasons compared with previous years, with influenza detections increasing in early-to-mid March and peaking in mid-to-late May. Chile (Figure 1) and Paraguay experienced later seasons compared with previous years, with influenza detections beginning in late April. Activity in Chile peaked at the beginning of August and in Paraguay in mid-June. During late October, all 4 countries were still reporting influenza detections. Overall, influenza A(H1N1)pdm09 virus was generally predominant throughout this region, with co-circulation of influenza B viruses throughout the season in Chile and Paraguay.


Morbidity and mortality

In Argentina, ILI activity was above the alert threshold from late April to mid-to-late June, peaking at the beginning of June, with a cumulative ILI rate through midOctober (2182 per 100 000) which was slightly higher than in 2015 but equivalent to that of the previous 5-year period. Surveillance of SARI showed that case reports were at warning and outbreak level thresholds for most of the year. The cumulative SARI rate through September (140 per 100 000) was 23% higher than in 2015, and 49% higher than in the previous 5-year period. Both ILI and SARI surveillance showed that the notification rate was highest in children aged <5 years.

In Chile, SARI surveillance data during January–October 2016 indicated that of 1711 hospitalized SARI cases with a respiratory virus detection, 493 (28.8%) tested positive for influenza A and 773 (45.2%) tested positive for respiratory syncytial virus (RSV). Among the SARI cases, 287 (16.8%) required admission to an ICU. A total of 63 deaths were reported through SARI surveillance, of which 47 (74.6%) were due to influenza, with 37 (78.7%) of the influenza-related deaths attributed to influenza A(H1N1)pdm09. The SARI mortality rate was higher than in 2015 (35 cases) and 2013 (40 cases) when influenza A(H1N1)pdm09 also predominated. ILI activity reached the alert threshold in midto-late July and peaked in mid-August. The pattern of hospital emergency visits for ILI was similar to that observed in 2015. In Paraguay, 5953 SARI cases were hospitalized at sentinel surveillance sites from January to September 2016. The age groups most affected were children aged <2 years and adults aged >60 years, a pattern similar to that of the previous year. Among these cases, 377 deaths were recorded, 103 of which were due to respiratory viruses; in particular, 37 tested positive for influenza A(H1N1)pdm09, 36 for respiratory syncytial virus, and 12 for influenza B. In comparison with 2015 (4909 cases and 253 deaths), the cumulative number of SARI cases and the proportion of fatal cases were higher this year. ILI activity reached the alert threshold in late April and remained at this level during the entire season; the peak occurred in mid-to-late June. In comparison with previous years, ILI activity was higher this year and peaked earlier. Among all hospitalizations in Uruguay, the proportion of SARI hospitalizations was approximately 2.8%, similar to that seen in 2015 (2.5%). This proportion peaked in mid-June, earlier than in previous years. The proportion of SARI hospitalizations admitted to the ICU peaked in mid-to-late May; in comparison with previous years, the peak occurred earlier and was higher (32.8% in 2016 compared with 14% in 2014 and 10% in 2015). SARI surveillance indicated that 262 people had a positive respiratory virus detection from January to October 2016. Of these cases, 66 (25.2%) tested positive for influenza A and 170 (64.9%) tested positive for RSV.


South Africa

Timing of the season and circulating influenza viruses

In 2016, South Africa reported sporadic detections of influenza from January to May, the majority of which (24/27, 89%) were influenza B viruses. The influenza season started in the second week of May when the influenza virus detection rate rose above 10% in the Viral Watch sentinel surveillance programme (outpatients attending private practitioners). The average onset of the South African influenza season, calculated using Viral Watch data from 11 previous years, was around the last week of May. For the past 2 years the season has started earlier than this. The detection rate rose above 50% at the end of June and remained at that level until mid-September, after which it dropped rapidly, with the season ending in early October. The season started with influenza B viruses predominating, accounting for the majority of detections until mid-July. After this time, influenza A(H3N2) accounted for the majority of influenza virus detections. The proportion of influenza A(H1N1)pdm09 virus detections also increased during the second half of the season. Of the total samples tested in 3 surveillance programmes (nationwide pneumonia surveillance, ILI in public clinics, and ILI in the private sector) between 1 January and 31 October 2016, 978 of 5418 (18.1%) were positive for influenza. The majority of influenza positive specimens in all 3 programmes were influenza B viruses (427/978, 43.7%), followed by influenza A(H3N2) viruses (307/978, 31.4%), while influenza A(H1N1)pdm09 viruses accounted for the remaining 244/978 (24.9%). In 2014 and 2015, influenza B accounted for 20–21% of all influenza detections. The last time influenza B accounted for the majority of detections in South Africa was in 2010 and, prior to that, in 1999.


