Postlicensure Monitoring of Intussusception After RotaTeq Vaccination in the United States, February 1, 2006, to September 25, 2007
BACKGROUND. In 1999, a previous rotavirus vaccine (RotaShield; Wyeth Laboratories, Marietta, PA) was withdrawn from the US market after postlicensure monitoring identified an association with intussusception. Although the new rotavirus vaccine (RotaTeq; Merck, West Point, PA) introduced in 2006 was not associated with intussusception in prelicensure trials, additional monitoring is important to ensure a complete safety profile.
METHODS. We assessed intussusception reports after RotaTeq vaccination by using data from the Vaccine Adverse Event Reporting System and the Vaccine Safety Datalink, a cohort of children enrolled in managed care. Observed versus expected rate ratios were determined by using vaccine dose distribution data and Vaccine Safety Datalink background intussusception rates.
RESULTS. Between February 1, 2006, and September 25, 2007, the Vaccine Adverse Event Reporting System received 160 intussusception reports after RotaTeq vaccination. With the assumptions that reporting completeness was 75% and that 75% of the distributed doses of RotaTeq were administered, the observed versus expected rate ratios were 0.53 and 0.91 for the 1–21 and 1–7 day interval after vaccination, respectively. In the Vaccine Safety Datalink, 3 intussusception cases occurred within 30 days after 111521 RotaTeq vaccinations, compared with 6 cases after 186722 non–RotaTeq vaccinations during the same period. If, like RotaShield, RotaTeq had a 37-fold increased risk of intussusception within 3 to 7 days after vaccination, then 8 intussusception cases would be expected within 3 to 7 days among the ∼84000 infants vaccinated with the first dose of RotaTeq in the Vaccine Safety Datalink (N = 49902) and the prelicensure trial (N = 34035) combined, whereas no cases have been observed.
CONCLUSIONS. Available data do not indicate that RotaTeq is associated with intussusception. Although an intussusception risk similar in magnitude to that of RotaShield can be excluded, continued monitoring is necessary for complete assessment of the safety profile of RotaTeq.
In February 2006, a new human-bovine reassortant rotavirus vaccine, RotaTeq (Merck, West Point, PA), was licensed by the US Food and Drug Administration (FDA). The Advisory Committee on Immunization Practices recommended routine vaccination of US infants with 3 doses of RotaTeq, administered orally at ages 2, 4, and 6 months.1 In 1999, a previous rhesus-human reassortant rotavirus vaccine, RotaShield (Wyeth Laboratories, Marietta, PA), was withdrawn from the US market 9 months after its introduction because it was associated with intussusception.2,3 The link between RotaShield and intussusception was first identified through the US Vaccine Adverse Event Reporting System (VAERS).4
Because of this proven association of a previous oral rotavirus vaccine with intussusception, the safety of the new RotaTeq vaccine was evaluated in a large, prelicensure, clinical trial involving 34035 infants who received vaccine and 34003 who received placebo.5 In that controlled trial, which was powered to exclude a 10-fold increase in the risk of intussusception during a 42-day period after RotaTeq vaccination, a statistically significantly elevated risk for intussusception was not observed (6 cases among vaccine recipients and 5 among placebo recipients). We present data from the first 19 months of postlicensure monitoring for intussusception after routine childhood vaccination with RotaTeq in the United States.
Vaccine Safety Datalink
The Vaccine Safety Datalink (VSD) is a collaborative project between the Centers for Disease Control and Prevention and several large US health maintenance organizations, in which computerized vaccination data can be linked to medical outcomes.6,7 Children who received RotaTeq vaccination at 4 to 48 weeks of age at 1 of 6 participating VSD sites were monitored for intussusception for 30 days after vaccination, by using methods described previously.6 For comparison, infants in the same age range who received any licensed vaccine other than RotaTeq during the same period were also monitored. We conducted a 1-time analysis to assess the crude intussusception relative risk and 95% confidence interval (CI) by using exact methods comparing infants vaccinated with RotaTeq and infants vaccinated with any other licensed vaccine.
