Background
An increase in the incidence of pertussis, particularly among older children, adolescents and adults, was observed since the 1980s in the United States and since the 1990s in Canada and several European countries, all with high childhood pertussis vaccination coverage [
1‐
5]. Pertussis is most severe among infants, who have the highest risk of complications, hospitalization and death [
6‐
8]. Although the disease is frequently milder among older children, adolescents and adults, a protracted course is the rule and complications such as pneumonia, urinary incontinence, weight loss and rib fractures may occur [
9‐
12]. Several factors contribute to the persistent circulation of the infectious agent,
Bordetella pertussis, despite immunization programs, the most important of which is waning of pertussis-specific immunity after about 7 to 20 years after natural infection and 4 to 12 years after immunization [
13,
14]. Furthermore,
B. pertussis is highly contagious, with a basic reproductive number of 12–17 [
15]. Because of the non-specific initial presentation of the illness and because the infection may not cause the full-blown clinical picture in immunized individuals or those with previous natural infection, the diagnosis may be delayed or overlooked, thereby leading to prolonged transmission. Thus the susceptibility to and transmission of the disease is influenced by current and past vaccination coverage, vaccination schedule, type of vaccine in use and social mixing patterns.
Vaccination recommendations, resulting vaccination coverage and the incidence of pertussis differed markedly in the former German Democratic Republic (GDR) and the former West Germany (FWG), with differences persisting for some time after reunification. Therefore, the aim of our paper is to compare the epidemiology of pertussis in the two parts of Germany to better understand patterns of disease transmission and optimize prevention efforts.
Methods
Official pertussis vaccination recommendations in the GDR, FWG and the reunited Germany were reviewed. The Standing Committee on Vaccination (STIKO) was responsible for vaccination recommendations in FWG since 1972, and since 1991 also in the reunited Germany. Prior to 1972, a "Pertussis-Committee" of pediatric and infectious disease specialists made recommendations in FWG [
16]. Recommendations are generally adopted by the individual federal states, which can, however, deviate from them. In the former GDR the ministry of health was responsible for making vaccination recommendations [
16].
Data on pertussis vaccination coverage in Germany routinely obtained from obligatory medical examinations performed by the regional health authorities at school entry (children aged 5 to 7 years) have been made available to RKI by the federal states since 1998. In addition, Medline, Embase and Embase Alert data bases were searched for vaccination coverage studies in Germany using the free text search terms [(pertussis or whooping cough) AND Germany AND (vaccination coverage or vaccination rate)].
The following official data sources were used to obtain data on the incidence of pertussis: Pertussis was statutorily notifiable from 1947 to 1961 in Germany and data were published by the Federal Statistical Office [
17]. In FWG, notification was no longer required as of 1962; thus surveillance data are lacking. In GDR pertussis was a statutorily notifiable disease and aggregate data for 1962–1990 were published by the Statistical Office of the German Democratic Republic. Reporting continued after reunification in five former East German states (referred to as FEG: Brandenburg (BB), Mecklenburg-Western Pomerania (MV), Saxony (SN), Saxony-Anhalt (ST), Thuringia (TH), but not East Berlin, counted as part of FWG after reunification in 1991). These 5 states transmitted case-based data with a minimal variable set to the Robert Koch-Institute (RKI), the federal institution responsible for disease control and prevention, from 1995 to 2000, enabling a basic demographic analysis. Pertussis was not made a notifiable disease under the new German Protection Against Infection Law implemented in 2001. However, FEG continued to require statutory reporting of pertussis according to state-specific regulations. From 2001 onwards notified case-based data were transmitted to RKI, although these were incomplete in 2001.
A uniform case definition was applied from 2002 onwards [
18], thus a detailed analysis of FEG surveillance data was undertaken for the period January 2002 to December 2007 according to region, age and sex, hospitalization and vaccination status, as well as outbreaks. According to the surveillance case definition in place from 2002–2008, a clinical case must manifest at least one of the following symptoms for at least 2 weeks: i. cough attacks, ii. whooping, iii. post-tussive vomiting or – in infants – iv. episodes of apnea. Laboratory diagnosis, initiated at the discretion of the physician, consists of cultural isolation or detection of
B. pertussis by PCR in nasopharyngeal swabs or secretions or serological diagnosis by means of an elevated pertussis-specific IgA- antibody concentration in a single serum sample or a 4-fold increase in IgG- or IgA- pertussis-specific antibodies using any commercially available test kit (e.g. Serion ELISA
classic measuring IgG-[PT, FHA (filamentous hemagglutinin)] and IgA- (PT, FHA) antibodies, Genzyme Virotech GmbH ELISA measuring IgG- and IgA-antibodies against PT, or Bordetella pertussis IgG Virastripe
®, an immunoassay also measuring IgG-antibodies against PT and FHA (produced by Viramed) [
19]. The majority of the 18,080 cases notified to RKI between January 2002 and December 2007 were laboratory confirmed (93.8%); 1.9% fulfilled the clinical case definition and were epidemiologically linked to a laboratory confirmed case and 4.4% solely fulfilled the clinical case definition. Of the laboratory confirmed cases, 172 (1.0%) were diagnosed by culture, 1769 (9.8%) by PCR, 4861 (26.9%) by an increase in pertussis-specific antibodies, and 10,153 (56.2%) by a high concentration of pertussis-specific IgA-antibodies in a single serum sample. Detailed data on the vaccination status of cases were available from 2004 onwards for 78% of notified cases. A case was defined as adequately vaccinated if 3 doses of pertussis vaccine had been given with the third dose less than one year prior to disease onset or if at least 4 doses had been given with the last dose = 10 years prior to disease onset. Doses received within 3 weeks prior to disease onset were not counted.
