 
|
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Journal of the New Zealand Medical Association,
05-May-2006, Vol 119 No 1233
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Antibiotic use for upper respiratory tract infections before
and after a education campaign as reported by general practitioners in New
Zealand
Linda Sung, Justine Arroll, Bruce Arroll, Felicity
Goodyear-Smith, Ngaire Kerse, Pauline Norris
Many patients presenting to their general practitioners
(GPs) still receive antibiotics regardless of efficacy.1,2 Indeed, antibiotics
are considered to be over-prescribed.3 A USA study found that the antibiotic
prescribing rates for upper respiratory tract infections (URTIs) may be as high
as 63% when all drug information is analysed. Another study has shown that
prescribing rates for URTIs of presumed viral aetiology ranged from 17 to 60% in
the UK and US, respectively.4
There is also a general trend toward increased use of
broad-spectrum agents. In the USA, for example, there has been an increasing
trend toward the use of broad-spectrum antimicrobials and decreasing rates of
narrower-spectrum antimicrobials from 1980–1992.5 In one study, the rate
of broad-spectrum used increased from 24% to 48% of antibiotic prescriptions in
adults (p<0.001), and from 23% to 40% in children (p<0.001).6
By 1998–1999, 22% of adult and 14% of paediatric
prescriptions for broad-spectrum antibiotics were for viral URTIs. Indeed,
physicians are increasingly turning to expensive, broad-spectrum agents, even
when there is little clinical rationale for their use.7
In 1999, PHARMAC (New Zealand’s Pharmaceutical
Management Agency which is responsible for nationwide funding of
pharmaceuticals) launched the Wise Use of
Antibiotics campaign aimed at reducing antibiotic use by educating the
public that antibiotics are ineffective against viruses. The campaign involved
posters in family practice waiting-rooms and pharmacies, leaflets given to
patients in pharmacies and primary health care surgeries, plus small group
training for GPs.8 The campaign was endorsed by the Royal New Zealand College of
General Practitioners.
As a result, PHARMAC reported a decrease in the national
antibiotic drug bill from $NZ36 million in 1996 to 14.5 million in 2003. This
reduction is a combination of decreased volume (one-third) and price
(two-thirds) of antibiotics prescribed. Additionally, from 1995 to 2002, there
was also a national reduction from 7% to 3.5% (p<0.05) in penicillin
resistance among pneumococci.9
The aim of our study was to determine any change in reported
antibiotic use for URTIs by GPs in the Auckland region between 1998 and 2002,
before and after the educational campaign.
MethodsOne hundred GPs were randomly selected from a list of
Auckland-based practitioners supplied by the local diagnostic laboratory in
1998. In 2002–3, 65 of the initial group were available to participate in
a subsequent interview. A further 35 were randomly selected to make the sample
to 100. GPs were contacted by telephone or fax and asked to participate in
research into primary care prescription of antibiotics for URTIs.
The total populations of GPs in the Auckland region is
approximately 1000, hence 100 represents 10% of the population. From previous
surveys, the authors have found statistically significant differences with such
a sample size, and 100 practitioners were within the resources of the study.
Questions asked included the conditions under which
they would prescribe antibiotics; their use of “as-needed” or
delayed prescriptions, and the specific antibiotics they would prescribe. Data
from the questionnaires was entered into a Microsoft Excel spreadsheet and
analysed using Stat-Sak and SPSS version 11 statistical packages with
Chi-squared statistical analysis.
ResultsFrom the initial randomised list of 179 GPs in the Auckland
region chosen in 1998, 16 were unable to be contacted at the number given.
Fifty-two GPs declined to participate, and a further 11 failed to call back
within the response period. Interviews were discontinued after 100 had been
conducted. This gave a response rate of 61%. Of 51 additional GPs approached in
2002–2003, 35 (69%) agreed to be interviewed.
