Estimation of excess mortality associated with influenza-epidemics by age and cause specific death in Japan, 1975-1999

OBJECTIVES: The purpose of this study was to assess the effects of recent influenza epidemics on mortality in Japan. METHODS: We applied a new definition of excess mortality associated with influenza epidemics and a new estimation method (new method) proposed in our previous paper to the national vital statistics for 1975-1999 (ICD8-ICD10 had been adopted) in Japan. This new method has the advantages of removing a source of random variations in excess mortality and of being applicable to shifting trends in mortality rates from different causes of death in response to the revision of ICD. The monthly rates of death from all causes other than accidents (all causes) and some cause-specific deaths such as pneumonia, malignant neoplasm, heart disease, cerebrovascular disease(C.V.D) and diabetes(D.M.) were analyzed by total and by five age groups: 0-4 years, 5-24 years, 25-44 years, 45-64 years, and 65 years old or older. RESULTS: The following findings were noted: 1. For each epidemic in every other year since 1993, large-scale excess mortality of over 10,000 deaths was observed and the effect of those epidemics could be frequently detected in mortality even among young persons, i.e., 0-4 years or 5-25 years. 2. Excess mortality associated with influenza epidemics influenced mortality by some chronic diseases such as pneumonia, heart disease, C.V.D., D.M., etc. For some epidemic years since 1978, excess mortality rates were detected even in mortality by malignant neoplasm. CONCLUSIONS: It has been definitely shown by applying the new method to the national vital statistics for 1975-1999 in Japan that influenza epidemics in recent years exerted an influence on overall mortality, increasing the number of deaths among the elderly and the younger generation. Monitoring of the trends in excess mortality associated with influenza epidemics should be continued.     

Mortality in children from influenza and respiratory syncytial virus

STUDY OBJECTIVE: To quantify mortality attributable to influenza and respiratory syncytial virus (RSV) infection in children. DESIGN AND METHODS: Comparison of death rates (all cause and certified respiratory) in England over winters 1989/90 to 1999/00 during and outside influenza and RSV circulation periods. Virus active weeks were defined from clinical and virological surveillance data. Excess deaths associated with weeks of either influenza or RSV activity over virus non-active weeks were estimated in each winter for age groups 1-12 months, 1-4, 5-9, and 10-14 years. The estimate obtained was allotted to influenza and RSV in the proportion derived from independent separate calculations for each virus. MAIN RESULTS: Average winter respiratory deaths attributed to influenza in children 1 month-14 years were 22 and to RSV 28; and all cause deaths to influenza 78 and to RSV 79. All cause RSV attributed deaths in infants 1-12 months exceeded those for influenza every year except 1989/90; the average RSV and influenza attributed death rates were 8.4 and 6.7 per 100 000 population respectively. Corresponding rates for children 1-4 years were 0.9 and 0.8 and for older children all rates were 0.2 or less, except for an influenza rate of 0.4 in children 10-14 years. CONCLUSIONS: Influenza and RSV account for similar numbers of deaths in children. The impact varies by winter and between age groups and is considerably underestimated if analysis is restricted to respiratory certified deaths. Summing the impact over the 11 winters studied, compared with influenza RSV is associated with more deaths in infants less than 12 months, almost equal numbers in children 1-4 years, and fewer in older children. Improved information systems are needed to investigate paediatric deaths.     

Comparative study of new method and the Kawai and Fukutomi methods for estimating excess mortality associated with influenza-epidemics, based upon national vital statistics from 1975 to 1997]

PURPOSE: In our previous paper, we proposed a new definition and method for estimating excess mortality associated with influenza epidemics. In this paper, we applied this new method to the national vital statistics for 1975-1997 in Japan and compared the estimates obtained with those generated with the Kawai and Fukutomi method. METHODS: The monthly rates of death from all causes other than accidents (all-causes) and deaths attributed to pneumonia between 1975-1997 in Japan were analyzed using our new method. Epidemic periods were identified by examining the monthly rates for deaths attributable to influenza and associated excess mortality was then estimated for the 23-year period using defined criteria. Finally, the estimates obtained using the new method were compared with those obtained using the Kawai and Fukutomi method. RESULTS: 1) An increase in observed over expected mortality (i.e. excess mortality under the old definition) was detected even for months when influenza epidemics did not occur. 2) Estimates made using the Kawai and Fukutomi method were between 2,000-14,000 higher for deaths from all-causes and about 500-3,000 higher for those from pneumonia for each of the epidemic periods, compared to the relevant figures obtained using the new method. This finding provided a good indication of the methodological difference with the new method, which considers the range of random variation in seasonal mortality. Overall, the two methods differed in their estimates of which month had the highest excess monthly mortality rate for the year and which year had the highest excess annual mortality rate. CONCLUSIONS: By comparing estimates obtained using the new method and the Kawai and Fukutomi method, we demonstrated that the former provides a more reasonable estimate of excess mortality rates, regardless of whether or not the period in question occurred during an influenza epidemic.     

Influenza vaccine coverage, influenza-associated morbidity and all-cause mortality in Catalonia (Spain)

The objective of this work was to study the behaviour of influenza with respect to morbidity and all-cause mortality in Catalonia, and their association with influenza vaccination coverage. The study was carried out over 13 influenza seasons, from epidemiological week 40 of 1994 to week 20 of 2007, and included confirmed cases of influenza and all-cause mortality. Two generalized linear models were fitted: influenza-associated morbidity was modelled by Poisson regression and all-cause mortality by negative binomial regression. The seasonal component was modelled with the periodic function formed by the sum of the sinus and cosines. Expected influenza mortality during periods of influenza virus circulation was estimated by Poisson regression and its confidence intervals using the Bootstrap approach. Vaccination coverage was associated with a reduction in influenza-associated morbidity (p<0.001), but not with a reduction in all-cause mortality (p=0.149). In the case of influenza-associated morbidity, an increase of 5% in vaccination coverage represented a reduction of 3% in the incidence rate of influenza. There was a positive association between influenza-associated morbidity and all-cause mortality. Excess mortality attributable to influenza epidemics was estimated as 34.4 (95% CI: 28.4-40.8) weekly deaths. In conclusion, all-cause mortality is a good indicator of influenza surveillance and vaccination coverage is associated with a reduction in influenza-associated morbidity but not with all-cause mortality.     

Respiratory syncytial virus--the unrecognised cause of health and economic burden among young children in Australia

Respiratory syncytial virus (RSV) presents very similar to influenza and is the principle cause of bronchiolitis in infants and young children worldwide. Yet, there is no systematic monitoring of RSV activity in Australia. This study uses existing published data sources to estimate incidence, hospitalisation rates, and associated costs of RSV among young children in Australia. Published reports from the Laboratory Virology and Serology Reporting Scheme, a passive voluntary surveillance system, and the National Hospital Morbidity Dataset were used to estimate RSV-related age-specific hospitalisation rates in New South Wales and Australia. These estimates and national USA estimates of RSV-related hospitalisation rates were applied to Australian population data to estimate RSV incidence in Australia. Direct economic burden was estimated by applying cost estimates used to derive economic cost associated with the influenza virus. The estimated RSV-related hospitalisation rates ranged from 2.2-4.5 per 1,000 among children less than 5 years of age to 8.7-17.4 per 1,000 among infants. Incidence ranged from 110.0-226.5 per 1,000 among the under five age group to 435.0-869.0 per 1,000 among infants. The total annual direct healthcare cost was estimated to be between $24 million and $50 million. Comparison with the health burdens attributed to the influenza virus and rotavirus suggests that the disease burden caused by RSV is potentially much higher. The limitations associated with using a passive surveillance system to estimate disease burden, and the need to explore further assessments and to monitor RSV activity are discussed.     