In addition to official data on pertussis disease burden, Medline, Embase and Embase Alert data bases were searched for further sources on pertussis incidence, such as sentinel studies or regional surveys, in Germany using the search terms [(pertussis or whooping cough) AND Germany AND (incidence or disease burden or morbidity)].
Discussion
This overview describes markedly different development of pertussis epidemiology in the two parts of Germany prior to reunification in 1991. The divergent vaccination scenarios in the 1970s and 1980s led to a much lower pertussis incidence in FEG than FWG. These differences can be ascribed to limited use of pertussis vaccine in FWG due to concerns about adverse effects in contrast to continued high vaccination coverage in FEG. Despite the fact that severe adverse effects due to pertussis whole cell vaccine were never substantiated, reports to this effect led to reduced vaccine uptake and an increase in pertussis morbidity in other countries as well [
41].
Analysis of pertussis epidemiology in Germany is limited in that notification data after 1961 are accessible only from FEG. Only a partial assessment of the situation in FWG is possible based on hospital discharge and mortality data. Moreover, the higher incidence of pertussis cases in FEG in infants based on hospital discharge statistics than based on notified cases until 2007 suggests a significant degree of underreporting in the FEG statutory surveillance system. Results of sentinel surveillance in a FEG city [
40] also suggest that disease burden in adults is markedly underestimated by routine surveillance data, likely due to atypical presentation in persons with pre-existing natural or vaccine-induced immunity. A further potential limitation of routine surveillance data from FEG lies in the high proportion of cases diagnosed solely serologically. A recent study showed that the performance of the most frequently used ELISAs in Germany was variable regarding antigens used for testing, sensitivity (60–95%), and quantitative measurement of antibody levels [
42]. Specificity was not reported, but other studies have shown high specificities of >90% for ELISAs based on PT- and FHA-antigens [
43,
44]. The observation that only 4.1% of all notified cases received a dose of pertussis vaccine < 3 years prior to illness onset suggests that recent vaccination is not an explanation for positive serology in the majority of cases. As well, an analysis of only those cases diagnosed by culture, PCR or a rise in antibody concentrations showed similar age distributions and trends over time as did the analysis of all cases presented here (data not shown).
In addition, the observed epidemiologic patterns, such as the higher pertussis incidence observed in women than men in FEG, are consistent with other studies: High pertussis-specific IgG-antibodies indicative of recent infection were found more frequently among women than men in a seroepidemiological study in England and Wales [
45], and in a household contact study performed in Germany, secondary pertussis cases also occurred more frequently among women than men [
46]. Furthermore, the recent increase in pertussis incidence in FEG associated with a shift to older age groups mirrors similar developments in pertussis epidemiology in other western countries with longstanding vaccination programs [
47]. For instance, in the United States, an increase in incidence was observed beginning in the 1980s and, starting in the early 1990s, particularly among 10 to 19 year olds [
10,
48,
49] despite improved vaccination uptake of the preschool booster [
49], thus suggesting waning immunity 5–10 years after the preschool booster.
Correspondingly, despite high vaccination coverage with the primary vaccination series among FEG preschool children in recent years, the incidence of pertussis continued to increase in the four FEG states that did not recommend a booster vaccine dose prior to school entry until 2006. The increase in incidence since 2004 was most marked in children between 5 and 14 years of age, a high proportion of whom had received 4 doses of pertussis vaccine in infancy, suggesting that waning immunity was playing a role. This is supported by an outbreak investigation performed in a highly vaccinated school population in Mecklenburg Western-Pomerania in 2006, which showed a markedly higher attack rate in children who had received their fourth vaccine dose >5 years previously [
19]. Increased vaccination coverage may lead to a period of decreased natural boosting, which may "unmask" the effect of waning vaccine-induced immunity, causing incidence to increase despite a lower level of transmission. Modeling studies suggest that the severity of disease – and thus the diagnostic threshold – depends on the level of transmission, with a higher incidence of severe disease expected at intermediate transmission intensities [
50]. Thus an overall reduction in transmission initiated through increased vaccination coverage could theoretically lead to a higher incidence of more severe – and therefore recognized – disease through a combination of waning immunity and less natural boosting.
