The community impact of the 2009 influenza pandemic in the WHO European region: a comparison with historical seasonal data from 28 countries

Machine learning-derived prediction of in-hospital mortality in patients with severe acute respiratory infection: analysis of claims data from the German-wide Helios hospital network

BACKGROUND

Severe acute respiratory infections (SARI) are the most common infectious causes of death. Previous work regarding mortality prediction models for SARI using machine learning (ML) algorithms that can be useful for both individual risk stratification and quality of care assessment is scarce. We aimed to develop reliable models for mortality prediction in SARI patients utilizing ML algorithms and compare its performances with a classic regression analysis approach.

METHODS

Administrative data (dataset randomly split 75%/25% for model training/testing) from years 2016-2019 of 86 German Helios hospitals was retrospectively analyzed. Inpatient SARI cases were defined by ICD-codes J09-J22. Three ML algorithms were evaluated and its performance compared to generalized linear models (GLM) by computing receiver operating characteristic area under the curve (AUC) and area under the precision-recall curve (AUPRC).

RESULTS

The dataset contained 241,988 inpatient SARI cases (75 years or older: 49%; male 56.2%). In-hospital mortality was 11.6%. AUC and AUPRC in the testing dataset were 0.83 and 0.372 for GLM, 0.831 and 0.384 for random forest (RF), 0.834 and 0.382 for single layer neural network (NNET) and 0.834 and 0.389 for extreme gradient boosting (XGBoost). Statistical comparison of ROC AUCs revealed a better performance of NNET and XGBoost as compared to GLM.

CONCLUSION

ML algorithms for predicting in-hospital mortality were trained and tested on a large real-world administrative dataset of SARI patients and showed good discriminatory performances. Broad application of our models in clinical routine practice can contribute to patients' risk assessment and quality management.

Age-Specific Seasonal Influenza Vaccine Effectiveness against Different Influenza Subtypes in the Hospitalized Population in Lithuania during the 2015-2019 Influenza Seasons

Background: Continuous monitoring of seasonal influenza vaccine effectiveness (SIVE) is needed due to the changing nature of influenza viruses and it supports the decision on the annual update of vaccine composition. Age-specific SIVE was evaluated against different influenza subtypes in the hospitalized population in Lithuania during four influenza seasons. Methods: A test-negative case-control study design was used. SIVE and its 95% confidence intervals (95% CI) were calculated as (1 – odds ratio (OR)) × 100%. Results: Adjusted SIVE in 18–64-year-old individuals against influenza A, A(H1N1)pdm09 and B/Yamagata were 78.0% (95% CI: 1.7; 95.1%), 88.6% (95% CI: −47.4; 99.1%), and 76.8% (95% CI: −109.9; 97.4%), respectively. Adjusted SIVE in individuals aged 65 years and older against influenza A, influenza B, and B/Yamagata were 22.6% (95% CI: −36.5; 56.1%), 75.3% (95% CI: 12.2; 93.1%) and 73.1% (95% CI: 3.2; 92.5%), respectively. Unadjusted SIVE against influenza A(H3N2) among 18–64-year-old patients was 44.8% (95% CI: −171.0; 88.8%) and among those aged 65 years and older was 5.0% (95% CI: −74.5; 48.3%). Conclusions: Point estimates suggest high SIVE against influenza A in 18–64-year-old participants, and against influenza B and B/Yamagata in those 65 years old and older.

Analysis of multi-level spatial data reveals strong synchrony in seasonal influenza epidemics across Norway, Sweden, and Denmark

