- Email: [email protected]
Long-term follow-up of chronic suppurative otitis media in a high-risk children cohort
International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Long-term follow-up of chronic suppurative otitis media in a high-risk children cohort Ramon Gordon Jensen a,b,*, Preben Homøe a, Mikael Andersson b, Anders Koch b a b
Department of Otolaryngology, Head & Neck Surgery, Rigshospitalet, University Hospital of Copenhagen, Denmark Department of Epidemiology Research Statens Serum Institut, Copenhagen, Denmark
A R T I C L E I N F O
A B S T R A C T
Article history: Received 31 January 2011 Received in revised form 20 April 2011 Accepted 22 April 2011 Available online 17 May 2011
Objective: Chronic suppurative otitis media (CSOM) is the leading cause of mild to moderate hearing impairment in children worldwide and a major public health problem in many indigenous populations. There is a lack of basic epidemiological facts and knowledge on the development of CSOM, as the disease primarily affects developing countries where research capacities often are limited. The purpose of this study was to determine the long-term outcome of CSOM in a high-risk population and to identify risk factors. Methods: Follow-up study (2008) on a population-based cohort of 465 children in Greenland, initially examined (1996–8) between the ages 0 and 4 years. Follow-up was attempted among 307 children living in the two major towns. Binomial logistic regression analysis was made to identify risk factors for developing CSOM and for maintaining disease in to adolescence (odds ratios). Log linear binomial regression was used to estimate risk ratios and absolute risks. Results: At follow-up 236 participated (77% of those available). The prevalence of CSOM was 32/236 (14%) at age group 0–4 years and 21/236 (9%) at age group 11–15 years. Thirteen had disease debut after the initial study. Of those with CSOM in the initial study 24/32 (75%) healed spontaneously. Risk factors for the development of CSOM at any time in childhood was the mother’s history of CSOM OR 2.55 (95% CI 1.14–5.70; p = 0.02), and mothers with low levels of schooling OR 1.57 (1.03–2.40; p = 0.04). Once CSOM had developed boys were more likely to have persistent disease OR 5.46 (95% CI 1.47–20.37; p = 0.01). The absolute risk of CSOM if the mother had both a history of CSOM and low schooling was for boys 45.4% (95% CI 26.5–77.7) and for girls 30.7% (95% CI 17.8–53.10). The cumulative risk of CSOM was 19% at follow-up. Conclusions: Even though a large number of CSOM cases seemed to heal spontaneously, the prevalence of untreated CSOM among school-age children in Greenland remained high as new cases were found at follow-up. Increased focus on prevention and identification of children at special risk could reduce the high prevalence of CSOM. ß 2011 Elsevier Ireland Ltd. All rights reserved.
Keywords: Chronic suppurative otitis media Follow-up Cohort study Risk factor Tympanic membrane perforation Developing countries
1. Introduction Chronic suppurative otitis media (CSOM) is the leading cause of mild to moderate hearing impairment in children worldwide and a major health problem in many indigenous populations around the world. In 2004 the WHO estimated that CSOM affects between 65 and 330 million people worldwide, causing 28,000 deaths and leaving 39–200 million with hearing loss [1]. Even though it is the most severe form of otitis media, epidemiological research on this
* Corresponding author at: Department of Otolaryngology, Head & Neck Surgery F2071, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark. Fax: +45 35452690. E-mail address: [email protected] (R.G. Jensen). 0165-5876/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2011.04.017
disease is limited compared with the abundant literature on acute otitis media and otitis media with effusion. CSOM is a chronic inflammation of the middle ear and mastoid mucosa, with discharge through a chronic perforation of the tympanic membrane. A chronic perforation without evident discharge is included under this definition as it is considered to be a different stage of the same disease, where the tympanic membrane fails to heal, leaving it vulnerable to recurrent infections and causing conductive hearing loss [2]. CSOM is thought to develop in early childhood and has the potential of persisting in to adulthood causing repeated periods with draining ears that can last for months or years. The associated hearing loss may cause educational difficulties in children [3–5]. Managing CSOM is challenging as there is no consensus on the optimal medical treatment [6]. Furthermore, the lack of specialists in Greenland limits access to surgical treatment and follow-up [7].
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
Fifteen thousand of the total population of 56,000 live in the capital, Nuuk, where the only ear nose and throat specialist is based. The remaining part of the population is visited once per year by Danish otologists. The WHO states that a CSOM prevalence above 4% is a massive public health problem which needs urgent attention [2]. During the last decades several epidemiological studies have confirmed a continuously high prevalence of CSOM in Greenland (7–12%) as well as in other Inuit populations in the Arctic [8,9]. Risk factors for upper respiratory tract infections (URTI) and acute and chronic suppurative otitis media have been identified [9–11]. However, the knowledge on the course of CSOM and long-term outcomes in high-risk populations is scarce and has not been described in Greenland before. The lack of an effective medical treatment regime and low access to surgical treatment in the area underline the need for knowledge of factors that may predict CSOM and possible longterm sequelae of CSOM, to identify children at special risk, as well as the need for early treatment and follow-up. Therefore, we conducted a follow-up study on a populationbased cohort of children in order to describe the long-term outcome of CSOM, determine risk factors for development of CSOM and for persistence of CSOM. To our knowledge this is the first long-term population-based cohort study on CSOM.
949
carried out. Ear examinations were performed by the first author with a Carl Zeiss Otomicroscope (Opmi Pico ENT) and tympanometry with a Madsen Otoflex 100, GN Otometrics. All findings were reported in a standardized scheme modeled from the WHO WHO/ PDH Ear and Hearing Disorders Examination Form Version 7.1A [12]. Information about previous otological disease, upper respiratory tract infections (URTI), former ear surgery, background information (family history, housing conditions, parents’ educational level, childcare attendance and breast feeding) was obtained through a questionnaire. In Greenland all contacts to the healthcare system and prescription of medicine are recorded in the patient’s medical files which follow the patient throughout life. Recordings are made by medical doctors, nurses or trained health personnel. The date of contact, treatment, otologic findings as otorrhea or perforations and duration of disease, were collected from the records, in order to define otitis media cases according to the study definitions. Information about ear surgery was also collected from the medical files. Written informed consent was obtained by parents or guardians of all participating children. The study fulfilled the Helsinki II Declaration and was ethically approved by the Commission for Scientific Research in Greenland. 2.3. Definition of tympanic membrane pathology in the follow-up study
2. Materials and methods The present study is a follow-up study conducted in October and November 2008 of an open cohort of 465 children established between April 1996 and December 1998, when the children were between 0 and 4 years of age. The children were at that time living in the West Greenlandic town of Sisimiut which has 5400 inhabitants and is the second largest town in Greenland. 2.1. Initial study 1996–1998 The specific objectives for the cohort study 1996–1998 were to describe incidence and prevalence of upper and lower respiratory tract infections and to identify risk factors for such. The cohort formation has previously been described in detail [9]. In brief, all children in Sisimiut and two adjacent settlements below 3 years of age were asked to participate. The 465 participating children (86% of children in the community) were followed at regular intervals for a 2-year period until their 3rd or 4th year birthday, and episodes of respiratory tract infections were recorded based on medical history and clinical examinations including otoscopy and tympanometry. Background information was obtained at enrolment through standardized interviews. CSOM was defined as either ear discharge from a tympanic membrane perforation or a dry perforation both persisting more than 14 days [11]. Diagnosis of CSOM could be made at any point in point during the study period. 2.2. Follow-up 2008 In August 2008 the addresses for the 465 children in the initial cohort study were sought in the Civil Registration System (CRS) that contains updated vital information on all Greenlanders. Towns in Greenland are scattered over a vast area and for logistic reasons we selected the 307 children who still lived in Sisimiut inclusive of the two settlements Itilleq and Sarfannguaq, or the capital Nuuk. The remaining children had moved to other towns in Greenland or to Denmark. A letter of invitation was sent to the selected families, followed by up to three phone calls, to invite children to participate. The clinical examinations were carried out at the local hospital where collection of questionnaire information was
