- Email: [email protected]
Risk factors for invasive disease among children in Spain
Journal of Infection (2004) 48, 320–329
www.elsevierhealth.com/journals/jinf
Risk factors for invasive disease among children in Spain ´a,*, Javier Dı´ez-domingoa, Leopoldo Segarrab, Inmaculada Pereiro Alfredo Ballestera, Amparo Alberta, Amparo Moranta a
Vaccine Institute of Valencia (VIVA), Primary Care, Area 4 Generalitat, Valenciana, Spain Hospital Arnau de Vilanova, Valenciana, Spain
b
KEYWORDS Invasive disease; Risk factors; Neisseria meningitidis; Haemophilus influenzae; Streptococcus pneumoniae
Summary Objective. To identify the risk factors related to invasive disease caused by Haemophilus influenzae type b (Hib), Neisseria meningitidis, and Streptococcus pneumoniae. Methods. Case –control study. All hospitals of the region of Valencia in Spain (covering 95% of the population of Valencia). The patients are children aged less than 15 years in whom Hib, N. meningitidis or S. pneumoniae are isolated from normally sterile sites. Results. From 1995 to 1998, 275 cases of invasive disease were analysed, and 243 hospital controls were selected in the month after the onset of the case. The paediatrician completed a survey administered to the relatives at the time of admission. The risk factors related to invasive disease by Hib were exposure to tobacco smoke (number of smokers, adjusted OR (aOR) 1.74, 95% confidence interval (CI) 1.02– 2.96) and living with more than four people (aOR 3.72, 95% CI 1.3 –3.7). For N. meningitidis, there is a dose-response relationship; if more than 60 cigarettes are smoked daily at home, the aOR is 3.61 (95% CI 1.04– 12.57). If there are more than four people living in the household, aOR 1.69 (95% CI 1.01 –2.85). In children under two years of age, having siblings less than 15 years of age (OR 1.76, 95% CI 0.75 –4.17) and attending a day nursery represents a risk for suffering invasive pneumococcal disease (aOR 4.21, 95% CI 1.28– 13.83). Conclusions. Among the variables tested, the modifiable risk factor is smoking; if smoking was reduced at home, the number of cases of invasive disease could be reduced in children, mainly in those under 5 years of age. Identification and vaccination of these risk groups would significantly reduce these diseases. Q 2003 The British Infection Society. Published by Elsevier Ltd. All rights reserved.
Introduction Invasive diseases in children cause a high morbidity and mortality in spite of the use of potent antimicrobial agents. The organisms causing most *Corresponding author. Address: C/Diputacio ´n 12, 46520 Puerto-Sagunto (Valencia), Spain. Tel.: þ34-962675400; fax: þ 34-963301264. E-mail address: [email protected]
cases of invasive disease in children over one month of age in developed countries include Neisseria meningitidis, Haemophilus influenzae type b (Hib), and Streptococcus pneumoniae. The epidemiology and incidence of these diseases varies substantially from one country to another, even between neighbouring countries with a similar socioeconomic level, for unknown reasons. Prospective studies performed in Spain found that the incidence rate of invasive disease by
0163-4453/$30.00 Q 2003 The British Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2003.10.015
Risk factors for invasive disease among children in Spain
N. meningitidis in children under 5 years of age was 22.3 cases per 100 000 children;1 the respective figures for Hib (in non-vaccinated children) and S. pneumoniae were 18.2 2 and 10.5 cases per 100 000.3 At the end of 1997, due to the increased number of cases of disease caused by serogroup C meningococcus in Spain, a universal vaccination campaign was performed with the meningococcal A þ C polysaccharide vaccine which dramatically reduced infection by this organism.1 The Hib vaccine has also been administered routinely since 1998, although the vaccine was filed for registration at the end of 1994, and was sold in retail pharmacies on prescription by paediatricians before 1988. As there are no records of vaccine coverage in Spain between registration of the vaccine and its incorporation into the regular vaccination scheme, the efficacy of the vaccine in this period is unknown. It is important to identify those groups which have a higher risk of developing these diseases, and to act on any modifiable risk factors. This is especially true when no vaccine exists.4 While the relationship between invasive disease and risk factors has been widely studied for Hib, risk factors for N. meningitidis and S. pneumoniae have been studied less. An increased risk has been reported for passive smoking, low socio-economic status, crowding, day nursery and individuals with a history of viral infection.5 – 18 No studies have examined how risk factors are modified after the introduction of vaccines.19 This study was performed in Valencia, a region in the Mediterranean area having climatic and socioeconomic characteristics different from those reported in published studies. The incidence of disease in different regions is variable and this may reflect differences in risk factors in these regions. The objective of this study is, therefore, to identify risk factors for suffering invasive disease caused by N. meningitidis, H. influenzae type b, and Streptococcus pneumoniae in children aged less than 15 years in the region of Valencia.
Material and methods Beginning 1st 1995, data were collected from the coordinated epidemiological surveillance program. An active search was made for cases of invasive diseases (sepsis/bacteraemia, meningitis, pneumonia, epiglottitis, arthritis or cellulitis) caused by Hib, N. meningitidis and S. pneumoniae in the 19 public hospitals of the region of Valencia, which has
321
630 525 inhabitants aged less than 15 years (1996 Census). To study risk factors, a case – control study was designed, with longitudinal selection of the controls in the month following the occurrence of the case (concurrent or contemporary controls). A case was defined as any child aged under 15 years of age in whom Hib, N. meningitidis or S. pneumoniae was isolated from any normally sterile site (blood, CSF, pleural fluid, joint fluid or subcutaneous aspirate). All Hib cases were serotype b, the isolates were sent to a reference laboratory for serotyping and when necessary to the reference H. Influenzae laboratory at the Instituto de Salud Carlos III. Madrid and/or to the Haemophilus Reference Laboratory (Public Health Laboratory Service, Oxford, England) for polymerase chain reaction (PCR) confirmation. In the case of meningococcal disease, when cultures were negative, a case was defined as the presence of a clinical condition compatible with meningococcal disease with hemorrhagic exanthema. Children who had no infectious disease or respiratory condition were used as controls. They were recruited from wards or outpatient clinics from the hospital where the case had been admitted. At least one control was selected for each case of the same age, except in two hospitals where no controls were recruited. All controls were therefore used for the study of each organism. A paediatrician from each hospital was responsible for completing a clinical – epidemiological survey for both cases and controls. This survey was performed by direct questioning of relatives in the first hours following admission. The following information was obtained: age, sex, number of siblings under 15 years, number of people living with the child in the past 15 days, number of bedrooms, smokers living with the child and number of cigarettes smoked at home, attendance at day nursery and current breastfeeding. They were also asked about the presence of chronic diseases: immunodeficiency, pre-maturity, asplenia, chronic renal disease, diabetes mellitus, or chronic pulmonary or heart disease. Finally, information was obtained on vaccine status. The analysis was stratified by age groups. Most variables were recoded into qualitative variables with two categories, except for the number of cigarettes, for which four categories were used. A univariate analysis was performed using SPSS statistical software (version 7), calculating the crude OR. To establish the independence of the different risk factors, a logistic regression was made. All variables which in the univariate analysis were associated with the disease or variables
322
considered as potential confounding variables were introduced in the multivariate analysis using the ENTER method. The dose-response relationship was investigated in variables with more than two categories.
Results During the study period, data were collected from 275 cases of invasive disease. Haemophilus influenzae type b was the organism causing the disease in 31 cases, whereas 63 cases were caused by the pneumococcus, and 181 by the meningococcus, of there 138 were culture proven. The controls included a total of 243 children under 15 years of age. There were 50 cases of injuries or traumatologic surgery, 41 cases of general surgery and the rest miscellaneous conditions: digestive, cardiac, endocrine, allergy, nephrourology, etc. The three groups included cases where data collection was not complete, or data were difficult to interpret, so that some variables are not consistent with the number of children. The cases of Hib disease occurred in the first two years of the study, before the vaccine was universally administered. Therefore, controls are only used for those years. The description of the risk factors studied and the univariate and multivariate analyses are shown in the tables by organism (Tables 1 and 2 H. influenzae type b, Tables 3 – 5 pneumococcus and Tables 6 – 8 meningococcus). Age and sex. The age and sex were similar between the cases and the controls, except for cases of pneumococcal disease. For those with Hib disease, all children were under five years of age, with a median age of 13.4 months; 54.1% were boys. For those with meningococcal disease, the median age was 36.8 months, and 54% were boys. For those with pneumococcal disease, significant age differences were seen (median age of cases: 19.8 months, median age of controls: 36 months); however, in the group aged less than two years, ages were similar between cases and controls. Sixty-one percent were boys. Exposure to smoking at home. Children with invasive disease caused by Hib or meningococcus were more exposed to tobacco smoke than the controls. In cases caused by Hib, parents smoked more and there was a significantly higher number of smokers at home with an adjusted OR of 1.74 (95% CI 1.02 – 2.96). For meningococcal disease, the mean number of smokers was higher for cases than for controls,
