Pertussis surveillance in Switzerland, 1992 to 1997 – a large epidemic in 1994. What next?

H. C. Matter 1 and Sentinella Arbeitsgemeinschaft 1,2
1. Swiss Federal Office of Public Health, Epidemiology and Infectious Diseases, Viral Diseases and Sentinel Systems, Bern, Switzerland
2. Fakultäre Instanz für Allgemeinmedizin (FIAM), University of Bern, Bern, Switzerland

IntroductionCases of whooping cough have been recorded by the Swiss Sentinel Surveillance Network ‘Sentinella’ (SSSN) since June 1991 to monitor pertussis trends in Switzerland and to assess the current vaccination strategy (1,2). The Sentinella network consists of a sample of about 200 general practitioners, specialists in internal medicine, and paediatricians in private practice who report voluntarily morbidity data each week to the Swiss Federal Office of Public Health (SFOPH). The participating physicians account for about 3.5% of all physicians under 65 years of age in each of the three specialities in Switzerland.

Cases of whooping cough are defined as patients with a cough persisting at least 14 days – either characterised by acute episodes, barking quality, or vomiting after a coughing fit (‘sporadic’ cases), or who have been in contact with another confirmed or suspected case of whooping cough (‘epidemic’ cases). Since 1994, the Children’s Hospital in Basel has offered testing of nasopharyngeal secretions for Bordetella pertussis by polymerase chain reaction (PCR) for all cases notified to the SSSN free of charge (3, 4). Before 1994 no standard system for the microbiological confirmation ofB. pertussis was available to the SSSN.

The Swiss immunisation schedule recommends primary immunisation against whooping cough (three injections of diphtheria, tetanus, and pertussis (DTP) with either whole cell pertussis or acellular pertussis vaccine (DTPa)) of infants at the ages of 2, 4, and 6 months. Boosters are recommended at the age of 15 to 24 months (DTP or DTPa) (fourth dose) and school entry (4-7 years, DTPa) (fifth dose) (5). Until 1996, most infants and children received a locally produced whole cell vaccine, which was first registered in 1950. Since 1996, various acellular vaccines introduced in Switzerland have been approved for primary vaccination, the fourth and fifth doses, or both. Two national studies of vaccine coverage in children aged 27 to 35 months in 1991 and 1998 found that about 90% of toddlers had received at least three doses of a pertussis vaccine (6,7).

Results

Incidence

Whooping cough is endemic in Switzerland. The annual incidence of whooping cough, based on Sentinella surveillance data was less than 100 cases per 100 000 population in 1992 and 1993 but rose in the next two years (table 1). The increase was especially marked from July to October 1994, when cases reached epidemic levels (figure). We estimated that 40 000 clinical cases of pertussis (12 500 confirmed by PCR) occurred during the 1994-1995 epidemic in Switzerland. Infants and children up to the age of 6 years were most affected (2).

Table 1: Cases of whooping cough reported in the Swiss Sentinel Surveillance Network, 1992-1997

1992

1993

1994

1995

1996

1997

Number of clinical cases reported

142

146

879

652

384

476

Estimated annual incidence for Switzerland (per 100 000 population)

70

70

370

280

165

185

Estimated total number of clinical cases in Switzerland

5000

5000

26 000

20 000

11 500

13 000

Laboratory test result available * (%)

0

0

718 (81.7)

547 (83.9)

296 (77.1)

351 (73.7)

Positive laboratory test result* (%)

0

0

244 (34.0)

105 (19.2)

58 (19.6)

76 (21.7)

Percentage of female patients

62.7

53.4

55.5

57.8

55.5

53.6

Percentage of female patients among adults (20 years or more)

94.1

82.8

70.1

67.3

63.3

67.6

Median age (years, if under 20 y)

Male

5.5

6.7

5.8

5.8

5.2

6.0

Female

5.8

5.8

5.6

5.4

5.7

6.6

Total

5.5

6.3

5.7

5.7

5.5

6.3

Identification of Bordetella pertussis in nasopharyngeal swabs by PCR 

0412-023-01

An estimated 11 500 and 13 000 cases arose in Switzerland in 1996 and 1997 (with 2000-2500 and 2500-3000, respectively, confirmed by PCR). The estimated incidence of clinical pertussis in the two years was 165/100 000 (1996) and 185/100 000 (1997) given that the total population of Switzerland in 1997 was 7.08 million (8).

