10
Death

The Vaccine Safety Committee was charged with assessing a causal relation between all of the vaccines that it reviewed and death. The myriad possible causes of death made this a difficult task. To facilitate review of this serious adverse event, the committee used a categorization scheme, which is discussed below. Because the vast majority of reports of death following vaccination reside in passive surveillance systems, the committee used the scheme to analyze data from a currently operating one, the Vaccine Adverse Event Reporting System (VAERS). A description of the system is included in this chapter, and the results of this analysis are also discussed. All evidence reviewed by the committee (published literature and reports from the passive surveillance systems of the Centers for Disease Control and Prevention [CDC] and U.S. Food and Drug Administration [FDA]) regarding deaths in association with immunization is discussed later in this chapter.

To answer the question "Can any deaths be attributed to the use of the vaccines discussed in this review?," the committee categorized reports of death following immunization into seven categories:

  • deaths for which the available data were insufficient to allow a judgment of cause;

  • deaths associated with vaccine administration but attributable to inappropriate handling, contamination, production error, or error of medical care;



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality 10 Death The Vaccine Safety Committee was charged with assessing a causal relation between all of the vaccines that it reviewed and death. The myriad possible causes of death made this a difficult task. To facilitate review of this serious adverse event, the committee used a categorization scheme, which is discussed below. Because the vast majority of reports of death following vaccination reside in passive surveillance systems, the committee used the scheme to analyze data from a currently operating one, the Vaccine Adverse Event Reporting System (VAERS). A description of the system is included in this chapter, and the results of this analysis are also discussed. All evidence reviewed by the committee (published literature and reports from the passive surveillance systems of the Centers for Disease Control and Prevention [CDC] and U.S. Food and Drug Administration [FDA]) regarding deaths in association with immunization is discussed later in this chapter. To answer the question "Can any deaths be attributed to the use of the vaccines discussed in this review?," the committee categorized reports of death following immunization into seven categories: deaths for which the available data were insufficient to allow a judgment of cause; deaths associated with vaccine administration but attributable to inappropriate handling, contamination, production error, or error of medical care;

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality deaths temporally associated with vaccine administration but clearly caused by something other than the vaccine; deaths classified as sudden infant death syndrome (SIDS); deaths that are a consequence of vaccine-strain viral infection (applies to measles, mumps, or oral polio vaccine [OPV] for this report); deaths that are a consequence of an adverse event that itself is causally related to a vaccine reviewed in this report; and deaths temporally associated with vaccine administration and the cause of death is other than those listed above. EXAMPLES Deaths for Which the Available Data Were Insufficient to Allow a Judgment of Cause The reports from passive surveillance systems accessed by the committee vary in the quantity and quality of the information that they contain. Many of the reports contained phrases such as "died" or "found dead at baby-sitter's." Reports with more information frequently had no additional documentation submitted with them, so assessment of the diagnosis was not possible. The information in VAERS is discussed below in great detail. The published literature contains reports of deaths following immunization that lack sufficient information for a causality assessment as well. This is most common in uncontrolled observational studies intended to give a broad picture of the results of immunization campaigns. Deaths Associated With Vaccine Administration but Attributable to Inappropriate Handling, Contamination, Production Error, or Error of Medical Care This category includes a wide range of contamination or handling problems; vaccines, like any other pharmaceutical agent, are subject to mishandling that might, in extreme cases, lead to death. The Cutter incident is a well-known example of vaccine contamination caused by errors in quality control by the manufacturer and lack of clear guidelines from a regulatory agency; 60 vaccine recipients and 89 contacts of recipients contracted polio as a result of contamination of two production pools of inactivated polio vaccine (IPV) with live virus in 1955 (Nathanson and Langmuir, 1963). Contamination can occur at a more local level. Sokhey (1991) reported several deaths following administration of measles vaccine in India. The report lists the cause of death as "toxic shock syndrome" and notes that contamination was likely because syringes and needles were reused and the sterilization procedures were unsatisfactory. Staphylococcus aureus was

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality isolated from a few available implicated vials. That report is discussed later in this chapter. Deaths Temporally Associated with Vaccine Administration but Clearly Caused by Something Other Than the Vaccine Passive surveillance systems contain many reports that fall into this category. Reports to the manufacturer or to the government regarding the death of a vaccine recipient in temporal relation to vaccination can be made before a cause of death is established. Once an autopsy is performed, it is sometimes clear that the death was temporally but not causally related to vaccination. An example of such a death is one that was reported to VAERS. This report describes the death of a 5-year-old 10 days after receipt of diphtheria and tetanus toxoids and pertussis vaccine (DPT), OPV, and measles-mumps-rubella vaccine (MMR). The cause of death was Haemophilus influenzae type b meningitis, which did not appear to be vaccine related. Deaths Classified as SIDS For many years, the standard immunization schedule in the first year of life (the period in which most cases of SIDS occur) included only DPT and polio vaccine. Use of hepatitis B and H. influenzae type b vaccines during the first year of life is increasing rapidly. Although the scientific question of interest is, "Does vaccination increase an infant's probability of dying of SIDS," the research has focused on the role of DPT, even though polio vaccine is often given with DPT. The previous Institute of Medicine report on rubella and pertussis vaccines (Institute of Medicine, 1991) concluded that the evidence favors rejection of a causal relation between DPT and SIDS. "Studies showing a temporal relation between these events are consistent with the expected occurrence of SIDS over the age range in which DPT immunization typically occurs" (Institute of Medicine, 1991, p. 141). A few studies that primarily investigated the role of DPT in SIDS also examined the role of polio vaccine (Bouvier-Colle et al., 1989; Hoffman et al., 1987; Taylor and Emery, 1982; Walker et al., 1987). Only Hoffman and colleagues (1987) report odds ratio estimates of the relative risk of a SIDS case infant being immunized with OPV (0.57 for age-matched controls and 0.61 for age-, race-, and low-birth-weight-matched controls). These odds ratios were very similar to those obtained for the relative risk from DPT immunization. The committee's evaluation of the causal relation between diphtheria and tetanus toxoids for pediatric use (DT) and SIDS is discussed later in this chapter. Passive surveillance systems such as the Monitoring System for Adverse Events Following Immunization (MSAEFI) and VAERS contain many