I. Pereiro ´ et al.
there was also a higher number of parents who smoke, and when there were other persons also living in the home, they smoke more than those living with controls (Table 7). Strong association was found when those people living with children smoke over 20 cigarettes daily, and persists when controlling by number of people living with them (Table 7). There is also a dose-response effect with the number of cigarettes smoked at home, which is strongest in children aged less than five years. All variables related to smoking are more strongly associated in the group of children aged less than five years (Table 7). After adjusting for other variables, exposure to smoking is shown to be a significant risk factor for invasive disease, mainly in children aged less than five years (Table 8). Number of household members. Living in a household with four or more residents is an independent risk factor for invasive disease caused by Hib (Table 2) or meningococcus in children aged under 5 years (Table 8). Number of siblings. No association was found between the disease and having at least one sibling aged under 15 years of age living in the same household, except for children with pneumococcal disease aged less than one year. In this group of 17 cases and 51 controls, 82.4% have siblings under 15 years of age, vs. 37.3% in the controls (OR 7.51, 95% CI 1.88 – 30.08). Household crowding. As regards crowding (number of people/bedrooms), when the value is above 1.5 there is a significant association with meningococcal disease, for all age groups. A higher OR is found in children aged from 5 to 15 years (Table 7). Attendance at day nursery. This analysis has been performed in children aged less than five and two years. A significant association has only been found in cases with pneumococcal disease aged less than two years (Table 5). Breastfeeding. In this study, breastfeeding does not appear to be a protective factor. In both, cases and controls, the percentages of breastfed children are very low: 26.3% in children under 3 months of age, 20.9% at 6 months, and 13.4% at 12 months. No significant differences have been found between cases and controls by age group or organism. Vaccination. Children previously vaccinated for Hib are protected against this disease with an adjusted OR of 0.16 (Table 2), which represents a vaccination efficacy of 84%. Coverage with the meningococcal polysaccharide vaccine in the third year of the study, after the vaccination campaign, was 75%, there were only 2 cases of meningococcus C disease and had not received the vaccine.1 Other risk factors. The variables related to
Risk factors for invasive disease among children in Spain
323
Table 1 Risk factors for invasive Haemophilus influenzae disease among children under 5 years old in Spain 1995–1997 Cases ðn ¼ 31Þ Siblings under 15 years No 9 Yes 16 Household members ,4 6 $4 17 Crowding , 1.5 14 $ 1.5 9 Maternal smoking No 9 Yes 14 Paternal smoking No 6 Yes 17 Other smoking No 19 Yes 4 No. of cigarettes smoked at home None 6 1–29 7 30–59 8 . 59 2 No. of smokers (mean:SD) 1.48:1.08 Day nursery No 17 Yes 8 Previous invasive disease No 25 Yes 5 Breastfeeding in under 24 months (21 cases, 67 controls) Yes 3 No 16 Others risk factors No 22 Yes 4 Hib vaccination Yes 2 No 29
previous invasive infections or pneumoniae were analysed, and no relationship with the present invasive disease was seen in any case. Similarly, no significant relationship was found between the
Controls ðn ¼ 123Þ
p
OR
66 54
0.08
2.17
64 56
0.02
3.24
95 24
0.05
2.54
69 50
0.09
2.14
59 60
0.04
2.79
101 18
0.78
1.18
48 42 26 3 0.97:0.89
0.63 0.13 0.09
1.33 2.46 5.33
0.02
1.79
76 44
0.66
1.23
119 6
0.08
3.24
4 60
0.20
2.81
94 24
0.56
0.71
50 70
,0.001
0.09
occurrence of an invasive disease and the presence of the chronic diseases analysed.
Discussion Table 2 Logistic regression analysis for risk factors of invasive Haemophilus influenzae disease among children under 5 years old OR crude OR adjusted 95% CI No. of smokers at home 1.79 Household members ,4 3.24 $4 Hib vaccination Yes 0.09 No
1.74
1.02–2.96
3.72
1.30–10.73
0.16
0.03–0.76
The results of this study suggest that exposure to tobacco smoke is a significant risk factor for invasive disease by meningococcus or Hib. Living with more than four people in the same household also increases the risk of suffering these diseases, while having siblings aged less than 15 years and going to a day nursery increases the risk of pneumococcal disease. These associations persist after adjusting for the confounding effect of other known risk factors. This is the first study of invasive disease caused
324
I. Pereiro ´ et al.
Table 3 Risk factors for invasive pneumococcal disease among children under 15 years old in Spain 1995–1998 Variable
Cases ðn ¼ 63Þ
Siblings under 15 years No 27 Yes 28 Household members ,4 23 $4 32 Crowding , 1.5 45 $ 1.5 10 Maternal smoking No 44 Yes 19 Paternal smoking No 42 Yes 21 Other smoking No 55 Yes 8 No. of cigarettes smoked at home None 33 0– 29 20 30– 59 8 . 59 2 No. of smokers (mean:SD) 0.84:1.25 Day nursery No 29 Yes 35 Previous invasive disease No 64 Yes 2 Breastfeeding in under 12 months (20 cases, 51 controls) Yes 3 No 17 Other risk factors No 51 Yes 11
by Hib, pneumococcus and meningococcus and their risk factors, carried out in a Mediterranean region where the social and climatic characteristics differ from those of countries where similar studies have been done. All controls were hospital patients who had neither the disease under study nor related diseases, which makes a selection bias unlikely.20 With regard to passive smoking, several studies show a close relationship between this and the development of respiratory diseases, particularly in childhood.21,22 A recent review has shown a causal relationship between smoking parents, particularly the mother, and acute respiratory disease in children, mainly in the first years of life, with a clear dose-response relationship. High nicotine levels have also been found in children with persistent bronchial infections and a smoking mother.23 In the case of Hib disease, the mechanism by
Controls ðn ¼ 243Þ
p
OR
109 130
0.64
0.87
101 136
0.91
1.03
186 47
0.002
0.82
148 87
0.31
0.73
128 107
0.08
0.60
200 35
0.66
0.83
102 78 47 8 0.89:0.88
0.47 0.14 0.75
0.79 0.53 0.77
0.73
0.95
137 102
0.09
1.62
237 13
0.45
0.56
5 46
0.53
2.70
198 39
0.81
1.10
which passive exposure to tobacco smoke increases susceptibility to the disease is unknown. Takala14 suggests that excessive exposure to tobacco smoke can damage the respiratory mucosa, increasing the potential for invasion by Hib and the resulting risk of disease. Some studies have provided further evidence for a relationship between smoking and Hib disease,10,15,16 however, this relationship may not persist after adjusting for other variables. In other studies,17,18 the adjusted OR is up to six in children living with at least two smokers in the same household, while other authors have found no relationship between passive smoking and disease caused by Hib.19,24 We provide support for the hypothesis that smoking is a risk factor, and we have found that both the father and the mother of cases smoke more than those of controls, with an adjusted OR of 1.74 for each smoker in the household. As regards invasive meningococcal disease,
Risk factors for invasive disease among children in Spain
325
Table 4 Risk factors for invasive pneumococcal disease by age. Univariate analysis All ages* OR Siblings under 15 years No 0.87 Yes Household members ,4 1.03 $4 Crowding , 1.5 0.82 $ 1.5 Maternal smoking No 0.73 Yes Paternal smoking No 0.60 Yes Other smoking No 0.83 Yes No. of cigarettes smoked at home None 0.79 1–29 0.53 30–59 0.77 . 59 No. of smokers (mean:SD) 0.95 Day nursery No 1.62 Yes Other risk factors No 1.10 Yes
,2 y**
2 –15 y***
95% CI
OR
95% CI
OR
95% CI
0.48–1.56
1.41
0.63–3.16
0.65
0.26–1.60
0.57–1.87
1.51
0.67–3.39
1.08
0.41–2.83
0.43–1.58
0.78
0.32–1.90
0.87
0.32–2.38
0.40–1.34
0.68
0.29–1.59
0.95
0.39–2.31
0.33–1.07
0.54
0.24–1.22
0.61
0.25–1.52
0.36–1.89
0.77
0.20–2.98
1.18
0.40–3.44
0.42–1.48 0.23–1.24 0.15–3.82
0.68 0.44 1.02
0.26–1.77 0.15–1.34 0.88–11.94
1.20 0.73 0.98
0.47–3.05 0.18–2.91 0.10–9.06
0.70–1.27
0.75
0.47–1.18
1.24
0.86–1.80
0.93–2.82
2.73
0.96–7.78
2.02
0.72–5.67
0.52–2.29
1.35
0.52–3.52
0.74
0.20–2.68
*Cases, 63; controls, 243. **Cases, 35; controls, 93. ***Cases, 28; controls, 150. In four cases age was missed.