Of the 65 GPs who were interviewed at both periods (1998 and
2002–2003), there was a decrease (from 82% in 1998 to 57% in 2003) in the
numbers of GPs agreeing that most patients who see a GP for an URTI expect to be
given antibiotics. Of the total of 100 GPs, 77% said they were less likely to
prescribe antibiotics for URTI; 2% more likely, and 21% felt
unchanged.
Similar percentage occurred in terms of patients wanting
antibiotics (71% less likely, 7% more likely, and 22% unchanged). Over a quarter
of the GPs believed that the change was due to education of both the doctors and
patients (12% doctor from education, 14% patient from education).
When asked what would encourage them to prescribe
antibiotics, there were some interesting changes over the period of time. The
direction of change was the same for both the original 65 GPs and the total of
100 GPs. We report the 65 GPs’ results which give greater statistical
power due to use of paired comparisons (McNemar’s test)—see Table 1,
and the 100 GPs’ results—see Table 2.
Table 1. Comparison of reasons for GPs prescribing
antibiotics in 1998 and 2002–3 (N=65)
Furthermore, there was a significant increase in the giving
of antibiotics to patients for the following reasons: smoker; symptoms of
sinusitis; older patients; patients expecting antibiotics; patients planning
imminent overseas trip; green-coloured sputum; and purulent nasal discharge. The
only significant reason for decrease in prescribing antibiotics was for otitis
media.
All others showed no significant change. However a third of
the doctors still reported they would prescribe antibiotics for fear that
patients would otherwise go to another GP, and two-thirds were willing to give
antibiotics to patients who expected and asked for them.
100 GPs interviewed in 2002-3 reported giving “as
needed” or “delayed” prescriptions as follows: always 50%,
often 25%, sometimes 36%, rarely 16%, and never 5%. In 1998, the proportions
were 0%, 13%, 52%, 30%, and 5% respectively. There was a statistically
significant increase in the number of GPs reporting that they often prescribed
“as needed’ or “delayed” prescriptions” between
the two studies (p=0.017). Thirty-nine percent of the GPs in 2002–3 said
they had increased the number of delayed prescriptions while 12% had decreased
them and 46% had made no change.
The most common first-line antibiotic used by the doctors
was still amoxicillin (28% vs 78%) followed by amoxicillin clavulantate (21% vs
4%) and tetracyclines (14% vs 5%) for 1998 and 2002–3 respectively (all
p<0.05). Only 12% vs 6% (p>0.05) used penicillin as first choice.
Table 2. Comparison of reasons for the 100 GPs
prescribing antibiotics in the 1998 and 2002–3 studies (N=100)
DiscussionIt is encouraging to find that 77% of the original 65 GPs
were less likely to prescribe antibiotics after 5 years. A similar percentage
felt there was a reduction in patients wanting antibiotics. This might indicate
that patients and GPs were more correctly informed about the effectiveness of
antibiotics.
It is difficult to ascertain how much the
Wise Use of Antibiotics campaign
contributed to this reduction. However the national figures suggest a one-third
reduction in antibiotic prescribing during the course of the campaign. National
figures show a reduction in amoxicillin clavulanate and an increase in
amoxicillin use consistent with our data. The higher response to patient
expectations may be a response to the “patient-centred medicine”
approach which has received more attention in recent years.10
There is no new literature suggesting that coloured sputum
may be responsive to antibiotic therapy although there is some for acute
purulent rhinitis.11 Ironically, for sinusitis, new guidelines suggest not
treating mild cases, hence the increase in giving antibiotics to these patients
is difficult to explain.12 Two systematic reviews conclude that antibiotic use
does not significantly affect the resolution of acute cough nor change the
course of illness and any modest benefits may be outweighed by the side
effects.13,14
Reduction in antibiotic use for acute otitis media is
consistent with studies indicating that delayed prescribing is an effective
means of reducing antibiotic use in children over the age of 6 months.15
Apparent contradictions may result from GPs seeing patients with more severe
symptoms (those with minor symptoms now being less likely to visit their doctor)
and hence more likely to prescribe antibiotics.