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality reports of SIDS that occurred within 24 or 48 hours following vaccination, but these case reports often contain inadequate information to substantiate the diagnosis, and they are not necessarily evidence of a causal relation, because cases will occur in the 24- to 48-hour period following vaccination by chance alone. The licensure and recommendations for administration in the first year of life of vaccines to prevent hepatitis B and H. influenzae type b infections would suggest that studies of a possible role of vaccines in SIDS will continue. There are no controlled studies of a relation between either of these two newer vaccines and SIDS. MMR is customarily given after the first year of life in the United States; therefore, a relation between MMR and SIDS has not been investigated. Deaths That Are a Consequence of Vaccine-Strain Vital Infection Measles and mumps vaccines and OPV are made up of live attenuated viruses. It is plausible that administration of a live viral vaccine could cause a systemic infection in the recipient that could, in certain circumstances, be fatal. The committee identified a few such reports; they are discussed later in this chapter. Deaths That Are a Consequence of an Adverse Event That Itself Is Causally Related to a Vaccine Reviewed in This Report Several of the adverse events studied by the committee can lead to death. If the adverse event was caused by the vaccine and led to death, then one may say that the vaccine caused the death. An example would be a patient who suffers fatal anaphylaxis associated with vaccine administration. If the evidence favors the acceptance of (or establishes) a causal relation between a vaccine and an adverse event and that adverse event can be fatal, then in the committee's judgment the evidence favors the acceptance of (or establishes) a causal relation between the vaccine and death from that adverse event. The case fatality rate for adverse events associated with vaccines (other than, possibly, the live viral vaccines) should not be different from that for adverse events associated with all other causes. For example, anaphylaxis caused by a vaccine should be no more or less likely to result in death than anaphylaxis precipitated by any other antigen. It is plausible, however, that some adverse events caused by an attenuated viral vaccine might be milder than that same adverse event caused by the wild-type virus. In later sections of this chapter, the committee discusses the data regarding death from adverse events that are causally related to vaccines. If the evidence is inadequate to accept or reject a causal relation between a vaccine and an adverse event, then in the committee's judgment

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality the evidence is inadequate to accept or reject a causal relation between the vaccine and death resulting from that adverse event. Deaths Temporally Associated with Vaccine Administration and the Cause of Death Is Other Than Those Listed above The committee explored the possibility that vaccines may cause death by mechanisms other than vaccine-strain viral infection or an adverse event that itself is causally related to vaccine administration. The committee considered whether it might have overlooked possible vaccine-related mechanisms or pathways that could lead to death. The committee was unable to hypothesize such causes. However, had the committee identified reports of death following vaccination that did not fall into any of the other six categories, de facto those reports would have been placed into this category and causality would have been assessed for those reports. The committee found no reports of death that could be placed in this category, either in theory or by exclusion from the other causes listed above. REPORTS OF DEATH IDENTIFIED FROM VAERS The preponderance of data concerning death as an adverse consequence of vaccination comes from passive surveillance systems. The committee made a concerted effort to evaluate these data. The number of reports in VAERS of death in temporal association with vaccination has been the topic of presentations to the Advisory Commission on Childhood Vaccines (ACCV), the Advisory Committee on Immunization Practices, an FDA-sponsored workshop on contraindications to vaccination, a public session regarding changes to the Vaccine Injury Table, and the Vaccine Safety Committee of the Institute of Medicine. Because of this interest, because VAERS has a number of advantages over its predecessor data bases (the Spontaneous Reporting System [SRS], which was run by the FDA, and MSAEFI, which was run by the CDC), and because VAERS contains reports of reactions to hepatitis B vaccine (particularly in infants and children) and H. influenzae type b vaccine (which would be scarce in MSAEFI and SRS because of the recent licensure of these vaccines), the committee focused its investigations of reports to the U.S. government-run passive surveillance systems of death after vaccination on the information contained in VAERS. A description of VAERS and a discussion of its strengths and weaknesses will help in the analysis of the data that follow.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality History and Background VAERS was established by the National Childhood Vaccine Injury Act of 1986 (P.L. 99-660) to collect reports of adverse reactions following vaccination. The database is managed by a contractor under the aegises of both the CDC and the FDA. Reporting forms are available in drugstores and elsewhere, and anyone may file a report. A blank form appears in Appendix B (Figure B-1). A toll-free phone number (800-822-7967) can be used instead of the form. Physicians and nurses may report an adverse event, as may the parents, relatives, or neighbors of a patient. Health care providers are obligated to report specific adverse reactions to vaccines covered by the National Vaccine Injury Compensation Program (see box entitled The Vaccine Injury Table; see Chapter 1 for more information on the compensation program); however, reports can be made regarding any reaction to any vaccine. Several people may submit reports about one patient. The person reporting the adverse event may supply additional information, but no supporting information is required. The form asks for the following items: the patient's initials, age, and sex; the time of event onset; description of events; general assessment of condition (dead, hospitalized, disabled, etc.); relevant diagnostic tests or laboratory data; date of vaccination; and the vaccines that were given. Once submitted to the contractor, the data are entered into a computerized database. The narrative regarding the description of the adverse event is used to determine the index terms used to categorize the adverse reaction. Up to eight terms can be assigned to each report. The indexing system used is the Coding Symbol for Thesaurus of Adverse Reaction Terms (COSTART) system. VAERS began operation in November 1990. By July 31, 1992, there were over 17,000 reports in VAERS, almost 11,000 of which concerned vaccines covered by the National Vaccine Injury Compensation Program. Of the total number of reports, just over 2,500 of them were considered to be "serious," which is defined as the following: the patient died, suffered a life-threatening illness, or suffered a reaction that resulted in, or prolonged, hospitalization or that resulted in permanent disability. Strengths and Weaknesses The ease and accessibility of the reporting system might be useful in allowing VAERS to cast a large net, bringing attention to any possible adverse event. The negative aspect of this open system is that data may be inaccurate, poorly documented, or incomplete and that records may be duplicated. Although VAERS may be useful as a monitor for detecting ad-