CSOM was defined as more than two weeks of ear discharge through a perforated tympanic membrane or a chronic perforation. A chronic perforation was defined as a tympanic membrane perforation lasting for more than 3 months, as this is the time when non-chronic tympanic membrane perforations (e.g. after acute otitis media) are expected to have healed spontaneously [13]. Participants with perforation at examination were asked about symptoms of otitis media or ear trauma during the previous three months and participants with discharge were asked about the duration. Otitis media with effusion (OME) was defined as a type B tympanogram with an ear canal volume (ECV) between 0.5 ml and 1.5 ml as this was considered normal in this age group [14]. ECV under 0.5 ml indicated misplacement of the immitance probe. ECV over 1.5 indicated perforation and was confirmed by otomicroscopy. Tubal dysfunction was defined as either type C1 or C2 tympanogram, Myringosclerosis was defined as calcified plaques and atrophy as localised areas of thinning of the tympanic membrane. Retraction was defined as a medialized area of the tympanic membrane and fibrosis was defined as areas of the tympanic membrane without transluscency, with no other signs of oedema due to middle-ear effusion or inflammation. 2.4. Statistics Statistical analyses were performed using SAS v 9.2 (SAS Institute Inc., Cary, NC, USA). Fisher’s exact x2 test was used for test for proportions and Kruskal–Wallis test for testing medians between groups. The associations between potential risk factors and CSOM or ear surgery for CSOM at any time as outcome were analyzed separately using binomial logistic regression analysis adjusting for gender. All variables were analyzed as categorical variables with p-values indicating tests for homogeneity, but if estimates showed a consistent change in size over categories (e.g. maternal schooling), categories were given successive values (e.g. 1–4) and the variable tested for trend as continuous. Based on these univariate results a multivariate regression model was created with all risk factors with p-values < 0.1 included in the starting model. Insignificant factors were eliminated by backwards elimination until all factors in the final model had reached p < 0.05. Results are given in odds ratios. Direct estimations of risk
950
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
and relative risk (RR) were made using log linear binomial regression. Similar risk factor analyses were carried out among children who developed CSOM at any time, and had persistent CSOM at follow-up. One minus the Kaplan–Meier survival method was used to calculate the cumulative risk. 3. Results 3.1. Participation Of the 307 children who were contacted in total, 236 (77%), now aged 11–15 years, participated. In Sisimiut 254 children were contacted, while 11 had recently moved out of town. Of the remaining children, 209 (86%) participated. Nine refused to participate, one due to medical reasons not related to CSOM. We were not able to trace the correct addresses or phone numbers of 25 children. Participation from the two settlements Itilleq and Sarfanguaq was low, 5 of 24 (21%), as the study did not cover transportation costs. In the capital, Nuuk, 22 of 29 (76%) participated while we were not able to establish contact with the rest. Fig. 1 shows participation and CSOM findings in the initial study and the follow-up study. Characteristics of the participants, non-participants and the children lost to follow up are shown in Table 1. There were no significant differences in distribution of age, gender or history of CSOM between these groups. There were fewer Danes among the participants compared with the children lost to follow-up. 3.2. Otological findings Table 2 shows the otomicroscopical findings at follow-up in October/November 2008. Wax removal was not possible in 5 ears leaving 467 ears for description of tympanic membrane pathologies. A total of 16 ears (3.4%) in 16 children were found with perforation, 37 (7.9%) with myringosclerosis and 42 (9.0%) with atrophy. Five girls and 11 boys had CSOM. Additionally four boys had been surgically treated with tympanoplasty at the ages 11–15 years, and one girl with a history of CSOM from early childhood had a cholesteatoma resected at the age of nine. Thus, a total of 21 children had CSOM present at follow-up or had been surgically treated, of whom 13 were new cases with disease debut after the initial study. Fig. 2 shows the long-term tympanic membrane findings in the 32 participants who were diagnosed with CSOM. The proportion of spontaneous healing per ear was 87% (39/45). Thus, 8 participants
Table 1 Demographic characteristics of a population-based cohort of 465 children, followed-up 2008 in Sisimiut and Nuuk, Greenland. Lost to follow-up (n = 158) Age (median) Gender Girl Boy Etnicitya Inuit Danish Mixed Unknown CSOMb in first study
13
Participants (n = 236) 13
Non-participants (n = 71) 13
77 81
49% 51%
128 108
54% 46%
32 39
45% 55%
127 10 13 8 16
80% 6% 8% 5% 10%
202 4 25 5 32
86% 2% 11% 2% 14%
69 0 1 1 7
97% 0% 1% 1% 10%
a Inuits, both parents born in Greenland. Danish, both parents born in Denmark. Mixed, one parent born in Greenland, one parent born in Denmark or other countries. Unknown, one or both parents place of birth unknown. b CSOM: chronic suppurative otitis media, dry chronic perforations included.
(25%) had CSOM that proceeded from the initial study to follow-up either in the same or the contralateral ear, and none had bilateral disease. The proportion of spontaneous healing was higher among girls than boys (94% vs. 53%; p = 0.013; Fishers exact test). The proportion of participants without perforations at follow-up but with CSOM in the initial study was 75% (24/32). Among the 236 children who participated at follow-up, the median age at which they left the initial study was 37 months. For the 24 children who healed spontaneously the median age at the final examination in the initial study was 38 months and for the 8 children with persistent CSOM at follow-up and the 13 new cases of CSOM the median ages were 39 and 27 months, respectively (p = 0.75; Kruskal–Wallis test). In 7 of the 24 children who had healed spontaneously at followup, signs of perforation or otorrhea were already absent at the final examination in the initial study. The proportion of children with otorrhea at their final examination was 4/24 (17%) for those who had healed and 3/8 (38%) in the group that did not heal (p = 0.33, Fisher’s exact test). Four participants with dry perforations at follow-up were neither themselves nor were their parents aware of ear problems and had no information in their medical files about CSOM. None of these had experienced traumas to their ears or otitis media three months prior to examination that could explain the perforations. Nine (20%) of the 45 children found with CSOM in the initial study or at follow-up had never received antibiotic treatment for any type of otitis media, according to medical records, and 20 (49%) had received antibiotics less than three times. Among the 21 children found with CSOM/surgery at follow-up 3 (14%) had never received antibiotics and 10 (48%) had received antibiotics less than three times. None of the four children found with CSOM and present ear discharge were under antibiotic treatment at the time of examination. 3.3. Incidence, prevalence and risk factor analyses
Fig. 1. Participation in initial study and follow-up. Number in parenthesis indicates number of participants diagnosed with chronic suppurative otitis media (CSOM) in the initial study (CSOM-IS).
Based on the 45 participants with CSOM at any point until follow-up in 2008 the cumulative incidence of CSOM at 11–15 years of age was 19%. According to data from the initial study, medical files and questionnaires, all had their disease debut before the age of 7, except in 4 cases for whom no information was found. The cumulative incidence among the participants at follow-up was 14% at age 0–4 years, which corresponds to results found in the original study. The prevalence of untreated CSOM among children (aged 11–15 years) in this cohort was 7%, and when including children who had undergone surgery, 9%.
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
951
Table 2 Otomicroscopic and tympanometric findings in a population-based cohort of 236 children with a median age of 13 years followed up 2008 in Sisimiut and Nuuk, Greenland.a Chronic otitis media b
Boys Girls Sum % of total a b c d e f g h i
CSOM
Perf.
2 2 4 1.7
9 3 12 5.1
c
Sequelae otitidis Operation
d
4 1 5 2.1
Myringo.
e
5 18 23 9.7
OMEf
Tub.dys.g
Fibrosish
Normali
Total
7 2 9 3.8
1 1 2 0.8
19 28 47 19.9
49 65 114 48.3
104 132 236 100
Atrophy 8 12 20 8.5
Participants were categorized according to most severe pathology found in one ear at otomicroscopy and tympanometry. Chronic suppurative otitis media. Chronic perforation. Operations: four tympanoplasties; one cholesteatoma resection. Myringosclerosis. Otitis media with effusion. Tubal dysfunction. Fibrosis and flaccida retractions. Includes 3 participants with wax (cerumen impacta), were removal was not possible.
45 ears with CSOM-IS
19 ears without CSOM-IS
6 (13%) Ears with CSOM at follow-up
2 (11%) Ears with CSOM at follow-up
10 (22%) Sclerosis
3 (16%) Sclerosis
7 (16%) Atrophy
7 (36%) Atrophy
10 (22%) Fibrosis
2 (11%) Fibrosis
10 (22%) Normal A
5 (26%) Normal
A
In two participants the tympanic membrane was not seen adequately. However, tympanometry was normal.
Fig. 2. Otological findings at 10-year follow-up among 32 participants (64 ears) who had had chronic suppurative otitis media in the initial study (CSOM-IS).
