Descriptive epidemiology of exanthems in the Rotterdam region January 1997 to June 1998

O. Ronveaux 1,2, A. Bosman3, R. Reintjes 2,3, M. Conyn-van Spaendonck2
1. European Programme for Intervention Epidemiology Training (EPIET, a programme funded by DGV of Commission of European Communities).
2. Department of Infectious Diseases Epidemiology, National Institute of Public Health and the Environment (RIVM, The Netherlands).
3. Public Health Service (GGD), Rotterdam, The Netherlands.

IntroductionThe European Advisory Group on Immunisation has recommended that measles should be eliminated from Europe by the year 2007 (1), a target accepted by National Immunisation Programme Managers for the World Health Organization (WHO) European Region countries. In the run-up to measles elimination, it is important to study the epidemiology of exanthems (rashes associated with febrile illness). Measles was formerly the commonest exanthem disease of childhood in Europe, but it is seen less and less as vaccine coverage increases. Vaccine coverage in the Netherlands has been stable for many years at 93% for the two measles, mumps, and rubella doses included in the national vaccination programme (at 14 months and 9 years) and, in 1997, the notified national incidence of measles was only 0.14/100 000 (2). At the same time, the proportion of measles cases among vaccinated individuals is increasing, with the result that less typical presentations (milder rashes) are likely to be seen in an increasing proportion of patients (3). Thus general practitioners (GPs) may become less familiar with the diagnosis (4). We studied the epidemiology of non-vesicular exanthems in the Rotterdam region to estimate the proportion of such illnesses caused by measles virus infection. We made use of two available surveillance systems that report diagnostic information on exanthems: the Rotterdam GP sentinel network and the regional school network. In addition to describing the epidemiology of exanthems, we wanted to assess the value of these surveillance systems in the context of elimination of measles.

Methods

Description of the networks

Since 1965, the public health service (GGD Gemeenschappelijke Gezondheidsdienst) in Rotterdam (1995 population: 772 913 inhabitants with 350 GPs) has run a sentinel GP network project. During the study period (1 January 1997 to 30 June 1998), the network consisted of 16 GPs, covering a population of 45 000, who voluntarily sent weekly reports to GGD on a paper form. All exanthems (with and without vesicles) were reported using standardised case definitions for the commonest infectious exanthems. Clinical criteria for the diagnosis of measles included fever > 38ºC and coryza or conjunctivitis or cough and rash (morbilliform) or Koplik’s spots. GPs rarely seek laboratory confirmation of exanthems, therefore most diagnoses were recorded as presumptive.

The primary schools of the Rotterdam region (247 schools) form another network. The city obliges schools and nurseries to notify to the GGD (local decree) cases of selected diseases among pupils aged 4 to 12 years, including cases of exanthems. Diagnoses are reported by teachers, who are often informed by parents (during absence leave, or after the child’s return). Diagnoses reported were not endorsed systematically by either a GP or the GGD.

Data management

Data were entered anonymously onto two different Epi-Info databases (Version 6.03b) and analysed. The GP sentinel network database collects information on patient’s age, sex, main diagnosis, date of diagnosis, and remarks about the case (including laboratory confirmation). The list of people registered with each GP was available; the denominator used to calculate incidence was the sum of the people registered with the GPs who reported each week.

The school network database collects information on patient’s sex, year of birth, school grade, main diagnosis, date and source of notification, and remarks about the case.

Results

The GP sentinel network

From 1 January 1997 to 30 June 1998, an average of 13.7 reports were sent from the GP sentinel practices each week, for an average participation rate of 85%. A total of
2 875 912 patient-weeks or 36 635 patients per week on average were covered.

A total of 410 non-vesicular exanthems were reported, equivalent to an overall incidence of 741 episodes per 100 000 population (95% confidence interval (CI) 688-796). The average incidence evolved from 12.7/100 000 per week for the first six months of 1997 to 12.8 for the second six months of 1997, and 17.4 in the first six months of 1998. A presumptive diagnosis was reported in 61% of exanthems (251/410; table 1). Not a single case of measles was identified. Cases were aged between 0 and 95 years, and children under 5 years of age represented 44% of cases (n=174). Cases of exanthem subitum occurred in the youngest, while other infectious diseases occurred mainly in older children. Allergic reactions were commonest in older people (table 1).