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality THE VACCINE INJURY TABLE Vaccine/Toxoid Event Interval from Immunization DPT, pertussis vaccine A. Anaphylaxis or anaphylactic shock 24 hours DPT/poliovirus combined       B. Encephalopathy (encephalitis)a 3 days   C. Shock-collapse or hypotonic-hyporesponsive collapseb 3 days   D. Residual seizure disorderc 3 days   E. Any acute complication or sequela (including death) of above event which arose within the time period prescribed No limit Measles, mumps, and rubella; DT, Td, tetanus toxoid A. Anaphylaxis or anaphylactic shock 24 hours   B. Encephalopathy (or encephalitis)a 15 days for measles, mumps and rubella vaccines; 3 days for DT, Td and tetanus toxoids   C. Residual seizure disorderc 15 days for measles, mumps and rubella vaccines; 3 days for DT, Td and tetanus toxoids   D. Any acute complication or sequela (including death) of above event which arose within the time period prescribed No limit Oral poliovirus vaccine A. Paralytic poliomyelitis     —in a nonimmunodeficient recipient 30 days   —in an immunodeficient recipient 6 months   —in a vaccine-associated community case Not applicable

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality   B. Any acute complication or sequela (including death) of above event which arose within the time period prescribed No limit Inactivated poliovirus vaccine A. Anaphylaxis or anaphylactic shock 24 hours   B. Any acute complication or sequela (including death) of above event which arose within the time period prescribed No limit NOTE: Events listed are required by law to be reported by health care providers to the U.S. Department of Health and Human Services; however, VAERS will accept all reports of suspected adverse events after the administration of any vaccine. Aids to interpretation follow as footnotes. a Encephalopathy means any substantial acquired abnormality of, injury to, or impairment of brain function. Among the frequent manifestations of encephalopathy are focal and diffuse neurologic signs, increased intracranial pressure, or changes lasting ≥6 hours in level of consciousness, with or without convulsions. The neurologic signs and symptoms of encephalopathy may be temporary with complete recovery, or they may result in various degrees of permanent impairment. Signs and symptoms such as high-pitched and unusual screaming, persistent uncontrollable crying, and bulging fontanel are compatible with an encephalopathy, but in and of themselves are not conclusive evidence of encephalopathy. Encephalopathy usually can be documented by slow wave activity on an electroencephalogram. b Shock-collapse or hypotonic-hyporesponsive collapse may be evidenced by signs or symptoms such as decrease in or loss of muscle tone, paralysis (partial or complete), hemiplegia, hemiparesis, loss of color or change of color to pale white or blue, unresponsiveness to environmental stimuli, depression of or loss of consciousness, prolonged sleeping with difficulty arousing, or cardiovascular or respiratory arrest. c Residual seizure disorder may be considered to have occurred if no other seizure or convulsion unaccompanied by fever or accompanied by a fever of <102ºF occurred before the first seizure or convulsion after the administration of the vaccine involved, AND, if in the case of measles-, mumps-, or rubella-containing vaccines, the first seizure or convulsion occurred within 15 days after vaccination OR in the case of any other vaccine, the first seizure or convulsion occurred within 3 days after the vaccination, AND, if two or more seizures or convulsions unaccompanied by fever or accompanied by a fever of <102ºF occurred within 1 year after vaccination. The terms seizure and convulsion include grand mal, petit mal, absence, myoclonic, tonic-clonic, and focal motor seizures and signs. SOURCE: Adapted from Public Law 99-660.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality verse events, it is less useful for scientific analysis and assessments of causality (Chen et al., submitted for publication). Two issues, data quality and record duplication, limit the usefulness of the VAERS data set for analysis. Under the category of data quality are the concerns of validity, documentation, and completeness of information. Because the forms can be submitted by medical and nonmedical personnel, there is bound to be variability in the quality (and availability) of information. Even among health care providers there can be great variation in the use of diagnostic terms. Although this variation in diagnosis may be random (and not biased by the reporter's concern about the contributory effect of the vaccine), death is such a serious outcome that one would like to have information that confirms or clarifies the VAERS reports. Linkage to hospital records, autopsy reports, laboratory files, and death certificates could help validate the information. Besides accurate data regarding the cause of death, one would wish to confirm the validity of key variables such as the age of the patient at the time of vaccination, type of vaccination, calendar date (and perhaps time of day) of vaccination, date of onset of adverse event, and date of death. These would be minimal data requirements for a preliminary analysis of the overall pattern of deaths following vaccination. Table 10-1 illustrates some of the problems associated with VAERS reports. The committee assembled a table of 28 samples of VAERS reports of deaths. The sample was not necessarily random or representative, but examples were chosen to illustrate the problem of data quality. In the majority of cases, the cause of death was not stated, and there was no information to indicate whether an autopsy was performed. The VAERS data by themselves frequently do not supply enough information to allow informative data analysis. Some VAERS reports contain enough reliable information that the report can be evaluated and weighed as one would a published case report. For example, one specific report was filed by a nurse, the cause of death was stated to be H. influenzae type b meningitis, and laboratory tests were performed. The death occurred 10 days following immunization with DPT, OPV, and MMR. Such VAERS reports were often very useful to the committee in its assessments of all adverse reactions reviewed, not just the reports of death. Multiple reports relating to the same adverse event in a given patient are an obvious disadvantage of analyzing the VAERS database. Many duplicates can be found by going through the VAERS reports by hand, but one cannot be certain that all duplicate reports have been identified. Verification of data quality would probably help to eliminate more duplicate records. There are other problematic aspects of VAERS, most notably, the problem of underreporting. The accessibility of VAERS should encourage re-