case – control studies have show that being a passive smoker is associated with carriage,25 – 27 which has led to the hypothesis that children growing up in smoking environments were more susceptible to developing meningococcal disease in later years. Our data demonstrate that passive smoking increases the risk of suffering meningococcal disease, mainly in children aged less than 5 years, with a possible dose-response relationship. The role played by other smoking members of the family, besides parents, should be noted. A possible
Table 5 Logistic regression analysis for risk factors of invasive pneumococcal disease among children under 2 years old OR crude OR adjusted 95% CI Day nursery (yes/no) 2.73 Siblings under 15 years 1.41 (yes/no) Age (in months) 0.99
4.21 1.76
1.28 –13.83 0.75 –4.17
0.97
0.97 –1.05
explanation could be that caregivers, mostly grandparents, are those spending more time at home with the children, and it is in the group under five years of age where contact with these people is closer and more continued. Other European studies also relate exposure to tobacco smoke with a risk of invasive meningococcal disease in children,6,7,28,29 and report very high associations (OR 7.5), with a dose-response, independent of social status.6 In the United States, it was found that having a smoking mother was the independent risk factor most strongly associated with invasive meningococcal disease (OR 3.8) in children aged less than 18 years.8 A recent study conducted in New Zealand30 found a weaker association, in that the presence of at least one smoker in the household represented an OR of 1.4 (95% CI 1.0 – 1.8) in children aged less than 8 years. Various authors state that a possible explanation for these findings is that tobacco smoke reduces the defences of the nasopharyngeal mucosa against meningococcal invasion.6 – 8,26
326
I. Pereiro ´ et al.
Table 6 Risk factors for meningococcal disease among children under 15 years old in Spain 1995–1998 Cases ðn ¼ 181Þ Siblings under 15 years No 75 Yes 103 Household members ,4 53 $4 122 Crowding , 1.5 121 $ 1.5 49 Maternal smoking No 96 Yes 78 Paternal smoking No 82 Yes 91 Other smoking No 135 Yes 40 No. of cigarettes smoked at home None 53 1– 29 68 30– 59 43 . 59 9 No. of smokers (mean:SD) 1.32:1.25 Day nursery No 77 Yes 107 Breastfeeding in under 12 months (36 cases 51 controls) Yes 5 No 31 Previous meningitis No 195 Yes 1 Other risk factors No 170 Yes 22 Meningococcal polysaccharide vaccine Yes 23 No 166
For cases of a pneumococcal disease, we found no association with exposure to tobacco smoke, in agreement with other authors.21 – 33 As regards the number of people living in the same household, there is an association which persists after adjusting for other variables in children aged less than 5 years both for diseases caused by the meningococcus and by Hib. For meningococcal disease, similar results are reported in another study,6 but for more than six people living in the same household. In this study, crowding, defined as more than 1.5 persons per room, showed a positive association in children under 5 years of age. In the Fischer study, living in a house with three or more residents aged less than 18 years was an independent risk factor.10 Crowding is the most important risk factor seen by another
Controls ðn ¼ 243Þ
P
OR
109 130
0.48
1.15
101 136
0.01
1.71
186 47
0.04
1.60
148 87
0.11
1.38
128 107
0.16
1.33
200 35
0.04
1.69
102 78 47 8 0.89:0.88
0.07 0.03 0.04 0.13 0.0001
1.00 1.68 1.76 2.16 1.50
137 102
0.86
0.97
5 46
0.42
1.1
237 1
0.89
1.21
198 39
0.14
1.52
31 207
0.79
1.1
author,30 with an OR of 10.7. Studies on Hib cases found similar results.15,16 In our study, it is more important to live with siblings regardless of the number of rooms in the house since crowding is not associated with disease when other factors are included in the multivariate analysis. Having older school-aged siblings has been identified in several studies as a risk factor for disease caused by Hib15,18 and pneumococcus.31 In our study, this is a risk factor for cases caused by pneumococcus in children aged less than one year. Attendance at day nursery has been considered a significant risk factor for suffering Hib disease, mainly in younger children.18 A dose-response effect has also been reported for this risk depending on the number of hours spent in the day nursery,11 – 13 and is higher in the first month of
Risk factors for invasive disease among children in Spain
327
Table 7 Risk factors for meningococcal disease by age. Univariate analysis All ages* OR Crowding (,1.5/$1.5) 1.60 Household members (,4/$4) 1.71 Siblings under 15 years (yes/no) 1.15 No. of smokers (mean:SD) 1.50 Maternal smoking (no/yes) 1.38 Paternal smoking (no/yes) 1.33 Other smoking (no/yes) 1.69 No of cigarettes smoked by other at home None , 10 0.97 10–20 2.29 . 20 11.82 No. of cigarettes smoked at home None 1–29 1.68 30–59 1.76 . 59 2.16 Other risk factors (no/yes) 1.52
,5 y**
5–15 y***
95% CI
OR
95% CI
OR
95% CI
1.01–2.54 1.13–2.58 0.78–1.70 1.22–1.84 0.93–2.06 0.89–1.97 1.02–2.80
1.40 1.73 1.07 1.64 1.50 1.55 1.99
0.78–2.53 1.05–2.83 0.66–1.73 1.21–2.12 0.92–2.46 0.95–2.55 1.07–3.68
1.94 1.58 1.23 1.32 1.24 1.08 1.24
0.92–4.10 0.68–3.62 0.60–2.53 0.91–1.92 0.62–2.48 0.56–2.14 0.52–2.99
0.49–1.92 1.04–5.04 1.46–95.38
1.02 2.57 –
0.44–2.36 0.96–6.89 –
0.80 0.71 1.50
0.05–13.14 0.03–14.30 0.07–31.6
1.05–2.67 1.03–2.99 0.79–5.93 0.87–2.67
2.07 2.16 3.91 1.60
1.13–3.77 1.11–4.19 1.17–13.03 0.80–3.20
1.28 1.38 0.46 1.33
0.60–2.70 0.55–3.47 0.45–4.69 0.50–3.53
*Cases, 181; controls, 243.**Cases, 119; controls, 164.***Cases, 57; controls, 79. In five cases age was missed.
schooling.14 However, many studies, including ours, do not find this relationship,17,24 and even find the opposite, that this has a protective effect.19 In day nurseries, children are exposed to early contact with diverse of microbial flora, and inflammation of the respiratory mucosa is frequent; therefore, there is an increased possibility of contact with Hib, followed by invasive disease. Different findings in several studies suggest that, in the absence of bias, the conditions in day nurseries may change the incidence of disease. Thus, in Spain children spend much of the day in day nursery gardens, so that the number of hours in classrooms is lower than in countries with colder or more rainy climates. Similarly, the mild climate allows that young children not going to school to spend many hours outdoor, and houses are ventilated daily, and therefore, the exposure of children to tobacco
smoke, and the contact with potential carriers in closed places may be lower. This may explain why the association found is lower than that previously described. Socioeconomic variables were not been controlled in our study; however, it should be noted that some variables associated with low socioeconomic status in other countries such as smoking and day nursery attendance are equally distributed among socioeconomic classes as stated by the National Health Survey in our setting.34 Only crowding is correlated with a low socioeconomic level in this survey, but this variable is systematically excluded from our multivariate analysis. The efficacy of the Hib vaccine is similar to that reported in other developed countries. It must be noted that this vaccine was administered to a very high percentage of children (up to 60% of children aged less than 5 years in the second year of the
Table 8 Logistic regression analysis for risk factors of meningococcal disease among children under 5 years old
No. of cigarettes smoked at home None 10–29 30–59 . 59 Household members (,4 $ 4) Meningococcal polysaccharide vaccine Yes No
OR crude
OR adjusted
95% CI
2.07 2.16 3.91 1.73
2.35 2.18 3.61 1.69
1.26–4.37 1.01–4.34 1.04–12.57 1.01–2.85
0.92
0.69
0.24–2.01
328
study), despite the fact that it was not included in the vaccination scheme. Since the coverage of the vaccine was not known in the first years, the only way to establish its efficacy are case – control studies such as this one. The high Hib vaccine coverage reached significantly reduced the incidence of the disease even in the unvaccinated2 by decreasing nasopharingeal carriage. This might affect the results of our study as there was a lower Hib disease risk during this period that could hide other risk factors. As the vaccine had to be partially paid by parents it is possible that the unvaccinated were from a different economic status that the vaccines, however, the most deprived population reached the vaccine for free. This could possible bias the study, but a previous study revealed that the vaccine coverage was most influenced by the attitude of the paediatrician towards vaccination.35 Based on the results obtained in our study, also supported by previous investigations, it appears that the incidence of invasive disease caused by meningococcus or Hib would be decreased if smoking at home was reduced, mainly in children aged less than five years. Additionally, identification and vaccination of these risk groups would probably significantly reduce these diseases.
