skip navigation

OffLine Spring 1999

Volume 11 Number 4

Immunizations rank among the biggest public health success stories of the twentieth century. Since their advent, many previously common diseases, such as measles, polio, and diphtheria have decreased dramatically in the U.S. and most countries worldwide. Measles incidence, for example, has plummeted from a pre-vaccine high of 894,000 in the U.S. in 1941 to just 488 in 1996. The irrefutable effectiveness of many vaccines has led to an increase in their use, with approximately 16 million doses currently administered each year to children under 2 in the U.S. Because vaccines are not perfectly safe, increased use translates into a higher number of associated adverse reactions—true reactions as well as those that are merely coincidental in time with the immunization. As a result, issues related to vaccine safety, particularly among children, have received increasing attention in recent years.

Seizure disorders, aseptic meningitis and diabetes are just a few of the conditions that have been postulated as adverse outcomes linked to childhood vaccinations. The majority of these potential adverse effects have been identified through "passive surveillance systems" which rely on spontaneous reports from physicians. Such systems, however, are inherently flawed in their ability to assess the size of risk or vaccine causality because of critical shortcomings such as biased reporting, under-reporting, and lack of denominators or comparison groups. The inadequacy of existing systems prompted The Institute of Medicine in the late 1980's to call for improved research to address "gaps in knowledge" about vaccine safety.

To address these knowledge gaps, the Centers for Disease control awarded contracts to 4 HMOs in 1991–1992 to prospectively study vaccine safety among children aged 0 to 6 years. These HMOs—Group Health, Southern California Kaiser, Northern California Kaiser and Northwest Kaiser—together enroll about 2 to 3  percent of all children in this age range in the U.S. at any point in time. The project was novel in its use of "large linked databases" to link computerized information on immunizations with computerized diagnoses of subsequent adverse outcomes. The advantage of these large databases, which contain information on more than 600,000 children under age 7, is that they permit the study of relatively rare adverse events. (Many of the adverse events potentially associated with vaccination probably occur in fewer than one child in 50,000). It was hoped that this rigorous method of monitoring vaccine safety would instill public confidence, produce new knowledge of vaccine side effects, and foster the development of safer vaccines.

This project, eventually named the Vaccine Safety Datalink (VSD), required enormous upfront effort to establish databases that met the contract needs. Group Health Center for Health Studies Investigator Robert S. Thompson, Principal Investigator of the Group Health portion of the VSD, coordinated a Coop-wide effort to this end. One of the results was a computerized immunization tracking system that has had a far-reaching impact on preventive care at Group Health. "Exceptional reports" generated from these systems enable Coop staff to proactively contact children who are not in compliance with their vaccinations. As a result, the proportion of continuously enrolled 2-year-olds with up-to-date immunization status increased from 63 percent in the first quarter of 1993 to 90 percent in the first quarter of 1995, at which level or higher it has remained.

Since the initial CDC contract in 1991, CHS' immunization-related research program has blossomed. The Immunization Studies Program (ISP), a broadened and renamed VSD headed by Thompson, supports 5 investigators and 20 staff members to conduct research on vaccines and vaccine-preventable diseases. The ISP continues its CDC-funded research on vaccine safety, expanded in 1995 to include adolescents and adults. In addition, the CDC and other funding sources, including private industry, sponsor work related to the testing of new combination vaccines, post-marketing surveillance, infectious disease epidemiology and immunization rates

Seizure disorders are adverse events that have been potentially linked to two common childhood immunizations, diphtheria-tetanus-pertussis (DTP) and measles-mumps-rubella (MMR). Evidence of these relationships, however, has been either minimal or inconsistent. CHS investigators Bob Davis and Bill Barlow and colleagues used the VSD data to further examine the relationship between first-time seizures and those vaccines in childhood. Their analyses distinguished between febrile seizures, those accompanied with fever, and non-febrile seizures. The latter, which include epilepsy, are much less common than febrile seizures.

The investigators found that DPT was associated with more than a 5-fold elevation in risk for febrile seizures on the day of the vaccine while MMR, a live virus vaccine, was associated with a 2.8 fold increase in febrile seizure risk 8 to 14 days following vaccination. These rates translate into approximately 5 additional febrile seizures per million DTP vaccinations administered, and 14 additional febrile seizures per million MMR vaccinations. They found no increased risk for non-febrile seizures or epilepsy. Although DPT and MMR vaccinations were associated with a transient increased febrile seizure risk, the investigators caution that this small risk should be weighed against the overall public health benefits of vaccines.

This overall risk-benefit analysis, however, does not allay parents' concerns when their child experiences a febrile seizure brought on by the DTP or MMR vaccine. While it is known that most children with febrile seizures have few long term disabilities, it has been unclear whether children who experience these seizures following vaccination have different prognosis from children with febrile seizures not temporarily related to immunizations. Some have postulated that febrile seizures attributed to vaccinations are the first signs of chronic neurological disorders such as epilepsy, autism or other learning and behavior disorders. Davis, Barlow and others compared children who have febrile seizures following vaccination with other children with febrile seizures and found that they are at no greater risk for subsequent seizure, developing epilepsy, or for learning, behavioral or other psychological problems. This was a reassuring finding given the generally good outcomes associated with febrile seizures overall.