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality porting, but underreporting of events may still occur. The possible incompleteness of the data hampers any attempt to analyze the data. VAERS is a passive reporting system that could serve as a sentinel for as yet unsuspected adverse events that occur following immunization. If linked to other detailed and accurate information, such a surveillance system could be a database for scientific analysis. Follow-up of Reports of Death to VAERS The FDA followed up all reports of death following immunization submitted to VAERS during the time period July 1, 1990, to September 30, 1991 (R. P. Wise, FDA, presentation to the ACCV, December 1992). Of 235 reports, 29 were duplicates, leaving 206 unduplicated reports. The submitted data were considered to be adequate for 78 of those reports. Complete follow-up was achieved for an additional 81 reports. Thus, the number of reports with completed assessments was 159. Follow-up was done by FDA staff. Most queries were by telephone, and a few were by mail. There were 3.5 calls per clarified report and 2.5 completed calls per clarified report. The typical succession of calls was first to the reporting physician's office and then one or more referrals to the hospital medical records department, the coroner or medical examiner's office, and occasionally, the county recorder or state public health vaccination department. As a result, for 159 VAERS reports the diagnosis was confirmed or clarified by the FDA. Before analyzing the reports, the committee excluded 17 VAERS reports because for 16 reports the deaths followed administration of a vaccine or vaccines not under consideration by the present committee and one report concerned a miscarriage, which the committee classified as an adverse event in temporal relation to vaccination experienced by the woman who was vaccinated. All of the 90 cases of SIDS reported to VAERS during this 15-month period occurred within 28 days of vaccination, and 78 occurred within 7 days of vaccination. There have been no changes in the incidence of SIDS over recent years; each year, on the order of 5,000 deaths in the United States are classified as SIDS (Little and Peterson, 1990). The committee calculated the number of SIDS deaths that might be expected to occur within 7 days of vaccination if vaccination had no effect on the incidence of SIDS. Assuming that all infants receive three separate immunizations during the first year of life, there are three 7-day periods when a SIDS death would occur within 7 days of vaccination by chance alone. Of the SIDS deaths in a 1-year period, 288 deaths (21/365 of the total 5,000 SIDS deaths in 1 year) would occur within 7 days following a vaccination. If it is assumed that only half of all children receive their immunizations (a low estimate), one might expect 144 SIDS deaths in the 7 days following vaccination.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality TABLE 10-1 Quality and Quantity of Data in a Sample of VAERS Reports of Death Following Immunization Report No. Was Vaccine Suspected? Is Report by an MD or RN? Is Cause of Death Stated? Was an Autopsy Performed? 1 No information No No information No information 2 No No Yes, meningitis No 3 No information Yes—MD No No information 4 No information Yes—RN No No information 5 No information No information No No information 6 No information No information No No information 7 No information Yes—MD No No information 8 No information Yes—RN No No information 9 No information No information No No information 10 No No information No Yes 11 No information Yes—MD No No information 12 No information Yes—MD No No information 13 No information Yes—RN No No information 14 No information No information No No information 15 No information Yes—RN No No information 16 No information Yes—RN No No information 17 No information Yes—MD No No information 18 No information Yes—RN Yes, H. influenzae type b meningitis No information 19 No information No information No No information 20 No information Yes—MD No Yes 21 No information No information No No information 22 No information Yes—RN No No information 23 No information; drug company summary Yes No No information 24 Not sure; drug company report No information No Yes 25 No; drug company report No No No information 26 No information; drug company report No No No information 27 No information No information Yes, meningitis No information 28 No; drug company duplicate reports No information No No information

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality vaccine strain was unidentified). Apparently, all 12 children had adverse reactions and 6 survived. The ages of those who died ranged from 9 to 18 months. This extraordinary mortality rate was attributed by the author to TSS and, specifically, the use of unhygienic conditions in administering the vaccine. A second incident involved the death of the only child (age 8 months) immunized at a particular location. Death occurred within 24 hours of vaccine administration and was attributed to TSS. In a third incident, 1 of 33 immunized children died following measles vaccination. This child was 23 months old and died of TSS 17 hours after immunization. A fourth incident involved the death of one of seven immunized children. The 16-month-old child became semiconscious within 30 minutes of receiving the vaccine and died approximately 40 hours later. The death was attributed to underlying hemorrhagic diathesis. Deaths Temporally Associated with Vaccine Administration, and the Cause of Death Is Other Than Those Listed Above Miller (1982) described a case series of 10,035 children who were vaccinated between 1970 and 1980 in Oxford, England, with the Beckenham 31 and Schwarz measles vaccine strains. One 12-month-old child died 36 hours after the vaccination. The report indicated that the coroner's death certificate stated "sudden death in infancy and acute bronchiolitis" (Miller, 1982, p. 535). Causality Argument The data relating death and measles or mumps vaccine are from case reports and case series. The largest series comes from India, but TSS caused by the unhygienic conditions involved in the immunization program was the apparent cause of death reported for eight of nine patients. Evidence based on RNA sequencing techniques has linked measles vaccine and measles infection to subsequent death in some severely immunocompromised children. In contrast, studies of the immunogenic response to measles vaccine in children infected with human immunodeficiency virus (HIV), which causes acquired immune deficiency syndrome (AIDS), have not recorded any deaths from measles vaccine-strain viral infection. The evidence favors the acceptance of a causal relation between measles vaccine and anaphylaxis. The evidence establishes a causal relation between MMR and thrombocytopenia and anaphylaxis (see Chapter 6). Anaphylaxis and thrombocytopenia can be fatal. Although there is no direct evidence of death as a consequence of measles vaccine-related anaphylaxis or of MMR-related thrombocytopenia or anaphylaxis, in the committee's judgment measles vaccine could cause fatal anaphylaxis and MMR could cause fatal thrombocytopenia or anaphylaxis. There is no evidence or reason to believe that the case fatality rate from measles vaccine-related

