Giuseppe
Emeritus
Update: WHO-confirmed human cases of avian influenza A(H5N1) infection, November 2003–May 2008 [WHO WER]
Full PDF text: http://www.who.int/wer/2008/wer8346.pdf
From WEEKLY EPIDEMIOLOGICAL RECORD - 14 NOVEMBER 2008, 83rd YEAR - No. 46, 2008, 83, 413–420
http://www.who.int/wer
Update: WHO-confirmed human cases of avian influenza A(H5N1) infection, November 2003–May 2008
Introduction
Since November 2003, WHO has been recording laboratory-confirmed cases of human infection with avian influenza A(H5N1) viruses. As of the end of May 2008, 383 confirmed human cases had been reported to WHO.
Although the main epidemiological characteristics of human A(H5N1) infection have remained almost unchanged since the last description by WHO,1 there are some differences among affected countries.
Methods
This report includes information on all laboratory-confirmed human cases of infection with H5N1 officially reported to WHO between November 2003 and May 2008. Cases were analysed by time of onset, place (country), demographic characteristics (age and sex), clinical history (time from onset to hospitalization) and clinical outcome. Odds ratios (OR) were calculated when comparing
dichotomous data. The t-test on ranked values was used to compare continuous data; the Wilcoxon rank-sum test was used to compare subsets of continuous data. Country-specific incidence rates were calculated using the 2006 revision of World population prospects.2 Countries with >25 cases were further analysed and selected variables were compared.
Results
Temporal and geographical distribution
Between November 2003 and May 2008, 383 laboratory confirmed cases of influenza A(H5N1) infection were reported to WHO from 15 countries (Table 1). The 5 countries that accounted for 90% of cases were Indonesia (133), Viet Nam (106), Egypt (50), China (30) and Thailand (25). The number of countries that reported human cases increased from 2 in 2003 to 9 in both 2006 and 2007,
following the geographical spread of A(H5N1) outbreaks in poultry. The number of cases (Fig. 1) peaked each year during December–March; the fewest cases were always reported during the third quarter of each year.
Gender and age distribution
Cases of human infection with influenza A(H5N1) were reported equally in males and females. However, the proportion of female cases varied by country, ranging from 36% in Thailand to 68% in Egypt (Table 2). In some countries, the distribution by sex changed over time. For example, in Indonesia the proportion of cases among females was significantly higher in 2007 than in other years
(OR, 2.6; 95% confidence interval [CI], 1.2–5.6) (Table 2). The mean age of cases was 21.7 years; the median age was 20.0 years. Cases ranged in age from <1 year to 81 years. Mean age ranged from 17.4 years in Egypt to 28.1 years in China (Table 3). In Thailand, the mean age of female cases was significantly higher than the mean age of male cases (Table 3). Cases among those aged 0–9 years accounted for 28% of all cases; cases among those aged 10–19 years accounted for 24% of all cases; cases among those aged 20–29 years accounted for 25% of all cases. In Egypt and Thailand, those aged 0–9 years accounted for the highest percentage of cases; in China, Indonesia and Viet Nam, slightly more cases occurred among those aged 20–29 years. Analysis of the age-specific incidence using country-wide age-distribution did not change this pattern.
Outcome
The overall case-fatality rate (CFR) of laboratory-confirmed cases of H5N1 infection was 63%. However, CFRs varied somewhat by year and among countries. Among the 5 most seriously affected countries, the CFR ranged from 44% in Egypt to 80% in Indonesia (Table 4). When data from all countries were combined, the CFR was highest (78%) in the 10–19 year age group; there was no consistent pattern by age. Overall, females were less likely to survive than males (OR, 1.6; 95% CI, 1.06–2.46). In Egypt, women were 10 times less likely to survive than men (OR, 10; 95% CI, 1.71–75.8).
Time from onset of symptoms to hospitalization
The mean time from onset of illness to hospitalization was 4.6 days, with a median of 4 days and a range of 0–22 days. The mean ranged from 2.2 days in Egypt to 5.7 days in Indonesia (Table 5). Although people who survived had been hospitalized earlier after illness onset (mean, 3.9 days) than those who died (mean, 5 days) (P=0.02), among the 5 most seriously affected countries, this difference was significant only in Egypt (Table 5). Not surprisingly, the CFR increased with increasing time from disease onset to hospitalization: the CFR was 12% for 0 days between onset and hospitalization, 47% for 1 day, 55% for 2 days and >70% for 4–6 days (χ2 for trend, P=5.6 x 10-6).