Attributable deaths due to influenza: a comparative study of seasonal and pandemic influenza

Influenza epidemics lead to an increase in hospitalizations and deaths. Up to now the overall impact of attributable deaths due to seasonal and pandemic influenza viruses in Austria has not been investigated in detail. Therefore we compared the number and age distribution of influenza associated deaths during ten influenza epidemic seasons to those observed during the pandemic influenza A(H1N1)2009 season. A Poisson model, relating age and daily deaths to week of influenza season using national mortality and viral surveillance data adjusted for the confounding effect of co-circulating Respiratory Syncytial Virus was used. We estimated an average of 316 influenza associated deaths per seasonal influenza epidemic (1999/2000-2008/2009) and 264 for the pandemic influenza season 2009/2010 in the area of Vienna, Austria. Comparing the mortality data for seasonal and pandemic influenza viruses in different age groups revealed a statistically significant increase in mortality for pandemic A(H1N1)2009 influenza virus in the age groups below 34 years of age and a significant decrease in mortality in those above 55 years. Our data adjusted for co-circulating RSV confirm the different mortality pattern of seasonal and pandemic influenza A(H1N1)2009 virus in different age groups.     

Impact of influenza epidemics on mortality in the United States from October 1972 to May 1985.

Baseline levels of mortality in the United States in the absence of influenza epidemics were estimated using cyclical regression models applied to national vital statistics from October 1972 to May 1985. Models were developed by age and by region. Results from 1983 to 1985 are preliminary. More mortality than predicted by the theoretical baseline occurred during nine influenza seasons. Epidemics with high morbidity in children and young adults occurred in 1976/77 and 1978/79. Regional differences in the impact of influenza occurred occasionally. Total influenza-associated excess mortality in six epidemics from the winters of 1972/73 to 1980/81 was about 200,000. About 80-90 per cent of excess mortality occurred in persons over 64 years old.     

Population estimates of persons presenting to general practitioners with influenza-like illness, 1987-96: a study of the demography of influenza-like illness in sentinel practice networks in England and Wales, and in The Netherlands

Incidence data by age of new episodes of influenza-like illness reported by sentinel general practice networks in England and Wales and in The Netherlands over a 10-year period were examined to provide estimates of the consulting population during influenza epidemic periods. Baseline levels of recording in each age group were calculated from weeks in which influenza viruses were not circulating and the excess over baseline calculated to provide the population estimates during influenza epidemics. Influenza A/H3N2 epidemics were associated with higher population estimates for consultations than influenza B, especially in the age groups 0-4 and 65 years and over. In the intervening age groups, population estimates were more consistent regardless of the virus type. Both networks reported simultaneous peaking of incidence rates in all of the age groups. There were substantial increases in the number of persons reporting other respiratory illnesses during influenza epidemics. Population estimates of the consulting population provide the only secure basis for which health services resource utilization during influenza epidemics can be estimated.     

Short-term effects of air pollution on total and cardiovascular mortality: the confounding effect of influenza epidemics

BACKGROUND: Air pollution is associated with total mortality. This association may be confounded by uncontrolled time-varying risk factors such as influenza epidemics. METHODS: We analyzed independent data on influenza epidemics from 7 European cities that also had data on mortality associated with particulates (PM10). We used 10 methods to control for epidemics (5 derived from influenza data and 5 from respiratory mortality series) and compared those results with analyses that did not control for these epidemics. RESULTS: Adjustment for influenza epidemics increased the PM10 effect estimate in most cases (% change in the pooled regression coefficient: -1.9 to 38.9 for total mortality and 1.3 to 25.5 for cardiovascular mortality). A 10-microg/m increase in PM10 concentrations (lag 0-1) was associated with a 0.48% (95% confidence interval=0.27-0.70%) increase in daily mortality unadjusted for influenza epidemics, whereas under the various methods to control for epidemics the increase ranged from 0.45% (0.26-0.69%) to 0.67% (0.46-0.89%). The corresponding figures for cardiovascular mortality were 0.85% (0.53-1.18%) with no adjustment and from 0.86% (0.53-1.19%) to 1.06% (0.74-1.39%) with the methods of control. CONCLUSIONS: The association between air pollution and mortality is not weakened by control for influenza epidemic irrespective of the method used. To adjust for influenza epidemics, one can use methods based on respiratory mortality counts instead of counts of influenza cases if the latter are not available. However, adjustment for influenza by any method tested did not markedly alter the air pollution effect estimate.     

