Research

Antibiotic use in the Australian community, 1990-1995

Peter McManus, M Leigh Hammond, Susan D Whicker, John G Primrose, Andrea Mant and Steven R Fairall

Abstract
P
att erns of antibiotic prescribing in
Australia between 1987 and 1989
were reported previously by the Objective: To determine the pattern of antibiotic use in the Australian community,
Drug Utilization Sub-Committee 1990-1995, and compare it with the pattern in other developed countries.
(DUSC) of the Pharmaceutical Benefits Design: Survey of data from the national database on drugs dispensed in Australia
Advisory Committee, Department of (1990-1995), an international database on retail drug sales (1985-1994), and
Health and Family Services.' At the Australian prescriber surveys (1994, 1995).
time, the Sub-Committee expressed
concern over evidence of "inappropriate Main outcome measures: National and international retail sales of oral antibiotics
and unnecessarily expensive" practice, (defined daily doses [000s)/1000 population/day) and antibiotic prescriptions
which was "directed too heavily towards dispensed through community pharmacies by drug type; antibiotic prescribing profiles
the use of broad spectrum agents and for common conditions.
newer more expensive drugs". I Results: Antibiotic use in Australia remained steady between 1990 and 1995, with an
To update these data, DUSC con- estimated 24.7 000s/1000 population/day dispensed through community
vened a working group in 1995, com- pharmacies in 1990 and 24.8 000s/1000 population/day in 1995. Amoxycillin,
prising representatives of DUSC, the although declining in use, remained the most dispensed antibiotic. Compared with
Australian Pharmaceutical Manufac- the other countries surveyed, Australia had the highest percentage use of
turers Association and the Therapeutics tetracyclines, such as doxycycline, and the lowest percentage use of
Resource and Educational Network for fluoroquinolones. Use of trimethoprim-sulfamethoxazole and flucloxacillin declined in
Doctors (TREND) project of the Royal Australia. In new cases of upper respiratory tract infection or pharyngitis, an antibiotic
Australian College of General Practi- prescription was recorded for 57% of urban patient encounters and 73% of rural
tioners. patient encounters.
This group reviewed Australian and Conclusions: Antibiotic use in Australia is high, as in many other developed
international data on antibiotic sales and countries, but did not increase between 1990 and 1995. The overall profile of
dispensing to determine patterns of antibiotic use in Australia by drug class was similar to that in the United Kingdom.
antibiotic use in Australia between Antibiotics were still commonly prescribed for upper respiratory tract infection (which
1990 and 1995, and to compare these is usually viral), more commonly by rural than by urban general practitioners.
with patterns in similar developed coun-
tries. To understand trends in drug use, MJA 1997; 167: 124-127
the group also analysed prescriber sur-
veys of the indications for antibiotic use. monitors dispensing of prescription tions from an ongoing survey of a
medicines through community pharma- sample of about 250 community phar-
cies in Australia. These data include dis- macies.> Measurement units are either
Methods pensing to some private hospitals, but prescription volumes or number of
Data were derived from four sources: not public hospitals. The database con- defined daily doses (DDDs)/1000 pop-
tains information on all subsidised pre- ulation/day. The DDD is based on the
Prescription dispensing data: These scriptions processed by the Health assumed average daily dose of the drug
were obtained from the database main- Insurance Commission, together with an when used for its main indication by
tained by the DUSC secretariat which estimate of non-subsidised prescrip- adults. It is the unit approved by the
World Health Organization for drug use
For editorial comment, see page 116 studies and allows for comparisons inde-
Department of Health and Family Services, Canberra, ACT. pendent of differences in price, prepara-
Peter McManus, BPharm, MMedSc, Secretariat, Drug Utilization Sub-Committee; tion and quantity per prescription. 3
John G Primrose, FRACR, Medical Advisor, Health Benefits Division.
SmithKline Beecham (Aust), Melbourne, VIC. Sales data: Data on sales of oral
M Leigh Hammond, FRCPA, Medical Director.
antibiotics to retail and hospital markets
Therapeutics Unit, Royal Australian College of General Practitioners, Sydney, NSW.
Susan D Whicker, PhD, Scientific Director. in a number of countries were obtained
Drug and Alcohol Program, Eastern Sydney Area Health Service, Sydney, NSW. from Intercontinental Medical Statistics
Andrea Mant, MD, FRACGP, Clinical Director. (IMS), Melbourne. This is a commer-
GPS Business Unit, Eli Lilly, Sydney, NSW. cial market research organisation which
Steven R Fairall, BScAg, Associate Director.
Reprints will not be available from the authors. Correspondence: Mr P McManus, Drug Utilization
collects sales data and surveys general
Sub-Committee, Department of Health and Family Services, GPO Box 9848, Canberra, ACT 2601. practitioner (GP) prescribing in 33
E-mail: [email protected] countries.' Data were retrieved as unit