Discussion
This follow-up paper provides a brief update on the demographic and epidemiological characteristics of laboratory-confirmed human cases of influenza A(H5N1) infection reported to WHO during 4-1/2 years. Human infections with influenza A(H5N1) are still rare, and the epidemiological characteristics of this infection in humans are similar to those reported in 2007. So far, the A(H5N1) viruses that have infected people have retained avian influenza virus genes. There has been no sustained human-to-human transmission; limited human-to-human transmission has been seen in only a few settings, usually involving family members.3 The previous epidemiological description1 of the first 256 cases found a median age of 18 years, a sex ratio of 1, a CFR of 60 (with higher mortality in the 10–19 year age group), a higher mortality in women and a mean of 4 days from disease onset to hospitalization. We found significant variations in patterns among
countries. These national variations might reflect differences in exposure related to social behaviour, cultural and religious practices, access to care, case ascertainment, treatment or virulence of the viruses. This analysis includes only laboratory-confirmed symptomatic cases, identified mainly through passive surveillance and officially reported to WHO; thus these cases potentially represent both an underestimate of all cases and the more severe spectrum of cases. However, the extent to which asymptomatic and minimally symptomatic cases of human A(H5N1) infection occur is not clear. Among the few available seroepidemiological studies, the results of most suggest that asymptomatic human influenza A(H5N1) infection is relatively rare. However, an important unanswered question is why 90% of human cases (of the 383 cases reported from November 2003 until the end of May 2008) have been reported from only 5 countries when animal outbreaks have occurred in 60 countries since 2003. Although many of these countries have had only limited and rapidly contained outbreaks among animals, others have had ongoing circulation or repeated incursions of A(H5N1), suggesting there has been ample opportunity for humans to be exposed to infected animals. The lack of human cases in these countries may reflect national
differences in risk factors or exposures, control or prevention strategies, human or viral genetic factors, or the sensitivity of surveillance or sampling strategies, and should be further explored. Similar questions may be posed about why these 383 individuals have become ill and not others who have been similarly exposed. More than 90% of cases in clusters have occurred among blood-related family members, which might suggest a possible explanation.4, 5
In the northern hemisphere, an increase in human cases was observed repeatedly during the cooler months, paralleling the pattern seen in animal outbreaks of A(H5N1) infection as well as seasonal influenza. In the tropics, the degree to which A(H5N1) and seasonal influenza patterns parallel each other is less clear. To help address these and other issues, global collaborative efforts to collect and analyse standardized data on exposure, clinical features and virological features would improve understanding and identifi cation of significant factors. Greater use of standardized tools, such as the WHO clinical summary form,6 would facilitate this effort. Virological and epidemiological surveillance data must be supplemented by data from studies, such as serological surveys, environmental studies and more detailed case–control studies of potential risk factors and exposures. WHO, together with its partners, is studying potential factors related to exposure as well as the role of markets where live poultry are sold.
To facilitate such studies, WHO plans to establish an influenza research agenda that reflects gaps in public health knowledge to help prioritize and guide worldwide efforts to address these gaps in the knowledge base.
As outbreaks of avian influenza A(H5N1) continue to occur, the risk for human exposure and the emergence of a pandemic influenza strain remain prominent concerns. All efforts to comprehensively assess the epidemiological, ecological and virological characteristics of this situation should continue, especially since such efforts may lead to a better understanding of infections with novel infuenza viruses and their pandemic potential
1) See No. 6, 2007, pp. 41–47.
2) World population prospects: the 2006 revision. Population database. New York, United Nations Population Division, 2006 (http://esa.un.org/unpp/index.asp?panel=1; accessed July 2008).
3) See No. 40, 2008, pp. 359–364.
4) Writing committee of the Second WHO Consultation on Clinical Aspects of Human Infection with Avian Influenza A (H5N1) virus. Update on avian influenza A(H5N1) virus infection in humans. New England Journal of Medicine, 2008, 358:261–273.
5) Nyoman K, et al. Factors associated with case fatality of human H5N1 virus infections in Indonesia: a case series. Lancet, 2008, 372:744–749.
6) WHO H5N1 clinical case summary form. Geneva, WHO Global Infl uenza Programme, 2007 (http://www.who.int/csr/disease/avian_infl uenza/guidelines/SummaryForm07.pdf; accessed November 2008).
