Anaphylaxis is a systemic hypersensitivity reaction to a foreign stimulus with rapid onset and release of cytokines and inflammatory mediators, which can progress rapidly to multiorgan dysfunction leading to a severe and life-threatening condition. The incidence of anaphylaxis from all causes in the United Kingdom has been estimated to be 8.4 per 100,000 person-years.1
Anaphylaxis occurring as an adverse event following immunization (AEFI) is reported to be a very rare event although the calculated risk in children and adolescence varied widely between 0.65 cases per 1,000,000 doses and 22.4 cases per 100,000 doses administered in previous studies.2–6 The aim of this study was to estimate the annual frequency and incidence of anaphylaxis occurring within 48 hours after immunization in children and adolescents in Germany leading to hospitalization. Anaphylaxis is a feared adverse event because of the fact that it might be life threatening. These first estimates from Germany may help addressing many physicians’ and parents’ concerns.
MATERIALS AND METHODS
Children and adolescents younger than 18 years with suspected anaphylaxis after immunization were reported to the German surveillance unit for rare pediatric diseases [Erhebungseinheit für seltene pädiatrische Erkrankungen in Deutschland (ESPED)].7
A reporting card with a list of rare disorders is sent to pediatric hospitals each month. Alternatively, hospitals can opt for online reporting.
Pediatricians return the (electronic) card, notifying ESPED of any cases or indicating “nothing to report.” In case of a positive notification, follow-up information is requested to complete the individual case report form regarding the case presentation, diagnosis, management and outcome.
Hospital pediatricians were asked to report all children and adolescents who may have experienced anaphylaxis within 48 hours after immunization over 24 months between June 1, 2008, and May 31, 2010. All reports were classified by medically qualified staff according to the Brighton Collaboration Case Definition (BCCD) for anaphylaxis, which uses a grading according to diagnostic certainty irrespective of the probability of causal relation or of allergic etiology.8 BCCD levels 1–3 identify events that “meet the case definition,” BCCD level 4 specifies events of “reported anaphylaxis with insufficient evidence to meet the case definition,” and BCCD level 5 excludes anaphylaxis. Cases fulfilling BCCD level 1–3 were eligible. If a clear alternative diagnosis was reported to account for the combination of symptoms, the case was reassigned level 5. The study was approved by the Ethics Committee at the University Hospital Mannheim (2008-274E-MA).
Spontaneous Case Reports of Anaphylaxis After Immunization
According to the German Act on the Prevention and Control of Infectious Diseases, it is mandatory for physicians to report AEFI to the local health authorities, which have to notify the national competent authority, the Paul-Ehrlich-Institut (PEI). Marketing authorization holders have to report suspected serious adverse reactions to the PEI according to the German drug law. Case reports are stored in the PEI adverse drug reaction database. A broad search strategy was applied to ensure identification of all relevant individual case safety reports. Standardized Medical Dictionary for Regulatory Activities (MedDRA) queries (SMQs) including narrow and/or broad terms have been used.9 Narrow terms are those that are highly likely to represent the condition of interest. A structured query language (SQL) database search was performed for case reports regarding pediatric and adolescent vaccinees with administration date from June 1, 2008, through May 31, 2010 matching “Preferred Terms” covered by the SMQs “Anaphylactic reaction narrow,” “Anaphylactic reaction broad,” “Hypersensitivity narrow” and “Hypersensitivity broad.” The retrieved case reports were reviewed and validated according to BCCD for anaphylaxis as an AEFI8 by medically qualified staff. BCCD level 1–3 case reports with an anaphylactic reaction occurring within 48 hours after immunization leading to hospitalization were eligible.
The capture–recapture method10 is used in epidemiology to estimate the size of populations11–13 by utilizing data from overlapping lists of cases that originate from various sources to estimate the number of missing cases and the total affected. We used the Chapman capture–recapture methods,14 which have been applied previously to evaluate the reporting completeness of a passive safety surveillance system.15
Data linkage of retrieved case reports from ESPED and PEI was performed manually on several variables including year and month of birth, gender, vaccination date, vaccine trade name and reported reactions.