I. Pereiro ´ et al.
9.
10.
11.
12.
13.
14.
15.
16.
17.
References 18. 1. Pereiro ´ I, Dı´ez-Domingo J, Morant A, Gimeno C, Lerma M, San-Martı´n M, Gonza ´lez A. Epidemiological pattern of meningococcal disease in Valencia, Spain. Impact of a mass immunization campaign with meningococcal C polysaccharide vaccine. Scand J Infect Dis 2001;33:581—584. 2. Dı´ez-Domingo J, Pereiro ´ I, Morant A, Gimeno C, San-Martı´n M, Gonza ´lez A. Impact of non-routine vaccination on the incidence of invasive Haemophilus influenzae type b (Hib) disease: experience in the autonomous region of Valencia, Spain. J Infect 2001. 3. Diez-Domingo J, Pereiro ´ I, Morant A, Gimeno C, Lerma M, Oyagu ´lez A. Epidemiology of invasive Strepto¨ez I, Gonza coccus pneumoniae infections in children in Spain, 1996— 1998. J Infect 2002;45:139—143. 4. Wilson N, Baker M, Martin D, Lennon D, O’Hallahan J, Jones N, et al. Meningococcal disease epidemiology and control in New Zealand. N Z Med J 1995;108:437—442. 5. Jones D. Epidemiology of meningococcal disease in Europa and the USA. In: Cartwright K, editor. Meningococcal disease. Chichester: Wiley; 1995. p. 147—157. 6. Stanwell-Smith RE, Stuart JM, Hugues AO, Robinson P, Griffin MB, Cartwright K. Smoking, the environment and meningococcal disease: a case control study. Epidemiol Infect 1994; 112:315—328. 7. Haneberg B, Toujum T, Rodabil K. Factors preceding the onset of meningococcal disease with special emphasis on passive smoking, stressful events physical fitness and general symptoms of sill health. NIPH Ann 1983;6:169—174. 8. Fischer M, Hedberg K, Cardosi P, Plikaytis B, Hoesly F,
19.
20. 21.
22.
23.
24.
25. 26.
Steingart K, et al. Tobacco smoke as a risk factor for meningococcal disease. Pediatr Infect Dis J 1997;16: 979—983. Strachan DP, Cook DG. Health effects of passive smoking. Parenteral smoking and lower respiratory illness in infancy and early childhood. Thorax 1997;52:905—914. Cochi SL, Fleming DW, Hightower AW, Limpakarnjanarat K, Facklam RR, Smith JD, et al. Primary invasive Haemophilus influenzae type b disease: a population-based assessment of risk factors. J Pediatr 1986;108:887—896. Fogarty J, Moloney AC, Newell JB. The epidemiology of Haemophilus influenzae type b disease in the Republic of Ireland. Epidemiol Infect 1995;114:451—463. Istre G, Coner J, Broome C, Hightower A, Hopkins R. Risk factors for primary invasive Haemophilus influenzae disease: increased risk from day care attendance and schoolaged household members. J Pediatr 1985;106:190—195. Redmond SR, Pichichero ME. Haemophilus influenzae type B disease. An epidemiologic study with special reference to day-care centers. JAMA 1984;252:2581—2584. Takala AK, Eskola J, Palmgren J, Ro ¨nnberg PR, Kela E, Rekola P, et al. Risk factors of invasive Haemophilus influenzae type b disease among children in Finland. J Pediatr 1989;115: 694—701. Clements DA, Weigle KA, Gilbert GL. A case—control study risk factors for invasive Haemophilus influenzae type b disease in Victoria, Australia 1988—90. J Paediatr Child Health 1995;31:513—518. Takala AK, Clements DA. Socioeconomic risk factors for invasive Haemophilus influenzae type b disease. J Infect Dis 1992;165(Suppl 1):S11—S15. Vadheim CM, Greenberg DP, Bordenave N, Ziont L, Christenson P, Waterman SH, et al. Risk factors for invasive Haemophilus influenzae type b in Los Angeles county children 18—60 months of age. Am J Epidemiol 1992;136: 221—235. Arnold C, Makintube S, Istre G. Day care attendance and other risk factors for invasive Haemophilus influenzae type b disease. Am J Epidemiol 1993;138:333—340. Mu ¨hlemann K, Alexander ER, Weiss NS, Pepe M, Schopfer K, The Swiss H. Influenzae Study Group. Risk factors for invasive Haemophilus influenzae disease among children 2—16 years of age in the vaccine era, Switzerland 1991—1993. Int J Epidemiol 1996;25:1280—1285. Lilienfeld A, Lilienfeld D. Fundamentos de epidemiologı´a, Me ´xico. Fondo educativo interamericano 1983;177—178. Holdberg C, Wright A, Martinez F, Morgan W, Taussing L, Group Health Medical Associates. Child day Care, smoking by caregivers, and lower tract illness in the first 3 years of life. Pediatrics 1993;91:885—892. Pedreira F, Guandolo V, Feroli E, Mella G, Weiss I. Involuntary smoking and incidence of respiratory illness during the first year of life. Pediatrics 1985;75:594—597. Cook DG, Whincup PH, Jarvis MJ, Strachan DP, Papacosta O, Bryant A. Passive exposure to tobacco smoke in children aged 5—7 years: individual, family, and community factors. Br Med J 1994;308:384—389. Silfverdal SA, Bodin L, Hugosson S, Garpenholt O, Werner B, Esbjo ¨rner E, et al. Protective effect of breastfeeding on invasive Haemophilus influenzae infection: a case—control study in Swedish preschool children. Int J Epidemiol 1997; 26:443—450. Stuart J, Cartwright K, Robinson P, Noah N. Effect of smoking on meningococcal carriage. Lancet 1989;23:723—725. Davies AL, O’Flanagan D, Salmon RL, Coleman TJ. Risk factors for Neisseria meningitidis carriage in a school during
Risk factors for invasive disease among children in Spain
27.
28.
29.
30.
31.
a community outbreak of meningococcal infection. Epidemiol Infect 1996;117:259—266. Imrey P, Jackson L, Ludwinski P, England III A, Fella G, Fox B, et al. Meningococcal carriage, alcohol consumption, and campus bar patronage in a serogroup C meningococcal disease outbreak. J Clin Microbiol 1995;33:3133—3137. Stuart J, Cartwright K, Dawson A, Rickard J, Noah N. Risk factors for meningococcal disease: a case control study in south west England. Comm Med 1988;10:139—145. Kriz P, Bobak M, Kriz B. Parenteral smoking, socioeconomic factors, and risk of invasive meningococcal disease in children: a population based case—control study. Arch Dis Child 2000;83:117—121. Baker M, Mcnicholas A, Garret N, Jones N, Stewart J, Koberstein V, et al. Household crowding a major risk factor for epidemic meningococcal disease in Auckland children. Pediatr Infect Dis J 2000;19:983—990. Takala A, Jero J, Kela E, Ro ¨nnberg P, Koskenniemi E, Eskola J. Risk factors for primary invasive pneumococcal disease among children in Finland. JAMA 1995;273:859—864.