Morbidity and mortality

A programme collecting anonymous outpatient consultations and hospitalization data from one private hospital group in 7 provinces, showed that the proportions of patients with a diagnosis of P & I were highest in early August for inpatients and mid-August for outpatients. The proportion of inpatients with P & I peaked at 5.5% (historical 5-year range: 4.9–6.1%) whereas for outpatients the peak was 2.6% (range: 2.3–3.1%).



Timing of the season and circulating influenza viruses

In Australia, influenza detections began to increase from baseline levels in July and peaked in early September. Influenza activity decreased to low levels by October 2016. The onset of the 2016 influenza season was later than the 2015 season, but similar to the 2014 season. Influenza A viruses predominated, accounting for 90% of laboratory-confirmed notified cases [73% A(not subtyped), 11% A(H3N2), and 6% A(H1N1) pdm09]. Influenza B viruses accounted for 10% of notifications, in contrast to the 2015 season, when around 60% of notified influenza cases were due to influenza B viruses.9 Of 999 ILI specimens tested by the Australian Sentinel Practices Research Network (ASPREN) as of mid-September 2016, 29.8% were positive for influenza; 25.7% were due to influenza A (4.5% A(not subtyped), 16.9% A(H3N2) and 4.3% A(H1N1)pdm09] and 4.1% due to influenza B.
Morbidity and mortality Between January and the end of October 2016, a total of 82 987 laboratory-confirmed cases of influenza were reported to the National Notifiable Diseases Surveillance System (NNDSS), higher than in all other previous years with the exception of 2015. The rate of influenza notifications was highest in adults aged ≥75 years, followed by a smaller peak in children aged <5 years. The peak rate of ILI consultations in general practice in 2016 (15.0 per 1000 consultations in mid-August) was lower than that in the 2015 season (19.3 per 1000 consultations in late August). ILI activity at general practices followed a pattern seen in previous years, with the exception of 2012 when the peak in activity occurred approximately one month earlier. Hospital admissions due to influenza peaked in early September. Fewer admissions were reported in the peak week in 2016 than in the previous 2 seasons. From early April to the end of October, the Influenza Complications Alert Network (FluCAN) reported that 9.5% of patients admitted with confirmed influenza were admitted directly to the ICU, a rate consistent with previous moderate years. A higher proportion of Aboriginal and Torres Strait Island peoples (16%) and pregnant women (15%) were admitted directly to the ICU. Consistent with detections in other systems, the majority of influenza admissions to sentinel hospitals in 2016 were due to influenza A infections (93%).


New Zealand

Timing of the season and circulating influenza viruses

In New Zealand, influenza detections in 2016 began to increase from baseline levels in late June, peaking in late August. However, as of late September, there was still an active outbreak of influenza virus occurring, as evidenced by the lack of consistent decline since the first week after the peak. With regards to temporality, the timing of this year’s influenza season in New Zealand was similar to previous years, both in onset and duration.

From January to late August 2016, 1813 influenza viruses were detected through both sentinel ILI and SARI and laboratory-based hospital inpatient and outpatient surveillance. Of those influenza viruses, 1181 (65%) were identified as A(H3N2), 329 (18%) as A(H1N1) pdm09, 18 (1%) as influenza influenza B/Yamagata lineage, and 13 (1%) as influenza B/Victoria lineage; 72 (4%) were identified as influenza B (lineage not determined) typed, and 200 (11%) were a non-subtyped influenza A virus.