International literature suggests other antibiotic campaigns
have been effective in lowering the use of antibiotics for URTIs and
subsequently leading to reduction in resistance to commonly used antibiotics.
For instance, a nationwide Finnish programme involved recommendations in
response to concern about increasing resistance to group A streptococci. In that
programme, a relative risk reduction of 42% was found in daily doses of
erythromycin which translated to a 7.9% reduction in the frequency of antibiotic
resistance among group A streptococci.16
In response to penicillin-resistant pneumococci increasing
from 2.3% in 1989 to 20% in 1993, an Iceland initiative used radio, television,
and newspaper articles as well as targeting the medical community using
infectious disease experts in a publicity campaign regarding antibiotic
overuse.17 Penicillin-resistant pneumococci subsequently dropped to 16.9%.
Morever, in 1994, a Swedish programme responded to concern
over increasing antibiotic resistance by producing a guideline on how to deal
with penicillin-non-susceptible pneumococci.18 This led to a relative risk
reduction of 39%. National programmes in USA, Canada, Belgium and Australia
aimed at controlling antibiotic use and resistance have also reported success.19
The Canadian programme increased use of “appropriate” first-line
antibiotics for URTIs and the Belgium programme resulted in significant but
transient reduction in retail antibiotics from 17% to 9%. This is a similar
order of magnitude to that achieved in New Zealand.
A Dutch randomised trial of GP peer-group education with
monitoring and feedback as well as pharmacist and patient education found that
this multiple intervention reduced prescribing rates of antibiotics for URTIs
without decreasing patient satisfaction.20 A Spanish quasi-experimental
intervention study of the effects of GP education and feedback had similar
results.21
The strength of our study is that it reports on a random
selection of the GPs comparing two time periods between which there was a
campaign to reduce antibiotic use. Two-thirds of the participants were involved
in both time periods, thus allowing paired statistical analysis. We are not
aware of any other such study in the international literature. The response rate
was 69% for the additional GPs, which is acceptable for this type of study.
A limitation of the study is that it relies on asking
doctors what they do, rather than measuring what they do. However this study is
a step towards explaining the changes in antibiotic use over the duration of the
Wise Use of Antibiotics campaign. What
the doctors report they do is consistent with actual national data. Given a 61%
response rate from the GPs initially studied in 1998, a selection bias is
possible. A causal relationship between decreased GP prescribing and the
educational campaign cannot be proved.
In conclusion, the global response of GPs stating that they
are less likely to prescribe antibiotics is consistent with the reduction in
antibiotic use nationally. This may be related to the national campaign. The
reduction may be a combination of combined GP and patient change. There was a
significant reduction in the use of amoxicillin clavulanate. The apparent
increase in antibiotic use for specific conditions may relate to patients
presenting with more serious conditions/symptoms.
There are no conflicts of
interest in this study.
Author information:
Linda Sung, Medical Student University of Otago, Dunedin, Justine Arroll, Law
Student, University of Auckland, Auckland, Felicity Goodyear-Smith, Senior
Lecturer, Department of General Practice and Primary Health Care, School of
Population Health, University of Auckland, Auckland; Bruce Arroll, Associate
Professor, Department of General Practice and Primary Health Care, School of
Population Health, University of Auckland, Auckland; Ngaire Kerse, Associate
Professor, School of Population Health, University of Auckland, Auckland;
Pauline Norris, Senior Lecturer, School of Pharmacy, University of Otago,
Dunedin
Acknowledgments: We
thank the Charitable Trust of the Auckland Faculty of the Royal New Zealand
College of General Practitioners for their funding of this study.
Correspondence: Dr F
Goodyear-Smith, Dept of General Practice and Primary Health Care, School of
Population Health, University of Auckland, Private Bag 92019, Auckland. Fax:
(09) 373 7624; email: f.goodyear-smith@auckland.ac.nz
References:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| | Current
issue | Search journal |
Archived issues | Classifieds
| Hotline (free ads) Subscribe | Contribute | Advertise | Contact Us | Copyright | Other Journals |
|