329
32. Gessner B, Ussery X, Parkinson A, Breiman R. Risk factors for invasive disease caused by Streptococcus pneumoniae among Alaska native children younger than two years of age. Pediatr Infect Dis J 1995;14:123—128. 33. Levine O, Churchill J, Farley M, Lefkowitz L, Harrison L, Mcgeer A, et al. Risk factors for invasive pneumococcal disease in children: a population-based case—control study. First International Symposium on Pneumococci and Pneumococcal diseases. Helsingor, 1998. 34. Consellerı´a de Sanitat i Consum, Encuesta de salud de la Comunidad Valenciana. Valenciana: Generalitat Valenciana; 1993. 35. Diez-Domingo J, Pereiro ´ I, Ferrer A, Fullana A, Momparler P, Ballester A, Morant A, Saurı´ V. Cobertura de las vacunas administradas sistema ´ticamente y de la vacuna frente a Haemophilus influenzae tipo b previa a su incorporacio ´n al calendario vacunal en la Comunidad Valenciana. An Esp Pediatr 1998;49:568—570.
www.elsevierhealth.com/journals/jinf
Risk factors for invasive disease among children in Spain ´a,*, Javier Dı´ez-domingoa, Leopoldo Segarrab, Inmaculada Pereiro Alfredo Ballestera, Amparo Alberta, Amparo Moranta a
Vaccine Institute of Valencia (VIVA), Primary Care, Area 4 Generalitat, Valenciana, Spain Hospital Arnau de Vilanova, Valenciana, Spain
b
KEYWORDS Invasive disease; Risk factors; Neisseria meningitidis; Haemophilus influenzae; Streptococcus pneumoniae
Summary Objective. To identify the risk factors related to invasive disease caused by Haemophilus influenzae type b (Hib), Neisseria meningitidis, and Streptococcus pneumoniae. Methods. Case –control study. All hospitals of the region of Valencia in Spain (covering 95% of the population of Valencia). The patients are children aged less than 15 years in whom Hib, N. meningitidis or S. pneumoniae are isolated from normally sterile sites. Results. From 1995 to 1998, 275 cases of invasive disease were analysed, and 243 hospital controls were selected in the month after the onset of the case. The paediatrician completed a survey administered to the relatives at the time of admission. The risk factors related to invasive disease by Hib were exposure to tobacco smoke (number of smokers, adjusted OR (aOR) 1.74, 95% confidence interval (CI) 1.02– 2.96) and living with more than four people (aOR 3.72, 95% CI 1.3 –3.7). For N. meningitidis, there is a dose-response relationship; if more than 60 cigarettes are smoked daily at home, the aOR is 3.61 (95% CI 1.04– 12.57). If there are more than four people living in the household, aOR 1.69 (95% CI 1.01 –2.85). In children under two years of age, having siblings less than 15 years of age (OR 1.76, 95% CI 0.75 –4.17) and attending a day nursery represents a risk for suffering invasive pneumococcal disease (aOR 4.21, 95% CI 1.28– 13.83). Conclusions. Among the variables tested, the modifiable risk factor is smoking; if smoking was reduced at home, the number of cases of invasive disease could be reduced in children, mainly in those under 5 years of age. Identification and vaccination of these risk groups would significantly reduce these diseases. Q 2003 The British Infection Society. Published by Elsevier Ltd. All rights reserved.
Introduction Invasive diseases in children cause a high morbidity and mortality in spite of the use of potent antimicrobial agents. The organisms causing most *Corresponding author. Address: C/Diputacio ´n 12, 46520 Puerto-Sagunto (Valencia), Spain. Tel.: þ34-962675400; fax: þ 34-963301264. E-mail address: [email protected]
cases of invasive disease in children over one month of age in developed countries include Neisseria meningitidis, Haemophilus influenzae type b (Hib), and Streptococcus pneumoniae. The epidemiology and incidence of these diseases varies substantially from one country to another, even between neighbouring countries with a similar socioeconomic level, for unknown reasons. Prospective studies performed in Spain found that the incidence rate of invasive disease by
0163-4453/$30.00 Q 2003 The British Infection Society. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.jinf.2003.10.015
Risk factors for invasive disease among children in Spain
N. meningitidis in children under 5 years of age was 22.3 cases per 100 000 children;1 the respective figures for Hib (in non-vaccinated children) and S. pneumoniae were 18.2 2 and 10.5 cases per 100 000.3 At the end of 1997, due to the increased number of cases of disease caused by serogroup C meningococcus in Spain, a universal vaccination campaign was performed with the meningococcal A þ C polysaccharide vaccine which dramatically reduced infection by this organism.1 The Hib vaccine has also been administered routinely since 1998, although the vaccine was filed for registration at the end of 1994, and was sold in retail pharmacies on prescription by paediatricians before 1988. As there are no records of vaccine coverage in Spain between registration of the vaccine and its incorporation into the regular vaccination scheme, the efficacy of the vaccine in this period is unknown. It is important to identify those groups which have a higher risk of developing these diseases, and to act on any modifiable risk factors. This is especially true when no vaccine exists.4 While the relationship between invasive disease and risk factors has been widely studied for Hib, risk factors for N. meningitidis and S. pneumoniae have been studied less. An increased risk has been reported for passive smoking, low socio-economic status, crowding, day nursery and individuals with a history of viral infection.5 – 18 No studies have examined how risk factors are modified after the introduction of vaccines.19 This study was performed in Valencia, a region in the Mediterranean area having climatic and socioeconomic characteristics different from those reported in published studies. The incidence of disease in different regions is variable and this may reflect differences in risk factors in these regions. The objective of this study is, therefore, to identify risk factors for suffering invasive disease caused by N. meningitidis, H. influenzae type b, and Streptococcus pneumoniae in children aged less than 15 years in the region of Valencia.
Material and methods Beginning 1st 1995, data were collected from the coordinated epidemiological surveillance program. An active search was made for cases of invasive diseases (sepsis/bacteraemia, meningitis, pneumonia, epiglottitis, arthritis or cellulitis) caused by Hib, N. meningitidis and S. pneumoniae in the 19 public hospitals of the region of Valencia, which has
321
630 525 inhabitants aged less than 15 years (1996 Census). To study risk factors, a case – control study was designed, with longitudinal selection of the controls in the month following the occurrence of the case (concurrent or contemporary controls). A case was defined as any child aged under 15 years of age in whom Hib, N. meningitidis or S. pneumoniae was isolated from any normally sterile site (blood, CSF, pleural fluid, joint fluid or subcutaneous aspirate). All Hib cases were serotype b, the isolates were sent to a reference laboratory for serotyping and when necessary to the reference H. Influenzae laboratory at the Instituto de Salud Carlos III. Madrid and/or to the Haemophilus Reference Laboratory (Public Health Laboratory Service, Oxford, England) for polymerase chain reaction (PCR) confirmation. In the case of meningococcal disease, when cultures were negative, a case was defined as the presence of a clinical condition compatible with meningococcal disease with hemorrhagic exanthema. Children who had no infectious disease or respiratory condition were used as controls. They were recruited from wards or outpatient clinics from the hospital where the case had been admitted. At least one control was selected for each case of the same age, except in two hospitals where no controls were recruited. All controls were therefore used for the study of each organism. A paediatrician from each hospital was responsible for completing a clinical – epidemiological survey for both cases and controls. This survey was performed by direct questioning of relatives in the first hours following admission. The following information was obtained: age, sex, number of siblings under 15 years, number of people living with the child in the past 15 days, number of bedrooms, smokers living with the child and number of cigarettes smoked at home, attendance at day nursery and current breastfeeding. They were also asked about the presence of chronic diseases: immunodeficiency, pre-maturity, asplenia, chronic renal disease, diabetes mellitus, or chronic pulmonary or heart disease. Finally, information was obtained on vaccine status. The analysis was stratified by age groups. Most variables were recoded into qualitative variables with two categories, except for the number of cigarettes, for which four categories were used. A univariate analysis was performed using SPSS statistical software (version 7), calculating the crude OR. To establish the independence of the different risk factors, a logistic regression was made. All variables which in the univariate analysis were associated with the disease or variables
322
considered as potential confounding variables were introduced in the multivariate analysis using the ENTER method. The dose-response relationship was investigated in variables with more than two categories.
Results During the study period, data were collected from 275 cases of invasive disease. Haemophilus influenzae type b was the organism causing the disease in 31 cases, whereas 63 cases were caused by the pneumococcus, and 181 by the meningococcus, of there 138 were culture proven. The controls included a total of 243 children under 15 years of age. There were 50 cases of injuries or traumatologic surgery, 41 cases of general surgery and the rest miscellaneous conditions: digestive, cardiac, endocrine, allergy, nephrourology, etc. The three groups included cases where data collection was not complete, or data were difficult to interpret, so that some variables are not consistent with the number of children. The cases of Hib disease occurred in the first two years of the study, before the vaccine was universally administered. Therefore, controls are only used for those years. The description of the risk factors studied and the univariate and multivariate analyses are shown in the tables by organism (Tables 1 and 2 H. influenzae type b, Tables 3 – 5 pneumococcus and Tables 6 – 8 meningococcus). Age and sex. The age and sex were similar between the cases and the controls, except for cases of pneumococcal disease. For those with Hib disease, all children were under five years of age, with a median age of 13.4 months; 54.1% were boys. For those with meningococcal disease, the median age was 36.8 months, and 54% were boys. For those with pneumococcal disease, significant age differences were seen (median age of cases: 19.8 months, median age of controls: 36 months); however, in the group aged less than two years, ages were similar between cases and controls. Sixty-one percent were boys. Exposure to smoking at home. Children with invasive disease caused by Hib or meningococcus were more exposed to tobacco smoke than the controls. In cases caused by Hib, parents smoked more and there was a significantly higher number of smokers at home with an adjusted OR of 1.74 (95% CI 1.02 – 2.96). For meningococcal disease, the mean number of smokers was higher for cases than for controls,