For the AS03 adjuvanted A/H1N1 pandemic influenza vaccine exposure, estimates were available for individuals younger than 18 years.16
Health insurance in Germany is compulsory and divided between statutory and private schemes. The majority of residents are covered by statutory health insurance (Gesetzliche Krankenversicherung). Employees with a higher income, self-employed residents, students and civil servants for complementary coverage can opt for private health insurance (private Krankenversicherung). Only the statutory health insurance includes family members without own income (husband/wife and children) of the insured person free of charge.
Regarding vaccines recommended for routine childhood vaccination in Germany, figures on doses administered were derived from the “Associations of Statutory Health Insurance Physicians (ASHIP) vaccination monitoring project” at Robert Koch-Institute. The accuracy of vaccination coverage estimated from ASHIP data was previously demonstrated in an extensive validation procedure with various primary data sources like kindergarten and school entrance examinations.17
ASHIPs receive insurance refund claims from physicians for ambulatory medical services provided to statutory health insurees. These data include vaccination specific codes for nationally recommended vaccinations. As described previously,17 data are extracted from the ASHIPs’ databases, anonymized and transferred to Robert Koch-Institute. At vaccine-type level, we calculated the total of vaccinations administered from June 2008 until May 2010 to patients <18 years of age. For a total of 6 months within the study period, data on claimed vaccinations were not available from 2 out of the 17 ASHIPs but extrapolated from other ASHIPs. Because recommended vaccinations are generally also covered by private health insurances, we projected the estimate for the statutory health insured population to the total German population <18 years of age according to the proportion of statutory health insurees of 84.5%, as derived from statistics of statutory health insurees by the German Ministry of Health and population statistics by the Federal Statistics Office. For seasonal influenza and rotavirus vaccines exposure, estimates were not available.
In addition, we estimated the proportion of anaphylaxis as an AEFI among the mean annual number of cases hospitalized from 2008 to 2010 because of anaphylaxis from all causes in children and adolescents. Data on the frequency of hospitalization because of anaphylaxis from the main causes (International Classification of Diseases, ICD-10 codes: T78.0, “anaphylactic shock due to adverse food reaction”; T78.2, “anaphylactic shock, unspecified”; T80.5, “anaphylactic shock due to serum” and T88.6, “adverse effect of correct medicinal substance properly administered”) from 2008 to 2010 was provided by the Federal Statistics Office. The data were collected for age groups, ie, “under 1,” “1–4,” “5–9,” “10–14” and “15–19” by the hospital statistics.
Within the scope of sensitivity analyses, hospitalized BCCD level 4 cases with time to onset within 48 hours after immunization were included.
The capture–recapture point estimator was calculated according to Chapman et al.14 We estimated the variance of the Chapman estimate according to Seber18 and Wittens19 and hence the standard error. The 95% confidence interval (CI) based on the standard error was computed by a log transformation method (formula 12).20 If no cases were captured in either the ESPED only or the PEI only cell, a continuity correction of adding 0.1 to the zero cells was performed. Exact binomial CIs (Clopper–Pearson exact confidence limits) were calculated for proportions (Table 3). Statistical analyses were performed using the software package SAS version 9.3 (SAS Institute Inc., Cary, NC).
In both the ESPED and the PEI investigation, the reports were checked for eligibility in the following way: after removal of duplicates, the remaining reports were validated according to the criteria of the BCCD. Cases assigned level 5 were excluded from further analysis, and cases assigned level 4 were exclusively used for sensitivity analyses. Case reports with a time to symptoms onset exceeding 48 hours were excluded as well. And so were reports without recorded hospitalization. The remaining reports were eligible.
In 2008, 370 out of the 373 children’s hospitals in Germany participated in ESPED (coverage 99.2%) with a total of 473 ESPED collaborators being contacted (240 of whom received a reporting card per month and 233 opted for electronic reporting). The response rate was 96% (reporting card, 96%; electronic reporting, 95%). During the study period of 24 months, ESPED received a small number (n = 13) of case reports of suspected anaphylaxis after immunization, the majority (n = 9) of which were eligible (Fig. 1).