124 MJA Vol 167 4 August 1997

 Research

>.
<ll
35 o 1985 RACGP Quality Assurance and Con- 1: Sales of oral antibiotics to hospitals
32 01 989 tinuing Education Program. Partici- in 1994
§ 30
~
:;
o 1994 pating GPs recorded prospectively their DDDs/1000/day
c. 25 drug and non-drug management, Country (% of total sales')
8- including lifestyle advice, investigations West Germany 6.6 (38%)
8 20 and referrals, for 110 consecutive patient Italy 6,6 (26%)
encounters of all types (including France 5,8 (15%)
surgery and telephone consultations, United Kingdom 1,9 (11%)
Canada 2,0 (9%)
home, hospital and nursing home visits). Australia 1,8 (7%)
The GP sample was drawn from two United States 1,2 (5%)
urban and two rural areas in NSW DDD = defined daily dose,
France Australia United Canada lIaly United West
which were selected randomly from all 'Total sales = retail and hospital sales, Source:
States Kingdom Germany NSW Divisions of General Practice, Intercontinental Medical Statistics,

Figure 1: Retail sales of oral antibiotics in Aus- after matching within urban (16 Divi-
tralia and six major developed countries, sions) and rural (11 Divisions) strata. in 1990 and 24.8 DDDs/I000 popula-
1985-1994 (data from Intercontinental Medical Matching variables were the index of rel- tion/day in 1995. Pharmaceutical indus-
Statistics). ative socioeconomic disadvantage (from try sales data of 25.2 DDDs/I000
the 1991 Census") and GP-to-popula- population/day in 1989 and 24.8 DDDs/
sales by form and strength and drug tion ratios (from the Medicare claims
class and then converted to DDDs/I000 1000 population/day in 1994 (0.3% fall
database as at the end of June 1993) on compound annual growth rate) con-
population/day. (Gordon Calcino, Acting Director, firm this stability in overall use.
Prescriber survey: Diagnoses for Technical Support Section of the Gen-
which patients were prescribed anti- eral Practice Branch, Department of
International comparisons of
biotics and patient ages were obtained Health and Family Services, personal
antibiotic usage
from the Australian Medical Index communication). The rural areas
(AMI), the survey of GP prescribing included both major and outlying towns, Retail sales of oral antibiotics in Aus-
conducted by IMS. This uses a sample but not remote rural areas. All GPs from tralia and six major developed countries
of 420 GPs stratified in line with the the four areas were eligible for the in 1985, 1989 and 1994 are compared
total Australian GP population by met- survey, based on a list provided by the in Figure 1. Like most developed coun-
ropolitan/country location, age, year of General Practice Branch from the tries, Australia had a high use of oral
graduation, and practice size (patient Medicare claims database. Of the 783 antibiotics. In 1994 retail sales in Aus-
numbers). These GPs record all patient GPs approached, 303 (38.7%) com- tralia were second to those in France
encounters over seven consecutive days pleted the practice assessment in 1994. and followed closely by those in the
in each quarterly survey period. The Participants had a similar sex distribu- United States. However, between 1989
data are then projected to obtain quar- tion to the 1994 NSW GP population, and 1994 sales in Australia did not show
terly estimates of prescribing patterns by but a somewhat younger average age." the compound annual growth seen in all
diagnosis and by patient age and sex for the other countries except Canada -