<table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table>
<table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table>
<table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table>
<table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table>
Full PDF text: http://www.who.int/wer/2008/wer8346.pdf
From WEEKLY EPIDEMIOLOGICAL RECORD - 14 NOVEMBER 2008, 83rd YEAR - No. 46, 2008, 83, 413–420
http://www.who.int/wer
Update: WHO-confirmed human cases of avian influenza A(H5N1) infection, November 2003–May 2008
Introduction
Since November 2003, WHO has been recording laboratory-confirmed cases of human infection with avian influenza A(H5N1) viruses. As of the end of May 2008, 383 confirmed human cases had been reported to WHO.
Although the main epidemiological characteristics of human A(H5N1) infection have remained almost unchanged since the last description by WHO,1 there are some differences among affected countries.
Methods
This report includes information on all laboratory-confirmed human cases of infection with H5N1 officially reported to WHO between November 2003 and May 2008. Cases were analysed by time of onset, place (country), demographic characteristics (age and sex), clinical history (time from onset to hospitalization) and clinical outcome. Odds ratios (OR) were calculated when comparing
dichotomous data. The t-test on ranked values was used to compare continuous data; the Wilcoxon rank-sum test was used to compare subsets of continuous data. Country-specific incidence rates were calculated using the 2006 revision of World population prospects.2 Countries with >25 cases were further analysed and selected variables were compared.
Results
Temporal and geographical distribution
Between November 2003 and May 2008, 383 laboratory confirmed cases of influenza A(H5N1) infection were reported to WHO from 15 countries (Table 1). The 5 countries that accounted for 90% of cases were Indonesia (133), Viet Nam (106), Egypt (50), China (30) and Thailand (25). The number of countries that reported human cases increased from 2 in 2003 to 9 in both 2006 and 2007,
following the geographical spread of A(H5N1) outbreaks in poultry. The number of cases (Fig. 1) peaked each year during December–March; the fewest cases were always reported during the third quarter of each year.
Gender and age distribution
Cases of human infection with influenza A(H5N1) were reported equally in males and females. However, the proportion of female cases varied by country, ranging from 36% in Thailand to 68% in Egypt (Table 2). In some countries, the distribution by sex changed over time. For example, in Indonesia the proportion of cases among females was significantly higher in 2007 than in other years
(OR, 2.6; 95% confidence interval [CI], 1.2–5.6) (Table 2). The mean age of cases was 21.7 years; the median age was 20.0 years. Cases ranged in age from <1 year to 81 years. Mean age ranged from 17.4 years in Egypt to 28.1 years in China (Table 3). In Thailand, the mean age of female cases was significantly higher than the mean age of male cases (Table 3). Cases among those aged 0–9 years accounted for 28% of all cases; cases among those aged 10–19 years accounted for 24% of all cases; cases among those aged 20–29 years accounted for 25% of all cases. In Egypt and Thailand, those aged 0–9 years accounted for the highest percentage of cases; in China, Indonesia and Viet Nam, slightly more cases occurred among those aged 20–29 years. Analysis of the age-specific incidence using country-wide age-distribution did not change this pattern.
Outcome
The overall case-fatality rate (CFR) of laboratory-confirmed cases of H5N1 infection was 63%. However, CFRs varied somewhat by year and among countries. Among the 5 most seriously affected countries, the CFR ranged from 44% in Egypt to 80% in Indonesia (Table 4). When data from all countries were combined, the CFR was highest (78%) in the 10–19 year age group; there was no consistent pattern by age. Overall, females were less likely to survive than males (OR, 1.6; 95% CI, 1.06–2.46). In Egypt, women were 10 times less likely to survive than men (OR, 10; 95% CI, 1.71–75.8).
Time from onset of symptoms to hospitalization
The mean time from onset of illness to hospitalization was 4.6 days, with a median of 4 days and a range of 0–22 days. The mean ranged from 2.2 days in Egypt to 5.7 days in Indonesia (Table 5). Although people who survived had been hospitalized earlier after illness onset (mean, 3.9 days) than those who died (mean, 5 days) (P=0.02), among the 5 most seriously affected countries, this difference was significant only in Egypt (Table 5). Not surprisingly, the CFR increased with increasing time from disease onset to hospitalization: the CFR was 12% for 0 days between onset and hospitalization, 47% for 1 day, 55% for 2 days and >70% for 4–6 days (χ2 for trend, P=5.6 x 10-6).
