with students after they had answered the questionnaire, and full-scale testing of the questionnaire
in classroom settings27. The questions about perceived outcome of the dialogue with the school
nurse were specific for the Danish study. These questions were pilot tested twice before the final
data collection in order to ensure the students’ correct interpretation of the wording.
Measurements
The independent variable was frequency of four common symptoms: headache, stomach-ache,
difficulties getting to sleep and nervousness. The response key was recoded into three levels
(seldom or never, monthly, weekly or more often) and we excluded students with missing data.
The dependent variables were medicine use during the past month for the above four symptoms,
the response key was dichotomized into yes and no and we included the few missing
responses in the no-category. The analyses included three co-variates, sex, age group (grade)
and re-visit to the school nurse measured by the item ‘Did you visit the school nurse again’. The
response code was dichotomized into yes (yes right away + yes after some time) vs. no (no and
missing response).
Statistical procedures
We performed separate logistic regression analyses for each of the four kinds of medicine use. At
first, all analyses were carried out separately for boys and girls but since the pattern of associations
was similar among boys and girls we decided to combine the two sexes into the same analyses.
In each of the analyses we adjusted for sex, age group, frequency of the symptom for which
the medicine was taken, and re-visit to the school nurse. Further, we stratified the analyses into
students who did visit and did not re-visit the school nurse. Associations were reported as
odds ratios (OR) with 95 per cent confidence interval (CI). The final analyses included 4,911
students.
Ethical issues
There is no formal agency for ethical assessment and approval of questionnaire-based population
studies in Denmark. We have therefore obtained an ethical approval from each separate school’s
board of parents, school director, and students’ council. Further, we have informed the students –
orally and in writing – that participation was voluntary and anonymous.
Results
Table 1 shows the distribution of students according to age group, frequency of headache, stomachache,
difficulties getting to sleep, nervousness and re-visit to the school nurse. More girls than boys
reported medicine use for each of the four complaints and more girls than boys reported medicine
use for headache and stomach-ache. The percentage of boys who re-visited the school nurse was
9.3 per cent and of girls was 7.9 per cent (p = 0.0699).
Table 2 shows the results from the multivariate analyses. For each kind of medicine use there is
a strong and graded association with the complaint for which the medicine is taken. Further, students
who re-visited the school nurse had higher odds for medicine use for each of the four
complaints.
Table 3 shows the results from the stratified multivariate analyses. The two columns show students
who did not and who did re-visit the school nurse, respectively. Each cell in the table shows
the odds ratio values for medicine use for one of the four health complaints (headache, stomachache,
difficulties getting to sleep and nervousness). Each cell in the table shows a strong and
graded association between frequency of the complaint and medicine use for this complaint. This
is the case in both the left and right columns of the table. Further, the association is systematically
steeper in the left column, that is, among students who did not visit the school nurse again. The
differences in steepness are illustrated in Figure 1.
Discussion
Findings and their interpretation
The most important observations from this study are (1) that medicine use for common complaints
is strongly associated with the prevalence of the complaint for which the medicine is used, (2) that
students who re-visit the school nurse have higher odds for medicine use, and (3) that the association
between complaints and medicine use is less steep among students who re-visited the school nurse.
What it means should be interpreted with caution since we do not know what the re-visit related to.
One possible interpretation is that the students who visit the school nurse again are exposed to
many stressors such as bullying22 or health complaints23; stressors which prompt both re-visit to the
school nurse and medicine use.
Another possible interpretation is that the re-visit in itself is a buffer of the effect of complaints
on medicine use. This is supported by Larsson and Zaluha18 who found that Swedish students did
not contact the school nurse to get medication but rather for supportive discussion. This interpretation
is also supported by the findings by Schneider and collegues26. They report that only a few
visiting students were interested in medicine. Further support for this interpretation is the study by
Lightfoot and Bines12 who report that students use the school nurse as a confidant in drop-in visits,
that is, re-visits to the school nurse may contain a supportive discussion rather than medication. We
can only speculate about the content of the supportive discussion, but according to Reutzel and
Patel16 the school nurses seek advice among each other when there is no clear guideline related to
OTC medicine.
In Denmark, school nurses do not administer medicine. We do not know if the students get
medicine from other sources at school such as described by McCarthy17, where both school nurses
and school secretaries administer medicine. A recent study suggests, however, that the vast majority
of young adolescents who use medicine get it at home, not at school11.
A few other studies report reasons for students to make contact with the school nurse. Kay et al25
found alcohol, drugs and emotional concerns as reasons for visits. Sweeney and Sweeney24 mention
that certain needs of the students are being met. These needs could be vague symptoms or
complaints such as headache/stomach-ache, dizziness, chest pain, sore/painful limbs or in general
not feeling well. Barnes et al13 mention psychological concerns as the reason for visits and
Schneider et al26 found that visits were not explained by any specific symptoms. The role of the
school nurse as a confidant, as mentioned by Lightfoot and Bines12, is also supported by Noddings’
theory on caring as a relation rather than a set of specific behaviours as mentioned by Borup and
Holstein23. Noddings argues that we should teach students to care by creating a caring relation with
them, based on a dialogue approach. The aim of the health dialogue is to support the students in
making healthy choices and we argue that appropiate use of medicine is a healthy choice for a
teenager.
Strengths and limitations of the study
This study is a cross-sectorial study and therefore it is not possible to draw causal conclusions
but only to report associations. Nevertheless, we believe that a re-visit to the school nurse’s
office would result in less – and more appropriate – medicine use. Another limitation of the study
is that we do not know what happened during the visit: what issues were discussed, whether the
student received advice, and whether the student had a supportive dialogue with the school
nurse.
The large and representative study population and a high participation rate are strengths of this
study. The majority of the students were familiar with school health services and the concept of
health dialogue. The pilot test prior to the study showed that the students understood the items of
perceived outcome. The validity of the items on complaints has been tested and confirmed by
means of qualitative interviews29. The items about medicine use have also been tested and confirmed
by means of qualitative interviews30 and by a study which showed high agreement between
students’ and parents’ reports of the children’s medicine use.