I. Pereiro ´ et al.
there was also a higher number of parents who smoke, and when there were other persons also living in the home, they smoke more than those living with controls (Table 7). Strong association was found when those people living with children smoke over 20 cigarettes daily, and persists when controlling by number of people living with them (Table 7). There is also a dose-response effect with the number of cigarettes smoked at home, which is strongest in children aged less than five years. All variables related to smoking are more strongly associated in the group of children aged less than five years (Table 7). After adjusting for other variables, exposure to smoking is shown to be a significant risk factor for invasive disease, mainly in children aged less than five years (Table 8). Number of household members. Living in a household with four or more residents is an independent risk factor for invasive disease caused by Hib (Table 2) or meningococcus in children aged under 5 years (Table 8). Number of siblings. No association was found between the disease and having at least one sibling aged under 15 years of age living in the same household, except for children with pneumococcal disease aged less than one year. In this group of 17 cases and 51 controls, 82.4% have siblings under 15 years of age, vs. 37.3% in the controls (OR 7.51, 95% CI 1.88 – 30.08). Household crowding. As regards crowding (number of people/bedrooms), when the value is above 1.5 there is a significant association with meningococcal disease, for all age groups. A higher OR is found in children aged from 5 to 15 years (Table 7). Attendance at day nursery. This analysis has been performed in children aged less than five and two years. A significant association has only been found in cases with pneumococcal disease aged less than two years (Table 5). Breastfeeding. In this study, breastfeeding does not appear to be a protective factor. In both, cases and controls, the percentages of breastfed children are very low: 26.3% in children under 3 months of age, 20.9% at 6 months, and 13.4% at 12 months. No significant differences have been found between cases and controls by age group or organism. Vaccination. Children previously vaccinated for Hib are protected against this disease with an adjusted OR of 0.16 (Table 2), which represents a vaccination efficacy of 84%. Coverage with the meningococcal polysaccharide vaccine in the third year of the study, after the vaccination campaign, was 75%, there were only 2 cases of meningococcus C disease and had not received the vaccine.1 Other risk factors. The variables related to
Risk factors for invasive disease among children in Spain
323
Table 1 Risk factors for invasive Haemophilus influenzae disease among children under 5 years old in Spain 1995–1997 Cases ðn ¼ 31Þ Siblings under 15 years No 9 Yes 16 Household members ,4 6 $4 17 Crowding , 1.5 14 $ 1.5 9 Maternal smoking No 9 Yes 14 Paternal smoking No 6 Yes 17 Other smoking No 19 Yes 4 No. of cigarettes smoked at home None 6 1–29 7 30–59 8 . 59 2 No. of smokers (mean:SD) 1.48:1.08 Day nursery No 17 Yes 8 Previous invasive disease No 25 Yes 5 Breastfeeding in under 24 months (21 cases, 67 controls) Yes 3 No 16 Others risk factors No 22 Yes 4 Hib vaccination Yes 2 No 29
previous invasive infections or pneumoniae were analysed, and no relationship with the present invasive disease was seen in any case. Similarly, no significant relationship was found between the
Controls ðn ¼ 123Þ
p
OR
66 54
0.08
2.17
64 56
0.02
3.24
95 24
0.05
2.54
69 50
0.09
2.14
59 60
0.04
2.79
101 18
0.78
1.18
48 42 26 3 0.97:0.89
0.63 0.13 0.09
1.33 2.46 5.33
0.02
1.79
76 44
0.66
1.23
119 6
0.08
3.24
4 60
0.20
2.81
94 24
0.56
0.71
50 70
,0.001
0.09
occurrence of an invasive disease and the presence of the chronic diseases analysed.
Discussion Table 2 Logistic regression analysis for risk factors of invasive Haemophilus influenzae disease among children under 5 years old OR crude OR adjusted 95% CI No. of smokers at home 1.79 Household members ,4 3.24 $4 Hib vaccination Yes 0.09 No
1.74
1.02–2.96
3.72
1.30–10.73
0.16
0.03–0.76
The results of this study suggest that exposure to tobacco smoke is a significant risk factor for invasive disease by meningococcus or Hib. Living with more than four people in the same household also increases the risk of suffering these diseases, while having siblings aged less than 15 years and going to a day nursery increases the risk of pneumococcal disease. These associations persist after adjusting for the confounding effect of other known risk factors. This is the first study of invasive disease caused
324
I. Pereiro ´ et al.
Table 3 Risk factors for invasive pneumococcal disease among children under 15 years old in Spain 1995–1998 Variable
Cases ðn ¼ 63Þ
Siblings under 15 years No 27 Yes 28 Household members ,4 23 $4 32 Crowding , 1.5 45 $ 1.5 10 Maternal smoking No 44 Yes 19 Paternal smoking No 42 Yes 21 Other smoking No 55 Yes 8 No. of cigarettes smoked at home None 33 0– 29 20 30– 59 8 . 59 2 No. of smokers (mean:SD) 0.84:1.25 Day nursery No 29 Yes 35 Previous invasive disease No 64 Yes 2 Breastfeeding in under 12 months (20 cases, 51 controls) Yes 3 No 17 Other risk factors No 51 Yes 11
by Hib, pneumococcus and meningococcus and their risk factors, carried out in a Mediterranean region where the social and climatic characteristics differ from those of countries where similar studies have been done. All controls were hospital patients who had neither the disease under study nor related diseases, which makes a selection bias unlikely.20 With regard to passive smoking, several studies show a close relationship between this and the development of respiratory diseases, particularly in childhood.21,22 A recent review has shown a causal relationship between smoking parents, particularly the mother, and acute respiratory disease in children, mainly in the first years of life, with a clear dose-response relationship. High nicotine levels have also been found in children with persistent bronchial infections and a smoking mother.23 In the case of Hib disease, the mechanism by
Controls ðn ¼ 243Þ
p
OR
109 130
0.64
0.87
101 136
0.91
1.03
186 47
0.002
0.82
148 87
0.31
0.73
128 107
0.08
0.60
200 35
0.66
0.83
102 78 47 8 0.89:0.88
0.47 0.14 0.75
0.79 0.53 0.77
0.73
0.95
137 102
0.09
1.62
237 13
0.45
0.56
5 46
0.53
2.70
198 39
0.81
1.10
which passive exposure to tobacco smoke increases susceptibility to the disease is unknown. Takala14 suggests that excessive exposure to tobacco smoke can damage the respiratory mucosa, increasing the potential for invasion by Hib and the resulting risk of disease. Some studies have provided further evidence for a relationship between smoking and Hib disease,10,15,16 however, this relationship may not persist after adjusting for other variables. In other studies,17,18 the adjusted OR is up to six in children living with at least two smokers in the same household, while other authors have found no relationship between passive smoking and disease caused by Hib.19,24 We provide support for the hypothesis that smoking is a risk factor, and we have found that both the father and the mother of cases smoke more than those of controls, with an adjusted OR of 1.74 for each smoker in the household. As regards invasive meningococcal disease,
Risk factors for invasive disease among children in Spain
325
Table 4 Risk factors for invasive pneumococcal disease by age. Univariate analysis All ages* OR Siblings under 15 years No 0.87 Yes Household members ,4 1.03 $4 Crowding , 1.5 0.82 $ 1.5 Maternal smoking No 0.73 Yes Paternal smoking No 0.60 Yes Other smoking No 0.83 Yes No. of cigarettes smoked at home None 0.79 1–29 0.53 30–59 0.77 . 59 No. of smokers (mean:SD) 0.95 Day nursery No 1.62 Yes Other risk factors No 1.10 Yes
,2 y**
2 –15 y***
95% CI
OR
95% CI
OR
95% CI
0.48–1.56
1.41
0.63–3.16
0.65
0.26–1.60
0.57–1.87
1.51
0.67–3.39
1.08
0.41–2.83
0.43–1.58
0.78
0.32–1.90
0.87
0.32–2.38
0.40–1.34
0.68
0.29–1.59
0.95
0.39–2.31
0.33–1.07
0.54
0.24–1.22
0.61
0.25–1.52
0.36–1.89
0.77
0.20–2.98
1.18
0.40–3.44
0.42–1.48 0.23–1.24 0.15–3.82
0.68 0.44 1.02
0.26–1.77 0.15–1.34 0.88–11.94
1.20 0.73 0.98
0.47–3.05 0.18–2.91 0.10–9.06
0.70–1.27
0.75
0.47–1.18
1.24
0.86–1.80
0.93–2.82
2.73
0.96–7.78
2.02
0.72–5.67
0.52–2.29
1.35
0.52–3.52
0.74
0.20–2.68
*Cases, 63; controls, 243. **Cases, 35; controls, 93. ***Cases, 28; controls, 150. In four cases age was missed.