Regarding the PEI investigation, several reports were retrieved by more than 1 of the 4 SMQs, so that the total number of unique AEFI reports identified (n = 705) was smaller than the number of case reports retrieved by the 4 SMQs together (Fig. 2). After removal of duplicates (n = 24), exclusion of case reports assigned levels 4 (n = 8) and 5 (n = 646), exclusion of case reports with a time to symptoms onset exceeding 48 hours (n = 3) and exclusion of case reports without recorded hospitalization (n = 5) 19 eligible case reports were left.
The 5 case reports (3 males, 2 females) that were excluded because the patients had not been hospitalized were all received in 2009. In 3 cases, anaphylaxis occurred after administration of a pandemic vaccine (BCCD levels 1, 2, 2), and one patient had received a human papilloma virus vaccine (BCCD level 1) and another one a seasonal influenza vaccine (BCCD level 2).
Among the 22 eligible case reports, 8 (36.4%) were classified BCCD level 1, 13 (59.1%) level 2 and 1 (4.5%) level 3. Three cases had exclusively been reported to ESPED, 13 exclusively to PEI and 6 had been reported to both ESPED and PEI (Table 1). The median age of the affected individuals was 7.0 years (range: 2 months to 17 years). Four individuals (18.2%) were reported to have a preexisting allergy, of whom patient 1 had an allergy to antibiotics (amoxicillin and penicillin); patient 2 to mite; patient 3 to ambroxol hydrochloride sirup, birch, nuts, rye and grass and patient 4 to pollen, mite and food. Six children and adolescents were reported to have another preexisting medical condition. In 5 cases, premedication was notified, but only in one case, premedication included antiasthmatic drugs.
A total of 26 vaccines were reported to have been administrated before (suspected) anaphylaxis with AS03 adjuvanted A/H1N1 pandemic influenza vaccine being the vaccine with the highest number of notifications (n = 8; Table 2). In 3 patients (13.6%), concomitant vaccinations were reported: second dose combined diphtheria–tetanus–acellular pertussis, hepatitis B, poliovirus and Haemophilus influenzae type b vaccine + second dose pneumococcal 7-valent conjugate vaccine (diphtheria CRM197 protein); fourth dose combined diphtheria–tetanus–acellular pertussis, hepatitis B, poliovirus and Haemophilus influenzae type b vaccine + first dose pneumococcal 10-valent conjugate (protein D) vaccine; and second dose combined diphtheria–tetanus–acellular pertussis, hepatitis B, poliovirus and Haemophilus influenzae type b vaccine + second dose pneumococcal 7-valent conjugate vaccine (diphtheria CRM197 protein) + second dose 5-valent, live, oral rotavirus vaccine.
In half of the 26 vaccines administered, the patient received the first dose, in 9 vaccines, the second dose and in 1 vaccine, the fourth dose. For 3 vaccines, information on dosage was not available.
The most frequently reported symptoms were generalized urticaria (hives) or erythema, difficulty in breathing without wheeze or stridor, localized or generalized angioedema, tachycardia and nausea (see Table, Supplemental Digital Content 1, http://links.lww.com/INF/C391). Median time to symptoms onset was 5 minutes (range, 1 minutes to 24 hours).
In 86.4% of the reports, the patient received a specific therapy. A total of 68.2% patients were treated with corticosteroids, 50% with antihistamines and 40.9% with intravenous fluids (see Table, Supplemental Digital Content 2, http://links.lww.com/INF/C392). Therapy also included oxygen (27.3%), catecholamines (13.6%) and beta-2 sympathomimetics (4.5%). With the exception of 1 patient for whom the outcome was unknown, all patients had recovered at the time of reporting.
