Risk Factors for Severe Outcomes following 2009 Influenza A (H1N1) Infection: A Global Pooled Analysis

This study analyzes data from 19 countries (from April 2009 to Jan 2010), comprising some 70,000 hospitalized patients with severe H1N1 infection, to reveal risk factors for severe pandemic influenza, which include chronic illness, cardiac disease, chronic respiratory disease, and diabetes.


1
Standardized reporting of underlying risk conditions in human cases of pandemic (H1N1) 2009 1.1 Background, rational, and application of standardized reporting of underlying risk conditions Since the start of the 2009 H1N1 pandemic, several countries have published estimates of the proportion of pandemic H1N1 2009 cases with an underlying risk condition and have described high risk groups. However, compilation of data internationally, making comparisons between countries, and monitoring for changes over time have been difficult due to a lack of a standardized approach to data gathering and reporting of risk conditions. In collaboration with several partners, WHO has developed a recommended standardized format for reporting underlying risk conditions associated with cases of pandemic H1N1 2009.
Tracking the proportion of severe and fatal pandemic H1N1 2009 cases with underlying risk conditions over time is a critical part of monitoring the evolution of the pandemic, monitoring for changes in the epidemiology of the disease, and assessing changes in the severity of virus infection. A clear understanding of risk groups will allow policy makers to adapt recommendations regarding vaccination, antiviral use, and non pharmaceutical control strategies.
Standardization of reporting will facilitate: 1. Comparability of data across time and geographical areas to facilitate monitoring of risk groups for changes in attack patterns or virulence of pandemic influenza. 2. Development and refinement of targeted intervention strategies for groups at increased risk for poor outcomes. 3. Standardization of language for communication purposes.

1.2
Risk factors for severe outcomes with influenza infections Several conditions that increase the risk of severe outcomes from influenza have been described for seasonal influenza [1,2,3]. Early epidemiological data from the current pandemic suggest that these previously recognized underlying risk conditions for seasonal influenza may be similar for pandemic H1N1 2009 influenza. These risk factors can be grouped into three categories.
1. Chronic medical illnesses such as chronic lung disease and diabetes 2. Pregnancy

Extremes of age
As new data on the pandemic emerges, other conditions, such as obesity have been proposed as risk factors for severe disease and members of some indigenous populations have been noted to have higher rates of hospitalization and death. At the time of writing of this guidance, these factors have not yet been definitively shown to increase risk of severe outcomes independently of other previously accepted risk factors and so should be reported as a separate category. The separate reporting of these factors will greatly facilitate achieving a better understanding of their role in increasing the risk of poor outcome with pandemic influenza.
Finally, due to the high prevalence in some parts of the world of certain medical conditions for which the risk in relation to influenza is less well understood, (for example, tuberculosis and malnutrition) there is a need to quantify the risk associated with these conditions. Currently, reporting of risk factor data is sometimes confused by the inclusion of other chronic medical conditions that are commonly diagnosed but that have not been associated with severe outcomes such as hypertension in the absence of associated heart disease, smoking in the absence of associated lung disease, and hyperlipidemia in the absence of associated cardiovascular disease. Inclusion of these conditions in the reporting of underlying comorbid conditions complicates the interpretation of risk factor data.
The recommended standardized reporting format for underlying risk conditions is not intended to be an additional formal reporting requirement, but rather, an effort to provide guidance on standardized reporting of underlying risk conditions where that information exists or is already being collected and reported. WHO encourages countries that collect these data to report them using this format.

1.3
Recommended reporting format for recognized underlying risk conditions In order to facilitate WHO recommends the separate reporting of the categories of risk conditions as outlined above (chronic medical illness, pregnancy, age, and other) for different levels of poor outcome. Where data are available, the proportions with recognized underlying risk conditions among confirmed H1N1 cases should be reported for three severe outcome groups: those hospitalized, those requiring intensive care or mechanical ventilation, and those who die (Table S1). Therefore, risk conditions associated with these severe outcomes should be categorized as follows:  Proportion of patients in each of the three severe-outcome groups with at least one recognized underlying high-risk chronic medical illness.
 Proportion of patients in each of those poor outcome groups that are pregnant.
 Proportion of patients in each of the three severe outcome groups by age: <5 years old, 5 to <15 years old, 15 to < 25 years old, 25 to <50 years old, 50 to <65 years old, and ≥65 years old.
 Proportion in each of the severe outcome groups with other putative risk factors such as obesity, membership in an indigenous population or other suspect high-risk group, tuberculosis, malnutrition, or other groups suspected to be at high-risk by national health authorities.
The last group will differ by country and depend on the local interest of public health authorities. Help for definition on some conditions, such as malnutrition, can be found in other WHO publications.
It is expected that member states with adequate resources for doing detailed data collection and analysis will gather additional detailed data on risk factors and this is encouraged. The recommendations included in this annex and the tables that follow are meant to be used for summary reporting of risk factor data to WHO and internationally in a standard format.

Socially defined risk groups
Indigenous populations (self identified), racial and ethnic minorities, other identifiable groups with limited access to health care or high prevalence of chronic illness.

Tuberculosis
History of previous or current symptomatic tuberculosis requiring treatment.
Data on chronic conditions and all other risk factors among hospitalized, ICU and fatal H1N1pdm patients available by country are provided in Table S3, Table S4 and Table S5, respectively.
Obesity among H1N1pdm patients was categorized in four groups: 1) patients with body mass index (BMI) between 30 and 40, 2) BMI >40 (morbidly obese), 3) patients for which no BMI was measured but who were clinically judged to be obese, 4) BMI ≥30 or judged to be clinically obese.       Note: "x" indicates data was contributed by named country

Pooled Odds Ratio Calculations
Country-specific odds ratios (OR) and 95% confidence intervals (CI) for death were calculated separately for each risk factor (i.e., the odds of death given hospitalization and a specific risk factor), thereby comparing the odds of death in one group (for example, among hospitalized patients with asthma) with the odds of death in all others combined (for example, among hospitalized patients without asthma) (individual country ORs not shown). We then used the I² statistic to quantify the percentage of variation across countries that is due to true underlying heterogeneity in the OR rather than chance variability [59]. The I² statistics for all examined risk factors indicated that there was substantial true underlying variation between ORs from different countries. Thus, we undertook a random effects meta-analysis to describe the distribution of the OR estimates across the countries for which data were available for analysis.
Underlying this approach is the assumption that, although the individual countries give rise to different OR estimates, these estimates arise from a distribution with a central value, the estimate of which is referred to as the "pooled OR", and normally distributed variability around this value.
Countries with ≥25 confirmed hospitalized or fatal cases were included in the analysis. Countries included in gender pooled OR: Hong Kong, China, Thailand, Singapore, Canada, Madagascar, Spain, the Netherlands, Germany, Japan and US; Countries included in ≥1 condition pooled OR: