Skip Navigation
  • Text Size: A A A
  • Print
  • Email
  • Facebook
  • Tweet
  • Share

National Vaccine Advisory Committee

Report on 2009 H1N1 Vaccine Safety Risk Assessment

Approved by the National Vaccine Advisory Committee on April 23, 2010
Approved by the Assistant Secretary for Health on April 27, 2010

Background

The National Vaccine Advisory Committee (NVAC) established the H1N1 Vaccine Safety Risk Assessment Working Group (H1N1 VSRAWG) with the charge to conduct independent, rapid reviews of available safety monitoring data for the 2009 H1N1 influenza vaccines.  Since the Working Group was created, it has met eleven times to review available data from the federal vaccine safety monitoring systems listed in Table 1. Based on the review and discussion of H1N1 safety data available as of its meeting on April 19, 2010, the Working Group has provided the following assessment for NVAC’s consideration on April 23, 2010, via telephone conference call. 

Report

Since our last report, an additional 136,800 doses of inactivated H1N1 and 17,800 doses of live attenuated H1N1 vaccine have been distributed through the immunization program.  A total of 105,097,420 doses of inactivated H1N1 and 21,755,200 doses of live attenuated H1N1 vaccine have been distributed as of March 31, 2010. Most H1N1 vaccine safety monitoring systems report a substantial slowing of distribution and administration of H1N1 vaccine.

To place this report in context, a review of the guidelines for the Working Group is provided.  Under the guidelines, summary data are presented to the members at the Working Group’s regularly scheduled meetings, including evidence of any potential link between the receipt of an H1N1 vaccine and an adverse event.  The Working Group then can make an assessment as to whether there is evidence of:

  • A signal, defined as adverse event occurrence following receipt of vaccine at a rate greater than anticipated based on chance alone;
  • An association, defined as a stronger linkage between vaccination and the event, varying with exposure to vaccine, as well as greater strength of association;
  • Causality, defined as the assessment that the adverse event is a direct consequence of an exposure to vaccine using Bradford-Hill criteria.

In most of the surveillance systems reviewed by the Working Group, screening for a signal is conducted by comparing the rate of a number of pre-specified diagnoses, as coded from inpatient and outpatient medical encounters, with the rates seen in comparison situations.  The potential of a signal is indicated by the rate in the H1N1-exposed group crossing a pre-specified statistical threshold.  It should be noted that the detection of a signal does not necessarily indicate an association, and that, per protocol, several steps must be taken to validate the signal.  To be assured that the signal is valid, individual cases must be reviewed to check for coding errors and for other supporting evidence of the diagnosis.  In addition, it should be noted that the Working Group does not conduct its own analyses, but critically reviews the analyses conducted by the various federal agencies and departments, questions aspects of the results and suggests additional considerations.

At the Working Group’s meeting on April 5, preliminary results were seen from five reporting systems for three adverse events for H1N1 monovalent inactivated vaccine: Guillain-Barré syndrome (GBS), thrombocytopenia/idiopathic thrombocytopenic purpura (TP/ITP) and Bell’s palsy (BP).  The Working Group requested further investigation and follow-up of these findings, which were reviewed during its meeting on April 19. 

Based on the data presented in Table 1, the Working Group concluded that the data are adequate to assess the presence or absence of a signal.  With regard to the specific adverse events above, the Working Group concluded the following:

1.  Guillain-Barré syndrome

A potential weak signal between H1N1 vaccine exposure and GBS was initially detected in the Emerging Infections Program (EIP) data.  Additional analyses with updated information on estimates of vaccine coverage and accounting for variation in the estimation of vaccination coverage rates further weakened the signal.  GBS surveillance is also being conducted in five other systems.  Although some systems are reporting elevated relative risks, none have crossed the threshold for a signal.  The Working Group will continue to monitor these data including an updated analysis from the EIP with data from March (current data is through February).  Of importance is the fact that, even if an association between H1N1 vaccine exposure and GBS were substantiated, the estimate is that the vaccine would account for only one extra case of GBS per 1 million persons vaccinated based on currently available data.

2. Bell’s palsy

A weak signal linking H1N1 vaccine exposure and BP emerged in two monitoring systems.  In one system, several analyses to examine this finding yielded inconsistent results with some comparisons providing support for the signal while others did not. 

3. Thrombocytopenia//idiopathic thrombocytopenic purpura

A weak signal between H1N1 vaccine exposure and thrombocytopenia also emerged in three systems.  In these systems the cases are being reviewed to see if the diagnoses are valid.  More rigorous comparisons between cohorts with H1N1 vaccine exposure and other vaccines or no exposure are planned.

When assessing the “strength” of the signal, we evaluated factors that are typically considered in assessing the level of concern include: strength of the association (e.g. elevated relative risk in a controlled study), temporal relationship between the receipt of the product and onset of the event, consistency of findings across available data, evidence of a dose response effect, potential biologic mechanisms linking the vaccine and the adverse event, and the rigor of the methodology and analyses being employed.   Since many analyses in several systems are being conducted simultaneously, the possibility that temporal associations will arise by chance alone is important to recognize. As designated in Table 1, a “weak signal” implies a low level of risk and/or substantial methodological limitations in data or study design.  Before any assessment of the association of vaccine exposure and adverse event is possible, several steps are needed to assure the validity of the findings and to explore potential alternatives that might result in a spurious association.