There was 1 noneligible case report with a fatal outcome, which was originally assigned level 4 and later on reassigned level 5. This case of a 5-month-old male infant had been reported to both ESPED and PEI. The autopsy report that had been provided to the PEI by the public prosecutor upon request stated that histological examinations revealed a severe hemorrhagic pulmonary edema of both lungs possibly caused by an acute left heart failure. Furthermore, a mild lymphocytic inflammatory reaction at the heart possibly attributable to a preexisting myocarditis was detected. The forensic pathologists suspected a genetic defect because a sibling of the deceased infant had died in 2007 at a similar age. An autopsy had also been performed on the brother’s corpse. Main findings had included a hemorrhagic, nonreactive pneumonia and a lymphocytic myocarditis.
With respect to the AS03 adjuvanted A/H1N1 pandemic influenza vaccine, the estimated absolute number of cases with anaphylaxis as an AEFI was estimated to be 11.0 (95% CI: 8.5–26.2), which corresponds to an incidence of 11.8 (95% CI: 9.1–28.2) cases per 1,000,000 doses administered (Table 3). The incidence rates of anaphylaxis for all other vaccines were low, without clustering of cases for a specific vaccine (Table 3). Regarding vaccines other than AS03 A/H1N1 pandemic influenza vaccine excluding monovalent measles and monovalent rubella vaccine to avoid indication bias, the absolute annual number of anaphylactic reactions in children and adolescents leading to hospitalization within the observational period of 24 months was estimated to be 13.7 (95% CI: 12.3–21.8) cases corresponding to an annual frequency of 6.8 (95% CI: 6.1–10.9) cases. In relation to the mean (±standard deviation) annual number of hospitalized cases of anaphylaxis from the main causes in this age group (1246 ± 24.7), this corresponds to a percentage of 0.6% (95% CI: 0.5–0.9).
Inclusion of BCCD level 4 case reports in the analyses resulted in a total number of 29 reports, 5 of which originated exclusively from ESPED, 17 exclusively from the PEI and 7 from both ESPED and PEI.
As to the AS03 A/H1N1 pandemic influenza vaccine, the frequency of anaphylactic reactions was estimated to be 14.0 (95% CI: 10.7–33.4), which in relation to the number of individuals with uptake of pandemic vaccine younger than 18 years translates to an estimated incidence of 15.1 (95% CI: 11.5–35.9) per 1,000,000 doses administered.
Regarding vaccines other than AS03 A/H1N1 pandemic influenza vaccine excluding monovalent measles and monovalent rubella vaccine to avoid indication bias, the annual frequency was estimated to be 10.5 (95% CI: 9.0–16.8). In relation to the mean annual number of hospitalized cases of anaphylaxis from the main causes in this age group, this corresponds to a percentage of 0.8% (95% CI: 0.7–1.3).
Estimates for specific vaccines calculated within the scope of the sensitivity analyses were provided in Table 3.
Our study supports the notion that anaphylaxis after immunization in children and adolescents is a very rare event. The strength of this study is that hospitalized children with suspected anaphylaxis were prospectively analyzed by ESPED, which is an active surveillance system with a (external) coverage of 99.2% and a very good (internal) compliance (ESPED response rate in 2008: 96%) as well as retrospectively using all BCCD classified AEFI reports received by a passive surveillance system (PEI).
Reporting of adverse events after immunization to PEI as the competent authority is mandatory in Germany. Thus, the PEI is presumably in charge of the most complete collection of AEFI reports in Germany. Completeness of the PEI AEFI reporting system in general is unknown. Regarding this study, we used capture−recapture methods to calculate an estimator for completeness which was 78% (nearly unbiased estimator) and to adjust for underreporting. Using the SMQs “anaphylactic reaction” and “hypersensitivity” enabled a standardized and systematic database search for “Preferred Terms” specifying signs and symptoms that may occur during an anaphylactic or hypersensitivity reaction. Thus, the probability is low that eligible reports may have slipped the prescreening process. Because numerous case reports with differential diagnoses like vaso-vagal reactions or rashes were retrieved, in a second step, the case reports identified by the SMQs were validated according to the BCCD for anaphylaxis as an AEFI. In contrast, in the ESPED study, cases were prospectively identified and recorded by medical staff according to a study protocol and by using a standardized case report form addressing the Brighton Collaboration criteria.