We pooled data from the RotaTeq clinical trial and the VSD cohort on intussusception reports after dose 1 to assess more closely the risk in the narrow 3- to 7-day window after vaccination, which was identified as the period of greatest risk after RotaShield vaccination. We calculated the expected number of cases that the VSD and the prelicensure clinical trial would have identified if RotaTeq had an intussusception risk similar to that of RotaShield (ie, 37-fold elevated risk 3–7 days after the first dose).2 We multiplied the number of children vaccinated with the first dose of RotaTeq in the VSD and the clinical trial by the VSD age-stratified background intussusception rate (18 cases per 100000 person-years for 6–14 weeks of age) for the 3- to 7-day interval and then multiplied this number by the RotaShield risk of 37.
We reviewed reports of intussusception after RotaTeq vaccination received by the VAERS from February 1, 2006, through September 25, 2007. The VAERS is a national passive surveillance system (managed jointly by the FDA and the Centers for Disease Control and Prevention) that receives reports of adverse events after vaccination from various sources, including health care providers, state and local immunization programs, vaccine recipients, and manufacturers.8–10 For all intussusception cases, health care providers are contacted to verify the diagnosis and to obtain information on clinical features, treatment, hospital course, and vaccination history. In this analysis, intussusception cases were considered confirmed if they met the level 1 criteria included in the Brighton Collaboration case definition.11
Observed Versus Expected Calculations
We compared the number of intussusception cases reported to the VAERS after RotaTeq vaccination with the number of intussusception cases expected on the basis of the background rate of natural intussusception. Because the background intussusception rate and the number of vaccine doses administered vary substantially according to age, we stratified this analysis into 3 age groups, that is, 6 to 14 weeks, 15 to 23 weeks, and 24 to 35 weeks. This age grouping closely follows the Advisory Committee on Immunization Practices recommendation of administrating the first dose between 6 and 12 weeks of age and administering subsequent doses at 4- to 10-week intervals. In addition, because each of the 3 individual vaccine doses may pose a different risk of intussusception, we conducted separate dose-specific analyses. We evaluated risk in 2 primary time periods (1–21 days and 1–7 days) for the onset of symptoms after RotaTeq vaccination, on the basis of the potential risk periods identified from the experience with RotaShield.2 Also, we hypothesized that any possible risk of intussusception would be related temporally to replication of the vaccine virus in the intestinal tract.1,5
To calculate background rates of intussusception, we examined data from the VSD. Baseline rates of intussusception were determined from hospital discharges coded with the International Classification of Diseases, Ninth Revision, code for intussusception (code 560.0) at VSD study sites in 2000–2004, when no rotavirus vaccine was in use. The expected number of background cases for the risk periods were calculated by multiplying the background rate of intussusception for each age group by the estimated number of vaccine doses administered as dose 1, 2, or 3 to infants in that age group.
To estimate the numbers of vaccine doses administered, we assumed that (1) the number of administered doses of vaccine equaled the total number of doses of RotaTeq distributed by the manufacturer from February 1, 2006, through August 31, 2007, and (2) the proportions of vaccine doses administered to infants in each of the 3 age groups nationally equaled the proportions of vaccine doses administered to infants in each of the 3 age groups in the VSD during the same time period. Observed/expected rate ratios (RRs) with 95% CIs were calculated by using exact, age-stratified, Poisson tests (StatXact 6; Cytel Software, Cambridge, MA).
We used sensitivity analyses to assess the potential impact on our results of variability in 2 key assumptions, that is, the completeness of reporting to the VAERS and the proportions of distributed doses that were administered.
Between February 1, 2006, and September 25, 2007, the VAERS received 1901 reports of adverse events after RotaTeq vaccination, including 160 reports (8.4%) of confirmed intussusception (Fig 1). No deaths occurred among reported intussusception cases. Of the 160 intussusception reports, 47 cases (29%) occurred in infants within 1 to 21 days after vaccination, including 27 (17%) that occurred within 1 to 7 days after vaccination. For comparison, >60% of the intussusception cases reported after RotaShield vaccination occurred within 1 to 7 days after vaccination (Fig 2).