the Australian GP population. At the France (up 2.8%), United States (up
Results
end of a four-quarter cycle, GPs are 2.3%), Italy (up 2.9%), West Germany
invited to continue participating. About Between 1990 and 1995, there was little (up 4.4%) and the United Kingdom
half agree, and the remainder are change in the level of antibiotics dis- (up 3%).
replaced by new recruits. As GPs start in pensed through Australian community Sales of oral antibiotics to hospitals in
various quarters of the year, there is con- pharmacies, with an estimated 24.7 1994 and the percentage this repre-
tinual turnover; GPs are not usually kept DDDs/I000 population/day dispensed sented of the total market are shown in
in the survey for more than Box 1. The percentage was
two years. D Narrow spectrum
penicillin similar in Australia, Canada
TREND project: Use of Fluoroquinolone
and the UK (7%, 9% and
•
antibiotics for management of
(e.q, norfloxacin) 11%, respectively), lowest in
upper respiratory tract infec- Macrolide the US (5%) and highest in
5l 60 D (e.q. erythromycin) West Germany (38%).
tion (URTI)/pharyngitis and ::l Trimethoprim and
influenza was examined using
c
s;Q 40 D combina tions
(e.q. trimethoprim-
Figure 2 shows the percent-
the database of 33203 C sulfame thoxazo le) age split of the oral antibiotic
doctor-patient encounters
-c D Cephalosporin market by drug class in 1994.
20 Australia had the highest per-
recorded in August and Sep- Broad spect rum
tember 1994 by the TREND D penicillin centage use of tetracyclines
project. This project was part o France Australia United Canada lIaly United West D Tetracyciine (25.5% of total oral antibi-
States KingdomGermany
of the development of a new otics), but the lowest use of
practice assessment in thera- Figure 2: Percentage split of oral antibiotic sales by drug class in 1994 fluoroquinolones (2.2%) and
peutics option within the (data from Intercontinental Medical Statistics). mid range use of penicillins -

MJA Vol 167 4 August 1997 125

Research

10 antibiotics dispensed through Australian community pharmacies in 1990 and 1995 and their percentage
to prescriptions for common conditions in 1995
Millions of prescriptions' Percentage of prescriptions for each condition (1995)1
Rank (DDDs/1000 population/day)
Otitis Urinary tract
Antibiotic 1990 1995 1990 1995 Sinusitis Bronchitis media URTI infection! SSTI

Amoxycillin 1 1 6.6 (6.3) 5.5 (56) 11.6 18.1 210 29.5 6.9 3.9
Amoxyciltin-clavulanate 5 2 2.1 (1.8) 34 (29) 181 13.7 20.5 94 17.2 8.7
Cefaclor 14 3 2.7 (1.0) 15.1 15.2 35.5 10.2 1.4 2.2
Doxycycline 2 4 31 (5.1) 2.6 (4.5) 20.7 9.5 0.5 5.5 03 8.2
Cephalexin 6 5 20 (1.2) 24 (1.5) 43 4.5 5.2 3.3 18.9 20.8
Roxithrornycin 6 20 (1.5) 13.2 165 0.7 8.7 0.1 1.5
Erythromycin 3 7 27 (3.0) 19 (2.2) 28 8.0 33 8.2 0.3 113
Trimethoprim-
sulfamethoxazole 4 8 2.5 (30) 1.2 (1.0) 2.5 2.5 3.7 3.3 28.5 19
Phenoxymethylpenicillin 8 9 11 (0.7) 1 1 (0.7) 04 0.1 0.1 7.2 0 12
Flucloxacillin 7 10 13 (09) 09 (0.7) 0.3 0 0 0.1 0.1 233
Tetracycline-nystatin 9 19 0.6 (09)
Tetracycline 10 12 04 (0.7) 13 10 01 0.3 0 14
DDD = defined daily dose. URTI = upper respiratory tract infection SSTI ~ skin and soft tissue Infection NA ~ not availablefor retail
available
'Source: Drug Utilization Sub-Committee. Department of Health and Family Services t Source Australian Medica/Index. lntercontinental
* 11 .2% of prescnpnonsfor urinarytract infectionwere for trimethopnm
" c , ." .""w__ "",·"_·~_,,,,,,_,,,,g:, . _,_,,,,,,,-,.,__ ,_·_,..,=,,,,="'.,."''''",,''''.'''N'_': ..., ,,.ji