Tables 3 and 4 show the results of the univariate risk factor analyses for development of CSOM. Several risk factors were associated with the development of CSOM but only mother’s history of CSOM and school degree as a continuous variable reached significance. In the multivariate model these variables remained significant: mother’s history of CSOM OR 2.55 (95% CI 1.14–5.70; p = 0.02) and a low level of schooling OR 1.57 (95% CI 1.03–2.40; p = 0.04). Similar analyses for persistence of CSOM (failure to heal) among the 45 children with CSOM showed, that boys were more likely to have persistent CSOM compared with girls (OR 5.46; 95% CI 1.47– 20.37; p = 0.01). Age at follow-up did not influence healing rates. The previously mentioned potential risk factors were all tested on failure to heal but no further associations were found. Neither age at debut or at follow-up nor any of the other risk factors above were associated with healing. A second multivariate model to directly estimate adjusted risk ratios and absolute risks, using the previous mentioned risk factors showed that the RR for girls developing CSOM compared to boys was 0.9 (95% CI 0.53–1.51; p = 0.68). The RR for a child to develop CSOM if the mother had a history of CSOM compared with no history of CSOM was 1.9 (95% CI 1.04–3.22; p = 0.04) and if the mother had low levels of schooling compared with high levels the RR was 1.37 (95% CI 1–1.85; p = 0.05). The absolute risk of getting CSOM was 21.3% (95% CI 14.8–30.6) for boys and 17.2% (95% CI 11.8–25.1) for girls. The absolute risk of getting CSOM being a boy with a mother who had a history of CSOM was 35.4% (95% CI 21– 59.5), if the mother furthermore had low schooling, the risk increased to 45.4% (95% CI 26.5–77.7). With the same combina-
tions of risk factors for girls the absolute risk was 30.7% (95% CI 17.8–53.10) and 40.76% (95% CI 22.3–74.5), respectively. The RR for persistence of CSOM was 0.4 for girls compared with boys (95% CI 0.17–0.79; p = 0.01). 4. Discussion This is to our knowledge the first study looking at long-term outcome and spontaneous healing after CSOM in a cohort design, not only in Greenland, but also globally. The overall participation rate at follow-up in 2008 was high, up to 86% in Sisimiut, where the majority of participants came from. We did not find demographic differences between participants and children lost to follow-up or non-participants that could introduce bias to the study, and the groups did not differ in respect to initial occurrence of CSOM. Therefore, we consider our findings representative of the background population. However, 36 percent of the original cohort had moved to other towns and the high mobility among Greenlandic families makes it challenging to follow-up on child cohorts over time. We found a surprisingly high proportion of spontaneous healing at 75%, which is encouraging but should be interpreted with some caution. The initial study definition of CSOM included dry perforation of 14 days duration as a minimum. Some of these cases initially diagnosed as CSOM might, at the time of observation, have been cases of acute otitis media with perforation in healing, which may take longer than 14 days [13]. Although the majority were of substantially longer duration than 14 days [11], we cannot exclude that a few cases might have been cases of healing after severe acute otitis media thus leading to an overestimation of the proportion of healing after CSOM. According to a survey on CSOM in Greenland in 1994, 19% had healed at follow-up after one year [15]. In the present study boys and girls seemed to have an equal risk of getting CSOM, but girls had better chances of spontaneous healing. This gender difference in healing of the tympanic membrane has, to our knowledge, not been described before. A few studies have described boys having a higher risk of developing severe courses of otitis media [16,17], and otitis media with effusion is also found with higher incidences among boys [18,19]. Part of the explanation could be found in the larger pneumatization area of the mastoid among girls [20], but a difference due to chance or other factors cannot be excluded. A gender difference was found for lower respiratory tract infections in this study cohort, but not for upper respiratory tract infections [9]. Differences in endocrine– immune interactions may play a role in male susceptibility to infections [21]. We found mother’s level of schooling and history of CSOM as independent risk factors for CSOM development among their children, confirming findings risk factor analysis in the initial study
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
952
Table 3 Child-related and mother-related risk factors for the development of chronic suppurative otitis media (CSOM) at any time in childhood. A follow-up study of 236 children aged 11–15 years in Nuuk and Sisimiut, Greenland. Univariate risk factor analyses.a Risk factors (no of observations)
(No with CSOM/without CSOM)
OR
Male (22/106) Female (23/85)
1 0.77
Greenlandic (42/160) Danish/mixed (3/26)
1 0.44
8th degree (6/15) 9th–10th degree (16/49) 11th–12th degree (21/101) High school (2/22)
4.28 3.51 2.26 1
Long/medium theoretical (26/95) Short/practical education of medium length (2/14) Short courses/education not completed (2/15) None (12/52) Student (3/13)
1 0.51 0.46 0.81 0.78
No (32/163) Yes (12/24)
1 2.52
No (32/163) Yes (8/21)
1 1.92
6 months (18/88) >6 months (19/85)
1 0.97
95% CI
Gender (n = 236)
0.42 0.40–1.47
Ethnicity (n = 231)
0.2 0.13–1.54
Mother’s schooling (n = 232)b
0.25/0.04c 0.76–24.22 0.74–16.62 0.49–10.36
Mother’s education (n = 234)c
0.8 0.11–2.39 0.10–2.17 0.38–1.75 0.21–2.91
Mother’s history of COM (n = 231)
0.02 1.14–5.56
Siblings’ history of COM (n = 224)
0.16 0.78–4.73
Breast feeding (n = 210)
a b c
p valueb
0.93 0.47–1.98
All analyses adjusted for gender. Test for homogeneity. Test for trend.
Table 4 Environment-related risk factors for the development of chronic suppurative otitis media (CSOM) at any time in childhood. A follow-up study of 236 children aged 11–15 years in Nuuk and Sisimiut, Greenland. Univariate risk factor analyses.a Risk factors (no of observations)
(No with CSOM/without CSOM)
OR
4 (15/52) 1–3 (6/41) 5 (15/59) 6+ (9/39)
1 0.51 0.9 0.79
0–1 (5/24) 2 (19/98) 3+ (21/60)
1 0.98 1.77
No (9/38) Yes (36/153)
1 0.98
0 (9/38) 1 (7/45) 2+ (29/108)
1 0.63 1.13
6 months (8/14) 7–12 months (2/13) >12 months (4/29)
1 0.25 0.23
6 months (1/7) 7–24 months (8/44) 25–48 months (22/75) >48 months (11/45)
1 1.22 1.97 1.68
95% CI
No. of persons in household (n = 236)
0.63 0.18–1.43 0.40–2.02 0.31–2.00
No. of persons sleeping in same room (n = 227)
0.22 0.33–2.89 0.59–5.26
Smoking at home (n = 236)
0.97 0.44–2.22
No. of smokers at home (n = 236)
0.45 0.22–1.87 0.49–2.61
Time in daycare (n = 70)
0.07 0.04–1.43 0.06–0.91
Time in child-care center (n = 213)
a b
All analyses adjusted for gender. Test for homogeneity.