No upward shift in the ages of patients has been observed since 1992. Among patients under the age of 20 years, the median age remained between 5.5 years and 6.3 years. Among adult patients (20 years and over), women outnumbered men (table 1). No deaths due to B. pertussis infection were reported to the Swiss Federal Statistical Office between 1992 and 1996 and mortality data for 1997 are not yet available. Between 1994 and 1997, 2% of all reported cases and 16% of all cases under the age of 1 year were admitted to hospital. Thirty per cent of infants under 1 year of age with PCR-confirmed pertussis were admitted to hospital.

PCR-confirmed cases

Laboratory confirmation was available for 81% of all clinical cases reported between 1994 and 1997. Overall, 25% of clinical cases were PCR positive (34% in 1994, 19% in 1995, 20% in 1996, 22% in 1997). Over 40% of the ‘epidemic’ cases tested by PCR were positive, double the rate of confirmation for ‘sporadic’ cases (table 2).

Table 2: PCR results: Swiss Sentinel Surveillance Network, March 1994 – December 1997

 

Sporadic pertussis Epidemic pertussis Cough < 14 days at day of reporting*

No indication

Total

Without contact **

With contact

Cases reported (%)

1156 (49.1)

528 (22.4)

239 (10.1)

214 (9.1)

220*** (9.3)

2357 (100.0)

PCR-result available (%)

986 (85.3)

407 (77.1)

213 (89.1)

177 (82.7)

129 (58.6)

1912 (81.1)

PCR-positive cases (%)

195 (19.8)

168 (41.3)

40 (18.8)

63 (35.6)

17 (13.2)

483 (25.3)

*These cases did not fulfill the requested minimal duration of cough at the day of reporting (consultation) but were reported by the participating physicians.
**Contact to a pertussis case
***Also included are 12 cases of cough illness lasting more than 14 days for which neither paroxysms, inspiratory whoop, nor post-tussive vomiting nor an epidemiological link to a pertussis case was reported. Of these, nine cases were tested by PCR and one had a positive test result.Doctors reporting to SSSN completed questionnaires for 81% (1906) of all pertussis cases reported between 1994 and 1997. For 84% (1598) of these, PCR was performed and 28% (449) were positive. A logistic regression showed that requests for a laboratory analysis were associated with certain characteristics (table 3). Paediatricians submitted nasopharyngeal samples for PCR more often than general practitioners and specialists in internal medicine. Laboratory tests were requested more often if the diagnosis was uncertain, as in sporadic cases and/or if typical symptoms of whooping cough – such as vomiting, a barking cough, or cyanosis – were absent. Finally, the locality of the physician’s practice was significantly associated with the decision to request PCR. Age, sex, and vaccination status of the patient, the presence of paroxysms, inspiratory whoop, dyspnoea, disturbed sleep, fever, or catarrhal symptoms did not influence the decision to request PCR.

Table 3:  Characteristics associated with PCR testing

 

Variable

PCR result available* (%)

Crude OR (95% CI)**

Adjusted OR (95% CI)**

Vomiting

No

694 (89.3)

1

1

Yes

904 (80.1)

0.5 (0.4; 0.6)

0.5 (0.4; 0.7)

Barking

No

516 (87.8)

1

1

Yes

1082 (82.1)

0.6 (0.5; 0.9)

0.7 (0.5; 1.0)

Cyanosis

No

1488 (84.6)

1

1

Yes

110 (74.3)

0.5 (0.4; 0.8)

0.6 (0.4; 0.9)

Year of reporting

1994-95

1067 (88.3)

1

1

1995-96

531 (76.1)

0.4 (0.3; 0.5)

0.4 (0.3; 0.5)

Case definition

Epidemic

550 (79.1)

1

1

Sporadic

1048 (86.5)

1.7 (1.3; 2.2)

1.9 (1.5; 2.5)

Location of the practice
French/Italian part

525 (82.0)

1

1

German part

1073 (84.8)

