 
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
26-July-2002, Vol 115 No 1158
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Chlamydia
trachomatis prevalence and sexual behaviour in Christchurch high school
students
Paul Corwin, Gillian Abel, J. Elisabeth Wells, Edward
Coughlan, Sue Bagshaw, Margaret Sutherland, Libby Plumridge.
Chlamydia
trachomatis is the most common sexually transmitted bacterial pathogen in
New Zealand.1 It can cause symptomatic
urethritis and epididymitis in males and cervicitis, salpingitis and
endometritis in females. Unfortunately over 90% of infections in males and
females are asymptomatic and therefore often go
untreated.2 Untreated women can go on to
develop pelvic inflammatory disease and subsequent chronic pelvic pain,
infertility or ectopic pregnancy. Untreated infection in men can lead to
epididymitis. Because of the long-term complications, it is estimated in the
United States that C. trachomatis
infections cost approximately $2.4 billion each
year.3
The prevalence of C.
trachomatis in New Zealand is unknown. The only information currently
collected systematically is from sexual health clinics and gives the prevalence
of C. trachomatis among clinic
attendees. In 2000 the overall prevalence of
C. trachomatis in clinic attendees aged
15-19 years was 6.3%.4 Some indication of the
cumulative incidence of C. trachomatis
in the community is available from past cohort studies. The self-reported
cumulative incidence of C. trachomatis
infection in 21 year old members of the Dunedin Multidisciplinary Health and
Development Study (DMHDS) was 2.4% for males and 9.0% for
females.5 The Christchurch Health and
Development Study found that the self-reported cumulative incidence of sexually
transmitted diseases (STDs) in females at age 18 was
5.8%.6 The vast majority of these STDs were
C. trachomatis (David Fergusson,
Department of Psychological Medicine, Christchurch School of Medicine and Health
Sciences, personal communication). A 1999 estimate of
C. trachomatis incidence from community
laboratory positivity data in the Waikato and Bay of Plenty showed a 1.9%
C. trachomatis incidence in males and
females aged 15-19.7 Such studies may
underestimate prevalence, however. Community based studies indicate a higher
prevalence. One study in 1999 in New Orleans urban high schools found a baseline
prevalence of 11.5% in females and 6.2% in
males.2 A recent United Kingdom study found a
3.0% prevalence of C. trachomatis in
sexually active women aged under 25 years and a 2.7% prevalence in
men.8
The need for community based studies in New Zealand prompted
this study of prevalence, risk factors and help-seeking behaviour.
MethodsAll 26 public and private high
schools in the Christchurch urban area were approached by letter and telephone
and visited by one of the researchers to request participation. Information
about the study and a questionnaire was provided to principals and school
boards. Approval for the study was received from the Canterbury Ethics
Committee. To obviate the need for parental consent only students aged sixteen
or older in Years 12 and 13 were invited to participate. It was thought that
about 50% of these students would be sexually active based on data from the
DMHDS.9
Although it was planned to sample by classroom within participating schools,
some schools asked us to approach the students in senior assemblies. Each of the
participating schools was stratified into Year 12 and Year 13 classes. 50% of
classes from within each stratum were randomly selected for participation. The
cluster design of the study was factored into sample size calculations by
assuming a design effect of 1.5. It was expected that a sample of 1500 Year 12
and 13 students would yield about 750 sexually active students. We expected the
C. trachomatis rate to be between 2%
and 10% in sexually active students. Taking into account the design of the
study, the confidence intervals were predicted. It was calculated that if the
observed prevalence was 2.0% amongst sexually active participants, a sample of
750 students would result in a 95% confidence interval of between 1.0% and
3.6%.
Students at participating schools were either seen in a classroom setting or in larger groups in the school hall and students completed their questionnaires there under the supervision of the researchers. Data collection took place between June and September 2001. The students were supplied with an information sheet about the study and about C. trachomatis. The study was also verbally explained to them by a member of the research team. Participation was voluntary and written consent was obtained. Students were asked both to complete a questionnaire and to provide a urine sample. They were informed that urine would be tested only from those who reported that they were sexually active. If they had passed urine in the past hour they were asked to provide a urine sample later in the day. The questionnaire asked about demographic details including age, sex, gender and ethnicity. They were also asked to indicate whether they had ever had sexual intercourse and to answer questions regarding sexual behaviour. The questionnaire was piloted with two groups of young people and revised in accordance with their feedback. All questionnaires and urine samples were given an encrypted label. The research co-ordinator had a master list matching names, questionnaires and urine samples which was used for notification of results. Samples were transferred directly to the laboratory in insulated containers and handled according to the manufacturer's recommendations. The Cobas Amplicor (Roche Molecular Diagnostics Branchburg, NJ, USA) test based on a polymerase chain reaction system was used to detect C. trachomatis according to the manufacturer's instructions. About one week after the sample and data collection, each participating student was seen in private by a member of the research team and given a sealed envelope containing their results. Students with a positive test were given the opportunity to debrief and were directed to a range of free treatment providers for themselves and their sexual contacts. Students with negative tests and those whose urine samples were not tested were given an envelope containing basic information on the prevention of sexually transmitted diseases. Data were entered into EpiInfo 2000 version 1.1.2. After completion of data entry, a check on 10% of data from respondents was carried out which revealed an error rate of 0.2%. To take account of the cluster design of the study, analyses were carried out using the Complex Sample program in EpiInfo and SUDAAN 8.0. In addition, a weighting was taken into account in the analysis as one of the schools required that all, instead of half of their Year 13 students, participate. ResultsTwelve of the fifteen public high
schools and five of the eleven private schools agreed to take part in the study.