All 47 intussusception cases that occurred within 1 to 21 days after vaccination were confirmed radiographically (ie, through barium enema radiography or ultrasonography) or surgically. Twenty-two infants (47%) required surgery, including 11 who underwent intestinal resection; for 25 infants (53%), intussusception was resolved through enema reduction. Of those 47 cases, 16 (34%) occurred after vaccine dose 1, 21 (45%) after dose 2, and 10 (21%) after dose 3 (Fig 3). The mean ages at vaccination for these cases were 11, 19, and 28 weeks for doses 1, 2, and 3, respectively. Only 3 (19%) of 16 children with onset after dose 1 received their first vaccine dose at >90 days of age (exact ages of 97, 103, and 129 days).
As of August 31, 2007, the manufacturer had distributed ∼9120726 doses of RotaTeq (Merck, unpublished data, 2007). Of the administered doses in the VSD, 42%, 32%, and 25% were given to infants 6 to 14 weeks, 15 to 23 weeks, and 24 to 35 weeks of age, respectively; an estimated 45% were dose 1, 33% were dose 2, and 22% were dose 3. Ninety-two percent and 93% of first doses were administered to infants 6 to 12 weeks and 6 to 14 weeks of age, respectively.
On February 13, 2007, a vaccine label change, adding intussusception to the package insert as a reported adverse event after vaccination with RotaTeq, was issued by Merck, and an accompanying public health notification was released by the FDA.12,13 This received substantial media coverage and was widely distributed among pediatricians. An increase in the total number of the VAERS intussusception reports occurred after the FDA notification; of the 160 total cases of intussusception reported during the 19-month surveillance period, 50 (43%) were reported within 8 weeks after the FDA notification. Thirty-seven (74%) of the 50 cases reported during the 8 weeks after the notification occurred outside the 1- to 21-day interval. In contrast, only 18 (49%) of 35 intussusception cases reported before the FDA notification occurred outside the 1- to 21-day interval (difference in proportions: 25%; P = 03). Of the 50 cases that were reported after the FDA notification, 12 occurred before the date of the FDA notification, and 4 of those 12 occurred within 1 to 21 days after vaccination.
Observed Versus Expected Calculations
The VSD background rate of intussusception was 18.1 cases per 100000 person-years at 6 to 14 weeks of age, 32.5 cases per 100000 person-years at 15 to 23 weeks of age, and 42.5 cases per 100000 person-years at 24 to 35 weeks of age. Applying these background intussusception rates to the number of doses administered to each age group, we calculated an expected number of 151 intussusception cases for the period 1 to 21 days after vaccination; 50 of those cases would be expected to occur 1 to 7 days after vaccination. In contrast, 47 intussusception cases reported to the VAERS occurred within 1 to 21 days after vaccination, including 27 that occurred 1 to 7 days after vaccination (Table 1). With the assumptions that 75% of intussusception cases that occurred during these time windows were reported to the VAERS and that 75% of the vaccine doses distributed were administered, the observed rate of intussusception was not elevated above the age-adjusted background rate of intussusception for either 1 to 21 days (RR: 0.53; 95% CI: 0.37–0.75) or 1 to 7 days (RR: 0.91; 95% CI: 0.59–1.37) after RotaTeq vaccination (Table 2). Observed/expected RRs were similar when stratified according to dose or age group (ie, 6–14 weeks, 15–23 weeks, and 24–35 weeks) (Table 1).
With assumptions of 100% reporting completeness and 100% administration of distributed doses, the observed rates of intussusception were lower than expected for both 1 to 21 days (RR: 0.30; 95% CI: 0.20–0.44) and 1 to 7 days (RR: 0.51; 95% CI: 0.32–0.81) after RotaTeq vaccination (Table 2). Sensitivity analysis for the 1- to 7-day interval indicated that, if reporting completeness was 50%, then administration of 50% (4560363 doses) of the ∼9120726 distributed doses of RotaTeq would yield a statistically significant association (RR: 2.01; 95% CI: 1.39–2.90) between the vaccine and intussusception (Table 2). Limiting the sensitivity analysis to the first dose with the same conservative reporting and dose administration assumptions would yield a RR of 3.7 (95% CI: 1.9–6.9).