Population structure, spatial diffusion, and climatic conditions mediate the spatiotemporal spread of seasonal influenza in temperate regions. However, much of our knowledge of these dynamics stems from a few well-studied countries, such as the United States (US), and the extent to which this applies in different demographic and climatic environments is not fully understood. Using novel data from Norway, Sweden, and Denmark, we applied wavelet analysis and non-parametric spatial statistics to explore the spatiotemporal dynamics of influenza transmission at regional and international scales. We found the timing and amplitude of epidemics were highly synchronized both within and between countries, despite the geographical isolation of many areas in our study. Within Norway, this synchrony was most strongly modulated by population size, confirming previous findings that hierarchical spread between larger populations underlies seasonal influenza dynamics at regional levels. However, we found no such association when comparing across countries, suggesting that other factors become important at the international scale. Finally, to frame our results within a wider global context, we compared our findings from Norway to those from the US. After correcting for differences in geographic scale, we unexpectedly found higher levels of synchrony in Norway, despite its smaller population size. We hypothesize that this greater synchrony may be driven by more favorable and spatially uniform climatic conditions, although there are other likely factors we were unable to consider (such as reduced variation in school term times and differences in population movements). Overall, our results highlight the importance of comparing influenza spread at different spatial scales and across diverse geographic regions in order to better understand the complex mechanisms underlying disease dynamics.

The spatiotemporal characteristics of influenza A and B in the WHO European Region: can one define influenza transmission zones in Europe?

We aimed to assess the epidemiology and spatiotemporal patterns of influenza in the World Health Organization (WHO) European Region and evaluate the validity of partitioning the Region into five influenza transmission zones (ITZs) as proposed by the WHO. We used the FluNet database and included over 650,000 influenza cases from 2000 to 2015. We analysed the data by country and season (from July to the following June). We calculated the median proportion of cases caused by each virus type in a season, compared the timing of the primary peak between countries and used a range of cluster analysis methods to assess the degree of overlap between the WHO-defined and data-driven ITZs. Influenza A and B caused, respectively, a median of 83% and 17% cases in a season. There was a significant west-to-east and non-significant (p = 0.10) south-to-north gradient in the timing of influenza activity. Typically, influenza peaked in February and March; influenza A earlier than influenza B. Most countries in the WHO European Region would fit into two ITZs: ‘Western Europe’ and ‘Eastern Europe’; countries bordering Asia may be better placed into extra-European ITZs. Our findings have implications for the presentation of surveillance data and prevention and control measures in this large WHO Region.

Excess all-cause and influenza-attributable mortality in Europe, December 2016 to February 2017

Since December 2016, excess all-cause mortality was observed in many European countries, especially among people aged ≥ 65 years. We estimated all-cause and influenza-attributable mortality in 19 European countries/regions. Excess mortality was primarily explained by circulation of influenza virus A(H3N2). Cold weather snaps contributed in some countries. The pattern was similar to the last major influenza A(H3N2) season in 2014/15 in Europe, although starting earlier in line with the early influenza season start.

Seasonal influenza vaccine effectiveness against laboratory-confirmed influenza in 2015-2016: a hospital-based test-negative casecontrol study in Lithuania

OBJECTIVE

A case-control study was conducted to assess seasonal influenza vaccine effectiveness (SIVE) during the 2015-2016 influenza season.

METHODS

A study was performed in three departments in Lithuania between 1 December 2015 and 1 May 2016. Data on demographic and clinical characteristics including influenza vaccination status were collected from the patients recommended to receive the seasonal influenza vaccine. Influenza virus infection was confirmed by multiplex reverse transcription polymerase chain reaction (RT-PCR) .

RESULTS

Ninety-one (56.4%) of the 163 included subjects were ≥65 years old. Fifteen (9.2%) subjects were vaccinated against influenza at least 2 weeks before the onset of influenza symptoms, 12 of them were ≥65 years old. Of the 72 (44.2%) influenza virus positive cases, 65 (39.9%) were confirmed with influenza A (including 50 cases of influenza A(H1N1)pdm09), eight (4.9%) were confirmed with influenza B and one was a co-infection. Unadjusted SIVE against any influenza, influenza type A and influenza A(H1N1)pdm09 was 57% (95% CI -41% to 87%), 52% (95% CI -57% to 85%) and 70% (95% CI -43% to 94%) respectively.

CONCLUSION

Although SIVE estimates were not statistically significant the point estimates suggest moderate effectiveness against influenza type A.

Establishing an ICD-10 code based SARI-surveillance in Germany - description of the system and first results from five recent influenza seasons

BACKGROUND

Syndromic surveillance of severe acute respiratory infections (SARI) is important to assess seriousness of disease as recommended by WHO for influenza. In 2015 the Robert Koch Institute (RKI) started to collaborate with a private hospital network to develop a SARI surveillance system using case-based data on ICD-10 codes. This first-time description of the system shows its application to the analysis of five influenza seasons.