Of the twelve public schools eight were co-educational, two were girls’
schools and two boys’ schools. The five private schools consisted of two
co-educational, one girls’ and two boys’ schools. Over half the
participating students came from socio-economically advantaged schools with
decile ratings of between eight and ten. Although there appeared to be 2300
students on the school rolls for the classes selected only 1582 students were
present when the study was explained due to a number of factors including
absenteeism, outdated school rolls, students away on field trips and failure to
arrive at the school hall to be informed of the study. Of these 1582 students
1136 consented to take part in the study giving a participation rate of 72%.
Three questionnaires were excluded from further analysis as there were
inconsistencies in answering, leaving a total of 1133 participants.
Table 1. Description of the sample.
Of the sexually active students, 84% provided urine samples.
Non-suppliers were more likely to have been diagnosed with a previous sexually
transmitted infection (8.2%) than the suppliers (1.1%). The mean age of
participants was 16.7 years and there were slightly more male than female
participants. The ethnic distribution closely resembled the overall distribution
in the Christchurch area. 49% of students indicated past sexual intercourse
(Table 1).
Participants were asked about previous diagnosis of a
sexually transmitted disease and 2.3% of sexually active participants reported
having been diagnosed with a sexually transmitted disease at some stage. No male
participants and 2.4% of sexually active female participants indicated a
previous diagnosis of C. trachomatis.
The point prevalence of C. trachomatis
in the study population was 2.0% for all sexually active participants who
provided urine specimens. The rate of infection was similar for males and
females with 1.8% of males and 2.3% of females testing positive (Table
2).
Table 2. Sexually transmitted infection amongst
sexually active participants.
∗ Calculated
allowing for clustering in classes. †
Calculated using exact
formula.32
‡
Prevalence amongst all sexually active participants who provided urine.
STI = sexually transmitted infections
39% of sexually active participants indicated that they had
had only one partner in their lifetime and 53% indicated only one partner in the
last twelve months. 44% of sexually active participants reported that they
always used condoms (51% of males and 39% of females) and 63% reported using
condoms the last time that they had had sexual intercourse (69% of males and 57%
of females) (Table 3).
DiscussionAlthough there have been a few
reports from school clinics and one report from females in a single-sex school
on C. trachomatis rates we are only
aware of two other school based prevalence
studies.10 The point prevalence of
C. trachomatis infection in our study
was considerably below that reported from a New Orleans school-based
study2 which looked at rates in urban, public
high schools in very deprived areas. Our rates were slightly higher than the
1.4% reported from a recent study in Belgian female high school
students.11 A higher prevalence of
C. trachomatis was expected in this
Christchurch study as incidence rates of 1.9% in the Waikato and 2.4% locally
have been reported from laboratory data in males and females aged 15-19 years
(personal communication from Medlab South and Southern Community Laboratories
for Christchurch data).7 As it was assumed that
as the population denominator of 15-19 year olds would have included individuals
who were not sexually active, the prevalence in sexually active students within
this sample would have been higher than the laboratory positivity data.
Countering this would have been the fact that some of those getting laboratory
tests would have been symptomatic and our sample had a lower mean age than the
15-19 year old laboratory groups reported.
Several factors in this study would have had the effect of
lowering prevalence. The 96 sexually active participants who did not provide a
urine sample were more likely to have had a previously diagnosed sexually
transmitted infection which may have had a lowering effect on the overall
C. trachomatis prevalence rate. Some of
these students reported verbally to the researchers that they had recently been
treated for C. trachomatis and did not
want to provide a urine sample. That half of participating students reported
past sexual intercourse was to be expected as similar figures have been reported
in other studies of people of this age
group.9,11
Table 3. Sexual behaviour of sexually active participants.
∗ Calculated
allowing for clustering in classes.
