Conceived and designed the experiments: ACM MSC MC. Performed the experiments: ACM VAP. Analyzed the data: ACM MSC MC VAP. Contributed reagents/materials/analysis tools: ACM. Wrote the paper: ACM MSC.
The authors have declared that no competing interests exist.
Pandemic influenza A (H1N1) 2009 has posed a serious public health challenge world-wide. In absence of reliable information on severity of the disease, the nations are unable to decide on the appropriate response against this disease.
Based on the results of laboratory investigations, attendance in outpatient department, hospital admissions and mortality from the cases of influenza like illness from 1 August to 31 October 2009 in Pune urban agglomeration, risk of hospitalization and case fatality ratio were assessed to determine the severity of pandemic H1N1 and seasonal influenza-A infections.
Prevalence of pandemic H1N1 as well as seasonal-A cases were high in Pune urban agglomeration during the study period. The cases positive for pandemic H1N1 virus had significantly higher risk of hospitalization than those positive for seasonal influenza-A viruses (OR: 1.7). Of 93 influenza related deaths, 57 and 8 deaths from Pune (urban) and 27 and 1 death from Pune (rural) were from pandemic H1N1 positive and seasonal-A positive cases respectively. The case fatality ratio 0.86% for pandemic H1N1 was significantly higher than that of seasonal-A (0.13%) and it was in category 3 of the pandemic severity index of CDC, USA. The data on the cumulative fatality of rural and urban Pune revealed that with time the epidemic is spreading to rural areas.
The severity of the H1N1 influenza pandemic is less than that reported for ‘Spanish flu 1918’ but higher than other pandemics of the 20th century. Thus, pandemic influenza should be considered as serious health threat and unprecedented global response seems justified.
Emergence of the influenza pandemic (H1N1) 2009 (pandemic H1N1) posed new challenges to the public health systems and communities all over the world. Global actions by international agencies and highly vigilant media generated tremendous fear resulting into unprecedented response to the new pandemic by majority of the nations. However, it is not clear that the fear generated was in proportion to the risk that this virus presents.
As of 6 December 2009, worldwide more than 208 countries had reported laboratory confirmed cases of pandemic H1N1, including at least 9596 deaths
The pandemic has created tremendous hardship on already overburdened health system in developing nations but unprecedented surveillance has also provided opportunity to study different epidemiological parameters including disease severity of influenza viruses in tropical settings like India. Understanding of severity is central to health care planning for management of the pandemic in the future and to avoid wastage of resources in resource poor settings. Better estimates will also facilitate identification of risk factors between and within populations and regions. Policy regarding social distancing, school closures, diagnosis of patients, vaccination etc would be greatly influenced by these estimates.
Severity of infection can be assessed by different approaches. One is to explore different genetic markers of the virus that are known to be associated with severe influenza. The current circulating pandemic H1N1 does not contain the known molecular markers of pathogenicity and transmissibility
In India, after declaration of pandemic (phase 6) by World Health Organization (WHO) on 11 June
In this communication we have followed the epidemiological approach suggested by Garske et al (2009) to assess severity of infection in pandemic H1N1 and seasonal influenza-A patients.
Since the beginning of this pandemic, 11 May to 31 October 2009, a total of 607,117 persons presented themselves for screening to various health centers in Pune. Of these 34,917 persons were determined as suspected cases of ILI. From 1 August to 31 October, a total of 7,866 suspected cases were sampled for diagnosis of influenza etiology. Frequencies of pandemic H1N1 and seasonal-A cases were 17.8% and 16.3% respectively (
Cases | Total Influenza-A | IPD cases | OPD cases | ||||||
Suspect cases | Pandemic H1N1 (%) | Seasonal-A (%) | Suspect cases | Pandemic H1N1 (%) | Seasonal-A (%) | Suspect Cases | Pandemic H1N1 (%) | Seasonal-A (%) | |
Total | 7866 | 1404 (17.8) | 1286 (16.3) | 3300 | 569 (17.2) | 424 (12.8) | 2967 | 541 (18.2) | 687 (23.2) |
Range | 8–327 | 1–97 | 0–79 | 7–73 | 0–21 | 0–20 | 0–298 | 0–75 | 0–65 |
Daily median/Average | 85.5 | 15.3 | 14 | 35.9 | 6.2 | 4.61 | 32.3 | 5.9 | 7.5 |
95% Confidence interval | 17.0, 18.7 | 15.5, 17.2 | 16.0, 18.6 | 11.7, 14.0 | 16.9, 19.7 | 21.7, 24.7 |
After the first death on 3 August, regular local screening of all cases of suspected ILI was started. From the second week of September onward, only limited samples, collected at the discretion of the clinicians, were referred to the laboratory because suspected cases were administered Oseltamivir without laboratory confirmation. From 1 August to 31 October 2009, clinical samples from 2967 suspected cases tested in laboratory yielded 18.2% and 23.2% pandemic and seasonal-A influenza respectively (
Between 1 August to 31 October, a total of 3,300 suspected cases were hospitalized (
In IPD, from 1 August to 31 Ocober 2009, out of 993 positive cases, pandemic H1N1 cases (569, 57.3%) were significantly higher than seasonal-A cases (424, 42.7%). However, in OPD, out of 1228 positive cases, pandemic H1N1 cases (541, 44.06%) were significantly less than seasonal-A cases (687, 55.94%) (
Influenza-A | Pandemic H1N1 | Seasonal-A | |
OPD cases | 4870 | 1944 | 2926 |
IPD cases | 397 | 182 | 215 |
Total cases | 5267 | 2126 | 3141 |
Hospitalization ratio IPD/Total cases (%) | 7.5 | 8.6 | 6.8 |
A total of 93 deaths were recorded in Pune district from 1 August to 31 October 2009. Of these, 57 and 8 deaths from Pune (urban) and 27 and 1 from Pune (rural) were among pandemic H1N1 positive and seasonal-A positive cases respectively. Of the 9 deaths associated with the seasonal influenza, 2 were positive for seasonal H1N1 and 7 for H3N2 viruses.