Influenza- and RSV-associated hospitalizations among adults

We estimated influenza- and respiratory syncytial virus (RSV)-associated hospitalizations by age, high-risk status and outcome, during the 1996/1997-1999/2000 respiratory seasons among adults who did not receive influenza vaccine. Using three health maintenance organization (HMO) databases and local viral surveillance data, we identified weeks when influenza and RSV were circulating and estimated influenza- and RSV-associated hospitalizations. Persons aged > or = 65 years with and without high-risk conditions had significantly increased rates of influenza-associated hospitalizations for pneumonia and influenza, and circulatory and respiratory diseases. Persons aged > or = 65 years with high-risk conditions also had significantly increased rates of influenza-associated hospitalizations for cardiac conditions (16.9 per 10,000 person periods). Relative to the influenza estimates for high-risk persons > or = 65 years, we found lower rates of RSV-associated hospitalizations for pneumonia and influenza diseases (23.4 per 10,000 person periods), cardiac diseases (4.3 per 10,000 person periods) and circulatory and respiratory diseases (44.0 per 10,000 person periods). Among low-risk persons aged 50-64 years, we did not identify significantly elevated rates of influenza- or RSV-associated hospitalizations. Excess hospitalization estimates among adults aged > or = 65 years and high-risk 50-64 year olds during the influenza season suggest that these groups should have priority for influenza vaccine during vaccine shortages.     

The impact of influenza epidemics on mortality: introducing a severity index.

OBJECTIVES: The purpose of this study was to assess the impact of recent influenza epidemics on mortality in the United States and to develop an index for comparing the severity of individual epidemics. METHODS: A cyclical regression model was applied to weekly national vital statistics from 1972 through 1992 to estimate excesses in pneumonia and influenza mortality and all-cause mortality for each influenza season. Each season was categorized on the basis of increments of 2000 pneumonia and influenza excess deaths, and each of these severity categories was correlated with a range of all-cause excess mortality. RESULTS: Each of the 20 influenza seasons studied was associated with an average of 5600 pneumonia and influenza excess deaths (range, 0-11,800) and 21,300 all-cause excess deaths (range, 0-47,200). Most influenza A(H3N2) seasons fell into severity categories 4 to 6 (23,000-45,000 all-cause excess deaths), whereas most A(H1N1) and B seasons were ranked in categories 1 to 3 (0-23,000 such deaths). CONCLUSIONS: From 1972 through 1992, influenza epidemics accounted for a total of 426,000 deaths in the United States, many times more than those associated with recent pandemics. The influenza epidemic severity index was useful for categorizing severity and provided improved seasonal estimates of the total number of influenza-related deaths.     

Influenza deaths in Leicestershire during the 1989-90 epidemic: implications for prevention.

There is an association between excess winter mortality and epidemics of influenza and it has been suggested that annual influenza vaccination should be offered to all over 65 years old as in the United States. This paper identifies the number of people dying from influenza in Leicestershire UK during the 1989-90 epidemic and the factors associated with a fatal outcome. The findings show that deaths attributed to influenza occur predominantly in very elderly people with underlying ill-health. The risk of influenzal death is greater in residential patients and increases substantially with the number of underlying medical conditions. The estimated death rates in vaccinated and non-vaccinated groups were not significantly different, but there were trends towards protection in both residential and non-residential groups. Influenza vaccine is not reaching the principal target groups and improved methods of influenza control are required.     

Rates of hospitalisation for influenza, respiratory syncytial virus and human metapneumovirus among infants and young children

To inform the development of a national influenza immunisation programme and the potential role of antiviral drugs in young children, we studied 613 children aged 71 months or younger who attended Leicester Childrens' Hospital during winter 2001-2002. During periods of respiratory syncytial virus (RSV), influenza, and human metapneumovirus activity, an estimated 12.2% (95% CI: 11.4-13.1), 7.1% (6.3-7.9), and 2.5% (2.1-2.9), respectively, of all medical cases assessed in the hospital were associated with these infections. The respective rates of hospital assessments for RSV, influenza, and human metapneumovirus (HMPV) were 1042 (95% CI: 967-1021), 394 (348-443), and 223 (189-262) per 100,000 children, and for admissions were 517 (465-574), 144 (117-175), and 126 (101-156) per 100,000. Few children with influenza had a prior risk factor. Children with influenza were admitted a median of 4 days after onset of illness and none was coded at discharge as influenza. We conclude that antivirals have little role in the hospital management of children with influenza. Our data provide health economists with information to evaluate the place of universal immunisation of young children against influenza. Hospitalisation rates decreased markedly with referral age, so vaccine would need to be given in early infancy for maximum benefit.     