both narrow spectrum (7.1 %) and most prescribed antibiotic was amoxy- group. This group received 36.9% of
broad spectrum (35%). The overall pro- cillin (18.1 %), followed closely by antibiotic prescriptions, but made up
file of antibiotic use in Australia was roxithromycin (16.5%) and cefaclor only 28.7% of the population. However,
similar to that in the UK. (15.2%). In urinary tract infections, for specific conditions the age patterns
trim e tho prim -s u I fa m ethox azo I e varied. For example, the TREND pro-
(28.5%) was most commonly pre- ject showed that an antibiotic was less
Types of antibiotics used in Australia
scribed, followed by cephalexin (18.9%) often prescribed for children aged 0-5
and indications
and amoxycillin-clavulanate (17.2 %). years with URTI (34.1 %) than for
The top 10 antibiotics, on prescription Antibiotic prescribing for upper res- adults aged 50 years and over (62.5%).
volume, dispensed through Australian piratory tract infection (URTI)/pharyn- Figure 3 shows quarterly fluctuations
community pharmacies in 1990 and gitis and influenza was examined in the in dispensing of oral antibiotics com-
1995 are shown in Box 2. Amoxycillin TREND sample of GPs from the pared with oral ~-blockers - a drug
was the most-dispensed antibiotic in August-September 1994 audit; 11.6% type used on an ongoing basis. Antibi-
1990 and, although declining in use, of encounters were for URTIIpharyn- otic dispensing was markedly seasonal,
remained in this position in 1995. Dis- gitis and 1.2% for influenza. For new with higher levels in the winter quarters
pensing of doxycycline and erythro- cases of URTIIpharyngitis, an antibiotic (April to June and July to September),
mycin also declined, while that of prescription was recorded for 57% of while dispensing of oral ~-blockers was
amoxycillin-elavulanate increased. Cefa- urban patient encounters and for 73% of low at the beginning of the year and high
clor and roxithromycin were new addi- rural patient encounters. Corresponding at the end. The second pattern was typ-
tions to the top 10 dispensed antibiotics figures for antibiotic prescribing in new
ical of "safety net" fluctuations." These
in 1995. cases of influenza without pneumonia
result from Pharmaceutical Benefits
Converting prescription volume to were 30% of urban and 62% of rural
D D Dsll 000 population/day altered the patient encounters. The TREND pro- 40

relative ranking of some drugs. For ject did not evaluate whether the pre- 35

example, drugs in the tetracycline scription was to be filled only under "ID ~ 30
en:!;'
group, which have increased maximum certain conditions (e.g., if symptoms .g s 25
quantities available for treatment of had not improved after a certain inter- ~~
"00.
20