Between May 21, 2006, and September 25, 2007, a total of 111521 doses of RotaTeq were administered to infants in the VSD. Three cases of intussusception within 30 days after vaccination (1 case after dose 1 at day 20 and 2 cases after dose 2 at days 10 and 17) were reported among RotaTeq recipients, compared with 9 cases of intussusception among ∼186722 non-RotaTeq recipients, for a relative risk of 0.84 (95% CI: 0.14–3.92). If RotaTeq had an intussusception risk similar to that of RotaShield, then 8 cases of intussusception would be expected within 3 to 7 days after vaccination among the ∼84000 infants vaccinated with the first dose of RotaTeq in the VSD (N = 49902) and the prelicensure clinical trial (N = 34035) combined, whereas none has been observed.
Available data do not indicate that RotaTeq is associated with intussusception. The overall number of intussusception cases reported to the VAERS among infants vaccinated with RotaTeq did not exceed the number of expected background cases for either the 1- to 21-day period or the 1- to 7-day period after vaccination. However, some key limitations in interpretation of VAERS data, such as possible underreporting of intussusception cases and uncertainty regarding the number of distributed vaccine doses that were administered,9,10 preclude definitive assessment of risk with these data alone. The VSD provides more-precise information, and only 3 cases of intussusception were detected through the VSD surveillance within 30 days after vaccination with nearly 111521 doses of RotaTeq, in comparison with 9 cases among 163123 non–RotaTeq recipients. Because of the limited number of children in the VSD who were vaccinated with RotaTeq, however, it is not possible to exclude some increased risk of intussusception. Ongoing monitoring and epidemiological studies will be important for assessment of the safety profile of rotavirus vaccines.
Although the 47 intussusception cases reported to the VAERS within 1 to 21 days after RotaTeq vaccination did not exceed the expected number for this risk period, a majority of those cases (n = 27; 57%) occurred within 1 to 7 days after vaccination (Fig 1). Reporting of adverse events to the VAERS is known to be more complete for events occurring closer to the time of vaccination, but it is not known whether completeness of reporting decreases substantially for events occurring in the second and third weeks after vaccination, compared with the first week.14 If completeness of reporting is similar during the first 3 weeks after vaccination, then the excess number of reported cases during week 1, relative to weeks 2 and 3, might represent an increased risk of intussusception shortly after vaccination. Additional monitoring should facilitate interpretation of the relevance of these findings.
Direct comparison of the risks of intussusception associated with RotaShield and RotaTeq requires cautious interpretation, because of differences in their schedules of administration (ie, catch-up vaccination with the first dose of RotaShield was permitted for infants up to 6 months of age, whereas the first dose of RotaTeq has to be administered by 12 weeks of age) and differences in patterns of vaccine virus shedding. Nevertheless, if RotaTeq had an intussusception risk similar to that of RotaShield (ie, 37-fold elevated risk 3–7 days after dose 1),2 then 8 cases of intussusception would be expected within 3 to 7 days after vaccination in the VSD and the prelicensure clinical trial combined; no cases of intussusception during this risk period have been identified.
VAERS data merit cautious interpretation because of several key limitations. First, it is not known how many intussusception cases after RotaTeq vaccination were not reported to the VAERS. The high level of awareness of the association of the previous RotaShield vaccine with intussusception among pediatricians and family practitioners15 and the severe, notable, clinical presentation of intussusception likely enhanced reporting to the VAERS, especially during the first few weeks after vaccination.16,17 This conclusion is supported by the fact that the initial public health alert for RotaShield published on July 16, 1999,18 resulted in ∼83 stimulated reports of intussusception to the VAERS in 1 month, and 58 (70%) of those cases were within 1 to 21 days after vaccination (Fig 2).17 In contrast, although the February 2007 FDA notification did enhance VAERS reporting considerably the proportion of cases that occurred within 1 to 21 days decreased from 49% before the notification to 29% after the notification. Moreover, only 4 of 12 stimulated cases (ie, cases that occurred before the FDA notification) were within 1 to 21 days after vaccination. If substantial underreporting in the 1- to 21-day period existed, then the notification should have generated more reports of cases occurring in this potential period of risk, which leads us to suspect that reporting completeness was high within 1 to 21 days after vaccination.