p valueb
0.72 0.13–11.40 0.23–16.96 0.19–15.18
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
among children below 4 years of age [11]. This original risk factor analysis also found that the use of childcare centers and being Inuit increased the risk of CSOM, which was also found in the present study, although not statistically significant. A previous finding of an association between CSOM and smoking in the home was not confirmed. In spite of the possibility to find more cases during a longterm follow-up study compared with a cross sectional study, the number of participants and cases in the present study is still low limiting the statistical power of the multivariate risk factor analysis. As 45 of the 236 participants in follow-up had a verified episode of CSOM, nearly every 5th child has had CSOM at some point in childhood. At follow-up we found 21 participants with present disease or who had undergone surgery for CSOM. Thirteen of these children (62%) were diagnosed with CSOM after the initial study. This large proportion with an apparent late debut may partly be explained by the fact that they left the initial study between 2 and 3 years of age, rather than an unusual late disease debut. Secondly some of the children could have had severe otitis media or CSOM and have healed before inclusion in the initial study, subsequently reperforated at some time point before follow-up. However, as indicated in the cumulative incidence we believe the majority of CSOM debut are before 4 years of age, which also concurs with findings in the medical records and the general experience with the disease in Greenland. It is possible that other participants might have had CSOM debut after termination of the first study, but healed spontaneously before the follow-up. Hence, we consider the number of detected CSOM cases in this cohort as a conservative estimate of the amount of individuals affected by the disease through childhood in the background population. The most frequent pathologic findings at otomicroscopy in this study were atrophy and myringosclerosis which are also frequent at long-term follow-up among children with otitis media in Western countries [22]. The prevalence of both pathologies was 17% of all ears in this study, which is comparable with results from Danish follow-up studies. However, in the Danish studies the pathologic changes are most often caused by tympanostomy tube placement or myringotomy [23]. In this Greenlandic cohort the tympanic membrane scarring is predominantly caused by former otitis media as the use of tympanostomy tubes and myringotomy has not been a part of otitis media treatment. 4.1. Clinical implications The initial examinations of this cohort in 1996–98 and the follow-up in 2008 show that most cases of CSOM have their debut at early childhood and at least before school age. Four participants with dry perforations at follow-up were neither themselves nor were their parents aware of ear problems, which indicates that CSOM can be clinically silent after otitis media early in childhood and emphasizes the importance of routine screening of hearing loss before school start. Only 20% of children with CSOM had ever received antibiotic treatment for any type of otitis media, according to the recordings in the medical files. This could partly be explained by the widespread use of antiseptic irrigation of draining ears with an acetic acid solution as the sole treatment for CSOM. However, it might also reflect the fact that otorrhea is so common in this population that some parents see it as a ‘‘normal’’ part of childhood and do not seek treatment if the children have no other symptoms than drainage from the ear, which explains why none of the four participants with draining ears at follow-up were receiving any treatment. This acceptance of the disease has been observed in other countries with high prevalence of CSOM [1]. There seems to be a diagnostic and treatment potential in the Greenlandic primary health care sector to reduce the risk of hearing loss. It is expected that the degree to which hearing is compromised is proportional to
953
the damage of structures in the middle ear. Hence a focus on reducing numbers and lengths of episodes with otorrhea in CSOM is important [24]. The use of ototopical antibiotics should be considered in any case of prolonged otorrhea [25,26]. The fact that Greenlandic boys and girls, with mothers having a history of CSOM and low education, have a 45% and 41% risk of developing CSOM, respectively, should be taken into account when deciding which children should receive antibiotic treatment for acute otitis media. Age has previously been seen as one of the most important factors determining successful outcome for surgical management of CSOM, but a number of recent studies contradict this [27,28]. However, in settings where the surgical capacity is limited, the possibility of spontaneous healing should be considered when selecting young children for surgery. The change in CSOM prevalence from early childhood to adolescence due to spontaneous healing indicates that some precaution should be made when comparing CSOM prevalences between countries and regions, based on different age groups varying from preschool children to adults [1]. In conclusion we found a surprisingly high rate of spontaneous healing from CSOM. However, the prevalence of CSOM among school-age children (11–15 years) in Greenland remained high. Only maternal history of CSOM and education proved in this study to be risk factors for overall CSOM development. Girls healed significantly better than boys, but no other factors were identified as determinants for such. Increased focus on early diagnosis and treatment is essential to diminish the effects of hearing loss during early speech and cognitive development. Information about the disease and treatment possibilities to primary health care providers as well as parents could reduce the risk of CSOMrelated hearing loss in school. Conflict of interest None. Acknowledgements We would like in particular to thank the participating children and their families of Sisimiut and Nuuk as well as the staff at Sisimiut Health Center, Sisimiut, Dronning Ingrids Hospital, Nuuk and The Primary Health Care Center, Nuuk. Funding for this study was provided by The Commission for Scientific Research in Greenland (KVUG). References [1] WHO, Chronic Suppurative Otitis Media. Burden of Illness and Management Options, WHO, Geneva, Switzerland, 2004. [2] WHO, Report of a WHO/CIBA Foundation Workshop 1996, Prevention of Hearing Impairment from Chronic Otitis Media, 2000. [3] H. Winskel, The effects of an early history of otitis media on children’s language and literacy skill development, Br. J. Educ. Psychol. 76 (2006) 727–744. [4] F. Olatoke, F.E. Ologe, C.C. Nwawolo, M.J. Saka, The prevalence of hearing loss among schoolchildren with chronic suppurative otitis media in Nigeria, and its effect on academic performance, Ear. Nose. Throat J. 87 (2008) E19. [5] A. Golz, S.T. Westerman, L.M. Westerman, D.A. Gilbert, A. Netzer, Does otitis media in early childhood affect reading performance in later school years? Otolaryngol. Head Neck Surg. 134 (2006) 936–939. [6] M. Verhoeff, E.L. van der Veen, M.M. Rovers, E.A. Sanders, A.G. Schilder, Chronic suppurative otitis media: a review, Int. J. Pediatr. Otorhinolaryngol. 70 (2006) 1–12. [7] P. Homoe, G. Nikoghosyan, C. Siim, P. Bretlau, Hearing outcomes after mobile ear surgery for chronic otitis media in Greenland, Int. J. Circumpolar Health 67 (2008) 452–460. [8] P. Homoe, Otitis media in Greenland. Studies on historical, epidemiological, microbiological, and immunological aspects, Int. J. Circumpolar Health 60 (Suppl. 2) (2001) 1–54. [9] A. Koch, K. Molbak, P. Homoe, P. Sorensen, T. Hjuler, M.E. Olesen, et al., Risk factors for acute respiratory tract infections in young Greenlandic children, Am. J. Epidemiol. 158 (2003) 374–384.
954
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
[10] P. Homoe, R.B. Christensen, P. Bretlau, Acute otitis media and sociomedical risk factors among unselected children in Greenland, Int. J. Pediatr. Otorhinolaryngol. 49 (1999) 37–52. [11] A. Koch, P. Homoe, C. Pipper, T. Hjuler, M. Melbye, Chronic suppurative otitis media in a birth cohort of children in greenland: population-based study of incidence and risk factors, Pediatr. Infect. Dis. J. 30 (2011) 25–29. [12] WHO, WHO Ear and Hearing Disorders Survey. Protocol for a Population-based Survey of Prevalence and Causes of Deafness and Hearing Impairment and Other Ear Diseases, 1999. [13] G. Berger, Nature of spontaneous tympanic membrane perforation in acute otitis media in children, J. Laryngol. Otol. 103 (1989) 1150–1153. [14] J.J. Haapaniemi, Immittance findings in school-aged children, Ear Hear. 17 (1996) 19–27. [15] P. Homoe, R.B. Christensen, P. Bretlau, Prevalence of otitis media in a survey of 591 unselected Greenlandic children, Int. J. Pediatr. Otorhinolaryngol. 36 (1996) 215–230. [16] M. Hotomi, N. Yamanaka, T. Samukawa, M. Suzumot, A. Sakai, J. Shimada, et al., Treatment and outcome of severe and non-severe acute otitis media, Eur. J. Pediatr. 164 (2005) 3–8. [17] E. Vartiainen, Changes in the clinical presentation of chronic otitis media from the 1970s to the 1990s, J. Laryngol. Otol. 112 (1998) 1034–1037. [18] A.M. Hardy, M.G. Fowler, Child care arrangements and repeated ear infections in young children, Am. J. Public Health 83 (1993) 1321–1325. [19] J.L. Paradise, H.E. Rockette, D.K. Colborn, B.S. Bernard, C.G. Smith, M. Kurs-Lasky, et al., Otitis media in 2253 Pittsburgh-area infants: prevalence and risk factors during the first two years of life, Pediatrics 99 (1997) 318–333.
[20] M. Tos, S.E. Stangerup, Secretory otitis and pneumatization of the mastoid process: sexual differences in the size of mastoid cell system, Am. J. Otolaryngol. 6 (1985) 199–205. [21] S.L. Klein, The effects of hormones on sex differences in infection: from genes to behavior, Neurosci. Biobehav. Rev. 24 (2000) 627–638. [22] P. Caye-Thomasen, S.E. Stangerup, G. Jorgensen, D. Drozdziewic, P. Bonding, M. Tos, Myringotomy versus ventilation tubes in secretory otitis media: eardrum pathology, hearing, and eustachian tube function 25 years after treatment, Otol. Neurotol. 29 (2008) 649–657. [23] S.E. Stangerup, M. Tos, R. Arnesen, P. Larsen, A cohort study of point prevalence of eardrum pathology in children and teenagers from age 5 to age 16, Eur. Arch. Otorhinolaryngol. 251 (1994) 399–403. [24] L.F. Silveira Netto, S.S. da Costa, P. Sleifer, M.E. Braga, The impact of chronic suppurative otitis media on children’s and teenagers’ hearing, Int. J. Pediatr. Otorhinolaryngol. 73 (2009) 1751–1756. [25] C.A. Macfadyen, J.M. Acuin, C. Gamble, Topical antibiotics without steroids for chronically discharging ears with underlying eardrum perforations, Cochrane Database Syst. Rev. (2005), CD004618. [26] C.A. Macfadyen, J.M. Acuin, C. Gamble, Systemic antibiotics versus topical treatments for chronically discharging ears with underlying eardrum perforations, Cochrane Database Syst. Rev. (2006), CD005608. [27] A.C. Lin, A.H. Messner, Pediatric tympanoplasty: factors affecting success, Curr. Opin. Otolaryngol. Head Neck Surg. 16 (2008) 64–68. [28] S. Sarkar, A. Roychoudhury, B.K. Roychaudhuri, Tympanoplasty in children, Eur. Arch. Otorhinolaryngol. 266 (2009) 627–633.