Morbidity profiles of patients consulting during influenza and respiratory syncytial virus active periods

We compared the burden of illness due to a spectrum of respiratory diagnostic categories among persons presenting in a sentinel general practice network in England and Wales during periods of influenza and of respiratory syncytial virus (RSV) activity. During all periods of viral activity, incidence rates of influenza-like illness, bronchitis and common cold were elevated compared to those in baseline periods. Excess rates per 100,000 of acute bronchitis were greater in children aged <1 year (median difference 2702, 95% CI 929-4867) and in children aged 1-4 years (994, 95% CI 338-1747) during RSV active periods rather than influenza; estimates for the two viruses were similar in other age groups. Excess rates of influenza-like illness in all age groups were clearly associated with influenza virus activity. For common cold the estimates of median excess rates were significantly higher in RSV active periods for the age groups <1 year (3728, 95% CI 632-5867) and 5-14 years (339, 95% CI 59-768); estimates were similar in other age groups for the two viruses. The clinical burden of disease associated with RSV is as great if not greater than influenza in patients of all ages presenting to general practitioners.     

Mortality during influenza epidemics in the United States, 1967-1978.

Excess deaths due to pneumonia and influenza and excess deaths for all causes were estimated using a time-series analysis for each of the eight influenza epidemics in the United States that occurred between 1967-1978. The effects of different analytic methods and different methods of structuring data are compared. Future directions for estimating the impact of influenza on mortality include a combination of regression techniques and multiple time-series analyses of surveillance data.     

Estimating Influenza-Related Hospital Admissions in Children and Adults

Introduction

Influenza is associated with illnesses such as pneumonia and other respiratory conditions and in severe cases leads to death. The prevalence of these illnesses and deaths fluctuates with the seasons during the year, even in the absence of influenza. Although many studies have focussed on mortality associated with influenza epidemics, and some have examined hospitalizations in elderly patients, there are very few studies that have examined the effect of influenza epidemics on adults or children. This study seeks to determine the association between general practitioner (GP) consultations for influenza-like illnesses and hospital admissions of adults and children associated with influenza epidemics.

Methods

Structural Time Series Models with stochastic trend and seasonal components were developed for two age groups (children aged 0–15 years, and adults aged 16–50 years). Data from the Swiss Sentinel Surveillance Network on GP consultation rates for influenza-like illnesses, and data from Swiss hospital admissions, were obtained for the period 1987–1996. The explanatory variables (i.e., the percentage of GP consultations for influenza-like illnesses and a 1-week lag of this variable) were modeled against hospital admission rates for pneumonia and influenza and other respiratory conditions. Excess hospitalizations were calculated as the difference between predicted hospital admissions during influenza epidemics and expected hospital admissions in the absence of influenza epidemics.

Results

In these two age groups, there was an annual average of 1452 (range: 1000–1700) hospital admissions directly associated with influenza epidemics. Excess admission rates were substantially higher in children (pneumonia and influenza: 4.77 per 10 000 children per year, and other respiratory conditions: 2.29 per 10 000 children per year) compared with adults (pneumonia and influenza: 0.86 per 10 000 adults per year and other respiratory conditions: 0.68 per 10 000 adults per year). The models explained 56–84% of the variation in hospital admissions. The seasonal patterns were stable over the 10 years modeled and the variances of the trends were small.

Conclusion

The structural time series models is an ideal approach to model influenza-related hospitalizations as the models capture trends, seasonal variation, and the association with exogenous factors.