The BCCD for anaphylaxis as an AEFI8 is an established assessment tool that has been used previously to evaluate case reports of potential anaphylaxis in children and adolescents.4–6 Passive surveillance systems have often been blamed for underreporting and a lack of detailed information. This is why our strategy was to use the prospective active surveillance system ESPED, which independently collected reports of anaphylaxis as an AEFI in individuals younger than 18 years in the same observational period using the BCCD for anaphylaxis as an AEFI as a second data source. Adjusted for underreporting, the annual frequency of anaphylaxis as an AEFI in individuals younger than 18 years in Germany regarding vaccines (excluding AS03 pandemic vaccines as well as monovalent measles and rubella vaccines) was estimated to be 6.8 (95% CI: 6.1–10.9) cases, a magnitude that is in line with the number of cases of anaphylaxis after immunization in individuals younger than 16 years detected by the British Paediatric Surveillance Unit (n = 7), which is very similar to ESPED regarding aims (surveillance of rare pediatric disorders) and methods (reporting cards, follow-up questionnaires).4 In contrast to the study conducted in United Kingdom and Ireland, in Germany, anaphylaxis as an AEFI was also reported to occur in preschool children and infants with the youngest subject being just 2 months whereby the risk of anaphylaxis after administration of vaccines used in standard immunization programs for infants seems to be low. Our finding is, however, consistent with results from the US2 where 4 out of the 5 cases identified were below 2 years of age and Australia where 2 infants were reported to have suffered from anaphylaxis after concomitant receipt of hexavalent, pneumococcal and rotavirus vaccines.5
For monovalent measles and monovalent rubella vaccines, incidence rates of a higher magnitude were estimated (Table 3). These calculations were, however, based on 1 single case each which could be subject to bias by indication or to misinterpretation and therefore have to be interpreted with caution. In contrast, only 1 BCCD level 4 case of anaphylaxis was reported after receipt of measles, mumps and rubella, and no cases were reported after receipt of measles, mumps, rubella and varicella vaccines after administration of 924,314 and 2,072,157 doses, respectively. The incidence rates for measles and rubella strains containing vaccines reported in the literature varied considerably, a fact that may also be ascribed to bias by indication or misinterpretation: 18.9 cases per 100,000 doses3 based on 1−2 cases reported and 12.0 cases per 100,000 doses4 based on 2 cases reported for monovalent measles vaccine; 22.4 cases per 100,000 doses3 for monovalent rubella vaccine based on 2−3 cases reported; 1 case per 100,000 doses21 based on 30 cases reported, 1.25 per 100,000 doses5 based on 6 cases reported and 5.14 per 1,000,000 doses6 based on 3 cases reported for measles, mumps and rubella vaccines and 19.8 per 1,000,000 doses6 based on 2 cases reported for measles, mumps, rubella and varicella vaccines.
The observational period included the 2009 H1N1 pandemic mass vaccination campaign that was initiated in Germany on October 26, 2009. In our study, AS03 adjuvanted A/H1N1 pandemic influenza vaccine (Pandemrix, GlaxoSmithKline, Rixensart, Belgium) was the vaccine with the highest absolute and relative proportion among all vaccines administered before anaphylactic reactions (n = 8; 30.8%) although the vaccination coverage in Germany was low [<14 years, 7.8% (95% CI: 6.1–10.0); 14–17 years, 4.0% (95% CI: 2.6–6.4)].16 In this study, the incidence for anaphylaxis after administration of the AS03 adjuvanted A/H1N1 pandemic influenza vaccine was higher when compared with the world-wide passive reporting rate published by the manufacturer GlaxoSmithKline in 2011,22 which was 1.8 case of anaphylaxis as an AEFI fulfilling the criteria of BCCD level 1–3 per 1,000,000 doses administered to all age groups. In line with our study, Rouleau et al23 also using the BCCD found an incidence of anaphylaxis after administration of another AS03 adjuvanted A/H1N1 split influenza vaccine (Arepanrix, GlaxoSmithKline, Mississauga, Ontario/Canada) of 13 cases per 1,000,000 doses administered for patients of all ages in Quebec, Canada.