Morbidity and mortality

In New Zealand, the burden of the 2016 season was much lower than the average from previous years, with the rate of ILI consultations never above the seasonal threshold for the whole season, similar to the season in 2000. The ILI consultation rate peaked during midAugust at 22.2 per 100 000 population, nearly 7 times less than the rate in 2015 (150 per 100 000). Through the Southern Hemisphere Influenza and Vaccine Effectiveness Research & Surveillance (SHIVERS) project, SARI surveillance was performed for the Auckland and Counties Manukau District Health Boards, with a population of 905 622. From late April to late September 2016, of the 936 patients admitted that met the SARI case definition and were tested, 15.7% (147) were positive for an influenza virus. Among the hospitalized SARI cases, SARI-associated influenza incidence was highest among infants aged <1 year (74.0 per 100 000), followed by the age group of ≥80 years (81.1 per 100 000).


Review of influenza activity in the tropics  and subtropics in 2015–2016

Patterns vary in human influenza transmission in regions of the tropics and subtropics. However, a distinct seasonality pattern with 1 or 2 peaks can be established for most tropical and subtropical countries. Countries near the equator often have significant year-round activity. Moreover, countries with a large latitudinal spread have subnational variation in their seasonality pattern.10, 11


Central America and the Caribbean

Timing of the season and circulating influenza viruses

Influenza activity in most countries during 2015–2016 followed historical patterns with a few exceptions. In Honduras and Nicaragua the season peaked later than usual in December, and in Jamaica earlier than usual in February. In Martinique and Puerto Rico activity peaked in February. In El Salvador, a primary period of activity occurred from mid-April to mid-June with almost exclusively A(H1N1)pdm09 detections; a secondary period with influenza A(H3N2) viruses occurred in October and November 2015. Continuous, variable influenza activity was reported in Cuba during this reporting period. Throughout the region, influenza A(H1N1) pdm09 was reported as the predominant virus for most of this reporting period. Influenza A(H3N2) virus activity was reported in late 2015 in Costa Rica, Cuba, El Salvador, Haiti, Honduras and Panama; influenza B virus activity throughout the period was also reported, especially in Barbados, El Salvador, Dominican Republic, Guatemala and Honduras. In Mexico, influenza A(H3N2) and influenza B viruses co-circulated with influenza A(H1N1)pdm09 throughout the season from January to April.


Morbidity and mortality

In Costa Rica, peak SARI activity was observed at the end of 2015 with 8% of cumulative all-cause hospitalizations attributable to SARI. SARI cases accounted for 58% of cumulative ICU admissions, above the 2013–2014 peak, and 8.5% of all deaths. In Jamaica, activity was similar to previous years, with peak activity occurring in March 2016; 4.3% of total hospitalizations at this time were attributed to SARI. In Mexico, end-of season cumulative acute respiratory infection (ARI) and pneumonia/ bronchopneumonia counts were lower compared with the previous season. However, 6.4% of influenza cases (614/9580) were reportedly fatal – an increase from 2.9% in the 2014–2015 season. The majority (472/614) of influenza-related deaths were due to influenza A(H1N1) pdm09.


Tropical South America

Timing of the season and circulating influenza viruses

In Brazil and in the Plurinational State of Bolivia, the timing of influenza activity followed typical patterns with one peak in activity. Influenza activity in Colombia was year-round with 2 peaks in January and late-June. In Ecuador, the timing of influenza activity differed from previous years, with one period of activity as opposed to 2 distinct periods. Activity in Peru peaked in April, earlier than in previous years. The predominant virus detected in tropical South America was influenza A(H1N1)pdm09. To a lesser extent, influenza B circulated throughout the region, notably in Peru. During late October, the tropical South American countries were still reporting influenza detections.