METHODS

Since week 40/2015, weekly updated anonymized data on discharged patients overall and on patients with respiratory illness including ICD-10 codes of primary and secondary diagnoses are transferred from the network data center to RKI. Retrospective datasets were also provided. Our descriptive analysis is based on data of 47 sentinel hospitals collected between weeks 1/2012 to 20/2016. We applied three different SARI case definitions (CD) based on ICD-10 codes for discharge diagnoses of respiratory tract infections (J09 - J22): basic CD (BCD), using only primary diagnoses; sensitive CD (SCD), using primary and secondary diagnoses; timely CD (TCD), using only primary diagnoses of patients hospitalized up to one week. We compared the CD with regard to severity, age distribution and timeliness and with results from the national primary care sentinel system.

RESULTS

The 47 sentinel hospitals covered 3.6% of patients discharged from all German hospitals in 2013. The SCD comprised 2.2 times patients as the BCD, and 3.6 times as many as the TCD. Time course of SARI cases corresponded well to results from primary care surveillance and influenza virus circulation. The patients fulfilling the TCD had been completely reported after 3 weeks, which was fastest among the CD. The proportion of SARI cases among patients was highest in the youngest age group of below 5-year-olds. However, the age group 60 years and above contributed most SARI cases. This was irrespective of the CD used.

CONCLUSIONS

In general, available data and the implemented reporting system are appropriate to provide timely and reliable information on SARI in inpatients in Germany. Our ICD-10-based approach proved to be useful for fulfilling requirements for SARI surveillance. The exploratory approach gave valuable insights in data structure and emphasized the advantages of different CD.

Improving the representativeness of influenza viruses shared within the WHO Global Influenza Surveillance and Response System

Background

Sharing influenza viruses within the WHO Global Influenza Surveillance and Response System is crucial for monitoring evolution of influenza viruses.

Objectives

Analysis of timeliness and geographic representativeness of viruses shared by National Influenza Centres (NICs) in the WHO European Region with the London WHO Collaborating Centre for Reference and Research on Influenza for the Northern Hemisphere's 2010–2011 and 2011–2012 influenza seasons.

Materials and methods

Data from NICs on influenza‐positive specimens shared with WHO CC London for the above‐mentioned influenza seasons were analyzed for timeliness of sharing with respect to the February deadline (31 January) for inclusion in the WHO consultations on the composition of influenza virus vaccines for the Northern Hemisphere and geographic representativeness.

Results

The 2010–2011 and 2011–2012 seasons were different in terms of the seasonal pattern, the timing of the epidemic, and the dominant virus. Consistent patterns of virus sharing across the seasons were observed. Approximately half the viruses collected before the deadline were not shared within the deadline; the average delay between date of specimen collection and shipment receipt was 3 and 1·5 months for the first and second season, respectively.

Conclusion

A baseline was provided for future work on enhancement of specimen sharing in the WHO European Region and improving the vaccine virus selection process. Greater insight into virus selection criteria applied by countries and the causes of delays in shipment are needed to understand the representativeness of viruses shared and to assess the importance of this for vaccine strain selection.

The potential risks and impact of the start of the 2015-2016 influenza season in the WHO European Region: a rapid risk assessment

BACKGROUND

Countries in the World Health Organization (WHO) European Region are reporting more severe influenza activity in the 2015-2016 season compared to previous seasons.

OBJECTIVES

To conduct a rapid risk assessment to provide interim information on the severity of the current influenza season METHODS: Using the WHO manual for rapid risk assessment of acute public health events and surveillance data available from Flu News Europe, an assessment of the current influenza season from 28 September 2015 (week 40/2015) up to 31 January 2016 (week 04/2016) was made compared with the 4 previous seasons.

RESULTS

The current influenza season started around week 51/2015 with higher influenza activity reported in eastern Europe compared to Western Europe. There is a strong predominance of influenza A(H1N1)pdm09 compared to previous seasons, but the virus is antigenically similar to the strain included in the seasonal influenza vaccine. Compared to the 2014/2015 season, there was a rapid increase in the number of severe cases in eastern European countries with the majority of such cases occurring among adults aged <65 years.