case – control studies have show that being a passive smoker is associated with carriage,25 – 27 which has led to the hypothesis that children growing up in smoking environments were more susceptible to developing meningococcal disease in later years. Our data demonstrate that passive smoking increases the risk of suffering meningococcal disease, mainly in children aged less than 5 years, with a possible dose-response relationship. The role played by other smoking members of the family, besides parents, should be noted. A possible
Table 5 Logistic regression analysis for risk factors of invasive pneumococcal disease among children under 2 years old OR crude OR adjusted 95% CI Day nursery (yes/no) 2.73 Siblings under 15 years 1.41 (yes/no) Age (in months) 0.99
4.21 1.76
1.28 –13.83 0.75 –4.17
0.97
0.97 –1.05
explanation could be that caregivers, mostly grandparents, are those spending more time at home with the children, and it is in the group under five years of age where contact with these people is closer and more continued. Other European studies also relate exposure to tobacco smoke with a risk of invasive meningococcal disease in children,6,7,28,29 and report very high associations (OR 7.5), with a dose-response, independent of social status.6 In the United States, it was found that having a smoking mother was the independent risk factor most strongly associated with invasive meningococcal disease (OR 3.8) in children aged less than 18 years.8 A recent study conducted in New Zealand30 found a weaker association, in that the presence of at least one smoker in the household represented an OR of 1.4 (95% CI 1.0 – 1.8) in children aged less than 8 years. Various authors state that a possible explanation for these findings is that tobacco smoke reduces the defences of the nasopharyngeal mucosa against meningococcal invasion.6 – 8,26
326
I. Pereiro ´ et al.
Table 6 Risk factors for meningococcal disease among children under 15 years old in Spain 1995–1998 Cases ðn ¼ 181Þ Siblings under 15 years No 75 Yes 103 Household members ,4 53 $4 122 Crowding , 1.5 121 $ 1.5 49 Maternal smoking No 96 Yes 78 Paternal smoking No 82 Yes 91 Other smoking No 135 Yes 40 No. of cigarettes smoked at home None 53 1– 29 68 30– 59 43 . 59 9 No. of smokers (mean:SD) 1.32:1.25 Day nursery No 77 Yes 107 Breastfeeding in under 12 months (36 cases 51 controls) Yes 5 No 31 Previous meningitis No 195 Yes 1 Other risk factors No 170 Yes 22 Meningococcal polysaccharide vaccine Yes 23 No 166
For cases of a pneumococcal disease, we found no association with exposure to tobacco smoke, in agreement with other authors.21 – 33 As regards the number of people living in the same household, there is an association which persists after adjusting for other variables in children aged less than 5 years both for diseases caused by the meningococcus and by Hib. For meningococcal disease, similar results are reported in another study,6 but for more than six people living in the same household. In this study, crowding, defined as more than 1.5 persons per room, showed a positive association in children under 5 years of age. In the Fischer study, living in a house with three or more residents aged less than 18 years was an independent risk factor.10 Crowding is the most important risk factor seen by another
Controls ðn ¼ 243Þ
P
OR
109 130
0.48
1.15
101 136
0.01
1.71
186 47
0.04
1.60
148 87
0.11
1.38
128 107
0.16
1.33
200 35
0.04
1.69
102 78 47 8 0.89:0.88
0.07 0.03 0.04 0.13 0.0001
1.00 1.68 1.76 2.16 1.50
137 102
0.86
0.97
5 46
0.42
1.1
237 1
0.89
1.21
198 39
0.14
1.52
31 207
0.79
1.1
author,30 with an OR of 10.7. Studies on Hib cases found similar results.15,16 In our study, it is more important to live with siblings regardless of the number of rooms in the house since crowding is not associated with disease when other factors are included in the multivariate analysis. Having older school-aged siblings has been identified in several studies as a risk factor for disease caused by Hib15,18 and pneumococcus.31 In our study, this is a risk factor for cases caused by pneumococcus in children aged less than one year. Attendance at day nursery has been considered a significant risk factor for suffering Hib disease, mainly in younger children.18 A dose-response effect has also been reported for this risk depending on the number of hours spent in the day nursery,11 – 13 and is higher in the first month of
Risk factors for invasive disease among children in Spain
327
Table 7 Risk factors for meningococcal disease by age. Univariate analysis All ages* OR Crowding (,1.5/$1.5) 1.60 Household members (,4/$4) 1.71 Siblings under 15 years (yes/no) 1.15 No. of smokers (mean:SD) 1.50 Maternal smoking (no/yes) 1.38 Paternal smoking (no/yes) 1.33 Other smoking (no/yes) 1.69 No of cigarettes smoked by other at home None , 10 0.97 10–20 2.29 . 20 11.82 No. of cigarettes smoked at home None 1–29 1.68 30–59 1.76 . 59 2.16 Other risk factors (no/yes) 1.52
,5 y**
5–15 y***
95% CI
OR
95% CI
OR
95% CI
1.01–2.54 1.13–2.58 0.78–1.70 1.22–1.84 0.93–2.06 0.89–1.97 1.02–2.80
1.40 1.73 1.07 1.64 1.50 1.55 1.99
0.78–2.53 1.05–2.83 0.66–1.73 1.21–2.12 0.92–2.46 0.95–2.55 1.07–3.68
1.94 1.58 1.23 1.32 1.24 1.08 1.24
0.92–4.10 0.68–3.62 0.60–2.53 0.91–1.92 0.62–2.48 0.56–2.14 0.52–2.99
0.49–1.92 1.04–5.04 1.46–95.38
1.02 2.57 –
0.44–2.36 0.96–6.89 –
0.80 0.71 1.50
0.05–13.14 0.03–14.30 0.07–31.6
1.05–2.67 1.03–2.99 0.79–5.93 0.87–2.67
2.07 2.16 3.91 1.60
1.13–3.77 1.11–4.19 1.17–13.03 0.80–3.20
1.28 1.38 0.46 1.33
0.60–2.70 0.55–3.47 0.45–4.69 0.50–3.53
*Cases, 181; controls, 243.**Cases, 119; controls, 164.***Cases, 57; controls, 79. In five cases age was missed.
schooling.14 However, many studies, including ours, do not find this relationship,17,24 and even find the opposite, that this has a protective effect.19 In day nurseries, children are exposed to early contact with diverse of microbial flora, and inflammation of the respiratory mucosa is frequent; therefore, there is an increased possibility of contact with Hib, followed by invasive disease. Different findings in several studies suggest that, in the absence of bias, the conditions in day nurseries may change the incidence of disease. Thus, in Spain children spend much of the day in day nursery gardens, so that the number of hours in classrooms is lower than in countries with colder or more rainy climates. Similarly, the mild climate allows that young children not going to school to spend many hours outdoor, and houses are ventilated daily, and therefore, the exposure of children to tobacco
smoke, and the contact with potential carriers in closed places may be lower. This may explain why the association found is lower than that previously described. Socioeconomic variables were not been controlled in our study; however, it should be noted that some variables associated with low socioeconomic status in other countries such as smoking and day nursery attendance are equally distributed among socioeconomic classes as stated by the National Health Survey in our setting.34 Only crowding is correlated with a low socioeconomic level in this survey, but this variable is systematically excluded from our multivariate analysis. The efficacy of the Hib vaccine is similar to that reported in other developed countries. It must be noted that this vaccine was administered to a very high percentage of children (up to 60% of children aged less than 5 years in the second year of the
Table 8 Logistic regression analysis for risk factors of meningococcal disease among children under 5 years old
No. of cigarettes smoked at home None 10–29 30–59 . 59 Household members (,4 $ 4) Meningococcal polysaccharide vaccine Yes No
OR crude
OR adjusted
95% CI
2.07 2.16 3.91 1.73
2.35 2.18 3.61 1.69
1.26–4.37 1.01–4.34 1.04–12.57 1.01–2.85
0.92
0.69
0.24–2.01
328
study), despite the fact that it was not included in the vaccination scheme. Since the coverage of the vaccine was not known in the first years, the only way to establish its efficacy are case – control studies such as this one. The high Hib vaccine coverage reached significantly reduced the incidence of the disease even in the unvaccinated2 by decreasing nasopharingeal carriage. This might affect the results of our study as there was a lower Hib disease risk during this period that could hide other risk factors. As the vaccine had to be partially paid by parents it is possible that the unvaccinated were from a different economic status that the vaccines, however, the most deprived population reached the vaccine for free. This could possible bias the study, but a previous study revealed that the vaccine coverage was most influenced by the attitude of the paediatrician towards vaccination.35 Based on the results obtained in our study, also supported by previous investigations, it appears that the incidence of invasive disease caused by meningococcus or Hib would be decreased if smoking at home was reduced, mainly in children aged less than five years. Additionally, identification and vaccination of these risk groups would probably significantly reduce these diseases.