Morbidity and mortality

In Brazil, from January to October 2016, 50 558 SARI cases were reported – a fourfold increase from 2015 (11 092 cases). Influenza was confirmed in 11 446 (28.3%) of the SARI cases sampled (40 431). The median age of SARI classified cases was 39 (range 0–110 years). Among the notified SARI cases, 6511 deaths were reported, of which 32.3% were confirmed as positive for influenza, and 70.2% were reported to have underlying risk factors. The southwest region of Brazil reported the majority of SARI-attributed deaths, with the highest concentration in Sao Paulo (39.1%). The median age for influenza deaths was 53 (range 0–99 years).12 In Colombia, from January to October 2016, 5521 cases of ILI-SARI were reported by sentinel surveillance. In the same period, a total of 488 deaths of children aged <5 years due to ARI were reported. The percentage of deaths specifically due to influenza is not known. In Ecuador, the proportion of SARI cases per 100 hospitalized patients peaked at 9.22 in mid-May, higher than the past 2 seasons.


Tropical southern Asia

Timing of the season and circulating influenza viruses

In late 2015, an increase of influenza A(H1N1)pdm09 detections was reported from the Islamic Republic of Iran. In early 2016, an increase in predominantly influenza A(H3N2) detections was reported from Nepal and Sri Lanka and A(H1N1)pdm09 from Pakistan. For the remainder of 2016, influenza detections were generally low except for a small peak of activity in May and June due to activity in Bangladesh, and July and August due to activity in Bhutan, Nepal and Sri Lanka. Influenza A and B viruses all co-circulated in the region in 2016. India reported very little activity in 2016, in sharp contrast with 2015 when there was a peak of influenza activity (over 30% positivity) in February and March of 2015.


Tropical southeast Asia

Timing of the season and circulating influenza viruses

In Cambodia, influenza activity peaked in early September 2016, slightly later than in previous years. In southern China, as in previous seasons, influenza activity peaked in late-March 2016, and increased again in September. Influenza activity peaked in the Philippines in mid-July, but there was also a period of increased activity from January through to March. Influenza activity peaked in April in Indonesia, May in Viet Nam and September in Lao People’s Democratic Republic, following previous patterns where one seasonal peak is typical. Influenza activity in 2015–2016 also followed previous trends in countries such as Singapore and Thailand where 2 peaks per year are typical. In Malaysia, influenza activity usually occurs year-round with no seasonality. An apparent increase in activity in 2016 was likely due to a change in the influenza surveillance system.13 Influenza A(H1N1)pdm09, A(H3N2) and B viruses co-circulated in most of the countries in the region during this reporting period, in contrast to the previous year when influenza A(H3N2) was predominant. Influenza A(H3N2) detections were commonly detected in 2016 only in Singapore. Of the influenza B viruses characterized, the Victoria lineage was predominant in the region in 2016; Yamagata lineage viruses were detected only at the end of 2015, especially in Cambodia and southern China.


Morbidity and mortality

Peaks in influenza activity were slightly greater than in previous years in Cambodia, Singapore and Thailand. In Singapore, the average daily number of patients seeking treatment in the polyclinics for ARI peaked in February 2016 at the same magnitude as the previous season. In southern China, the peak in percentage of outpatient or emergency visits for ILI at national sentinel hospitals in February and March was higher than that in the previous season. A second increase in ILI activity generally seen in June and July was much lower than in the past few seasons.


Central African tropical region

Timing of the season and circulating influenza viruses

Reports of influenza detections from Ghana, followed by Côte d’Ivoire, accounted for most of the activity reported in western Africa. Other countries (Burkina Faso, Niger, Nigeria, Senegal and Togo) reported only sporadic activity of mostly influenza A viruses. In this region, influenza detections were predominantly A(H1N1)pdm09 and A(H3N2) from the beginning of 2016 until mid-July when influenza B viruses of both lineages accounted for the majority of detections. In eastern Africa, influenza detections from Ethiopia, Madagascar and Uganda accounted for most of the reported activity in the region. Sporadic activity was reported from several other countries (Kenya, Mozambique, the United Republic of Tanzania and Zambia). Influenza A and B viruses generally co-circulated throughout the reporting period in this region. The timing of peaks in activity generally followed historical patterns of seasonality. In middle Africa, Cameroon, the Central African Republic and the Democratic Republic of the Congo (DRC) reported influenza activity following historical patterns of timing. Influenza A(H3N2) was predominant in Cameroon and the Central African Republic while influenza A(H1N1)pdm09 and B/Yamagata lineage viruses were co-circulating in the DRC.