A second important limitation of our VAERS analysis is that we could not determine what proportion of distributed doses of RotaTeq vaccine had been administered. On average, ∼450000 doses of vaccine have been distributed monthly since May 2006, and contracts for rotavirus vaccine purchase through the Vaccines for Children program (which purchases vaccine for nearly 45% of US children)19 have been in place for >12 months; these facts indicate that conditions are favorable for vaccine uptake. Of note, with the previous RotaShield vaccine, data from the National Immunization Survey indicated that ∼66% of the distributed RotaShield doses were administered during the 9 months of use of the vaccine.20
A final key limitation is with regard to background rates of intussusception. Although our expected calculations used VSD data and might not represent the general population from which the VAERS data arise, the VSD is a robust data source and the VSD background rates for intussusception hospitalizations are similar to those in the published literature. However, the accuracy and completeness of these data and other national data sources have not been verified. A study conducted at a large children's hospital in Cincinnati, Ohio, found that ∼40% of infants with intussusception in 1995–2000 were treated with short-stay/emergency department care only; those patients would not have been identified through examination of hospital discharge data alone, leading to underestimation of the background incidence of intussusception.21 Our sensitivity analysis did not account for the possibility of higher background rates of intussusception and thus might have overestimated risk.
With >9 million doses of vaccine distributed during the first 19 months after licensure, postlicensure data from the VAERS and the VSD do not indicate that RotaTeq is associated with intussusception. Nevertheless, because of the relatively small number of infants vaccinated with RotaTeq in the VSD and the limitations and assumptions in the analysis of VAERS data, the possibility that an increased risk for intussusception, of smaller magnitude than that observed with RotaShield, might be documented over time cannot be excluded. Ongoing monitoring, especially evaluation of the pattern of onset after vaccination for intussusception cases reported to the VAERS and periodic examination of VSD data and other epidemiological data as vaccine uptake increases, will be important for full assessment of the safety profile of RotaTeq.
- Accepted February 22, 2008.
- Address correspondence to Manish Patel, MD, Centers for Disease Control and Prevention, National Center for Immunizations and Respiratory Diseases, 1600 Clifton Rd, MS A-47, Atlanta, GA 30333. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
Prelicensure clinical trials of RotaTeq excluded a 10-fold increase in the risk of intussusception during a 42-day period after RotaTeq vaccination.
What This Study Adds
Postlicensure safety monitoring supports the clinical trial findings and excludes a risk of a magnitude similar to that of RotaShield in a narrow time window after vaccination.
- ↵Food and Drug Administration. FDA public health notification, February 13, 2007: information on RotaTeq and intussusception. Available at: www.fda.gov/cber/safety/phnrota021307.htm. Accessed July 27, 2007
- ↵Kempe A, Daley MF, Parashar UD, et al. Will pediatricians adopt the new rotavirus vaccine? Pediatrics.2007;119 (1):1– 10
- ↵Verstraeten T, Baughman AL, Cadwell B, Zanardi L, Haber P, Chen RT. Enhancing vaccine safety surveillance: a capture-recapture analysis of intussusception after rotavirus vaccination. Am J Epidemiol.2001;154 (11):1006– 1012
- ↵Zanardi LR, Haber P, Mootrey GT, Niu MT, Wharton M. Intussusception among recipients of rotavirus vaccine: reports to the Vaccine Adverse Event Reporting System. Pediatrics.2001;107 (6). Available at: www.pediatrics.org/cgi/content/full/107/6/e97
- ↵Santoli JM, Rodewald LE, Maes EF, Battaglia MP, Coronado VG. Vaccines for Children program, United States, 1997. Pediatrics.1999;104 (2). Available at: www.pediatrics.org/cgi/content/full/104/2/e15
- ↵Staat MA, Roberts N, Bernstein DI. Epidemiology and clinical features of intussusception in children <24 months of age in Hamilton County, Ohio [abstract]. Pediatr Res.2001;49 (5):242A
- Copyright © 2008 by the American Academy of Pediatrics