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality thrombocytopenia or anaphylaxis would differ from the case fatality rates for these adverse events associated with any other cause. Conclusion The evidence establishes a causal relation between vaccine-strain measles virus infection and death. The conclusion is based on case reports in immunocompromised individuals and not on controlled studies. No relative risk can be calculated. However, the risk of death from measles vaccine-strain infection would seem to be extraordinarily low. The evidence establishes a causal relation between MMR and death from complications associated with severe thrombocytopenia. The evidence establishes a causal relation between MMR and death from anaphylaxis. There is no direct evidence for this; the conclusion is based on the potential of severe thrombocytopenia and anaphylaxis to be fatal. The risk would seem to be extraordinarily low. The evidence favors acceptance of a causal relation between measles vaccine and death from anaphylaxis. There is no direct evidence for this; the conclusion is based on the potential of anaphylaxis to be fatal. The risk would seem to be extraordinarily low. The evidence is inadequate to accept or reject a causal relation between measles and mumps vaccines and death from causes other than those listed above. Risk-Modifying Factors There is evidence that some severely immunocompromised children, such as those with severe combined immunodeficiency syndrome, dysgamma-globulinemia, or leukemia, are susceptible to overwhelming measles infection and subsequent death, even from attenuated measles vaccine. Infection with HIV has not been associated with death from measles vaccine-strain viral infection. Polio Vaccines Deaths Classified as SIDS For many years, the standard immunization schedule in the first year of life (the period in which SIDS occurs) included only DPT and polio vaccine. The research on SIDS has focused on DPT. A few studies that primarily investigated the role of DPT in SIDS also looked at polio vaccine (Bouvier-Colle et al., 1989; Hoffman et al., 1987; Taylor and Emery, 1982; Walker et al., 1987). Only Hoffman and colleagues (1987) report odds ratio

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality estimates of the relative risk of a SIDS case infant being immunized with OPV (0.57 for age-matched controls and 0.61 for age-, race-, and low-birth-weight-matched controls). These odds ratios were very similar to those obtained for the relative risk of SIDS after DPT immunization. A small case-control study conducted by Taylor and Emery (1982) concluded that infants who had received DT and polio vaccine were not more likely to have died "unexpectedly" than age-matched controls. Mobius and colleagues (1972) described 13 deaths that occurred in the 21 days following receipt of polio vaccine over the years 1959 to 1968 in the former East Germany. Three deaths labeled sudden death followed administration of OPV and one labeled fever and sudden death followed administration of IPV. Passive surveillance systems such as MSAEFI and VAERS contain many reports of SIDS occurring within 24 or 48 hours following vaccination, but these case reports are not necessarily evidence of an association or a causal relation, because cases will occur in the 24- to 48-hour period following vaccination by chance alone. Deaths That Are a Consequence of Vaccine-Strain Viral Infection One VAERS report states that molecular biologic techniques were used to identify vaccine-strain poliovirus from the myocardium of a 3.5-month-old baby who died of myocarditis 4 days after receiving his second doses of OPV and DPT. On the basis of details in the VAERS report, the committee believes this report concerns the same baby described to the committee at its May 1992 public meeting (see Appendix B). Deaths That Are a Consequence of an Adverse Event That Itself Is Causally Related to a Vaccine Reviewed in This Report The evidence favors acceptance of a causal relation between OPV and GBS. The evidence establishes a causal relation between OPV and paralytic poliomyelitis (see Chapter 7). The committee identified no reports of death following GBS that occurred in temporal relation to OPV immunization. In the committee's review of over 90 papers reporting cases of vaccine-associated polio, the committee found 6 papers that described deaths among patients with vaccine-associated polio. These deaths are summarized in Table 10-2. Two papers described the same case (Davis et al., 1977; Saulsbury et al., 1975) of a female infant with immunodeficiency associated with cartilage-hair hypoplasia who contracted polio but who died of pneumonia 3 months after the diagnosis of polio. Two other deaths reported in the United States (Chang et al., 1966; Gaebler et al., 1986) also occurred in immunodeficient individuals. The sixth paper describes the first year of Peru's vaccination