Antigenic characterization, genetic testing and antiviral susceptibility

Information on antigenic characterization, genetic testing and antiviral susceptibility can be found online at the archive of WHO recommendations on the composition of influenza virus vaccines.14


Vaccine effectiveness15

Final estimates for northern hemisphere 2015–2016

In general, final estimates for vaccine effectiveness (VE) from countries in the northern hemisphere indicated  that vaccines provided moderate protection against infection for all subtypes. In cases where data were available, final VE estimates for influenza A(H1N1) pdm09 were slightly lower than the previous season; for influenza A(H3N2) they were slightly higher compared to last season when there were antigenic differences between the circulating influenza A(H3N2) viruses and the vaccine virus. VE estimates for influenza B virus infection varied but were similar to 2014–2015 estimates in the UK16 and the USA.17 The VE estimates included below were obtained from adjusted odds ratios, unless noted otherwise. Where possible, the publications from which the estimates are derived are cited. Estimates for all types/subtypes were available from Canada, the USA and the UK. In Canada, overall estimates for all types/subtypes were 46% (95% CI: 32–57%).18 In the USA, estimates from the US Flu Vaccine Effectiveness Network19 were 47% (95% CI: 39–53%) for all ages, and 45% (95% CI:10–66%) for elderly adults. The highest VE estimate was 64% (95% CI: 44–77%) in the 9–17 year-old cohort and lowest in the 50–64 year-old cohort at 23% (95% CI: -3–43%). In the UK, the overall estimate for all subtypes and all ages was 52% (95% CI: 41–62%).

In the North American countries, outpatient VE estimates for influenza A(H1N1)pdm09 were very similar, with 43% (95% CI: 25–57%) in Canada18 and 41% (95% CI: 31–51%) in the USA.20 The UK was slightly higher at 55% (95% CI: 42–65%).21 In the USA, for children aged 6 months to 8 years, the VE estimate was 47% (95% CI: 25–63%), and for children aged 9–17 years, 62% (95% CI: 21–82%). In Canada, the VE estimate for A(H1N1) pdm09 was highest in children aged 1–19 years at 67% (95% CI: 31–84%) and substantially lower in adults aged 20–64 years at 35% (95% CI: 10–52%); for elderly adults aged ≥65 years it was non-significant. In the UK, the highest VE estimate was in the adult cohort (18–44 years) at 60% (95% CI: 36–75%); and lowest in the youngest cohort (2–17 years) at 49% (95% CI: 9–71%). For elderly adults aged ≥65 years, the VE estimates ranged from 68% (95% CI: 39–83%) in the USA to 56% (95% CI: 7–79%) in the UK. Most influenza A(H1N1)pdm09 viruses collected from the patients in these studies belonged to clade 6B.1 and were closely related to the vaccine virus.

Final VE estimates for influenza A(H3N2) were available from 3 studies of outpatient data in Canada and the USA and ranged from 45% (95% CI: 9–66%) to 55% (unadjusted, 95% CI: 9–78%), for all age groups. One age-specific VE estimate for children aged <18 years in the USA was 59% (95% CI: 27–77%), from data collected by the Naval Health Research Center.
There was large variability in final VE estimates for influenza B viruses available from several countries in the northern hemisphere. Outpatient studies in North America and Europe showed that final VE estimates for all ages and all influenza B viruses ranged from 42% (95% CI: 8–63%) in Spain, to 50% (95% CI: 31–63%) in Canada, to 54% (95% CI: 33–69%) in the UK and 54% (95% CI: 42–63%) in the USA. For children aged <18 years, significant estimates ranged from 59% (95% CI: 43–71%) in the USA, to 77% (95% CI: 42–91%) for the age group 2–17 years in the UK. For the B/lineagespecific estimates, there was good agreement between significant VE estimates from Canada, UK and the USA, indicating moderate protection (data not shown).