Contents lists available at ScienceDirect
International Journal of Pediatric Otorhinolaryngology journal homepage: www.elsevier.com/locate/ijporl
Long-term follow-up of chronic suppurative otitis media in a high-risk children cohort Ramon Gordon Jensen a,b,*, Preben Homøe a, Mikael Andersson b, Anders Koch b a b
Department of Otolaryngology, Head & Neck Surgery, Rigshospitalet, University Hospital of Copenhagen, Denmark Department of Epidemiology Research Statens Serum Institut, Copenhagen, Denmark
A R T I C L E I N F O
A B S T R A C T
Article history: Received 31 January 2011 Received in revised form 20 April 2011 Accepted 22 April 2011 Available online 17 May 2011
Objective: Chronic suppurative otitis media (CSOM) is the leading cause of mild to moderate hearing impairment in children worldwide and a major public health problem in many indigenous populations. There is a lack of basic epidemiological facts and knowledge on the development of CSOM, as the disease primarily affects developing countries where research capacities often are limited. The purpose of this study was to determine the long-term outcome of CSOM in a high-risk population and to identify risk factors. Methods: Follow-up study (2008) on a population-based cohort of 465 children in Greenland, initially examined (1996–8) between the ages 0 and 4 years. Follow-up was attempted among 307 children living in the two major towns. Binomial logistic regression analysis was made to identify risk factors for developing CSOM and for maintaining disease in to adolescence (odds ratios). Log linear binomial regression was used to estimate risk ratios and absolute risks. Results: At follow-up 236 participated (77% of those available). The prevalence of CSOM was 32/236 (14%) at age group 0–4 years and 21/236 (9%) at age group 11–15 years. Thirteen had disease debut after the initial study. Of those with CSOM in the initial study 24/32 (75%) healed spontaneously. Risk factors for the development of CSOM at any time in childhood was the mother’s history of CSOM OR 2.55 (95% CI 1.14–5.70; p = 0.02), and mothers with low levels of schooling OR 1.57 (1.03–2.40; p = 0.04). Once CSOM had developed boys were more likely to have persistent disease OR 5.46 (95% CI 1.47–20.37; p = 0.01). The absolute risk of CSOM if the mother had both a history of CSOM and low schooling was for boys 45.4% (95% CI 26.5–77.7) and for girls 30.7% (95% CI 17.8–53.10). The cumulative risk of CSOM was 19% at follow-up. Conclusions: Even though a large number of CSOM cases seemed to heal spontaneously, the prevalence of untreated CSOM among school-age children in Greenland remained high as new cases were found at follow-up. Increased focus on prevention and identification of children at special risk could reduce the high prevalence of CSOM. ß 2011 Elsevier Ireland Ltd. All rights reserved.
Keywords: Chronic suppurative otitis media Follow-up Cohort study Risk factor Tympanic membrane perforation Developing countries
1. Introduction Chronic suppurative otitis media (CSOM) is the leading cause of mild to moderate hearing impairment in children worldwide and a major health problem in many indigenous populations around the world. In 2004 the WHO estimated that CSOM affects between 65 and 330 million people worldwide, causing 28,000 deaths and leaving 39–200 million with hearing loss [1]. Even though it is the most severe form of otitis media, epidemiological research on this
* Corresponding author at: Department of Otolaryngology, Head & Neck Surgery F2071, Rigshospitalet, Blegdamsvej 9, 2100 Copenhagen, Denmark. Fax: +45 35452690. E-mail address: [email protected] (R.G. Jensen). 0165-5876/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2011.04.017
disease is limited compared with the abundant literature on acute otitis media and otitis media with effusion. CSOM is a chronic inflammation of the middle ear and mastoid mucosa, with discharge through a chronic perforation of the tympanic membrane. A chronic perforation without evident discharge is included under this definition as it is considered to be a different stage of the same disease, where the tympanic membrane fails to heal, leaving it vulnerable to recurrent infections and causing conductive hearing loss [2]. CSOM is thought to develop in early childhood and has the potential of persisting in to adulthood causing repeated periods with draining ears that can last for months or years. The associated hearing loss may cause educational difficulties in children [3–5]. Managing CSOM is challenging as there is no consensus on the optimal medical treatment [6]. Furthermore, the lack of specialists in Greenland limits access to surgical treatment and follow-up [7].
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
Fifteen thousand of the total population of 56,000 live in the capital, Nuuk, where the only ear nose and throat specialist is based. The remaining part of the population is visited once per year by Danish otologists. The WHO states that a CSOM prevalence above 4% is a massive public health problem which needs urgent attention [2]. During the last decades several epidemiological studies have confirmed a continuously high prevalence of CSOM in Greenland (7–12%) as well as in other Inuit populations in the Arctic [8,9]. Risk factors for upper respiratory tract infections (URTI) and acute and chronic suppurative otitis media have been identified [9–11]. However, the knowledge on the course of CSOM and long-term outcomes in high-risk populations is scarce and has not been described in Greenland before. The lack of an effective medical treatment regime and low access to surgical treatment in the area underline the need for knowledge of factors that may predict CSOM and possible longterm sequelae of CSOM, to identify children at special risk, as well as the need for early treatment and follow-up. Therefore, we conducted a follow-up study on a populationbased cohort of children in order to describe the long-term outcome of CSOM, determine risk factors for development of CSOM and for persistence of CSOM. To our knowledge this is the first long-term population-based cohort study on CSOM.
949
carried out. Ear examinations were performed by the first author with a Carl Zeiss Otomicroscope (Opmi Pico ENT) and tympanometry with a Madsen Otoflex 100, GN Otometrics. All findings were reported in a standardized scheme modeled from the WHO WHO/ PDH Ear and Hearing Disorders Examination Form Version 7.1A [12]. Information about previous otological disease, upper respiratory tract infections (URTI), former ear surgery, background information (family history, housing conditions, parents’ educational level, childcare attendance and breast feeding) was obtained through a questionnaire. In Greenland all contacts to the healthcare system and prescription of medicine are recorded in the patient’s medical files which follow the patient throughout life. Recordings are made by medical doctors, nurses or trained health personnel. The date of contact, treatment, otologic findings as otorrhea or perforations and duration of disease, were collected from the records, in order to define otitis media cases according to the study definitions. Information about ear surgery was also collected from the medical files. Written informed consent was obtained by parents or guardians of all participating children. The study fulfilled the Helsinki II Declaration and was ethically approved by the Commission for Scientific Research in Greenland. 2.3. Definition of tympanic membrane pathology in the follow-up study
2. Materials and methods The present study is a follow-up study conducted in October and November 2008 of an open cohort of 465 children established between April 1996 and December 1998, when the children were between 0 and 4 years of age. The children were at that time living in the West Greenlandic town of Sisimiut which has 5400 inhabitants and is the second largest town in Greenland. 2.1. Initial study 1996–1998 The specific objectives for the cohort study 1996–1998 were to describe incidence and prevalence of upper and lower respiratory tract infections and to identify risk factors for such. The cohort formation has previously been described in detail [9]. In brief, all children in Sisimiut and two adjacent settlements below 3 years of age were asked to participate. The 465 participating children (86% of children in the community) were followed at regular intervals for a 2-year period until their 3rd or 4th year birthday, and episodes of respiratory tract infections were recorded based on medical history and clinical examinations including otoscopy and tympanometry. Background information was obtained at enrolment through standardized interviews. CSOM was defined as either ear discharge from a tympanic membrane perforation or a dry perforation both persisting more than 14 days [11]. Diagnosis of CSOM could be made at any point in point during the study period. 2.2. Follow-up 2008 In August 2008 the addresses for the 465 children in the initial cohort study were sought in the Civil Registration System (CRS) that contains updated vital information on all Greenlanders. Towns in Greenland are scattered over a vast area and for logistic reasons we selected the 307 children who still lived in Sisimiut inclusive of the two settlements Itilleq and Sarfannguaq, or the capital Nuuk. The remaining children had moved to other towns in Greenland or to Denmark. A letter of invitation was sent to the selected families, followed by up to three phone calls, to invite children to participate. The clinical examinations were carried out at the local hospital where collection of questionnaire information was