CONCLUSIONS

The current influenza season is characterised by an early start in eastern European countries, with indications of a more severe season. Currently circulating influenza A(H1N1)pdm09 viruses are similar antigenically to those included in the seasonal influenza vaccine and the vaccine is expected to be effective. Authorities should provide information to the public and health providers about the current influenza season, recommendations for treatment of severe disease and effective public health measures to prevent influenza transmission. This article is protected by copyright. All rights reserved.

Return of pandemic H1N1 influenza virus

Background

Influenza pandemics are usually caused by the re-assortment of several influenza viruses, results in the emergence of new influenza virus strains that can infect the entire population. These pandemic strains, as well as seasonal influenza viruses, are subjected to extensive antigenic change that has, so far, prevented the generation of a universal vaccine.

Methods

Samples of patients hospitalized due to infection with the pandemic H1N1 influenza virus (A(H1N1)pdm09) from 2009, when the virus first appeared, until 2013 were analyzed.

Results

While many patients were hospitalized in 2009 due to infection with the pandemic H1N1 influenza virus, only small percentages of patients were hospitalized later in 2010–2012. Surprisingly, however in 2012–2013, we noticed that the percentages of patients hospitalized due to the pandemic H1N1 influenza infection increased significantly. Moreover, the ages of hospitalized patients differed throughout this entire period (2009–2013) and pregnant women were especially vulnerable to the infection.

Conclusions

High percentages of patients (especially pregnant women) were hospitalized in 2013 due to the A(H1N1)pdm09 infection, which may have been enabled by an antigenic drift from those which circulated at the onset of the pandemic.

Electronic supplementary material

The online version of this article (doi:10.1186/s12879-014-0710-1) contains supplementary material, which is available to authorized users.

A prospective study on ambulatory care provided by primary care pediatricians during influenza season

Aim of this study was to obtain a picture of the nature of the primary care pediatricians' visits during a winter season. We investigated reasons for visits, diagnosis, and pattern of prescription in 284 children. The reason for visit was a planned visit in 54% of cases, a well-being examination in 26%, and an urgent visit for an acute problem in 20% of cases. Cough was the most common symptom reported (61%). The most common pediatricians' diagnosis was flu-like syndrome (47%). No disease was found by pediatrician in 27% of children with a symptom reported by caregivers. Antibiotics were prescribed in 25% of children, the vast majority of which affected by viral respiratory infections. The unjustified access to physician's visit may lead to a inappropriate prescription of drugs.

Relationships between A(H1N1)pdm09 influenza infection and infections with other respiratory viruses

BACKGROUND

A(H1N1)pdm09, a new influenza pandemic virus emerged in 2009. The A(H1N1)pdm09 infection had several unique characteristics which included rapid transmissibility and high morbidity in obese individuals, pregnant women and individuals suffering from chronic diseases.

OBJECTIVES

To study the relationships between A(H1N1)pdm09 influenza infection and infections with other respiratory viruses such as respiratory syncytial virus (RSV), human metapneumo virus (hMPV), adenovirus and seasonal influenza.

METHODS

Samples (nasopharyngeal swabs or aspirates) collected between 2007 until 2012 from patients of various ages that were hospitalized due to respiratory virus infections were analyzed for the presence of various respiratory viruses, using qRT-PCR.

RESULTS

In 2009-2010, when the pandemic influenza A(H1N1)pdm09 first appeared, two major infection peaks were noted and individuals of various ages were infected. Following the decline of the A(H1N1)pdm09 virus infection, the percentages of patients infected with adenovirus and hMPV increased, while infection frequency with RSV B and with seasonal influenza virus decreased. Furthermore, RSV infections were delayed and very few percentages of patients were co-infected with more than one virus. Interestingly, the A(H1N1)pdm09 virus lost its dominancy when it reappeared in the winter of 2010-2011, and at this time, only the incidence of RSV infections was affected by the A(H1N1)pdm09 virus.