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality TABLE 10-2 Deaths Among Patients with Vaccine-Associated Polio Reference Case Description Davis et al., 1977 Female diagnosed with polio at age 8 mo, diagnosed with dwarfism because of cartilage-hair hypoplasia at 10 mo, died of pneumonia at age 11 mo Gross et al., 1987 40-yr-old man receiving long-term steroid therapy; he developed polio 51 days after daughter's immunization with OPV and died of cardiac arrest I year later Gaebler et al., 1986 Infant boy received typical course of immunizations and developed polio at age 7.5 mo; he died at age 21 mo: the authors considered boy to be immune deficient, but evidence of vaccine-associated polio was not definitive Saulsbury et al., 1975 Same as Davis et al., 1977 Roedenbeck and Diaz. 1967 Fourteen deaths among patients with polio occurring in the first year (1966) of Peru's vaccination program Chang et al., 1966 A 7-yr-old boy diagnosed with ''vaccine-like'' polio and then discovered to be hypogammaglobulinemic; he died of pneumococcal meningitis more than 1 yr following diagnosis of polio program (Roedenbeck and Diaz, 1967). The authors described 14 deaths that occurred in children who were recently immunized with OPV. This was at a time when wild-type poliovirus was still circulating; the cases of polio and the deaths may therefore have been associated with wild-type strains. A third case occurred in a person who had been on long-term steroids and might have been immunocompromised (Gross et al., 1987). The CDC reported five deaths that occurred in people with vaccine-associated polio over the years 1980 to 1991 (Division of Immunization, Centers for Disease Control, personal communication, 1993). Four of the deaths occurred in immunodeficient or immunosuppressed patients. The fifth patient was not considered to be immunosuppressed but was described in the published case report described above (Gross et al., 1987) as possibly being immunocompromised because of long-term steroid use. There may have been other deaths among people with vaccine-associated polio, but the case reports in the literature do not always follow the patients long enough. It is interesting to note that all the deaths in the United States reported in the literature occurred in immunodeficient or immune-suppressed people. This suggests that immunodeficiency is a risk-modifying factor that increases the case fatality rate of polio. In such immunodeficient patients with polio, other infections contributed to the patients' deaths. The immunodeficient patients may therefore be at increased risk of dying compared with the risk for other people who contract vaccine-associated polio.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality Death from Other Causes Following Immunization Available Data Were Insufficient to Allow a Judgment of Cause Summary statistics from MSAEFI for the years 1979 to 1990 include 1 report of death following administration of OPV only, 1 report following IPV only, 445 reports following OPV plus other vaccines, and 1 report following IPV and other vaccines. VAERS (reports submitted between November 1990 and July 1992) contained 1 report of death following OPV only, 267 following OPV plus other vaccines, and 1 following IPV and other vaccines. Many VAERS reports contain insufficient data to assess a causal relation between polio vaccines and death. Deaths Temporally Associated with Vaccine Administration and the Cause of Death Is Other Than Those Listed Above Ehrengut and Ehrengut-Lange (1969) reviewed the deaths in which illness started in the 4 weeks following vaccination that occurred in Bavaria, Germany, during the years 1962 to 1964. They found that 26 such deaths occurred among 702,348 children aged 0 to 3 years. In all cases, poliomyelitis or polyradiculitis was ruled out as the cause of death. The main causes of death were respiratory infections (pneumonia, bronchitis, etc.). The authors did not comment on the causes of death, but suggested that mortality following receipt of polio vaccine was similar to that following receipt of smallpox vaccine in the same years. Mobius and colleagues (1972) described 13 deaths that occurred in the 21 days following receipt of polio vaccine over the years 1959 to 1968 in the former East Germany. Eleven deaths occurred following receipt of OPV and two occurred following receipt of IPV. The causes of death following receipt of OPV were as follows: three were "found dead," three were "sudden death," and there was one each of epiglottitis, aspiration, pneumonia, and ''unclear." One of the fatalities was in a child who died of myocarditis. The authors report that the child had been administered oral polio vaccine stains 1 and 3, however polio virus type II was isolated. The causes of death following IPV were convulsions and fever/sudden death. Causality Argument The evidence favors acceptance of a causal relation between OPV and GBS. The evidence establishes a causal relation between OPV and paralytic poliomyelitis in recipients or contacts (see Chapter 7). GBS and paralytic poliomyelitis can be fatal. Although there is no direct evidence of death as a consequence of OPV-induced GBS, in the committee's judgment OPV could cause fatal GBS. There are data regarding death from vaccine-strain polio infection; the data derive primarily from immunocompromised