severe acne, rose in rank (1990). In con- val), but some GPs commented that ~8. 15
trast, cefaclor, which has a high DDD this was their advice to patients. For ~§ 10
o~
_ _ Antibiotics
_ I3-Blockers
(1. 5g) by Australian standards, fell symptomatic management, the most 5
from third to seventh position (1995). common choices were decongestants o 1-:-'-;;-'-;;:'7t7"'-;:;-'-;;-'-:+,-'-;;-7.-t.-'-;;-'~;-/-;-'-;;-'-;;-'-:;l
Box 2 also shows antibiotic prescrib- and analgesics.
ing profiles for various indications in AMI survey data showed that overall
Figure 3: Quarterly fluctutations in dispensing of
1995. For sinusitis, the most prescribed antibiotic usage by age broadly followed antibiotics and {3-blockers in Australia,
antibiotics were doxycycline (20.7%), the proportional representation of par- 1991-1995 (data from the Drug Utilization Sub-
amoxycillin-clavulanate (18.1 %) and ticular age groups in the community, Committee, Department of Health and Family
cefaclor (15.1 %). For bronchitis, the with the exception of the 0-19 years age Services).

126 MJA Vol 167 4 August 1997

 Research

Scheme (PBS) provisions that provide flucloxacillin II were probably responsible This issue is currently being explored in
drugs free, or at lower cost, to the for the marked fall in use of these drugs Australia by the Australian Group on
patient if they are dispensed after the and the rise in use of cephalexin, a sub- Antimicrobial Resistance and by the
patient has incurred a set expenditure on stitute for flucloxacillin in skin and soft- National Antimicrobial Resistance Sur-
PBS items in a calendar year (the tissue infections pending the availability veillance Program and internationally by
"safety net" threshold). Consequently, of dicloxacillin (PBS listed in 1997) .12 the Alexander Project.
patients tend to fill repeat prescriptions The winter increase in dispensing of
towards the end of the safety net year, oral antibiotics in Australia contrasted Acknowledgement
after they reach this threshold. Changes with the "safety net" fluctuations of the Members of the Antibiotic Working Group. which helped
to reduce these fluctuations were intro- oral ~-blockers, which are used on an prepare these data were:
Drug Utilization Sub-Committee: Julie Lindner, Peter
duced in November 1994. ongoing basis, and probably reflected McManus. John Marley. Andrew Parkes and John Primrose.
treatment of respiratory tract infections. Australian Pharmaceutical Manufacturers Association and
Discussion In fact, antibiotics were prescribed for pharmaceutical industry: Lee Ausburn. Roger Fraser.
Steven Fairall. Mendel Grobler, Leigh Hammond, Peter
This report is a unique attempt by 57%-73% of new cases of URTI. This Kofler and Jenny Winter

diverse groups to combine their data contrasts with peer consensus recom- Royal Australian College of General Practitioners Ther-
apeutics Resource and Educational Network for Doctors
sources to give a more complete picture mendations on the use of antimicrobial (TREND): Andrea Mant and Sue Whicker
of antibiotic use in Australia than has drugs in medical practice. The Antibiotic The TREND project was funded by a grant from the Phar-
previously been available. guidelines I J state that for URTI "the maceutical Education Program of the Department of Health
and Family Services.
The comparison between seven major cause is almost invariably viral" and
developed countries showed that com- "antibiotics are not indicated". Further-
more, TREND data show that rural GPs References
munity use of antibiotics in Australia
between 1985 and 1994 was high - are more likely to prescribe antibiotics 1 Birkett OJ. Mitchell AS. Godeck A. et at. Profiles of
antibacterial drug use In Australia and trends from
second only to France, with the US a for URTI than urban GPs. The differ- 1987 to 1989 A report from the Drug Utilization Sub-
close third - but did not increase ence may reflect rural GPs' concern for committee of the Pharmaceutical Benefits AdVISOry
Committee Med J Aust 1991.155 410-415
during the 1990s, as it did in most of the the greater patient travel and inconve-
2 Edmonds OJ. Dumbrell OM, Primrose JG. et a! Devel-
other countries. nience in visiting the doctor in the coun- opment of an Australian drug utilisation database. A