CSOM was defined as more than two weeks of ear discharge through a perforated tympanic membrane or a chronic perforation. A chronic perforation was defined as a tympanic membrane perforation lasting for more than 3 months, as this is the time when non-chronic tympanic membrane perforations (e.g. after acute otitis media) are expected to have healed spontaneously [13]. Participants with perforation at examination were asked about symptoms of otitis media or ear trauma during the previous three months and participants with discharge were asked about the duration. Otitis media with effusion (OME) was defined as a type B tympanogram with an ear canal volume (ECV) between 0.5 ml and 1.5 ml as this was considered normal in this age group [14]. ECV under 0.5 ml indicated misplacement of the immitance probe. ECV over 1.5 indicated perforation and was confirmed by otomicroscopy. Tubal dysfunction was defined as either type C1 or C2 tympanogram, Myringosclerosis was defined as calcified plaques and atrophy as localised areas of thinning of the tympanic membrane. Retraction was defined as a medialized area of the tympanic membrane and fibrosis was defined as areas of the tympanic membrane without transluscency, with no other signs of oedema due to middle-ear effusion or inflammation. 2.4. Statistics Statistical analyses were performed using SAS v 9.2 (SAS Institute Inc., Cary, NC, USA). Fisher’s exact x2 test was used for test for proportions and Kruskal–Wallis test for testing medians between groups. The associations between potential risk factors and CSOM or ear surgery for CSOM at any time as outcome were analyzed separately using binomial logistic regression analysis adjusting for gender. All variables were analyzed as categorical variables with p-values indicating tests for homogeneity, but if estimates showed a consistent change in size over categories (e.g. maternal schooling), categories were given successive values (e.g. 1–4) and the variable tested for trend as continuous. Based on these univariate results a multivariate regression model was created with all risk factors with p-values < 0.1 included in the starting model. Insignificant factors were eliminated by backwards elimination until all factors in the final model had reached p < 0.05. Results are given in odds ratios. Direct estimations of risk
950
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
and relative risk (RR) were made using log linear binomial regression. Similar risk factor analyses were carried out among children who developed CSOM at any time, and had persistent CSOM at follow-up. One minus the Kaplan–Meier survival method was used to calculate the cumulative risk. 3. Results 3.1. Participation Of the 307 children who were contacted in total, 236 (77%), now aged 11–15 years, participated. In Sisimiut 254 children were contacted, while 11 had recently moved out of town. Of the remaining children, 209 (86%) participated. Nine refused to participate, one due to medical reasons not related to CSOM. We were not able to trace the correct addresses or phone numbers of 25 children. Participation from the two settlements Itilleq and Sarfanguaq was low, 5 of 24 (21%), as the study did not cover transportation costs. In the capital, Nuuk, 22 of 29 (76%) participated while we were not able to establish contact with the rest. Fig. 1 shows participation and CSOM findings in the initial study and the follow-up study. Characteristics of the participants, non-participants and the children lost to follow up are shown in Table 1. There were no significant differences in distribution of age, gender or history of CSOM between these groups. There were fewer Danes among the participants compared with the children lost to follow-up. 3.2. Otological findings Table 2 shows the otomicroscopical findings at follow-up in October/November 2008. Wax removal was not possible in 5 ears leaving 467 ears for description of tympanic membrane pathologies. A total of 16 ears (3.4%) in 16 children were found with perforation, 37 (7.9%) with myringosclerosis and 42 (9.0%) with atrophy. Five girls and 11 boys had CSOM. Additionally four boys had been surgically treated with tympanoplasty at the ages 11–15 years, and one girl with a history of CSOM from early childhood had a cholesteatoma resected at the age of nine. Thus, a total of 21 children had CSOM present at follow-up or had been surgically treated, of whom 13 were new cases with disease debut after the initial study. Fig. 2 shows the long-term tympanic membrane findings in the 32 participants who were diagnosed with CSOM. The proportion of spontaneous healing per ear was 87% (39/45). Thus, 8 participants
Table 1 Demographic characteristics of a population-based cohort of 465 children, followed-up 2008 in Sisimiut and Nuuk, Greenland. Lost to follow-up (n = 158) Age (median) Gender Girl Boy Etnicitya Inuit Danish Mixed Unknown CSOMb in first study
13
Participants (n = 236) 13
Non-participants (n = 71) 13
77 81
49% 51%
128 108
54% 46%
32 39
45% 55%
127 10 13 8 16
80% 6% 8% 5% 10%
202 4 25 5 32
86% 2% 11% 2% 14%
69 0 1 1 7
97% 0% 1% 1% 10%
a Inuits, both parents born in Greenland. Danish, both parents born in Denmark. Mixed, one parent born in Greenland, one parent born in Denmark or other countries. Unknown, one or both parents place of birth unknown. b CSOM: chronic suppurative otitis media, dry chronic perforations included.
(25%) had CSOM that proceeded from the initial study to follow-up either in the same or the contralateral ear, and none had bilateral disease. The proportion of spontaneous healing was higher among girls than boys (94% vs. 53%; p = 0.013; Fishers exact test). The proportion of participants without perforations at follow-up but with CSOM in the initial study was 75% (24/32). Among the 236 children who participated at follow-up, the median age at which they left the initial study was 37 months. For the 24 children who healed spontaneously the median age at the final examination in the initial study was 38 months and for the 8 children with persistent CSOM at follow-up and the 13 new cases of CSOM the median ages were 39 and 27 months, respectively (p = 0.75; Kruskal–Wallis test). In 7 of the 24 children who had healed spontaneously at followup, signs of perforation or otorrhea were already absent at the final examination in the initial study. The proportion of children with otorrhea at their final examination was 4/24 (17%) for those who had healed and 3/8 (38%) in the group that did not heal (p = 0.33, Fisher’s exact test). Four participants with dry perforations at follow-up were neither themselves nor were their parents aware of ear problems and had no information in their medical files about CSOM. None of these had experienced traumas to their ears or otitis media three months prior to examination that could explain the perforations. Nine (20%) of the 45 children found with CSOM in the initial study or at follow-up had never received antibiotic treatment for any type of otitis media, according to medical records, and 20 (49%) had received antibiotics less than three times. Among the 21 children found with CSOM/surgery at follow-up 3 (14%) had never received antibiotics and 10 (48%) had received antibiotics less than three times. None of the four children found with CSOM and present ear discharge were under antibiotic treatment at the time of examination. 3.3. Incidence, prevalence and risk factor analyses
Fig. 1. Participation in initial study and follow-up. Number in parenthesis indicates number of participants diagnosed with chronic suppurative otitis media (CSOM) in the initial study (CSOM-IS).
Based on the 45 participants with CSOM at any point until follow-up in 2008 the cumulative incidence of CSOM at 11–15 years of age was 19%. According to data from the initial study, medical files and questionnaires, all had their disease debut before the age of 7, except in 4 cases for whom no information was found. The cumulative incidence among the participants at follow-up was 14% at age 0–4 years, which corresponds to results found in the original study. The prevalence of untreated CSOM among children (aged 11–15 years) in this cohort was 7%, and when including children who had undergone surgery, 9%.
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
951
Table 2 Otomicroscopic and tympanometric findings in a population-based cohort of 236 children with a median age of 13 years followed up 2008 in Sisimiut and Nuuk, Greenland.a Chronic otitis media b
Boys Girls Sum % of total a b c d e f g h i
CSOM
Perf.
2 2 4 1.7
9 3 12 5.1
c
Sequelae otitidis Operation
d
4 1 5 2.1
Myringo.
e
5 18 23 9.7
OMEf
Tub.dys.g
Fibrosish
Normali
Total
7 2 9 3.8
1 1 2 0.8
19 28 47 19.9
49 65 114 48.3
104 132 236 100
Atrophy 8 12 20 8.5
Participants were categorized according to most severe pathology found in one ear at otomicroscopy and tympanometry. Chronic suppurative otitis media. Chronic perforation. Operations: four tympanoplasties; one cholesteatoma resection. Myringosclerosis. Otitis media with effusion. Tubal dysfunction. Fibrosis and flaccida retractions. Includes 3 participants with wax (cerumen impacta), were removal was not possible.
45 ears with CSOM-IS
19 ears without CSOM-IS
6 (13%) Ears with CSOM at follow-up
2 (11%) Ears with CSOM at follow-up
10 (22%) Sclerosis
3 (16%) Sclerosis
7 (16%) Atrophy
7 (36%) Atrophy
10 (22%) Fibrosis
2 (11%) Fibrosis
10 (22%) Normal A
5 (26%) Normal
A
In two participants the tympanic membrane was not seen adequately. However, tympanometry was normal.
Fig. 2. Otological findings at 10-year follow-up among 32 participants (64 ears) who had had chronic suppurative otitis media in the initial study (CSOM-IS).