An increase in pertussis incidence in Saxony, where a preschool booster was recommended since 1998, occurred in 2007. Although all age groups were affected, the increase was most pronounced in children aged 10 to 14 years, suggesting that uptake of the adolescent booster vaccination may be insufficient or occurring too late. The Saxony state health authority also increased awareness for the diagnosis of pertussis among physicians by actively informing about the possibility of PCR diagnosis at the state-run microbiological laboratory in 2007 (personal communication, DB). This, along with the higher proportion of cases diagnosed < 3 years after the primary vaccination series in 2007, suggests that a change in diagnostic practices might partially explain the observed increase. The incidence in Saxony still remains lower than that in all other FEG states (Fig.
2) and the increase in 2007 may also in part be due to periodic variation in incidence.
The continued increase in pertussis incidence in FEG despite improved vaccine uptake after the mid-1990s also coincided with widespread introduction of acellular vaccines in Germany. Thus another possible explanation might be a shorter duration of immunity compared to vaccination with whole cell vaccines. In FEG, the pertussis component of the combined diphtheria, tetanus, pertussis vaccine consisted of alum-adsorbed
B. pertussis (30–40 × 10
9 bacteria/ml), equivalent to at least 4 international potency units per 0.5 ml dose [
20,
51], fulfilled international potency requirements of at least 4 IPU issued by the World Health Organization in 1964 [
52]. In general, available studies suggest a similar duration of immunity after vaccination with acellular and whole cell vaccines [
53], although comparison is difficult due to the heterogeneity of both vaccine types, and inability to control for circulating levels of pertussis in the population, which could influence immunity through natural boosting.
The higher incidence of hospitalization due to pertussis in infants in FWG than FEG suggests a higher level of disease transmission in FWG, in keeping with lower vaccination coverage particularly in older children and adolescents, who along with adults presumably act as a transmission reservoir for infants [
54‐
58]. However, the slightly higher incidence of pertussis-related hospitalizations in FEG in older children and adolescents, who have higher vaccination coverage than their counterparts in FWG, is surprising. A higher incidence of pertussis hospitalizations in FEG than FWG was also found during intensified hospital-based surveillance for pertussis complications in children under 16 years in 1997–1998 [
59]. Possible explanations could be a lower hospitalization threshold in FEG, perhaps due to a lower density of physicians in private practice than in FWG [
60] or a higher degree of awareness for the diagnosis.
Sentinel surveillance in two cities in FWG and FEG [
40] suggests that pertussis incidence in the two parts of Germany is similar in adults despite the lack of childhood vaccination in FWG prior to 1991. Adults in FEG vaccinated in childhood presumably experienced less natural boostering than non-vaccinated adults in FWG, who would have experienced natural infection – thought to induce longer lasting immunity than vaccination – in childhood more frequently than in FEG. Thus, FEG adults might be expected to be more susceptible to pertussis than FWG adults. This could explain the similar incidence in adults in the two parts of Germany despite evidence for a higher level of
Bordetella pertussis transmission in FWG. This is corroborated by a seroprevalence study performed in several European countries including FWG and FEG in the mid-1990s [
61,
62], which revealed that overall, adults were less likely to have high pertussis-toxin antibody (anti-PT) concentrations suggestive of recent acute infection than younger persons. However, in countries with a history of high vaccination coverage (The Netherlands, Finland and FEG), adults comprised a higher proportion of cases with recent pertussis infection than in countries with low vaccination coverage (France, FWG, United Kingdom, and Italy). Thus, in the mid-1990s adults were 0.6 times less likely to have high anti-PT than children and adolescents in FEG, but 0.3 times less likely in FWG. While this difference in age distribution in the two parts of Germany has likely narrowed, the markedly higher incidence of hospitalizations for pertussis in infants and lower vaccination coverage among adolescents in FWG suggest that differences in epidemiologic pattern persist. Furthermore, differences in contact patters, another determinant of infectious disease transmission [
63], likely also exist. For instance, one difference between FEG and FWG in this regard is the higher proportion of children attending day care in FEG, even close to 20 years after reunification. In FEG states, day care attendance of children < 3 years of age in 2007 ranged from 31.7% in Saxony to 51.4% in Saxony-Anhalt, but from 5,2% to 36,1% in FWG states [
64], with less extreme differences in older children. Such differences could influence disease transmission patterns between children as well as from children to adults. However, this remains hypothetical without robust surveillance data from both parts of Germany.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
WH performed the literature searches, collated the data from the different sources, performed most of the data analysis and drafted the manuscript. DB, EJ, ML, HO provided and contributed to the analysis of routine surveillance data, SR performed analysis of the school entry vaccination coverage data, and all authors contributed to the interpretation of the results and to the final version of the manuscript.