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality individuals. There is no evidence or reason to believe that the case fatality rate for GBS or vaccine-associated poliovirus infection (including that resulting in paralytic poliomyelitis) is less than that for these adverse events associated with any other cause. The possible causal relation between polio vaccines and SIDS has rarely been studied. The evidence is inadequate to accept or reject a causal relation between polio vaccines and SIDS. Conclusion The evidence establishes a causal relation between OPV and death from vaccine-strain poliovirus infection, including infection that results in paralytic poliomyelitis. The conclusion is based on case reports and not on controlled studies. No relative risk can be calculated. However, the risk of death from OPV-related polio infection would seem to be extraordinarily low. The evidence favors acceptance of a causal relation between OPV and death from GBS. There is no direct evidence for this; the conclusion is based on the potential of GBS to be fatal. The risk would appear to be extraordinarily low. The evidence is inadequate to accept or reject a causal relation between polio vaccines and SIDS. The evidence is inadequate to accept or reject a causal relation between OPV and death from causes other than those listed above. Risk-Modifying Factors Immunodeficient individuals may be at greater risk of dying than immunocompetent persons if they contract vaccine-associated poliovirus infection. Hepatitis B Vaccine Deaths Classified as SIDS Hepatitis B vaccine has only recently been used in infants under 1 year of age. A possible causal relation between hepatitis B vaccine and SIDS has not been studied. There are reports in VAERS (submitted between November 1990 and July 1992) of SIDS occurring in temporal relation to hepatitis B vaccination.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality Deaths That Are a Consequence of an Adverse Event That Itself Is Causally Related to a Vaccine Reviewed in This Report The committee did not identify any reports of fatal anaphylaxis following hepatitis B vaccination. Death from Other Causes Following Immunization Available Data Were Insufficient to Allow a Judgment of Cause Two deaths following hepatitis B vaccination were reported to MSAEFI for the years 1979 to 1990. Both deaths occurred in vaccinees who concurrently received DPT and OPV. No other data regarding deaths were obtained from MSAEFI. The committee identified nine reports in VAERS (submitted between November 1990 and July 1992) of death that occurred in association with administration of hepatitis B vaccines. Six of these were associated with administration of hepatitis B vaccine only; the other three were associated with administration of hepatitis B vaccine in conjunction with other vaccines. The ages of the vaccinees ranged from 1 month to 70 years, and the interval from vaccination to death ranged from 1 to 30 days. Causality Argument The evidence establishes a causal relation between hepatitis B vaccine and anaphylaxis (see Chapter 8). Anaphylaxis can be fatal. Although there is no direct evidence of fatal anaphylaxis following hepatitis B vaccination, in the committee's judgment hepatitis B vaccine could cause fatal anaphylaxis. There is no evidence or reason to believe that the case fatality rate for vaccine-associated anaphylaxis would differ from the case fatality rate for anaphylaxis associated with any other cause. Hepatitis B vaccine has only recently begun to be administered to the age group that is affected by SIDS. There are no published studies of a possible causal relation between hepatitis B vaccine and SIDS. There are reports in VAERS of SIDS following immunization with hepatitis B vaccine given in conjunction with other vaccines. Conclusion The evidence establishes a causal relation between hepatitis B vaccine and fatal anaphylaxis. There is no direct evidence for this; the conclusion is based on the potential for anaphylaxis to be fatal. The risk would appear to be extraordinarily low.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality The evidence is inadequate to accept or reject a causal relation between hepatitis B vaccine and SIDS. The evidence is inadequate to accept or reject a causal relation between hepatitis B vaccine and death from any cause other than those listed above. Haernophilus influenzae Type b Vaccine Deaths Classified as SIDS There are no published studies of the relation between H. influenzae type b (Hib) vaccines and SIDS. Hib vaccines are licensed for administration to children under 1 year of age. There are reports in VAERS of SIDS occurring in temporal relation to Hib vaccine administration. As part of an investigation of VAERS reports of death following administration of Hib vaccines, the CDC has preliminary data suggesting that Hib vaccines do not appear to be causally related to SIDS (R. T. Chen, presentation to ACCV, March 1993). Deaths That Are a Consequence of an Adverse Event That Itself Is Causally Related to a Vaccine Reviewed in This Report The committee found no reports of death caused by Hib disease that occurred within 7 days of immunization with the unconjugated (polyribosylribitol phosphate [PRP]) Hib vaccine in children over 18 months of age who receive their first Hib immunization with unconjugated PRP vaccine, although in 10 instances outcome information was not provided. Death from Other Causes Following Immunization Available Data Were Insufficient to Allow a Judgment of Cause The committee identified 223 reports of death in association with the administration of Hib vaccine from VAERS between November 1990 and August 1992. Of these 223 reports, 17 were of children who received Hib vaccine alone. The ages of the children in the 223 reports ranged from 2 months to at least 36 months. The interval from the time of vaccination to death ranged from 1 to 210 days. Because data were missing from some VAERS reports, more precise information about the age of the child or the time interval from vaccination to death is unavailable. There were no reports to MSAEFI regarding death in association with Hib vaccination from 1979 to 1984. There were reports to MSAEFI of two deaths following Hib vaccination from 1985 to 1990. The recent licensure of the Hib vaccine makes this finding not surprising.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality Causality Argument The evidence favors rejection of a causal relation between conjugated Hib vaccines and early-onset Hib disease. The evidence favors acceptance of a causal relation between PRP vaccine and early-onset Hib disease in children 18 months of age or older who receive their first Hib immunization with unconjugated PRP vaccine (see Chapter 9). Hib disease can be fatal. Although there is no direct evidence of death as a consequence of early-onset Hib disease in children 18 months of age or older who receive their first Hib immunization with PRP vaccine, in the committee's judgment PRP vaccine could cause fatal Hib disease in children 18 months of age or older who receive their first Hib immunization with PRP vaccine. There is no evidence or reason to believe that the case fatality rate from PRP-associated Hib disease would differ from the case fatality rate from Hib disease not associated with PRP vaccine. Conclusion The evidence favors acceptance of a causal relation between PRP vaccine and death from early-onset Hib disease in children 18 months of age or older who receive their first Hib immunization with unconjugated PRP vaccine. There is no direct evidence for this; the conclusion is based on the potential for Hib disease to be fatal. The risk would appear to be extraordinarily low. The evidence favors rejection of a causal relation between conjugated Hib vaccines and death from early-onset Hib disease. The evidence is inadequate to accept or reject a causal relation between Hib vaccines and SIDS. The evidence is inadequate to accept or reject a causal relation between Hib vaccine and death from causes other than those listed above. REFERENCES Alderslade R, Bellman MH, Rawson NS, Ross EM, Miller DL. The National Childhood Encephalopathy Study: a report on 1000 cases of serious neurological disorders in infants and young children from the NCES research team. In: Department of Health and Social Security. Whooping Cough: Reports from the Committee on the Safety of Medicines and the Joint Committee on Vaccination and Immunization. London: Her Majesty's Stationery Office; 1981. Bellini WJ, Rota JS, Greer FW, Zaki SR. Measles vaccination death in a child with severe combined immunodeficiency: report of a case. Laboratory Investigation 1992;66:19A. Bouvier-Colle MH, Flahaut A, Messiah A, Jougla E, Hatton F. Sudden infant death and immunization: an extensive epidemiological approach to the problem in France—winter 1986. International Journal of Epidemiology 1989; 18:121-126.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality Chang TW, Weinstein L, MacMahon HE. Paralytic poliomyelitis in a child with hypogamma-globulinemia: probable implication of type I vaccine strain. Pediatrics 1966;37:630-636. Chen RT, Rastogi SC, Mullen JR, Hayes SW, Cochi SL, Donlon JA, et al. The Vaccine Adverse Event Reporting System (VAERS). Atlanta: Centers for Disease Control; submitted for publication. Christensen PE. Side reactions to tetanus toxoid. Scientific Publication, Pan American Health Organization 1972;253:36-43. Coffin CM, Roberts RL, Monafo WJ, Haslam DB, Bellini W J, Zaki S, et al. Disseminated measles infection following vaccination in severe combined immune deficiency syndrome. Society for Pediatric Pathology 1993; 13:102. Davis LE, Bodian D, Price D, Butler IJ, Vickers JH. Chronic progressive poliomyelitis secondary to vaccination of an immunodeficient child. New England Journal of Medicine 1977;297:241-245. Ehrengut W, Ehrengut-Lange J. Interkurrente todliche Erkrankungen nach Polioschluckimpfung und Pockenschutzimpfung. [Intercurrent fatal diseases after oral polio vaccination and smallpox vaccination.] Munchener Medizinische Wochenschrift 1969;111:1092-1099. Gaebler JW, Kleiman MB, French ML, Chastain G, Barrett C, Griffin C. Neurologic complications in oral polio vaccine recipients. Journal of Pediatrics 1986;108:878-881. Garenne M, Leroy O, Beau J-P, Sene I. Child mortality after high titre measles vaccines: prospective study in Senegal. Lancet 1991;338:903-907. Gross TP, Khurana RK, Higgins T, Nkowane BS, Hirsch RL. Vaccine-associated poliomyelitis in a household contact with Netherton's syndrome receiving long-term steroid therapy. American Journal of Medicine 1987;83:797-800. Haun U, Ehrhardt G. Zur problematik postvakzinaler Komplikationen nach Masernschutzimpfung. [Postvaccinal complications following protective vaccination against measles.] Deutsche Gesundheitswesen 1973;28:1306-1308. Hirayama M. Measles vaccines used in Japan. Reviews of Infectious Diseases 1983;5:495-503. Hoffman HJ, Hunter JC, Damus K, Pakter J, Peterson DR, van Belle G, Hasselmeyer EG. Diphtheria-tetanus-pertussis immunization and sudden infant death: results of the National Institute of Child Health and Human Development Cooperative Epidemiological Study of Sudden Infant Death Syndrome Risk Factors. Pediatrics 1987;79:598-611. Hong R, Gilbert EF, Opitz JM. Omenn disease: termination in lymphoma. Pediatric Pathology 1985;3:143-154. Institute of Medicine. Adverse Effects of Pertussis and Rubella Vaccines. Washington, DC: National Academy Press; 1991. Korger G, Quast U, Dechert G. Tetanusimpfung: Vertraglichkeit und Vermeidung von Nebenreaktionen. [Tetanus vaccination: tolerance and avoidance of adverse reactions.] Klminische Wochenschrift 1986:64:767-775. Kovalskaya SI. Anafilaktogennye svoistva adsorbipovannoi i neadsorbirovannoi kokliusnodifterino-stolbniachnoi vaktsin. [The anaphylactogenic properties of absorbed and nonabsorbed pertussis-diphtheria-tetanus vaccines.] Zhurnal Mikrobiologii, Epidemiologii, i Immunobiologii 1967;44:105-109. Landrigan PJ, Witte JJ. Neurologic disorders following live measles-virus vaccination. Journal of the American Medical Association 1973;223:1459-1462. Little RE, Peterson DR. Sudden infant death syndrome epidemiology: a review and update. Epidemiologic Reviews 1990:12:241-246. Mawhinney H, Allen IV, Beare JM, Bridges JM, Connolly HH, Haire M, et al. Dysgamma-globulinaemia complicated by disseminated measles. British Medical Journal 1971;2:380-381. Miller CL. Surveillance after measles vaccination in children. The Practitioner 1982;226:535-537.