The number of cumulative deaths in pandemic H1N1 positive cases, presented in
(a) IPD cases and the deaths in Pune (Urban) (b) Death cases in Pune.
(a) Age-wise distribution of death cases in Pune (b) The normalized age specific CFR.
In Pune (urban), the overall CFR for pandemic H1N1 cases (0.86%) was significantly higher than that of seasonal-A cases (0.13%) [P<0.001, OR 5.8, 95% CI (2.71, 13.34)]. The odds of death in hospitalized pandemic H1N1cases was about 6 times higher than seasonal-A cases. As suggested by Garske et al (2009), the precision of the CFR estimates depends upon the sample size of the cases in the first stage of the study. To obtain 95% confidence interval, around 1100 total cases and 200 hospitalizations are required to estimate CFR in the range of 0.5 to 1.5%. Therefore, in this study, the sample size for influenza-A (397 hospitalization and 5267 total cases) and pandemic H1N1 (182 hospitalizations and 2126 total cases) meets the criteria and the estimated CFR of 0.49 and 0.86% respectively can be considered precise at 95% confidence interval. At least 1300 total cases and 500 hospitalizations are needed for estimation of CFR of about 0.2%. The precision at 0.2% CFR cannot be obtained with 200 hospitalizations, no matter how large is the initial sample. Thus, the estimated CFR for seasonal influenza though useful for comparison with the pandemic H1N1, may not represent the true estimate. The CFR for different months of the study period varied between 0.49 and 0.43 for influenza-A, 0.67 and 1.07 for pandemic H1N1 and 0 and 0.19 for seasonal-A (
Influenza-A | Pandemic H1N1 | Seasonal-A | Period | |
Hospitalization rate (Hr) (%) | 7.5 | 8.6 | 6.8 | 4 August to 3 September |
Deaths/IPD cases (D) | 29/397 | 23/185 | 6/212 | August 2009 |
CFR (%) |
0.55 | 1.07 | 0.19 | |
Death/IPD cases (D) | 23/398 | 21/217 | 2/181 | September 2009 |
CFR (%) |
0.43 | 0.83 | 0.08 | |
Death/IPD cases (D) | 13/198 | 13/167 | 0/31 | October 2009 |
CFR (%) |
0.49 | 0.67 | 0 | |
Total deaths/total IPD cases (D) | 65/993 | 57/569 | 8/424 | 1 August to 31 October 2009 |
Overall CFR (%) |
0.49 | 0.86 | 0.13 |
* CFR = Hr*D.
As observed in some countries of southern hemisphere
Clearly, both seasonal-A and pandemic H1N1 viruses were highly active and present in almost equal proportions during the study period. It provided unique opportunity to study impact of these viruses independently and also to compare their severity. The epidemic started in monsoon season and continued with cyclic pattern. This pattern is similar to that observed for seasonal influenza in tropics
We believe that underreporting during initial phase of the epidemic was minimal due to the highly vigilant health department which sprang into action after the first death in Pune. Significant underreporting of actual OPD cases after the first few weeks of the infection occurred due to change in policy but IPD admissions and fatality reporting continued properly even in the middle of the epidemic. This fact was obvious as two distinct waves were seen in IPD cases but second wave was diminished in OPD data. Hospitalization ratio reported here is in line with the reports from USA and Canada
The relationship between hospitalization and death cases observed by us and in USA and Canada
The significantly higher CFR for pandemic H1N1 in comparison to seasonal-A noted in this study indicates that pandemic H1N1 is significantly more severe than seasonal-A in tropical settings like India. The rate is much higher than that reported from Canada, USA and European countries but matches with the initial CFR observed in Mexico
The CDC proposed an important and new concept ‘Pandemic severity index’ (PSI) as a new pandemic influenza planning tool for planning at different places
CFR at country level is often derived using many generalized assumptions, the validity of which is rarely proved. In a tropical country like India, large variations have been noticed in the morbidity and mortality during this pandemic. Higher incidence of influenza is associated with rainy and winter seasons though activity continues throughout the year. Therefore, different areas may experience different levels of influenza related hospitalizations and fatalities at different times. It is imperative that systematic studies be conducted in multiple places to arrive at a reasonable estimate of severity of the pandemic and seasonal influenza viruses. Mapping of the risk factors to delineate risk zones may provide better country plans.