Table 1: Frequency of diseases reported among exanthems in the GP sentinel network and age distribution, Rotterdam, January 1997-June 1998

Disease

No.

Percent of total

Age (years)

Mean

P25

P50

P75

Exanthem subitum

57

13.9

2.1

1

1

3

Allergic reaction

52

12.7

15.5

3

9

26

Erythema infectiosum

46

11.2

7.4

4

5

8

Scarlet fever

32

7.8

5.3

3

4

6

Rubella

2

0.5

17.5

5

18

30

Other viral infections

13

3.2

9.5

1

4

8

Other infections

49

12.8

33.3

9

28

52

Unknown-not specified

159

38.8

16.7

3

6

26

Total

397ƒ

96.8

14.6

2

5

21

ƒ Age unknown for 13 casesOnly 4% of diagnoses were reported to be laboratory confirmed.

Among 140 children from 4 to 12 years of age, the commonest diseases were erythema infectiosum (n=31, 22%) and scarlet fever (n=20, 14%). The sex ratio (M:F) was 1:1.3, but the female excess was seen in older age groups.

Looking at the evolution of the incidence for selected infectious exanthems by half-year (figure 1), the most striking feature was the increase in parvovirus infections detected in 1998. This explained most of the increase in incidence observed during the first half of 1998.

0312-023-01

 

The school network

A total of 140 schools (57% of all schools) notified 211 exanthems (table 2). A very small proportion (7%) were endorsed medically. Only one of the five notified cases of measles was reported to have been seen by a GP, who rejected the diagnosis.

School grades, reported in 125 cases (59%), were all represented. The number of diseases notified decreased with increasing grade. The sex ratio overall (M:F) was 1.05:1.

The trends of scarlet fever cases reported each month by the school network showed a seasonal pattern with a peak in late winter in 1997 and 1998.

Table 2: Frequency of diseases reported among exanthems in the school network, Rotterdam, January 1997-June 1998

 

Disease

No.

Percent of total

Age (years)

Mean

P25

P50

P75

Exanthem subitum

57

13.9

2.1

1

1

3

Allergic reaction

52

12.7

15.5

3

9

26

Erythema infectiosum

46

11.2

7.4

4

5

8

Scarlet fever

32

7.8

5.3

3

4

6

Rubella

2

0.5

17.5

5

18

30

Other viral infections

13

3.2

9.5

1

4

8

Other infections

49

12.8

33.3

9

28

52

Unknown-not specified

159

38.8

16.7

3

6

26

Total

397ƒ

96.8

14.6

2

5

21

 Discussion

The results given by the two Rotterdam networks illustrate exanthem epidemiology in the region. The two surveillance systems gave two different pictures, however, for reasons that may include the different age distributions and sources of the subjects of both networks. The diseases reported by the school network are, by definition, limited to those which occurred in school-aged children, particularly contagious diseases. Hence, allergic reactions were scarcely reported by the schools, perceived only as a minor threat for the school, and exanthem subitum cases were not reported as they usually occur before 3 years of age. In the specific primary school-aged population, however, the GP network diagnosed proportionally many more erythema infectiosum cases than the school network. This discrepancy may reflect a problem of misdiagnosis.

The small number of cases of measles reported in the two Rotterdam networks is not surprising in view of the effective Dutch vaccination programme. Very few measles cases have been notified nationally for several years (2). The small proportion of exanthems whose diagnoses are laboratory confirmed should lead to a cautious interpretation of these data. The low confirmation rate, however, could be an underestimate if laboratory confirmation occurred after cases were reported, or if confirmation was not reported as such (it is not a specific field on the notification form)

Confirmed measles has been reported to account for around 5% of exanthems in countries with high measles vaccine coverage (5-7). This percentage was respectively 0% and 2.4% in the GP network and in the school network. This discrepancy may be due to cases of mild measles which were not recognised: the proportion of unknown diagnoses was high in the GP network. In addition to this, as elimination approaches, the measles clinical signs has a lower positive predictive value (5). Therefore, the need for laboratory confirmation of suspected measles cases is crucial. Measles diagnostic salivary tests have been developed, and are being used in the United Kingdom (6) and being tested in France (8); their advantages include speed of diagnosis, sensitivity, and cheapness. They are easy to perform, and could even be used in non-medical settings such as schools. In the context of elimination, they should be used to detect measles in exanthems of unknown origin.