1.2 (0.9; 1.6)

1.6 (1.2; 2.1)

Speciality of the physician

Paediatrics

918 (88.0)

1

1

General practice, internist

680 (78.8)

0.5 (0.4; 0.7)

0.5 (0.3; 0.6)

* PCR test availability (test performed [1]; no test performed [0]) was used as response variable
**OR Odds ratio ; OR ajusted: multiple logistic regression odds ratio
CI Confidence interval (as indicated by EPI-INFO and SYSTAT®)A second logistic regression was performed to identify factors associated with a positive PCR result (data not shown). Positive results were associated with the presence of specific symptoms (vomiting, inspiratory whoop), the epidemiological context of the case, and geographic and sociodemographic features of the physician’s practice (which may be used as a marker of the patients’ origin). We therefore identified three variables that had an influence on both requesting PCR testing and obtaining a positive test result. Sporadic cases, cases without vomiting, and cases from the German part of Switzerland were more often tested, but were less likely to be positive than cases who had been in contact with another pertussis case, cases who had vomited, and cases from the French and Italian parts of Switzerland. As a result, the total number of PCR-positive cases estimated from the proportion of positive PCR results is probably too low, since it seems likely that a higher proportion of those not tested would have been PCR positive than those who were tested.

Vaccination

Ten per cent of the cases reported in 1997 had completed primary vaccination (three doses) and received an additional booster (documented by vaccination certificate). Another 19% were described as having been vaccinated, but the number of doses was not specified (table 4). In the age categories 7 months to 4 years, 5 to 9 years, and 10 to 19 years 8.8%, 2.2%, and 7.0% had received at least one but less than three doses as documented by vaccination certificates. In these age categories 13.8%, 7.6%, and 2.3% were not vaccinated whereas 77.5%, 90.2%, and 90.7% had completed primary vaccination.

The proportion of vaccinated patients in the over 19 age group was distinctly lower – 50% with certificates and 27% overall. A much higher proportion of adult patients than children and adolescents (aged 1-19 years) had an unknown or undocumented vaccination status.

Table 4: Vaccination status of pertussis cases reported in the Swiss Sentinel Surveillance Network, 1997

 

Age

Vaccination status documented by certificate

Vaccination status not documented

Vaccination status not known

Total

Vaccinated

Not vaccinated

Vaccinated

Not vaccinated

> 3 doses

1-2 doses

Not known

> 3 doses

1-2 doses

Not known

0-2 months

0

0

0

1

0

0

0

3

0

4

3-6 months

1

5

0

3

0

1

2

0

0

12

7-11 months

4

4

0

1

1

0

0

2

0

12

1-4 years

58

3

1

10

10

0

14

17

6

119

5-9 years

83

2

9

7

20

0

24

14

6

165

10-14 years

36

3

1

1

9

0

18

2

6

76

15-19 years

3

0

0

0

4

0

5

0

2

14

> 20 years

2

0

0

2

5

0

11

17

37

74

Total

187

17

11

25

49

1

74

55

57

476

Recommendations for primary vaccination were fulfilled if infants between 3 and 4 months, 5 and 6 months, and above 6 months had received at least one, two, and three vaccine doses, respectively.Discussion

Pertussis activity is usually characterised by a three to five year cycle. SSSN began to monitor pertussis in Switzerland in 1992 after a long period when pertussis morbidity had not been systematically monitored. Compared with the assumed epidemiological cycle of the disease, the period of surveillance is rather short, and this complicates the interpretation of the curve. The reasons for the large outbreak in 1994/95 are not fully understood. Pertussis vaccine effectiveness in young children during this period was estimated by an indirect method (9). Defining a positive PCR result as the case criterion, vaccine effectiveness was about 90%. Effectiveness was lower in older than in younger children (2), but no in-depth analysis of the efficacy of pertussis vaccine has been performed in Switzerland. We also do not know if the strains of B. pertussis which were circulating at this time had markedly changed, a phenomenon which may have been responsible for outbreaks observed elsewhere (10).