Tables 3 and 4 show the results of the univariate risk factor analyses for development of CSOM. Several risk factors were associated with the development of CSOM but only mother’s history of CSOM and school degree as a continuous variable reached significance. In the multivariate model these variables remained significant: mother’s history of CSOM OR 2.55 (95% CI 1.14–5.70; p = 0.02) and a low level of schooling OR 1.57 (95% CI 1.03–2.40; p = 0.04). Similar analyses for persistence of CSOM (failure to heal) among the 45 children with CSOM showed, that boys were more likely to have persistent CSOM compared with girls (OR 5.46; 95% CI 1.47– 20.37; p = 0.01). Age at follow-up did not influence healing rates. The previously mentioned potential risk factors were all tested on failure to heal but no further associations were found. Neither age at debut or at follow-up nor any of the other risk factors above were associated with healing. A second multivariate model to directly estimate adjusted risk ratios and absolute risks, using the previous mentioned risk factors showed that the RR for girls developing CSOM compared to boys was 0.9 (95% CI 0.53–1.51; p = 0.68). The RR for a child to develop CSOM if the mother had a history of CSOM compared with no history of CSOM was 1.9 (95% CI 1.04–3.22; p = 0.04) and if the mother had low levels of schooling compared with high levels the RR was 1.37 (95% CI 1–1.85; p = 0.05). The absolute risk of getting CSOM was 21.3% (95% CI 14.8–30.6) for boys and 17.2% (95% CI 11.8–25.1) for girls. The absolute risk of getting CSOM being a boy with a mother who had a history of CSOM was 35.4% (95% CI 21– 59.5), if the mother furthermore had low schooling, the risk increased to 45.4% (95% CI 26.5–77.7). With the same combina-
tions of risk factors for girls the absolute risk was 30.7% (95% CI 17.8–53.10) and 40.76% (95% CI 22.3–74.5), respectively. The RR for persistence of CSOM was 0.4 for girls compared with boys (95% CI 0.17–0.79; p = 0.01). 4. Discussion This is to our knowledge the first study looking at long-term outcome and spontaneous healing after CSOM in a cohort design, not only in Greenland, but also globally. The overall participation rate at follow-up in 2008 was high, up to 86% in Sisimiut, where the majority of participants came from. We did not find demographic differences between participants and children lost to follow-up or non-participants that could introduce bias to the study, and the groups did not differ in respect to initial occurrence of CSOM. Therefore, we consider our findings representative of the background population. However, 36 percent of the original cohort had moved to other towns and the high mobility among Greenlandic families makes it challenging to follow-up on child cohorts over time. We found a surprisingly high proportion of spontaneous healing at 75%, which is encouraging but should be interpreted with some caution. The initial study definition of CSOM included dry perforation of 14 days duration as a minimum. Some of these cases initially diagnosed as CSOM might, at the time of observation, have been cases of acute otitis media with perforation in healing, which may take longer than 14 days [13]. Although the majority were of substantially longer duration than 14 days [11], we cannot exclude that a few cases might have been cases of healing after severe acute otitis media thus leading to an overestimation of the proportion of healing after CSOM. According to a survey on CSOM in Greenland in 1994, 19% had healed at follow-up after one year [15]. In the present study boys and girls seemed to have an equal risk of getting CSOM, but girls had better chances of spontaneous healing. This gender difference in healing of the tympanic membrane has, to our knowledge, not been described before. A few studies have described boys having a higher risk of developing severe courses of otitis media [16,17], and otitis media with effusion is also found with higher incidences among boys [18,19]. Part of the explanation could be found in the larger pneumatization area of the mastoid among girls [20], but a difference due to chance or other factors cannot be excluded. A gender difference was found for lower respiratory tract infections in this study cohort, but not for upper respiratory tract infections [9]. Differences in endocrine– immune interactions may play a role in male susceptibility to infections [21]. We found mother’s level of schooling and history of CSOM as independent risk factors for CSOM development among their children, confirming findings risk factor analysis in the initial study
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
952
Table 3 Child-related and mother-related risk factors for the development of chronic suppurative otitis media (CSOM) at any time in childhood. A follow-up study of 236 children aged 11–15 years in Nuuk and Sisimiut, Greenland. Univariate risk factor analyses.a Risk factors (no of observations)
(No with CSOM/without CSOM)
OR
Male (22/106) Female (23/85)
1 0.77
Greenlandic (42/160) Danish/mixed (3/26)
1 0.44
8th degree (6/15) 9th–10th degree (16/49) 11th–12th degree (21/101) High school (2/22)
4.28 3.51 2.26 1
Long/medium theoretical (26/95) Short/practical education of medium length (2/14) Short courses/education not completed (2/15) None (12/52) Student (3/13)
1 0.51 0.46 0.81 0.78
No (32/163) Yes (12/24)
1 2.52
No (32/163) Yes (8/21)
1 1.92
6 months (18/88) >6 months (19/85)
1 0.97
95% CI
Gender (n = 236)
0.42 0.40–1.47
Ethnicity (n = 231)
0.2 0.13–1.54
Mother’s schooling (n = 232)b
0.25/0.04c 0.76–24.22 0.74–16.62 0.49–10.36
Mother’s education (n = 234)c
0.8 0.11–2.39 0.10–2.17 0.38–1.75 0.21–2.91
Mother’s history of COM (n = 231)
0.02 1.14–5.56
Siblings’ history of COM (n = 224)
0.16 0.78–4.73
Breast feeding (n = 210)
a b c
p valueb
0.93 0.47–1.98
All analyses adjusted for gender. Test for homogeneity. Test for trend.
Table 4 Environment-related risk factors for the development of chronic suppurative otitis media (CSOM) at any time in childhood. A follow-up study of 236 children aged 11–15 years in Nuuk and Sisimiut, Greenland. Univariate risk factor analyses.a Risk factors (no of observations)
(No with CSOM/without CSOM)
OR
4 (15/52) 1–3 (6/41) 5 (15/59) 6+ (9/39)
1 0.51 0.9 0.79
0–1 (5/24) 2 (19/98) 3+ (21/60)
1 0.98 1.77
No (9/38) Yes (36/153)
1 0.98
0 (9/38) 1 (7/45) 2+ (29/108)
1 0.63 1.13
6 months (8/14) 7–12 months (2/13) >12 months (4/29)
1 0.25 0.23
6 months (1/7) 7–24 months (8/44) 25–48 months (22/75) >48 months (11/45)
1 1.22 1.97 1.68
95% CI
No. of persons in household (n = 236)
0.63 0.18–1.43 0.40–2.02 0.31–2.00
No. of persons sleeping in same room (n = 227)
0.22 0.33–2.89 0.59–5.26
Smoking at home (n = 236)
0.97 0.44–2.22
No. of smokers at home (n = 236)
0.45 0.22–1.87 0.49–2.61
Time in daycare (n = 70)
0.07 0.04–1.43 0.06–0.91
Time in child-care center (n = 213)
a b
All analyses adjusted for gender. Test for homogeneity.