This report on severity of influenza infection is based on actual systematically collected epidemiologic data. Such information could not be provided earlier due to the lack of dedicated surveillance of this magnitude. Earlier projections were based on estimates derived from parallel data collected from other parts of the world. The similarity of CFR between India, Mexico
In summary, in India, pandemic H1N1 virus was associated, with more severe disease outcomes both in terms of hospitalization and mortality. The severity of pandemic H1N1 is lower than that reported for ‘Spanish flu, 1918’ but much higher than reported for other pandemics of 20th century. Comparatively, the seasonal influenza produces milder disease with much less mortality.
Pune, located at 18° 31′ North latitude and 73° 51′ East longitude, is one of the 35 districts in Maharashtra state, India. The population of the district is about 7.2 million and of this about 4 million live in Pune Urban Agglomeration (urban). In the present study, data on hospitalization and samples tested in laboratory were obtained only for Pune (urban). However, the data on death cases were available for Pune (urban) and also the other areas in the district, referred as Pune (rural).
A suspected case of influenza like infection (ILI) was a person with acute respiratory illness who had fever or a recent history of fever and sore throat. A suspected case of pandemic H1N1 was a suspected case of ILI who also had an epidemiological link with a confirmed pandemic H1N1 virus infection. A confirmed case of pandemic H1N1 was a ILI case with laboratory confirmation of pandemic H1N1 virus by real time reverse transcriptase polymerase chain reaction (rRT-PCR). A confirmed case of seasonal-A influenza was a suspected case of ILI with laboratory confirmation of influenza-A by rRT-PCR but negative for pandemic H1N1. It included un-typable as well as seasonal H1N1 or H3N2 viruses. A confirmed case of influenza-A included the cases positive for pandemic H1N1 and seasonal -A.
Throat and nasal swabs from the suspected cases, presenting at outpatient departments (OPD) or admitted in-patient departments (IPD) of the respiratory units in city hospitals, both in public and private sectors were collected and transported to the laboratory in transport medium on ice. All samples were tested by rRT-PCR following the Centers for Disease Control and Prevention (CDC) protocol
Following the first death on 3 August, an unprecedented level of surveillance was mounted by the government to monitor influenza cases. Regular daily meetings involving senior government officials, health officials and hospital representatives were held by the district administrative authorities. Duly filled forms submitted by each hospital were reviewed and systematic reporting of number of patients screened, admitted and outcome for each hospitalized patient was ensured.
Initially, due to strong media hype and high level of government attention, the turnout of patients with any symptom of influenza was very large and almost all the affected persons presented themselves to the screening centers. Many such centers were opened across the city to facilitate active surveillance of the pandemic. Soon it was impossible to test all the cases and provide reports in a timely manner. This demanded changes in policy and the suspected patients were provided with Oseltamivir without waiting for laboratory results from the second week of September onwards. Thereafter, only limited samples, collected at the discretion of the clinicians, from OPD were submitted for testing. However, systematic sample collection from all the admitted patients was continued throughout the study period
CFR, the percentage of deaths out of the total confirmed cases of the disease
The normalized age-specific CFR was calculated following de Silva et.al
On occasions, when all the samples could not be presented to the laboratory or all the presented samples could not be tested, the number of positive cases was derived by multiplying the recorded cases with the ratio obtained in the samples tested in laboratory.
The ethical clearance for the study was not required since samples were referred to us for diagnosis as a public health response to mitigate the pandemic. For the data presented in this study, the participating patient information remained anonymous.
The authors gratefully acknowledge the support and hard work done by all the members of pandemic influenza teams both from the institute and the state government during the pandemic. We also acknowledge support received from Centre for Disease Control, Atlanta, USA for providing the diagnostic kits and positive controls.
NIV technical team: Buwa B, Chowdhury DT, Dakhave M, Guru Kumar, Hingulkar R, Jadhav S, Joseph B, Kadam A, Karambelkar SM, Kawade SN, Keng S, Kengale HK, Khirawale A, Kulkarni PB, Kulkarni VS, Kumbhar NS, More B, Naik SS, Patil GB, Phaltane RM, Rangole M, Salunkhe AS, Sonar N, Tikhe SA, Walimbe A.
Government of MaharashtraTeam: Chandrakant Dalvi, Collector Pune; Mahesh Zagade, Commissioner PMC; Asheesh Sharma, Commissioner PCMC; Dr. Ashok Ladha, Jt. Director State Health Services; Dr. U.H. Gawande, Deputy Director Pune Circle and Dr. Pradeep Awate, Assistant Director In-Charge state flu control room.