The networks illustrated classical seasonal patterns of scarlet fever (9), and the increase in erythema infectiosum coincided with what was seen nationally (10). Typically, erythema infectiosum epidemics occur at the end of winter and beginning of spring. Hence, both networks are sensitive in detecting trends of common diseases. The GP sentinel network is no longer relevant for the detection of a rare disease such as measles because so few cases now occur that they are likely to be seen by physicians who do not belong to the sentinel surveillance system.

The school network represents only one age-selected part of the population, but its role may be particularly important. The system covers all areas, and reported diseases have direct practical consequences (e.g., quarantine, exclusion from school, chemoprophylaxis in the event of outbreaks of meningitis, immunisation in the event of hepatitis A), which motivate the network’s participants. Moreover, although notification is a statutory requirement in Rotterdam, we did not evaluate the exhaustiveness of the system. However, the system should be improved by systematic medical support. Finally, as high measles immunization coverage is achieved, the age distribution is expected to shift to affect older people and the possibility of conducting surveillance among other age groups should be considered. Cases occurring at pre-vaccination age should be covered too.

In conclusion, this study yielded substantial data on the epidemiology of exanthems in the Rotterdam region. In the context of the elimination of measles, the sentinel surveillance of measles is poorly efficient, but the school network might play a role. For the elimination purpose, there is a need for exhaustive surveillance of measles and systematic laboratory confirmation.


References1. Begg N, Ramsay M. World Health Organization aims to eliminate measles in Europe by 2007. Eurosurveillance Weekly 1997; 1: 971127. (http://www.eurosurv.org)

2. van den Hof S, Conyn-van Spaendonck MAE, de Melker HE, Geubbels ELPE, Suijkerbuijk AWM, Talsma E, et al. In:‘The effects of vaccination, the incidence of the target diseases’. Bilthoven: RIVM, 1998: 91-6. (report 213676 008)

3. Edmonson MB, Addiss DG, McPherson JT, Berg JL, Circo SR, Davis JP. Mild measles and secondary vaccine failure during a sustained outbreak in a highly vaccinated population. JAMA 1990; 263: 2467-71.

4. Tait DR, Ward KN, Brown DWG, Miller E. Measles and rubella misdiagnosed in infants as exanthem subitum (roseola infantum). BMJ 1996; 312: 101-2.

5. Papania M, Bromberg K, Grabowsky M, Stewart JD. Differential diagnosis of febrile rash illness in children, Brooklyn 1994. In: Program and abstracts of the 36th Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington: American Society for Microbiology, 1996. (abstract no. K748, p.263)

6. Ramsay M, Brugha R, Brown D. Surveillance of measles in England and Wales: implications of a national saliva testing programme. Bull World Health Organ 1997; 75: 515-21

7. Gay N, Ramsay M, Cohen B, Hesketh L, Morgan-Capner P, Brown D, et al. The epidemiology of measles in England and Wales since the 1994 vaccination campaign. Commun Dis Rep CDR Rev1997; 7: R17-21.

8. Antona D. Résultats préliminaires d’une étude de faisabilité d’une surveillance exhaustive de la rougeole dans 3 départements français à couverture vaccinale élevée. (http://www.b3e.jussieu.fr/msp/publicat/rougeole/page11.html#res)

9.Benenson AS. Control of communicable diseases manual. Washington DC: American Public Health Association, 1995: 411-7.

10. Steenbergen JE van. Toename parvovirus. Infectieziekten Bulletin 1998; 9: 173-4.

Copyright(c) 2015 Eurosurv.All Rights Reserved