     

Deaths Associated with Respiratory Syncytial and Influenza Viruses among Persons ≥5 Years of Age in HIV-Prevalent Area,South Africa, 1998–20091

We estimated deaths attributable to influenza and respiratory syncytial virus (RSV) among persons >5 years of age in South Africa during 1998–2009 by applying regression models to monthly deaths and laboratory surveillance data. Rates were expressed per 100,000 person-years. The mean annual number of seasonal influenza–associated deaths was 9,093 (rate 21.6). Persons >65 years of age and HIV-positive persons accounted for 50% (n = 4,552) and 28% (n = 2,564) of overall seasonal influenza-associated deaths, respectively. In 2009, we estimated 4,113 (rate 9.2) influenza A(H1N1)pdm09–associated deaths. The mean of annual RSV-associated deaths during the study period was 511 (rate 1.2); no RSV-associated deaths were estimated in persons >45 years of age. Our findings support the recommendation for influenza vaccination of older persons and HIV-positive persons. Surveillance for RSV should be strengthened to clarify the public health implications and severity of illness associated with RSV infection in South Africa.     

Effectiveness of influenza vaccine in reducing hospital admissions in people with diabetes.

The effectiveness of influenza vaccination in reducing hospitalization of people with diabetes for influenza, pneumonia, or diabetic events during influenza epidemics was assessed in a case control study in Leicestershire, England. Cases were 80 patients on the Leicestershire Diabetes Register who were admitted and discharged from hospital with International Classification of Disease codes for pneumonia, bronchitis, influenza, diabetic ketoacidosis, coma and diabetes, without mention of complications, during the influenza epidemics of 1989-90 and 1993. One hundred and sixty-controls, who were not admitted to hospital during this period, were randomly selected from the Register. Immunization against influenza was assessed in 37 cases and 77 controls for whom consent was obtained to access their clinical notes and for whom notes were available. Significant association was detected between reduction in hospitalization and influenza vaccination during the period immediately preceding an epidemic. Multiple logistic regression analysis estimated that influenza vaccination reduced hospital admissions by 79% (95% CI 19-95%) during the two epidemics, after adjustment for potential confounders.     

Relationships Between Community Virus Activity and Cardiorespiratory Rehospitalizations From Post-Acute Care

ObjectivesQuantify the relationship between increasing influenza and respiratory syncytial virus (RSV) community viral activity and cardiorespiratory rehospitalizations among older adults discharged to skilled nursing facilities (SNFs).DesignRetrospective cohort.Setting and ParticipantsAdults aged ≥65 years who were hospitalized and then discharged to a US SNF between 2012 and 2015.MethodsWe linked Medicare Provider Analysis and Review claims to Minimum Data Set version 3.0 assessments, PRISM Climate Group data, and the Centers for Disease Control and Prevention viral testing data. All data were aggregated to US Department of Health and Human Services regions. Negative binomial regression models quantified the relationship between increasing viral activity for RSV and 3 influenza strains (H1N1pdm09, H3N2, and B) and cardiorespiratory rehospitalizations from SNFs. Incidence rate ratios described the relationship between a 5% increase in circulating virus and the rates of rehospitalization for cardiorespiratory outcomes. Analyses were repeated using the same model, but influenza and RSV were considered “in season” or “out of season” based on a 10% positive testing threshold.ResultsCardiorespiratory rehospitalization rates increased by approximately 1% for every 5% increase in circulating influenza A(H3N2), influenza B, and RSV, but decreased by 1% for every 5% increase in circulating influenza A(H1N1pdm09). When respiratory viruses were in season (vs out of season), cardiorespiratory rehospitalization rates increased by approximately 6% for influenza A(H3N2), 3% for influenza B, and 5% for RSV, but decreased by 6% for influenza A(H1N1pdm09).Conclusions and ImplicationsThe respiratory season is a particularly important period to implement interventions that reduce cardiorespiratory hospitalizations among SNF residents. Decreasing viral transmission in SNFs through practices such as influenza vaccination for residents and staff, use of personal protective equipment, improved environmental cleaning measures, screening and testing of residents and staff, surveillance of viral activity, and quarantining infected individuals may be potential strategies to limit viral infections and associated cardiorespiratory rehospitalizations.