I. Pereiro ´ et al.
9.
10.
11.
12.
13.
14.
15.
16.
17.
References 18. 1. Pereiro ´ I, Dı´ez-Domingo J, Morant A, Gimeno C, Lerma M, San-Martı´n M, Gonza ´lez A. Epidemiological pattern of meningococcal disease in Valencia, Spain. Impact of a mass immunization campaign with meningococcal C polysaccharide vaccine. Scand J Infect Dis 2001;33:581—584. 2. Dı´ez-Domingo J, Pereiro ´ I, Morant A, Gimeno C, San-Martı´n M, Gonza ´lez A. Impact of non-routine vaccination on the incidence of invasive Haemophilus influenzae type b (Hib) disease: experience in the autonomous region of Valencia, Spain. J Infect 2001. 3. Diez-Domingo J, Pereiro ´ I, Morant A, Gimeno C, Lerma M, Oyagu ´lez A. Epidemiology of invasive Strepto¨ez I, Gonza coccus pneumoniae infections in children in Spain, 1996— 1998. J Infect 2002;45:139—143. 4. Wilson N, Baker M, Martin D, Lennon D, O’Hallahan J, Jones N, et al. Meningococcal disease epidemiology and control in New Zealand. N Z Med J 1995;108:437—442. 5. Jones D. Epidemiology of meningococcal disease in Europa and the USA. In: Cartwright K, editor. Meningococcal disease. Chichester: Wiley; 1995. p. 147—157. 6. Stanwell-Smith RE, Stuart JM, Hugues AO, Robinson P, Griffin MB, Cartwright K. Smoking, the environment and meningococcal disease: a case control study. Epidemiol Infect 1994; 112:315—328. 7. Haneberg B, Toujum T, Rodabil K. Factors preceding the onset of meningococcal disease with special emphasis on passive smoking, stressful events physical fitness and general symptoms of sill health. NIPH Ann 1983;6:169—174. 8. Fischer M, Hedberg K, Cardosi P, Plikaytis B, Hoesly F,
19.
20. 21.
22.
23.
24.
25. 26.
Steingart K, et al. Tobacco smoke as a risk factor for meningococcal disease. Pediatr Infect Dis J 1997;16: 979—983. Strachan DP, Cook DG. Health effects of passive smoking. Parenteral smoking and lower respiratory illness in infancy and early childhood. Thorax 1997;52:905—914. Cochi SL, Fleming DW, Hightower AW, Limpakarnjanarat K, Facklam RR, Smith JD, et al. Primary invasive Haemophilus influenzae type b disease: a population-based assessment of risk factors. J Pediatr 1986;108:887—896. Fogarty J, Moloney AC, Newell JB. The epidemiology of Haemophilus influenzae type b disease in the Republic of Ireland. Epidemiol Infect 1995;114:451—463. Istre G, Coner J, Broome C, Hightower A, Hopkins R. Risk factors for primary invasive Haemophilus influenzae disease: increased risk from day care attendance and schoolaged household members. J Pediatr 1985;106:190—195. Redmond SR, Pichichero ME. Haemophilus influenzae type B disease. An epidemiologic study with special reference to day-care centers. JAMA 1984;252:2581—2584. Takala AK, Eskola J, Palmgren J, Ro ¨nnberg PR, Kela E, Rekola P, et al. Risk factors of invasive Haemophilus influenzae type b disease among children in Finland. J Pediatr 1989;115: 694—701. Clements DA, Weigle KA, Gilbert GL. A case—control study risk factors for invasive Haemophilus influenzae type b disease in Victoria, Australia 1988—90. J Paediatr Child Health 1995;31:513—518. Takala AK, Clements DA. Socioeconomic risk factors for invasive Haemophilus influenzae type b disease. J Infect Dis 1992;165(Suppl 1):S11—S15. Vadheim CM, Greenberg DP, Bordenave N, Ziont L, Christenson P, Waterman SH, et al. Risk factors for invasive Haemophilus influenzae type b in Los Angeles county children 18—60 months of age. Am J Epidemiol 1992;136: 221—235. Arnold C, Makintube S, Istre G. Day care attendance and other risk factors for invasive Haemophilus influenzae type b disease. Am J Epidemiol 1993;138:333—340. Mu ¨hlemann K, Alexander ER, Weiss NS, Pepe M, Schopfer K, The Swiss H. Influenzae Study Group. Risk factors for invasive Haemophilus influenzae disease among children 2—16 years of age in the vaccine era, Switzerland 1991—1993. Int J Epidemiol 1996;25:1280—1285. Lilienfeld A, Lilienfeld D. Fundamentos de epidemiologı´a, Me ´xico. Fondo educativo interamericano 1983;177—178. Holdberg C, Wright A, Martinez F, Morgan W, Taussing L, Group Health Medical Associates. Child day Care, smoking by caregivers, and lower tract illness in the first 3 years of life. Pediatrics 1993;91:885—892. Pedreira F, Guandolo V, Feroli E, Mella G, Weiss I. Involuntary smoking and incidence of respiratory illness during the first year of life. Pediatrics 1985;75:594—597. Cook DG, Whincup PH, Jarvis MJ, Strachan DP, Papacosta O, Bryant A. Passive exposure to tobacco smoke in children aged 5—7 years: individual, family, and community factors. Br Med J 1994;308:384—389. Silfverdal SA, Bodin L, Hugosson S, Garpenholt O, Werner B, Esbjo ¨rner E, et al. Protective effect of breastfeeding on invasive Haemophilus influenzae infection: a case—control study in Swedish preschool children. Int J Epidemiol 1997; 26:443—450. Stuart J, Cartwright K, Robinson P, Noah N. Effect of smoking on meningococcal carriage. Lancet 1989;23:723—725. Davies AL, O’Flanagan D, Salmon RL, Coleman TJ. Risk factors for Neisseria meningitidis carriage in a school during
Risk factors for invasive disease among children in Spain
27.
28.
29.
30.
31.
a community outbreak of meningococcal infection. Epidemiol Infect 1996;117:259—266. Imrey P, Jackson L, Ludwinski P, England III A, Fella G, Fox B, et al. Meningococcal carriage, alcohol consumption, and campus bar patronage in a serogroup C meningococcal disease outbreak. J Clin Microbiol 1995;33:3133—3137. Stuart J, Cartwright K, Dawson A, Rickard J, Noah N. Risk factors for meningococcal disease: a case control study in south west England. Comm Med 1988;10:139—145. Kriz P, Bobak M, Kriz B. Parenteral smoking, socioeconomic factors, and risk of invasive meningococcal disease in children: a population based case—control study. Arch Dis Child 2000;83:117—121. Baker M, Mcnicholas A, Garret N, Jones N, Stewart J, Koberstein V, et al. Household crowding a major risk factor for epidemic meningococcal disease in Auckland children. Pediatr Infect Dis J 2000;19:983—990. Takala A, Jero J, Kela E, Ro ¨nnberg P, Koskenniemi E, Eskola J. Risk factors for primary invasive pneumococcal disease among children in Finland. JAMA 1995;273:859—864.
329
32. Gessner B, Ussery X, Parkinson A, Breiman R. Risk factors for invasive disease caused by Streptococcus pneumoniae among Alaska native children younger than two years of age. Pediatr Infect Dis J 1995;14:123—128. 33. Levine O, Churchill J, Farley M, Lefkowitz L, Harrison L, Mcgeer A, et al. Risk factors for invasive pneumococcal disease in children: a population-based case—control study. First International Symposium on Pneumococci and Pneumococcal diseases. Helsingor, 1998. 34. Consellerı´a de Sanitat i Consum, Encuesta de salud de la Comunidad Valenciana. Valenciana: Generalitat Valenciana; 1993. 35. Diez-Domingo J, Pereiro ´ I, Ferrer A, Fullana A, Momparler P, Ballester A, Morant A, Saurı´ V. Cobertura de las vacunas administradas sistema ´ticamente y de la vacuna frente a Haemophilus influenzae tipo b previa a su incorporacio ´n al calendario vacunal en la Comunidad Valenciana. An Esp Pediatr 1998;49:568—570.