One the one hand, different from seasonal influenza vaccines, which are particularly recommended for pediatric patients suffering from chronic diseases like asthma who may have a higher frequency of underlying allergic conditions as well as hyperreactivity of skin and mucosal structures, the H1N1 pandemic influenza vaccine was recommended for all children and adolescents irrespective of underlying conditions (mass vaccination campaign). On the other hand, it has to be considered that Pandemrix was a new vaccine which was controversially discussed in the media (reactogenicity, innovative adjuvant). It is known that new products on the market may provoke increased reporting of adverse events; this has been called “Weber effect.”24
Interestingly, in half of the cases, anaphylaxis occurred after administration of the first dose, a fact that supports the notion that vaccine-associated anaphylaxis may not necessarily be because of sensitization after repeated vaccination or an IgE-mediated allergy. Instead, it may be induced by histamine release through other mechanisms.25 However, a vaccine does not only contain the antigen but numerous other components like antibiotics or solvents. So, even after the first dose an allergic reaction may occur if the child was previously sensitized to a constituent of the vaccine (eg, the antibiotic that was used in the manufacturing process).
In this study, no proven vaccine-associated cases of anaphylaxis with fatal outcome were identified. Bohlke et al2 did not identify any case of anaphylaxis that resulted in death after administration of over 7.6 million doses. Also in the studies conducted in United Kingdom and Ireland,4 Australia5 and the US,6 no fatal cases were reported. Of note, catecholamines were administered only sparsely in our study although guidelines recommend catecholamines as first-line treatment.26 Thus, we speculate this may be because of fear of side effects that may occur especially when given in overdose.
Detailed information was not available for all spontaneous case reports. Paucity of information may lead to an underestimation of the annual frequency of anaphylaxis as an AEFI. Therefore, we used a prospective active surveillance system as a second data source. It is reassuring that our estimates are in line with data from the United Kingdom and Ireland.
Our estimate is limited to hospitalized cases of anaphylaxis because ESPED exclusively collaborates with children hospitals. Excluding the patients who were exclusively treated in an outpatient setting may also lead to an underestimation of the annual frequency of anaphylaxis after immunization.
Fulfilling the BCCD for anaphylaxis as an AEFI, which was developed for immunization settings,6 is irrespective of causality and reflects per se only a temporal but not a causal association. Furthermore, the BCCD does not only include anaphylactic (IgE mediated) but also anaphylactoid (non IgE mediated) reactions and maybe also nonallergic events. In consequence, capture of case reports in this study is characterized by a high sensitivity. To conclude, this study confirms that anaphylaxis after immunization is a very rare event. Nevertheless, pediatricians should always be prepared to face and manage such a potentially life-threatening adverse event in their own practice.
We address special thanks to Barbara van Zandbergen, MD and Jens Rüggeberg, MD for critical reading of the manuscript and helpful suggestions. We especially appreciate contributions of Dr. Rüggeberg to initiate the study and of Beate Heinrich to manage reporting cards and follow-up information CRFs. Finally, we thank Claudia Pönisch for performing the Standardised MedDRA Queries (SMQs).