Interim estimates for the southern hemisphere 2016

In general, interim VE estimates for influenza A(H1N1) pdm09, A(H3N2) and B viruses were moderate during the 2016 season; however the confidence intervals for most estimates were wide and often included zero, making interpretation difficult. Wide confidence intervals are often due to small study sample sizes or low vaccine coverage in a population. Few age-group specific interim estimates were available.

Interim VE estimates were assessed in several countries in the southern hemisphere with data obtained early in the season (March–September, depending on the study). For outpatients in New Zealand, estimates for influenza A(H1N1)pdm09 were 57% (95% CI: -21–85%), 77% (95% CI: 39–91%) for adults in Australia, and 81% (95% CI: 55–92%) for patients of all age groups in Australia. Restricting Latin American data to the main data contributor, Chile, the VE estimate against influenza A(H1N1)pdm09-associated SARI was 35% (95% CI: -4–60%) among individuals ≥65 years old, and 25% (95% CI: -60–65%) among children aged 6 months to 5 years. For influenza A(H3N2), VE estimates in Australia were 40% (95% CI: 10–60%) for all outpatients, and 36% (95% CI: 0–59%) for adults aged 18–64 years. Most A(H3N2) viruses fell in clade 3c.2a. The clade-specific VE estimate was 41% (95% CI: -5–67%) for all patients.

For influenza B viruses, interim results were only available from Australia and Latin America (Chile, Colombia and Paraguay). For outpatients in Australia, the interim VE estimate was 27% (95% CI: 44–63%) and for inpatients in the 3 Latin American countries, the interim VE estimate was 21% (95% CI: -47–58%).



Most countries experienced influenza activity during the periods normally described. A few exceptions to this were: Chile, the Democratic People’s Republic of Korea, Japan, Paraguay and the USA, where activity was later than typically reported, and Argentina, South Africa and Uruguay where activity was earlier. Influenza A(H1N1)pdm09 viruses were predominant from October 2015 to July 2016. From January to midMay 2016 this was mainly driven by influenza A(H1N1) pdm09 virus activity in the northern temperate countries with the exception of northern China, where influenza A(H3N2) and B viruses predominated. From July onwards, the predominance of influenza A(H3N2) was driven mainly by activity in Oceania and South Africa, and by activity in temperate countries in South America by the end of August. All seasonal influenza viruses circulated in tropical areas. Overall there was no substantial genetic or antigenic diversification of the circulating viruses during this reporting period and the majority of the characterized viruses were similar to the strains included in the influenza vaccines. Influenza A(H1N1)pdm09 viruses differed slightly, antigenically, from human post-vaccination serum, and influenza A(H3N2) viruses remained difficult to characterize antigenically. Vaccine effectiveness studies generally indicated moderate protection against medically-attended influenza infection.

The intensity of transmission was variable. Compared with previous years, higher transmission was reported mainly from countries in temperate South America, western and south-eastern Asia, eastern Europe transmission zones, and northern China. Influenza activity was notably low in India and New Zealand.

As has been observeded in previous seasons3, 4, 8, 22, 23 in seasons where influenza A(H1N1)pdm09 viruses predominate, influenza-associated morbidity (ICU admissions) tends to be increased in adults aged <65 years compared with seasons when A(H3N2) viruses predominate. Reports of severe influenza outbreaks and clusters of severe disease and deaths associated with influenza were also reported from some countries in multiple regions. Assessment of virological and epidemiological characteristics is key to understanding the severity of these events. Comparison of epidemiological data (ILI, SARI, pneumonia and death rates) with historical data allows for an informed determination if an event is unusual. However, in many countries these baseline data are still lacking. In addition, rapid characterization of the virus in the national influenza centres and in the WHO Collaborating Centres for influenza of the WHO Global Influenza Surveillance and Response System (GISRS), allows for identification of potential changes in the virus as well as judgement on current vaccine coverage and susceptibility to antivirals.