Thus, the Working Group concludes that the evidence suggests a weak signal between receipt of H1N1 vaccine and the indicated adverse events that requires further validation.  This validation is underway and includes:

  • More detailed review of the cases of adverse events to establish that the diagnoses are accurate and meets agreed upon definitions for the adverse events.
  • Additional analyses of the rates of adverse events under different comparison situations.
  • Review of the updated results at the next meeting of the VSRAWG on May 3, 2010.

The Working Group does not view these results as necessitating any immediate response by NVAC, but wishes that the NVAC be aware of progress to date.  In addition, all relevant federal agencies and departments are aware of these results, as they participate in the analyses and/or review calls.

The Working Group recommends that the federal government continue to monitor H1N1 vaccine safety as the body of evidence accumulates. Moreover, the Working Group suggests that NVAC recommend that federal agencies and departments prepare and present analytic plans to:

  • Assess the validity of the observations noted,
  • Define end-of-season analyses for adverse events,
  • Conduct analyses across monitoring systems to have sufficient power to assess the effect of H1N1 exposure on relatively rare events and to assess the consistency of estimates across the different databases.

All recommendations of the NVAC are made to the Department’s Assistant Secretary for Health.  The recommendation on vaccine safety monitoring listed above will be formally transmitted to the Assistant Secretary for Health, who will review and consider it for potential implementation options to include communications with various components of the Department.

H1N1 Vaccine Safety Risk Assessment Working Group Membership:

Stephen Cantrill, Associate Professor of Emergency Medicine, University of Colorado
John Clements, Professor of Microbiology and Immunology, Tulane University School of Medicine
Vicky Debold, Director of Research and Patient Safety, National Vaccine Information Center
Kathryn Edwards, Professor of Pediatrics, Vanderbilt University
Theodore Eickhoff, Professor Emeritus, University of Colorado School of Medicine
Susan Ellenberg, Professor of Biostatistics, University of Pennsylvania
Marie McCormick*, NVAC member, Professor of Maternal and Child Health, Harvard School of Public Health, former Chair of the IOM Immunization Safety Review Committee
Laura Riley, Assistant Professor of Obstetrics, Gynecology and Reproductive Biology, Massachusetts General Hospital
Mark Sawyer, Professor of Clinical Pediatrics, University of California, San Diego
*Chair of the NVAC H1N1 Vaccine Safety Risk Assessment Working Group

Table 1: Number of Persons Exposed to H1N1 Vaccine in Monitoring Systems Reviewed by the H1N1 VSRAWG
(Gray shading indicated a signal has been detected)

Vaccine Safety Program

Outcomes Monitored

Population Monitored

H1N1 MIV1 Exposures Captured in System

H1N1 LAMV2 Exposures Captured in System

Total H1N1 Vaccine Exposures Captured in System

Current as of

Analyses

Results

H1N1 Vaccine Trials (Manufacturer)

All health events

10,852

10,352a

500a

10,852a

01/25/10

Adjudication and analysis of SAE

No SAE related to vaccine

H1N1 Vaccine Trials (NIAID)

All health events

4,589

-

-

-

05/04/10

Adjudication and analysis of SAE

-

Vaccine Adverse Event Reporting System (VAERS)

All health events

US Population

105,097,420b

21,755,200b

126,852,620b

03/31/10

Comparison of reports for H1N1 versus seasonal influenza vaccines

SAE reporting after H1N1 are comparable to seasonal influenza immunization

Data-mining with comparison to similar vaccines

No signal

Vaccine Safety Datalink (VSD)

Pre-specified outcomes

9.5 million

1,288,496 a

263,930a

1,552,426a

03/27/10

Rapid cycle analysis

Weak signal (Bell’s palsy)

Real-Time Immunization Monitoring System (RTIMS)

All health events

US Population

8,625

1,133

9,758a

03/14/10

Symptoms that trigger an alert

No signal

Defense Medical Surveillance System (DMSS)

Pre-specified outcomes

1.4 million

1,211,821a

76,532a

1,288,353a

04/13/10

Rapid cycle analysis

Weak signal (TP/ITP)

Veteran’s Affairs (VA) Databases

1.Enhanced VAERS

 

2.Signal Detection

All health events

1.2 million

846,670b

 

-

846,670b

03/15/10

Comparison of reports for H1N1 versus seasonal influenza vaccines

No signal

Pre-specified outcomes

918,000

296,177a

-

296,177a

03/13/10

Rapid cycle analysis

Weak signal (TP/ITP)

Centers for Medicare and Medicaid Services (CMS)

Guillain-Barré syndrome

38 million

-

-

2,691,429a

04/12/10

Rapid cycle analysis

No signal

Indian Health Service (IHS)

Pre-specified outcomes

1.4 million

208,128a

51,856a

270,140

03/02/10

Rapid cycle analysis

Weak signal (Bell’s palsy & TP/ITP)

Post-Licensure Rapid Immunization Monitoring System (PRISM)

Pre-specified outcomes

30 million (17 million registry enhanced)

2,135,749a

66,180a

2,201,929a

03/13/10

Rapid cycle analysis

No signal

Guillain-Barré syndrome (GBS) enhanced surveillance

Guillain-Barré syndrome

45 million

-

-

-

04/12/10

Observed versus Expected Rates of GBS;

GBS incidence in vaccinated versus unvaccinated

Potential signal (GBS)

*Vaccines and Medications in Pregnancy Surveillance System (VAMPSS)

Maternal and fetal outcomes

3,100

-

-

-

-

-

-

1H1N1 monovalent inactivated vaccine
2H1N1 live attenuated monovalent vaccine
*Data not yet available
a Exposed to vaccine; b Doses of vaccine distributed
¥ Total is greater than sum of H1N1 MIV and LAMV as some H1N1 exposures are of unknown type