OCR for page 274
Adverse Events Associated with Childhood Vaccines: Evidence Bearing on Causality Mitus A, Holloway A, Evans AE, Enders JF. Attenuated measles vaccine in children with acute leukemia. American Journal of Diseases of Children 1962;103:243-248. Mobius G, Wiedersberg H, Wunscher W, Backer F. Pathologisch-anatomische Befunde bei Todesfallen nach Poliomyelitis- und Dreifach-Schutzimpfung. [Pathological-anatomical findings in cases of death following poliomyelitis and diphtheria-pertussis-tetanus vaccination.] Deutsche Gesundheitswesen 1972;27:1382-1386. Nader PR, Warren RJ. Reported neurologic disorders following live measles vaccine. Pediatrics 1968;41:997-1001. Nathanson N, Langmuir AD. The Cutter incident: poliomyelitis following formaldehyde-inactivated poliovirus vaccination in the United States during the spring of 1955. I. Background. American Journal of Hygiene 1963;78:16-28. Pollock TM, Morris J. A 7-year survey of disorders attributed to vaccination in North West Thames region. Lancet 1983;1:753-757. Pollock TM, Miller E, Mortimer JY, Smith G. Symptoms after primary immunization with DTP and with DT vaccine. Lancet 1984;2:146-149. Regamey RH. Die Tetanus-Schutzimpfung. In: Herrlick A, ed. Handbuch der Schutzimpfungen. Berlin: Springer: 1965. Roedenbeck SD, Diaz C. Poliomielitis y vacuna oral. [Poliomyelitis and oral vaccine.] Revista de Neuro-Psiquiatria 1967;30:38-56. Saulsbury FT, Winkelstein JA, Davis LE, Hsu SH, D'Souza BJ, Gutcher GR, Butler, IJ. Combined immunodeficiency and vaccine-related poliomyelitis in a child with cartilage-hair hypoplasia. Journal of Pediatrics 1975;86:868-872. Sokhey J. Adverse events following immunization: 1990. Indian Pediatrics 1991;28:593-607. Staak M, Wirth E. Zur problematik anaphylaktischer Reaktionen nach aktiver Tetanus-Immunisierung. [Anaphylactic reactions following active tetanus immunization.] Deutsche Medizinische Wochenschrift 1973;98:110-111. Starke VG, Hlinak P, Nobel B, Winkler C, Kaesler G. Masernradikation—eine Moglichkeit? Ergebnisse und Erfahrungen mit der Masernschutzimpfung 1967 bis 1969 in der DDR [Measles eradication—a possibility? Results and experiences with the measles inoculation from 1967 to 1969 in the GDR.] Deutsche Gesundheitswesen 1970;25:2384-2390. Taylor EM, Emery JL. Immunization and cot deaths (letter). Lancet 1982;2:721. Walker AM, Jick H, Perera DR, Thompson RJ, Knauss TA. Diphtheria-tetanus-pertussis immunization and sudden infant death syndrome. American Journal of Public Health 1987;77:945-951.