Global Influenza Programme, Infectious Hazard Management (IHM), Health Emergencies Cluster, World Health Organization (Corresponding author: Katelijn Vandemaele,

The WHO Global Influenza Surveillance and Response System (GISRS), the Global Influenza Vaccine Effectiveness (GIVE) Collaboration and the WHO Regional offices



  1. FluNet can be accessed online at
  2. FluID can be accessed online at
  3. Davlin SL, Blanton L, Kniss K, Mustaquim D, Smith S, Kramer N et al; et al. Influenza Activity — United States, 2015–16 Season and Composition of the 2016–17  Influenza Vaccine. MMWR Morb Mortal Wkly Rep 2016; 65:567–575. Available at: ; accessed November 2016.
  4. FluWatch report: 14 to 27 August 2016 (weeks 33–34), Canada. Available at:  accessed November 2016.
  5. Flu News Europe | Home. Available at:; accessed December 2016.
  6. 6 Influenza season summaries [Internet]. WHO Regional Office for Europe. Available at:,-other-influenza-surveillance-bulletinsand-seasonal-summaries/influenza-season-summaries2; accessed November 2016.
  7. Armenia, Azerbaijan, Cyprus, Georgia, Israel and Turkey. See:
  8. Summarising the 2015–2016 influenza season in Europe. ECDC. Available at: ecdc.; accessed November 2016.
  9. See No. 48, 2015, pp. 645–660
  10. Alonso WJ, Yu C, Viboud C, Richard SA, Schuck-Paim C, Simonsen, L et al. A global map of hemispheric influenza vaccine recommendations based on local patterns of viral circulation. 2015. Sci. Rep. 5, 17214; doi: 10.1038/srep17214.
  11. Hirve S, Newman LP, Paget J, Azziz-Baumgartner E, Fitzner J, Bhat N, et al. Influenza Seasonality in the Tropics and Subtropics – When to Vaccinate? 2016. PLoS ONE 11(4): e0153003. doi:10.1371/journal.pone.0153003
  12. Influenza update EW 40 2016. Health Portal, Ministry of Health, Brazil. Available at; accessed November 2016.
  13. Kamaludin, F. Epidemiology of Influenza in Malaysia & Surveillance System [presentation]. APACI Influenza Workshop - Kuala Lumpur, Malaysia 2016. Available at:; accessed November 2016.
  14. Archive of WHO recommendations on the composition of influenza virus vaccines [website]. WHO Global Influenza Programme. Available at:; accessed November 2016.
  15. Vaccine effectiveness estimates for seasonal influenza vaccines [website]. WHO Global Influenza Programme. Available at: recommendations/201502_qanda_vaccineeffectiveness.pdf?ua=1; accessed
    November 2016.
  16. Pebody R, Warburton F, Andrews N, et al. Effectiveness of seasonal influenza vaccine in preventing laboratory-confirmed influenza in primary care in the United Kingdom: 2014/15 end of season results. Euro Surveill. 2015;20(36):pii=30013. DOI:
  17. Zimmerman, Richard K, Nowalk, Mary Patricia, Chung, Jessie et al. 2014–2015 Influenza Vaccine Effectiveness in the United States by Vaccine Type, 2016; Clinical Infectious Diseases. doi = 10.1093/cid/ciw635.
  18. Verbal communication, D Skowronski, BC Centre for Disease Control (BCCDC).
  19. Estimates from the USA are derived from data gathered by the US Flu VE Network, unless where specified otherwise.
  20. See
  21. See
  22. Epperson S, Blanton L, Kniss K, Mustaquim D, Steffens C, Wallis T, et al. Influenza Activity — United States, 2013–14 Season and Composition of the 2014–15 Influenza Vaccines. MMWR Morb Mortal Wkly Rep 2014; 63(22);483-490. Available at; accessed November 2016.
  23. FluWatch report: 16 to 29 August 2016 (weeks 33–34), Canada. Available at; accessed November 2016.


Keywords: WHO; Worldwide; Updates; Seasonal Influenza.