CONCLUSIONS

The A(H1N1)pdm09 virus had distinct effects on other respiratory viruses when it first appeared versus later, when it evolved from being a pandemic to a seasonal virus.

Epidemiology of the 2009 influenza pandemic in Spain. The Spanish Influenza Surveillance System

In accordance with European Centre for Disease Prevention and Control recommendations, the Spanish Influenza Surveillance System (SISS) maintained its activity during the summer of 2009, and since July 2009 the pandemic virus activity was monitored by the SISS. In this paper, we describe the epidemiological and virological characteristics of the 2009 pandemic in the Spain through the SISS. Spain experienced a transmission of the new A(H1N1)pdm09 influenza virus during the summer of 2009, which gradually increased, resulting in the pandemic wave in early autumn of that year. The reproductive number R0, estimated during the growth phase of the pandemic wave (1.32; 95% confidence interval [95%CI], 1.29-1.36), showed a transmissibility comparable to preceding pandemics. There was an almost complete replacement of the previous seasonal A(H1N1) influenza virus by the pandemic virus A(H1N1)pdm09. The pandemic virus produced a greater burden of illness than seasonal influenza in children younger than 15 years old, while the incidence in those older than 64 years was lower compared with previous inter-pandemic seasons. Nevertheless, in Spain the 2009 pandemic was characterized as mild, considering the duration of the pandemic period and the influenza detection rate, both in the range of those observed in previous inter-pandemic seasons. Also, the case fatality ratio (CFR) was estimated at 0.58 deaths/1,000 confirmed ILI cases (95%CI, 0.52-0.64), in the range of the two previous pandemics of 1957 and 1968, with the highest CFR observed in the older than 64 years age group. In the 2009 pandemic there was a higher percentage of pandemic confirmed deaths in the younger ages, compared to seasonal influenza, since only 28% of the reported deaths occurred in persons aged 64 years and older.

Deviations in influenza seasonality: odd coincidence or obscure consequence?

In temperate regions, influenza typically arrives with the onset of colder weather. Seasonal waves travel over large spaces covering many climatic zones in a relatively short period of time. The precise mechanism for this striking seasonal pattern is still not well understood, and the interplay of factors that influence the spread of infection and the emergence of new strains is largely unknown. The study of influenza seasonality has been fraught with problems. One of these is the ever-shifting description of illness resulting from influenza and the use of both the historical definitions and new definitions based on actual isolation of the virus. The compilation of records describing influenza oscillations on a local and global scale is massive, but the value of these data is a function of the definitions used. In this review, we argue that observations of both seasonality and deviation from the expected pattern stem from the nature of this disease. Heterogeneity in seasonal patterns may arise from differences in the behaviour of specific strains, the emergence of a novel strain, or cross-protection from previously observed strains. Most likely, the seasonal patterns emerge from interactions of individual factors behaving as coupled resonators. We emphasize that both seasonality and deviations from it may merely be reflections of our inability to disentangle signal from noise, because of ambiguity in measurement and/or terminology. We conclude the review with suggestions for new promising and realistic directions with tangible consequences for the modelling of complex influenza dynamics in order to effectively control infection.

The role of the WHO Regional Office for Europe in response to seasonal, avian, and pandemic influenza

Between 2005 and 2011, the WHO Regional Office for Europe assisted the member states of the WHO European Region to prepare and respond to outbreaks of avian influenza H5N1, the 2009 pandemic, and to enhance their capacities for the prevention and control of seasonal influenza. It did this through conducting a combination of regional and subregional meetings and trainings, establishing a regional network for influenza surveillance, providing operational guidance for implementing influenza surveillance and strengthening the capacities of National Influenza Centers, and through assistance at the country-level where needed. In all, close to 60 country-missions or country-level activities were conducted. These activities were conducted in close coordination with WHO headquarters, WHO European Region Country Offices, the European Commission, the European Centre for Disease Prevention and Control, and with other partner organizations, and were in line with the implementation of the International Health Regulations (2005). The results of activities as well as guidance documents were disseminated to a wide audience through publication on the WHO Regional Office for Europe Influenza website, on the EuroFlu website, and through peer-reviewed publications.