Discussion
This follow-up paper provides a brief update on the demographic and epidemiological characteristics of laboratory-confirmed human cases of influenza A(H5N1) infection reported to WHO during 4-1/2 years. Human infections with influenza A(H5N1) are still rare, and the epidemiological characteristics of this infection in humans are similar to those reported in 2007. So far, the A(H5N1) viruses that have infected people have retained avian influenza virus genes. There has been no sustained human-to-human transmission; limited human-to-human transmission has been seen in only a few settings, usually involving family members.3 The previous epidemiological description1 of the first 256 cases found a median age of 18 years, a sex ratio of 1, a CFR of 60 (with higher mortality in the 10–19 year age group), a higher mortality in women and a mean of 4 days from disease onset to hospitalization. We found significant variations in patterns among
countries. These national variations might reflect differences in exposure related to social behaviour, cultural and religious practices, access to care, case ascertainment, treatment or virulence of the viruses. This analysis includes only laboratory-confirmed symptomatic cases, identified mainly through passive surveillance and officially reported to WHO; thus these cases potentially represent both an underestimate of all cases and the more severe spectrum of cases. However, the extent to which asymptomatic and minimally symptomatic cases of human A(H5N1) infection occur is not clear. Among the few available seroepidemiological studies, the results of most suggest that asymptomatic human influenza A(H5N1) infection is relatively rare. However, an important unanswered question is why 90% of human cases (of the 383 cases reported from November 2003 until the end of May 2008) have been reported from only 5 countries when animal outbreaks have occurred in 60 countries since 2003. Although many of these countries have had only limited and rapidly contained outbreaks among animals, others have had ongoing circulation or repeated incursions of A(H5N1), suggesting there has been ample opportunity for humans to be exposed to infected animals. The lack of human cases in these countries may reflect national
differences in risk factors or exposures, control or prevention strategies, human or viral genetic factors, or the sensitivity of surveillance or sampling strategies, and should be further explored. Similar questions may be posed about why these 383 individuals have become ill and not others who have been similarly exposed. More than 90% of cases in clusters have occurred among blood-related family members, which might suggest a possible explanation.4, 5
In the northern hemisphere, an increase in human cases was observed repeatedly during the cooler months, paralleling the pattern seen in animal outbreaks of A(H5N1) infection as well as seasonal influenza. In the tropics, the degree to which A(H5N1) and seasonal influenza patterns parallel each other is less clear. To help address these and other issues, global collaborative efforts to collect and analyse standardized data on exposure, clinical features and virological features would improve understanding and identifi cation of significant factors. Greater use of standardized tools, such as the WHO clinical summary form,6 would facilitate this effort. Virological and epidemiological surveillance data must be supplemented by data from studies, such as serological surveys, environmental studies and more detailed case–control studies of potential risk factors and exposures. WHO, together with its partners, is studying potential factors related to exposure as well as the role of markets where live poultry are sold.
To facilitate such studies, WHO plans to establish an influenza research agenda that reflects gaps in public health knowledge to help prioritize and guide worldwide efforts to address these gaps in the knowledge base.
As outbreaks of avian influenza A(H5N1) continue to occur, the risk for human exposure and the emergence of a pandemic influenza strain remain prominent concerns. All efforts to comprehensively assess the epidemiological, ecological and virological characteristics of this situation should continue, especially since such efforts may lead to a better understanding of infections with novel infuenza viruses and their pandemic potential
1) See No. 6, 2007, pp. 41–47.
2) World population prospects: the 2006 revision. Population database. New York, United Nations Population Division, 2006 (http://esa.un.org/unpp/index.asp?panel=1; accessed July 2008).
3) See No. 40, 2008, pp. 359–364.
4) Writing committee of the Second WHO Consultation on Clinical Aspects of Human Infection with Avian Influenza A (H5N1) virus. Update on avian influenza A(H5N1) virus infection in humans. New England Journal of Medicine, 2008, 358:261–273.
5) Nyoman K, et al. Factors associated with case fatality of human H5N1 virus infections in Indonesia: a case series. Lancet, 2008, 372:744–749.
6) WHO H5N1 clinical case summary form. Geneva, WHO Global Infl uenza Programme, 2007 (http://www.who.int/csr/disease/avian_infl uenza/guidelines/SummaryForm07.pdf; accessed November 2008).
<table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table><table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table><table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table><table style="width: auto;"><tbody><tr><td>
</td></tr><tr><td style="font-family: arial,sans-serif; font-size: 11px; text-align: right;">From TABLES</td></tr></tbody></table>