The MMR vaccine was also the subject of another vaccine safety study spearheaded by Davis. Two national advisory groups had issued conflicting advice as to the timing of the second dose of this immunization (MMR2), with one group recommending MMR2 at age 4 to 6 and the second group at age 10 to 12 years. As a result, Group Health and Northern California Kaiser (NCK). two participating HMOs in the VSD, had differing immunization policies for the age of administering MMR2. Davis and colleagues took advantage of this situation by using the large-linked databases and medical record review to compare the frequency of clinical events following, and possibly related to, MMR2.

The investigators studied more than 8,000 children receiving MMR2 at age 4 to 6 at NCK and over 18,000 children receiving MMR2 at age 10 to 12 years at Group Health. To account for age-related differences in health care utilization patterns, clinical events in a 30-day time period following immunization were compared to those in a 30-day period prior to vaccination. Children 10–12 years of age were 50 percent more likely to have an adverse event following MMR2 than in the period prior to immunization, while children 4–6 years of age were less likely to have a visit for an event post-immunization compared to the pre-immunization period. The investigators concluded that the risk for adverse clinical events following MMR2 immunizations was greater in the 10–12 year age group.

Findings from another study conducted by ISP investigator Lisa Jackson and colleagues confirmed the safety and tolerability of pneumococcal revaccination for adults. (JAMA 1999;281:243–248). Pneumococcal disease causes 40,000 deaths and 500,000 cases of pneumonia in the U.S. each year. Vaccination is recommended for all persons over the age of 65 and for people under 65 with chronic conditions (e.g., heart and lung problems) that make them more susceptible to the disease. In addition, to improve protection, revaccination after at least 5 years is recommended for 2 groups: 1) persons over age 65 who were vaccinated before age 65; and 2)previously vaccinated people under age 65 who are immunocompromised because of underlying medical conditions such as leukemia or chronic renal failure. However, compliance with this second vaccination has been less than optimal, partly because of concerns that revaccination is more likely than first-time vaccination to cause adverse events, such as redness or swelling.

Jackson compared post-pneumococcal vaccination events among 1414 Group Health enrollees aged 50 to 74 years—901 received their first vaccination while 513 were revaccinated at least 5 years after receiving their first vaccination. While revaccination was associated with more adverse local site reactions than were first-time shots, in most cases this meant moderate limitation in use of the arm for a short period of time. Jackson concludes that, while physicians and patients should be aware of the risk of localized reactions, the benefits of revaccination clearly outweigh the risks.

Jackson also recently conducted a randomized trial to assess the safety and tolerability of a new influenza vaccine strategy. Influenza, a major cause of illness and death among people with chronic medical conditions and the elderly, can be combated somewhat by an annual flu shot. However, because these vaccinations have been shown to be less than 50 percent effective in protecting the elderly against flu-like illness, incentive exists to evaluate alternative vaccines or vaccine strategies.

Traditionally, influenza vaccines are "inactivated", meaning they contain dead organisms. One of the promising areas of research over the last two decades has involved the administration of "live" influenza vaccine (e.g., they contain the living organism) in combination with inactivated vaccines. Prior studies have indicated that this vaccine strategy has the potential to improve protection against influenza infection. Jackson conducted a safety study of this alternate influenza vaccine among the high-risk elderly—Group Health enrollees aged 65 and older with chronic heart or lung problems or diabetes. One hundred subjects were given "FluMist" (live influenza vaccine) intranasally in addition to an injection of inactivated influenza vaccine; an additional 100 received placebo instead of the FluMist. Jackson found that sore throat during the week following vaccination was more common among FluMist recipients compared to placebo subjects. However, this was the only adverse symptom associated with FluMist. Jackson concludes that, among this high risk group, FluMist is safe and well tolerated when administered with inactivated vaccine, and more research on efficacy is warranted.

Much of the work of the VSD is directed at studying the association between vaccinations and 34 pre-specified outcomes of interest. CHS Research Associate Kari Bolke, for example, is currently studying the association between vaccination and anaphylaxis (an allergic reaction) and other rare events (myocarditis, increased intercranial pressure, pancreatitis). Another strength of the VSD project is the ability to rapidly address new hypotheses that merge regarding vaccine safety. Historically, researchers have conducted expensive and time-consuming case-control studies in order to address "hot" topics, as was the case with the alleged relationship between swine influenza vaccine and Guillain-Barre syndrome. Since lingering uncertainty about safety issues can undermine public confidence in vaccines, speedy resolution of possible concerns is desirable. Davis recently led a study that used the VSD data to rapidly address two recent controversies: 1) does measles-containing vaccine increase the risk for developing Crohn's Disease or ulcerative colitis; and 2)do routine childhood vaccinations trigger acute ataxia (drunken gait). Reassuringly, Davis' analyses revealed no evidence supporting these hypotheses.

Summing up the ISP vaccine safety studies, Principal Investigator Robert S. Thompson says, "This program represents a prototype for large-scale collaborative research. We are benefiting health care delivery at Group Health and at the national level through the publication of new information generated on vaccines and vaccine safety".

Top