When looking at risk factors for acquiring a sexually
transmitted infection, the results obtained in this study are comparable to
those in other studies. The DMHDS reported a mean number of partners in the
previous twelve months for males of 1.7 and for females 1.5 at age 18
years.12 Although the mean age of the
participants in this study was lower than the Dunedin study, partner numbers
were slightly higher with a mean number of partners in the previous twelve
months of 1.9 for males and 1.8 for females.
Condom use in this study was similar to the DMHDS, which
reported 37.6% of sexually active females and 48% of sexually active males using
condoms “usually or always”.12
Condom use on the last occasion of sex was also similar to an Australian study
of school students, which reported 71.5% of males and 53.4% of females using
condoms at last sex.13 Condom use among females
is consistently reported as being lower than
males.5,14,15 This is most likely due to girls
often having older partners and relying on oral contraceptives to prevent
pregnancy, which may pose more of a threat than sexually transmitted
infections.13,16-18
Nevertheless, despite such behavioural differences the
infection rate in this study was similar in males and females. This is in
contrast to the results from the New Orleans study and underlines the importance
of focussing control efforts on both males and females.
Disease prevalence has a major bearing on potential benefits
from population screening. Does the 2% prevalence of
C. trachomatis in sexually active
Christchurch students warrant formal screening? All but two of the ten Wilson
and Junger criteria for successful screening programmes apply to screening for
C.
trachomatis.19 Two criteria that are
impossible to meet given our current level of knowledge about
C. trachomatis are the need to
understand the natural history of the condition and the need to balance the
costs of screening in relation to possible expenditure on medical care as a
whole. The natural history of C.
trachomatis is not well understood. It is not known what proportion of
cases of untreated C. trachomatis
infections lead on to the known sequelae of pelvic inflammatory disease, ectopic
pregnancy and infertility. Moreover it is not clear how long untreated
C. trachomatis infection remains in the
genital tract.20 Estimates of what proportion
of untreated infections go on to cause pelvic inflammatory disease range from
10% to 80%; estimates of what percentage of this untreated cohort goes on to
infertility range from 8% to 20%.21-24 The
Wilson and Junger criterion for economic 'balance' in relation to possible
expenditure on medical care as a whole is an even harder goal to meet in regards
to screening for C. trachomatis.
Although the costs of ectopic pregnancy or infertility treatment can be
quantified, what price can be put on infertility or chronic pelvic pain? It is
unsurprising that cost-benefit models and analyses for
C. trachomatis screening reach very
different conclusions.22-26
Given the wide variation of estimates for
C. trachomatis complications the
prevalence of C. trachomatis becomes
less of a determinant of screening costs. Moreover, the cost of screening for
C. trachomatis in lower prevalence (2%)
populations is reduced by 59% by using pooled urine and DNA amplification
techniques.27 It should be noted that DNA
amplification tests have much greater sensitivity than previously available
tests for C. trachomatis. The Cobas
Amplicor test we used has been shown to have sensitivities ranging from 89%-93%
for both urine and endocervical or urethral swabs and specificities of
99%.28,29
Although one randomised controlled trial of screening has
given support to screening, it has been criticised on methodological
grounds.30,31 It would be extremely difficult
on ethical and practical grounds to repeat a randomised controlled trial of
screening for C. trachomatis and
screening decisions must be made on inferences from existing evidence.
The similar rates of C.
trachomatis infection in males and females in our study suggest that
opportunistic screening should be offered to both males and females at risk. It
is also interesting that in our study although 2.4% of females had previously
been diagnosed with C. trachomatis no
males had, suggesting that males are not being tested either opportunistically
or as contacts of females with C.
trachomatis. We hope our results will help inform the current debate on
how best to tackle sexually transmitted diseases in New Zealand.
Author Information:
Paul Corwin, Senior Lecturer; Gillian Abel, Junior Research Fellow; J.
Elisabeth Wells, Senior Research Fellow, Department of Public Health and General
Practice, Christchurch School of Medicine; Edward Coughlan, Clinical Director,
Christchurch Sexual Health Centre; Sue Bagshaw, Senior Doctor, Youth Health
Centre, Christchurch; Margaret Sutherland, Research Co-ordinator; Libby
Plumridge, Senior Lecturer/Senior Research Fellow, Department of Public Health
and General Practice, Christchurch School of Medicine, Christchurch.
Acknowledgements: We
thank Dr Helen Moriarty for advice on screening costs and Alison Parsons and
other department members for assistance in administering the questionnaire,
urine sample collection and distribution of results. The study was funded by the
Canterbury Medical Research Foundation.
Correspondence: Dr P
Corwin, Department of Public Health and General Practice, Christchurch School of
Medicine, PO Box 4345, Christchurch. Email: paul.corwin@chmeds.ac.nz.
References:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|