p valueb
0.72 0.13–11.40 0.23–16.96 0.19–15.18
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
among children below 4 years of age [11]. This original risk factor analysis also found that the use of childcare centers and being Inuit increased the risk of CSOM, which was also found in the present study, although not statistically significant. A previous finding of an association between CSOM and smoking in the home was not confirmed. In spite of the possibility to find more cases during a longterm follow-up study compared with a cross sectional study, the number of participants and cases in the present study is still low limiting the statistical power of the multivariate risk factor analysis. As 45 of the 236 participants in follow-up had a verified episode of CSOM, nearly every 5th child has had CSOM at some point in childhood. At follow-up we found 21 participants with present disease or who had undergone surgery for CSOM. Thirteen of these children (62%) were diagnosed with CSOM after the initial study. This large proportion with an apparent late debut may partly be explained by the fact that they left the initial study between 2 and 3 years of age, rather than an unusual late disease debut. Secondly some of the children could have had severe otitis media or CSOM and have healed before inclusion in the initial study, subsequently reperforated at some time point before follow-up. However, as indicated in the cumulative incidence we believe the majority of CSOM debut are before 4 years of age, which also concurs with findings in the medical records and the general experience with the disease in Greenland. It is possible that other participants might have had CSOM debut after termination of the first study, but healed spontaneously before the follow-up. Hence, we consider the number of detected CSOM cases in this cohort as a conservative estimate of the amount of individuals affected by the disease through childhood in the background population. The most frequent pathologic findings at otomicroscopy in this study were atrophy and myringosclerosis which are also frequent at long-term follow-up among children with otitis media in Western countries [22]. The prevalence of both pathologies was 17% of all ears in this study, which is comparable with results from Danish follow-up studies. However, in the Danish studies the pathologic changes are most often caused by tympanostomy tube placement or myringotomy [23]. In this Greenlandic cohort the tympanic membrane scarring is predominantly caused by former otitis media as the use of tympanostomy tubes and myringotomy has not been a part of otitis media treatment. 4.1. Clinical implications The initial examinations of this cohort in 1996–98 and the follow-up in 2008 show that most cases of CSOM have their debut at early childhood and at least before school age. Four participants with dry perforations at follow-up were neither themselves nor were their parents aware of ear problems, which indicates that CSOM can be clinically silent after otitis media early in childhood and emphasizes the importance of routine screening of hearing loss before school start. Only 20% of children with CSOM had ever received antibiotic treatment for any type of otitis media, according to the recordings in the medical files. This could partly be explained by the widespread use of antiseptic irrigation of draining ears with an acetic acid solution as the sole treatment for CSOM. However, it might also reflect the fact that otorrhea is so common in this population that some parents see it as a ‘‘normal’’ part of childhood and do not seek treatment if the children have no other symptoms than drainage from the ear, which explains why none of the four participants with draining ears at follow-up were receiving any treatment. This acceptance of the disease has been observed in other countries with high prevalence of CSOM [1]. There seems to be a diagnostic and treatment potential in the Greenlandic primary health care sector to reduce the risk of hearing loss. It is expected that the degree to which hearing is compromised is proportional to
953
the damage of structures in the middle ear. Hence a focus on reducing numbers and lengths of episodes with otorrhea in CSOM is important [24]. The use of ototopical antibiotics should be considered in any case of prolonged otorrhea [25,26]. The fact that Greenlandic boys and girls, with mothers having a history of CSOM and low education, have a 45% and 41% risk of developing CSOM, respectively, should be taken into account when deciding which children should receive antibiotic treatment for acute otitis media. Age has previously been seen as one of the most important factors determining successful outcome for surgical management of CSOM, but a number of recent studies contradict this [27,28]. However, in settings where the surgical capacity is limited, the possibility of spontaneous healing should be considered when selecting young children for surgery. The change in CSOM prevalence from early childhood to adolescence due to spontaneous healing indicates that some precaution should be made when comparing CSOM prevalences between countries and regions, based on different age groups varying from preschool children to adults [1]. In conclusion we found a surprisingly high rate of spontaneous healing from CSOM. However, the prevalence of CSOM among school-age children (11–15 years) in Greenland remained high. Only maternal history of CSOM and education proved in this study to be risk factors for overall CSOM development. Girls healed significantly better than boys, but no other factors were identified as determinants for such. Increased focus on early diagnosis and treatment is essential to diminish the effects of hearing loss during early speech and cognitive development. Information about the disease and treatment possibilities to primary health care providers as well as parents could reduce the risk of CSOMrelated hearing loss in school. Conflict of interest None. Acknowledgements We would like in particular to thank the participating children and their families of Sisimiut and Nuuk as well as the staff at Sisimiut Health Center, Sisimiut, Dronning Ingrids Hospital, Nuuk and The Primary Health Care Center, Nuuk. Funding for this study was provided by The Commission for Scientific Research in Greenland (KVUG). References [1] WHO, Chronic Suppurative Otitis Media. Burden of Illness and Management Options, WHO, Geneva, Switzerland, 2004. [2] WHO, Report of a WHO/CIBA Foundation Workshop 1996, Prevention of Hearing Impairment from Chronic Otitis Media, 2000. [3] H. Winskel, The effects of an early history of otitis media on children’s language and literacy skill development, Br. J. Educ. Psychol. 76 (2006) 727–744. [4] F. Olatoke, F.E. Ologe, C.C. Nwawolo, M.J. Saka, The prevalence of hearing loss among schoolchildren with chronic suppurative otitis media in Nigeria, and its effect on academic performance, Ear. Nose. Throat J. 87 (2008) E19. [5] A. Golz, S.T. Westerman, L.M. Westerman, D.A. Gilbert, A. Netzer, Does otitis media in early childhood affect reading performance in later school years? Otolaryngol. Head Neck Surg. 134 (2006) 936–939. [6] M. Verhoeff, E.L. van der Veen, M.M. Rovers, E.A. Sanders, A.G. Schilder, Chronic suppurative otitis media: a review, Int. J. Pediatr. Otorhinolaryngol. 70 (2006) 1–12. [7] P. Homoe, G. Nikoghosyan, C. Siim, P. Bretlau, Hearing outcomes after mobile ear surgery for chronic otitis media in Greenland, Int. J. Circumpolar Health 67 (2008) 452–460. [8] P. Homoe, Otitis media in Greenland. Studies on historical, epidemiological, microbiological, and immunological aspects, Int. J. Circumpolar Health 60 (Suppl. 2) (2001) 1–54. [9] A. Koch, K. Molbak, P. Homoe, P. Sorensen, T. Hjuler, M.E. Olesen, et al., Risk factors for acute respiratory tract infections in young Greenlandic children, Am. J. Epidemiol. 158 (2003) 374–384.
954
R.G. Jensen et al. / International Journal of Pediatric Otorhinolaryngology 75 (2011) 948–954
[10] P. Homoe, R.B. Christensen, P. Bretlau, Acute otitis media and sociomedical risk factors among unselected children in Greenland, Int. J. Pediatr. Otorhinolaryngol. 49 (1999) 37–52. [11] A. Koch, P. Homoe, C. Pipper, T. Hjuler, M. Melbye, Chronic suppurative otitis media in a birth cohort of children in greenland: population-based study of incidence and risk factors, Pediatr. Infect. Dis. J. 30 (2011) 25–29. [12] WHO, WHO Ear and Hearing Disorders Survey. Protocol for a Population-based Survey of Prevalence and Causes of Deafness and Hearing Impairment and Other Ear Diseases, 1999. [13] G. Berger, Nature of spontaneous tympanic membrane perforation in acute otitis media in children, J. Laryngol. Otol. 103 (1989) 1150–1153. [14] J.J. Haapaniemi, Immittance findings in school-aged children, Ear Hear. 17 (1996) 19–27. [15] P. Homoe, R.B. Christensen, P. Bretlau, Prevalence of otitis media in a survey of 591 unselected Greenlandic children, Int. J. Pediatr. Otorhinolaryngol. 36 (1996) 215–230. [16] M. Hotomi, N. Yamanaka, T. Samukawa, M. Suzumot, A. Sakai, J. Shimada, et al., Treatment and outcome of severe and non-severe acute otitis media, Eur. J. Pediatr. 164 (2005) 3–8. [17] E. Vartiainen, Changes in the clinical presentation of chronic otitis media from the 1970s to the 1990s, J. Laryngol. Otol. 112 (1998) 1034–1037. [18] A.M. Hardy, M.G. Fowler, Child care arrangements and repeated ear infections in young children, Am. J. Public Health 83 (1993) 1321–1325. [19] J.L. Paradise, H.E. Rockette, D.K. Colborn, B.S. Bernard, C.G. Smith, M. Kurs-Lasky, et al., Otitis media in 2253 Pittsburgh-area infants: prevalence and risk factors during the first two years of life, Pediatrics 99 (1997) 318–333.
[20] M. Tos, S.E. Stangerup, Secretory otitis and pneumatization of the mastoid process: sexual differences in the size of mastoid cell system, Am. J. Otolaryngol. 6 (1985) 199–205. [21] S.L. Klein, The effects of hormones on sex differences in infection: from genes to behavior, Neurosci. Biobehav. Rev. 24 (2000) 627–638. [22] P. Caye-Thomasen, S.E. Stangerup, G. Jorgensen, D. Drozdziewic, P. Bonding, M. Tos, Myringotomy versus ventilation tubes in secretory otitis media: eardrum pathology, hearing, and eustachian tube function 25 years after treatment, Otol. Neurotol. 29 (2008) 649–657. [23] S.E. Stangerup, M. Tos, R. Arnesen, P. Larsen, A cohort study of point prevalence of eardrum pathology in children and teenagers from age 5 to age 16, Eur. Arch. Otorhinolaryngol. 251 (1994) 399–403. [24] L.F. Silveira Netto, S.S. da Costa, P. Sleifer, M.E. Braga, The impact of chronic suppurative otitis media on children’s and teenagers’ hearing, Int. J. Pediatr. Otorhinolaryngol. 73 (2009) 1751–1756. [25] C.A. Macfadyen, J.M. Acuin, C. Gamble, Topical antibiotics without steroids for chronically discharging ears with underlying eardrum perforations, Cochrane Database Syst. Rev. (2005), CD004618. [26] C.A. Macfadyen, J.M. Acuin, C. Gamble, Systemic antibiotics versus topical treatments for chronically discharging ears with underlying eardrum perforations, Cochrane Database Syst. Rev. (2006), CD005608. [27] A.C. Lin, A.H. Messner, Pediatric tympanoplasty: factors affecting success, Curr. Opin. Otolaryngol. Head Neck Surg. 16 (2008) 64–68. [28] S. Sarkar, A. Roychoudhury, B.K. Roychaudhuri, Tympanoplasty in children, Eur. Arch. Otorhinolaryngol. 266 (2009) 627–633.