For estimations over several years, the temporal effect associated with requesting PCR testing and a positive test result should also be taken into account. Two factors may be especially important. When a new test is introduced and offered free of charge, the participating physicians are initially motivated to confirm as many cases as possible in the laboratory; subsequently the numbers of specimens submitted gradually fall. Another factor explaining the higher proportion of positive tests in 1994 and 1995 than in 1996 and 1997 may be the pertussis epidemic that occurred in the autumn of 1994 and the beginning of 1995 (2,11).

Recent outbreaks in Switzerland and various European countries have confirmed the continuing epidemic potential of B. pertussis (12). Maintenance of a high-level vaccine coverage remains highly recommended (11,13). The present situation underlines the importance of a continuous monitoring of this disease in Switzerland combining clinical case records and laboratory analysis. Our results suggest that a sentinel surveillance approach using PCR-based laboratory confirmation is appropriate to draw the national epidemic curve and to detect major outbreaks of whooping cough.

Acknowledgements

The Swiss Sentinel Surveillance Network is a co-project of the Swiss Federal Office of Public Health and the ‘Fakultäre Instanz für Allgemeinmedizin’ at the University of Bern. It is funded by the SFOPH.


References1. Matter HC, Cloetta J, Zimmermann H. Das Meldesystem Sentinella in der Schweiz am Beispiel des Pertussismonitorings von 1991 bis 1993. PRAXIS 1995; 84: 690-7.

2. Matter HC, Schmidt-Schläpfer G, Zimmermann H, Sentinella-Arbeitsgemeinschaft. Monitoring of a pertussis epidemic in Switzerland 1994-1995 by the Swiss Sentinel Network ‘Sentinella’. Schweiz Med Wschr 1996; 126: 1423-32.

3. He QS, Schmidt-Schläpfer G, Just M, Matter HC, Nikkari S, et al. Impact of polymerase chain reaction on clinical pertussis research: Finnish and Swiss experiences. J Infect Dis 1996; 174:1288-95.

4. Schmidt-Schläpfer G, Bundesamt für Gesundheit. Polymerase chain reaction für Bordetella pertussis. Ein neues Instrument für das Keuchhustenmonitoring durch Sentinella in der Schweiz. Bulletin Bundesamt für Gesundheit; 1998; 9-13.

5. Bundesamt für Gesundheit, Schweizerische Kommission für Impffragen. Impfplan für routinemässige Schutzimpfungen. Stand Dezember 1997. In: Bundesamt für Gesundheit (ed).Infektionskrankheiten Diagnose und Bekämpfung Suppl VIII, Bern: BAG; 1998; 1-4.

6. Minder C, Steffen R . Kleinkinderimpfungen. Eine repräsentative Erhebung zur Durchimpfung in der Schweiz 1991. Bulletin Bundesamt für Gesundheit ;1992; 32: 504-7.

7. Bundesamt für Gesundheit. Kleinkinderimpfungen. Repräsentative Erhebung zur Durchimpfung in der Schweiz 1998. Bulletin Bundesamt für Gesundheit ;1999; 20: 356-61.

8. Swiss Federal Statistical Office. Statistical data on Switzerland 1999. Bern: Swiss Federal Statistical Office, 1999.

9. Orenstein WA, Bernier RH, Dondero TJ, Hinman AR , Marks JS, Bart KJ, et al. Field evaluation of vaccine efficacy. Bull World Health Organ 1985; 63:1055-68.

10. de Melker HE, Conyn-van Spaendonck MAE, Rümke HC, van Wijngaarden JK, Mooi FR, Schellekens JFP. Pertussis in the Netherlands: an outbreak despite high levels of immunization with whole-cell vaccine. Emerg Infect Dis;1997; 3: 175-8.

11. Matter HC. Whooping cough in Switzerland. Eurosurveillance Weekly 1997; 1: 971223. (http://www.eurosurv.org/1997/971223.html)

12. Matter HC, Sentinella Arbeitsgemeinschaft. Bisher keine Zunahme der Keuchhustenfälle in der Schweiz. Resultate der Sentinella-Erhebung (Stand Ende Oktober 1997). Bulletin Bundesamt für Gesundheit; 1997; 50:12-3.

13. World Health Organization. CD Update. Communicable Diseases Report 1997;1-4

Copyright(c) 2015 Eurosurv.All Rights Reserved