1. Peng MM, Jick H. A population-based study of the incidence, cause, and severity of anaphylaxis in the United Kingdom. Arch Intern Med. 2004;164:317–319
2. Bohlke K, Davis RL, Marcy SM, et al.Vaccine Safety Datalink Team. Risk of anaphylaxis after vaccination of children and adolescents. Pediatrics. 2003;112:815–820
3. Erlewyn-Lajeunesse M, Manek R, Lingam R, et al. Anaphylaxis following single component measles and rubella immunisation. Arch Dis Child. 2008;93:974–975
4. Erlewyn-Lajeunesse M, Hunt LP, Heath PT, et al. Anaphylaxis as an adverse event following immunisation in the UK and Ireland. Arch Dis Child. 2012;97:487–490
5. Cheng DR, Perrett KP, Choo S, et al. Pediatric anaphylactic adverse events following immunization
in Victoria, Australia from 2007 to 2013. Vaccine. 2015;33:1602–1607
6. McNeil MM, Weintraub ES, Duffy J, et al. Risk of anaphylaxis after vaccination in children and adults. J Allergy Clin Immunol. 2016;137:868–878
7. Berner R, Bialek R, Forster J. Erhebungseinheit für seltene pädiatrische Erkrankungen in Deutschland (ESPED) . Monatsschr Kinderheilk. 2004;152:7779
8. Rüggeberg JU, Gold MS, Bayas JM, et al.Brighton Collaboration Anaphylaxis Working Group. Anaphylaxis: case definition and guidelines for data collection, analysis, and presentation of immunization safety data. Vaccine. 2007;25:5675–5684
9. MedDRA Homepage. 2014Accessed March 30, 2014 Available at: http://www.meddra.org/how-to-use/tools/smqs
10. Petersen CGJ. The yearly immigration of young plaice into the Limfiord from the German Sea. Rep.Dan.Biol.Stn. 1896;6:5–84
11. Hook EB, Regal RR. Capture-recapture methods in epidemiology: methods and limitations. Epidemiol Rev. 1995;17:243–264
12. No authors listed. . Capture-recapture and multiple-record systems estimation I: History and theoretical development. International Working Group for Disease Monitoring and Forecasting. Am J Epidemiol. 1995;142:1047–1058
13. No authors listed. . Capture-recapture and multiple-record systems estimation II: Applications in human diseases. International Working Group for Disease Monitoring and Forecasting. Am J Epidemiol. 1995;142:1059–1068
14. Chapman DG Some Properties of the Hypergeometric Distribution with Applications to Zoological Sample Censuses, Vol. 1. 1951 Berkeley, CA: University of California Publication in Statistics:1131–1160
15. Huang WT, Huang WI, Huang YW, et al. The reporting completeness of a passive safety surveillance system for pandemic (H1N1) 2009 vaccines: a capture-recapture analysis. Vaccine. 2012;30:2168–2172
16. Walter D, Böhmer MM, Heiden Ma, et al. Monitoring pandemic influenza A(H1N1) vaccination coverage in Germany 2009/10 - results from thirteen consecutive cross-sectional surveys. Vaccine. 2011;29:4008–4012
17. Rieck T, Feig M, Eckmanns T, et al. Vaccination coverage among children in Germany estimated by analysis of health insurance claims data. Hum Vaccin Immunother. 2014;10:476–484
18. Seber G. The effects of trap response on tag-recapture estimates. Biometrika. 1970;26:13–22
19. Wittens JT. On the bias and estimated variance of chapman’s two-sample capture-recapture population estimate. Biometrics. 1972;28:592–597
20. Chao A. Estimating the population size for capture-recapture data with unequal catchability. Biometrics. 1987;43:783–791
21. Patja A, Mäkinen-Kiljunen S, Davidkin I, et al. Allergic reactions to measles-mumps-rubella vaccination. Pediatrics. 2001;107:E27
22. Tavares F, Delaigle A, Slavin D, et al. Anaphylaxis following H1N1 pandemic vaccines: safety data in perspective. Vaccine. 2011;29:6402–6407
23. Rouleau I, De Serres G, Drolet JP, et al. Increased risk of anaphylaxis following administration of 2009 AS03-adjuvanted monovalent pandemic A/H1N1 (H1N1pdm09) vaccine. Vaccine. 2013;31:5989–5996
24. Weber JCP. Epidemiology of adverse reactions to non-steroidal antiinflammatory drugs. Adv Inflam Res. 1984;6:1–7
25. Seitz CS, Bröcker EB, Trautmann A. Vaccination-associated anaphylaxis in adults: diagnostic testing ruling out IgE-mediated vaccine allergy. Vaccine. 2009;27:3885–3889
26. Vanlander A, Hoppenbrouwers K. Anaphylaxis after vaccination of children: review of literature and recommendations for vaccination in child and school health services in Belgium. Vaccine. 2014;32:3147–3154