Rapid risk assessment during the early weeks of the 2015‐2016 influenza season in Ukraine

Background Several eastern European countries reported a severe influenza season to the World Health Organization (WHO) during late 2015. A country‐specific rapid risk assessment for Ukraine was conducted to assess the season's severity and inform public health action. Methods The exposure and hazard were assessed using acute respiratory infection (ARI), severe acute respiratory infection (SARI), laboratory surveillance, virological and vaccine data from weeks 40/2015 to 7/2016 with comparison to 4 previous seasons to describe the influenza season start (5‐week consecutive increase in ARI or SARI), predominant virus types, geographical spread and affected age groups. Results The exposure was characterised by an earlier and steeper increase in SARI (week 1/2016) and ARI (week 2/2016) compared to the previous 4 seasons. Transmission was across Ukraine with an increase in ARI and SARI cases aged 30‐64 years compared to 2014/15. Laboratory‐confirmed deaths increased from 11 in 2014/2015 to 342 in 2015/2016; the majority were 30‐64 years old and unvaccinated; and 63.5% had underlying conditions. Total population vaccination coverage was 0.3%. The hazard assessment found influenza virus A(H1N1)pdm09 accounted for >95% of viruses detected. Ukrainian virus strains (n = 62) were antigenically similar to vaccine strains and susceptible to neuraminidase inhibitors. Conclusions The first weeks of the 2015/16 influenza season were more severe than previous seasons, with an earlier and steeper increase in severe cases and deaths, particularly in younger adults. Influenza A(H1N1)pdm09 was the predominant strain and was closely related to the seasonal vaccine strain with no evidence of resistance to antiviral drugs.


| INTRODUCTION
Several eastern European countries alerted the World Health Organization (WHO) Regional Office for Europe in late 2015 about a potentially severe influenza season. In response, a rapid risk assessment for the whole WHO European Region was conducted and published in February 2016. 1,2 There was a high level of concern and media interest regarding influenza in Ukraine, including speculation that the 2009 pandemic influenza virus A(H1N1)pdm09 had become more virulent. Ukraine is a lower-middle income country in eastern Europe with an estimated population of 45 million. 3 There has been conflict in eastern Ukraine since 2014, with an estimated 1.5 million internally displaced people within Ukraine.
As part of a WHO Global Outbreak Alert and Response Network (GOARN) response requested by the Ministry of Health of Ukraine, a country-specific risk assessment for Ukraine was conducted to ascertain the severity of the influenza season, describe the first affected regions and assess the potential impact of the season in order to inform public health action and risk communication for the current and future seasons.

| METHODS
This rapid risk assessment was based on the WHO guidelines on acute public health events. 4

| Data sources
The Ukrainian Centre for Disease Control and Monitoring of the Ministry of Health of Ukraine (UCDC) has conducted universal acute respiratory infection (ARI) surveillance since 1986. 5   Comparisons of characteristics of cases between seasons were calculated with chi-squared or Fisher's exact test (if values were fewer than 5), 10 with 1-or 2-sided tests as appropriate. Analysis was carried out using microsoft excel (2010), STATA release 12 11 and r (version 3.2.4). 12

| Universal ARI surveillance
Cases of ARI increased from week 2/2016 onwards, earlier than the previous 4 seasons ( Figure 1A). Furthermore, the increase in incidence of ARI was much steeper, peaking in week 4 compared to weeks [8][9][10][11] in the previous seasons. The incidence was 1036.6 per 100 000 population during the peak week 4/2016 compared to a peak week of 700.5 cases per 100 000 in week 9/2015. Up to week 7/2016, 4 035 623 ARI cases had been reported, accounting for approximately 9% of the Ukrainian population.
Total ARI rates up to week 7/2016 showed that Kyiv and the surrounding region of Kyiv had the highest incidence ( Figure 2). The region of Kharkov (north-east) had the lowest incidence followed by Zakarpattya and Ternopil (west). The incidence in all regions was higher than the same period for 2014/15 (data not shown).
The Vinnytsia region (south-west) started to show an increase in ARI from week 51/2015, followed by the Donetsk (east), Odessa   Of those that were positive, only 34.4% were influenza A (all untyped) and 65.6% were influenza B.

| Vaccine coverage
In 2015/16, an estimated 0.3% of the overall Ukraine population was vaccinated with the season's influenza vaccine. Vaccination was carried out mainly by local governments and varied by region from 0.01% in Zhytomyr up to 1.2% in the Dnipropetrovsk area.  Although an east-to-west pattern of transmission was noted across the European Region this season, 1 We did not conduct a full context assessment as part of this risk assessment and it was difficult to assess the role of the current conflict in eastern Ukraine on the severity of the influenza season. The surveillance systems used in this risk assessment did not cover the

| Limitations
Characteristics of influenza-positive SARI cases, and deaths were collected using a tick box form with only positives recorded. When analysing these data, the assumption was made that if the condition was not indicated, then it was not present, and these results should therefore be interpreted with care. Further analysis is required to ensure that case definitions and sampling strategies were correctly applied throughout this epidemic; however, to our knowledge, there was no change in the surveillance systems between 2014/15 and 2015/16. We were unable to ascertain the data completeness of risk factors by clinics to do a full risk factor analysis or further statistical analysis. The small numbers of deaths in the 2014/15 season precluded statistical comparisons of characteristics and risk factors for mortality. We also acknowledge there are different risk assessment methodologies and this approach did not provide an overall quantification of the level of risk.

| CONCLUSION AND RECOMMENDATIONS
The Ukraine 2015/16 influenza season was more severe than previous seasons. It was characterised by an earlier steep increase in severe cases and deaths than in the previous year, particularly in the younger adult population. The predominant circulating strain was A(H1N1) pdm09, which was closely related to the seasonal vaccine strains.
There was no evidence of resistance to influenza antiviral drugs.
However, vaccination coverage in Ukraine was low, which may have contributed to the higher number of severe cases.
Recommendations from this risk assessment were made and communicated at the time to support appropriate control measures. We