In 1994, hospitals accounted for only try, with prescriptions being supplied for report from the Drug Utilization Subcommittee of
the Pharmaceutical Benefits Advisory Committee.
7% of retail sales of oral antibiotics in use if symptoms fail to resolve after a PharmacoEconomics 1993.3 427-432.
Australia, similar to the percentages seen period of time. However, it may also 3. Nordic Council on Medicines Nordic Statistics on
reflect differences in access to continuing Medicines 1990-1992 NLN publication number 34
in Canada and the UK. Similarly, a Uppsala. Sweden NLN. 1993
survey of cardiovascular drug use in education and in industry promotion 4 Hurley SF. McNeil JJ. Berbatis CG Sources of Aus-
Australian public hospitals found that between rural and urban GPs. tralian pharmacoepicernioloqy data. Community Health
Studies 1988: 1282-96
hospital use accounted for less than 10% The data also provide insight into
5. Australian Bureau of Statistics. Index of relative
of the total use of all cardiovascular drug GPs' perceptions of antibiotic resistance socioeconomic disadvantage. Canberra: ABS, 1993.
groups." The highest percentage of among bacteria. For sinusitis, the most (Catalogue no 13560)
6. Australian Institute of Health and Welfare. Medical
antibiotic use in the hospital sector was prescribed antibiotics were doxycycline, labour force 1992-93 National Health Labour Force
found in West Germany, where capping amoxycillin-clavulanate and cefaclor, Bulletin NO.3 Canberra: Australian Institute of Health
which are recommended if resistance and Welfare, Jul 1995
of GP prescribing budgets was intro-
7. McManus P Drug utilisation [letter]. MedJAust 1993:
duced in 1993. It has been argued that to amoxycillin is suspected or proven. 13 158: 724.
these prescribing budgets for reim- For bronchitis, amoxycillin was most 8 Doecke C, Harvey R. Havas L. Cardiovascular drug
prescribed, followed closely by roxithro- use In Australian hospitals 1990 Summary report of a
bursable drugs have increased hospital survey conducted by the Society of Hospital Pharma-
budgets." mycin and cefaclor, which are recom- cists of Australia and the Australian Institute of Health
Australia had the highest percentage mended if a ~-lactamase producing for the Commonwealth Department of Community
Services and Health. Canberra: the Department,
use of tetracyclines among the seven organism is isolated or if the clinical Oct 1991.
major developed countries, possibly response is slow. Our results indicate a 9 German drug caps boost hospital costs SCRIP 1993:
strong perception among doctors that 1877 4
because of its use in managing acne. In
10. Adverse Drug Reactions Advisory Committee.
contrast, Australia had the lowest per- resistant organisms are a significant clin- Inmethopnrn-sutpnarnethoxazote warning on elderly
centage use of fluoroquinolones, most ical problem, at least for sinusitis and Aust Adverse Drug React Bull 1990 Fed

likely because of PBS prescribing restric- bronchitis. Reasons for this are unclear 11 Fairley CK, McNeil JJ, Desmond P et al. Risk factors
for development of Ilucloxac.lun associated jaundice.
tions ("authority required"). and deserve investigation; they may BMJ 1993 306 233-235
Amoxycillin remains the antibiotic include previous clinical experience of 12. Turnidge J What to use Instead of nucroxacunn
slow resolution of these infections with [editorial]. Aust Prescriber 1995: 18 54-55
most dispensed through community 13. victonan Medical Postgraduate Foundation. Antibiotic
pharmacies in Australia, but its use standard therapy, or awareness of the quidelines 1996/97 9th edition. Melbourne: The
declined between 1990 and 1995, prevalence of resistant bacteria in the Foundation. 1996

undoubtedly because of transfer pre- local community.

scribing to amoxycillin-clavulanate or A corollary to the study of antibiotic
other antibiotics. Concerns over an use is an examination of the patterns of
increased risk of severe adverse reactions antibiotic resistance in bacteria, which
with trimethoprim-sulfamethoxazole in may be associated with levels of use of
the elderly'? and of hepatotoxicity with particular antibiotics and with total use. (Received 19 Dec 1996. accepted 29 Apr 1997) 0

MJA Vol 167 4 August 1997 127