Research Paper

Non-prescription medicines and Australian community

pharmacy interventions: rates and clinical significance
Kylie A. Williamsa, Lynne M. Emmertonb, Richard Taylorc, Joel Wernerd and Shalom Isaac Benrimoja
a
Faculty of Pharmacy, The University of Sydney, Sydney, bSchool of Pharmacy, The University of Queensland, Woolloongabba, Queensland, cSchool of
Public Health and Community Medicine, Faculty of Medicine, The University of New South Wales, Sydney and dc/o Faculty of Pharmacy, The University
of Sydney, Sydney, Australia

Keywords
Australia; community pharmacy; drug-related
problems; interventions; non-prescription; over
the counter
Correspondence
Dr Kylie A. Williams, Faculty of Pharmacy, A15,
The University of Sydney, Sydney, NSW, 2006,
Australia.
E-mail: kylie.williams@sydney.edu.au
Received March 4, 2010
Accepted December 15, 2010
doi: 10.1111/j.2042-7174.2010.00091.x
Abstract
Objective To quantify pharmacy intervention rates for non-prescription medica-
tions (pharmacist-only and pharmacy medicines), to document the clinical signifi-
cance of these interventions and to determine the adverse health consequences and
subsequent health care avoided as a result of the interventions.
Methods Non-prescription medicines interventions undertaken by community
pharmacy staff were recorded in two field studies: a study of all Australian pharma-
cies to determine incidence rates for low-incidence, highly significant interventions,
and a study of a sample of pharmacies to collect data on all non-prescription inter-
ventions. Recorded interventions were assessed by a clinical panel for clinical signifi-
cance, potential adverse health consequence avoided, probability and likely duration
of the adverse health consequence.
Key findings The rate of professional intervention that occurs in Australia for
pharmacist-only and pharmacy medicines is 5.66 per 1000 unit sales (95% confi-
dence interval 4.796.64). Rates of intervention varied by clinical significance. When
considering health care avoided, the main impact of the interventions was avoidance
of urgent general practitioner (GP) visits, followed by avoidance of regular GP visits
and accident and emergency treatment. The most common adverse health conse-
quences avoided were exacerbations of an existing condition (e.g. hypertension,
asthma) and adverse drug effects.
Conclusions This study demonstrates the way in which community pharmacy
encourages appropriate non-prescription medicine use and prevents harm through
intervening at the point of supply. It was estimated that Australian pharmacies
perform 485 912 interventions per annum when dealing with non-prescription
medicines, with 101 324 per annum being interventions that avert emergency
medical attention or serious harm, or which are potentially life saving.

Introduction
An increase in consumer use of non-prescription medicines
has occurred due to a number of factors, including greater
reliance on self-care practices and an increasing number of
products becoming available without prescription.[1,2] If all
non-prescription medicines were used appropriately and
effectively by consumers, this would presumably result in
positive health outcomes for patients and benefits to the
healthcare system. However, inappropriate use of these medi-
cines could lead to significant morbidity and even mortality,
as well as increased use of the healthcare system.[3,4] Phar-
macists have a key responsibility, in many countries embed-
ded in legislation, to provide consumers with appropriate
advice relating to non-prescription products.[58] Given their
expert pharmaceutical knowledge and accessibility to con-
sumers, pharmacists and pharmacy assistants are in an ideal
position to influence the use of these medicines. Community
pharmacists face interesting challenges, as their role as a
healthcare professional has to be practised in a retail setting.
The monopoly of some non-prescription medicines, and the
business arising from this monopoly situation, leads to

2011 The Authors. IJPP 2011

156 Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165
Kylie A. Williams et al.
external challenges and questioning from governments and
other retailers.
Medicines are scheduled according to the level of control
and consumer access needed to ensure public health and
safety.[9] In Australia, for medicines that do not require a pre-
scription, there are currently three groupings: unscheduled
medicines, which are available from any retail outlet; phar-
macy medicines (Schedule 2 or S2), which must be sold from
a pharmacy; and pharmacist-only medicines (Schedule 3 or
S3), where the pharmacist must consult with the consumer
prior to purchase.[8]
In 1999, a review of drugs, poisons and controlled sub-
stances was commissioned by the Council of Australian Gov-
ernments. One of the requirements of this review was to
determine whether there is a net benefit to the community as a
whole in retaining one or both of the non-prescription sched-
ules (pharmacy and pharmacist-only medicines).[10] Due to a
lack of available evidence in this area, the review recom-
mended the collection and evaluation of data on the effective-
ness of the controls on access to non-prescription medicines
and professional standards in relation to these medicines. As a
result of this recommendation, The Cost-Benefit Analysis and
Risk Assessment of Pharmacist Only (S3) and Pharmacy
Medicines (S2) and Risk-Based Evaluation of the Standards
Project was commissioned to evaluate the current scheduling
model for non-prescription products in Australia.
To undertake the cost-benefit analysis and to determine
the benefit that pharmacy staff provide in the area of non-
prescription medicines, an investigation of the identifica-
tion of actual or potential drug-related problems (DRPs)
and the pharmacy response to these (clinical intervention)
was required. From the literature, few studies were iden-
tified that evaluated pharmacy interventions in the non-
prescription area or tried to calculate the benefit that
such interventions made to the health of the community.
Several studies explored DRPs and pharmacy interventions
in both the hospital[11,12] and community sectors,[13,14]
but these mainly related to prescription medicines, or
non-prescription medicines were not differentiated from
prescription medicines.
Others have evaluated pharmacy staff responses to stan-
dardised scenarios (pseudo-patients or mystery shoppers)
with various behavioural practice outcomes; however, none
of this work has measured potential adverse effects.[1519]
Further studies considered how often pharmacy staff inter-
acted with consumers,[20,21] rather than specifically consider-
ing interventions related to non-prescription DRPs. Still
others focused on DRPs and pharmacy interventions for a
specific drug or drug class,[22] disease state,[23,24] patient
group[25,26] or level of intervention significance[27] only. In
many cases the DRPs and/or their classification was the main
focus of the research, rather than pharmacists interventions
to resolve these problems.
2011 The Authors. IJPP 2011
Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165
Non-prescription medicines interventions
Four studies that specifically examined non-prescription
interventions were identified, from South Africa (1999),[28]
Sweden (2001),[29] the UK (2004)[30] and Germany (2007)[31].
However, all four studies utilised different definitions of
DRPs and pharmacy interventions, and none used a well-
defined objective measure of the clinical significance of these
interventions, nor endeavoured to estimate the rate of non-
prescription interventions on a national level. Thus a com-
prehensive evaluation of the impact that pharmacy staff make
when intervening in non-prescription medicine purchases
does not appear to have been published or perhaps even
attempted.
This paper reports on the two studies that were undertaken
as part of the larger cost-benefit analysis[32] to evaluate the
impact of pharmacy staff in terms of clinical interventions
related to pharmacy and pharmacist-only medicines.
Study objectives
The objectives addressed in this study were: to quantify
the intervention rates of community pharmacies for non-
prescription medications (pharmacist-only (S3) and phar-
macy (S2) medicines), to document the clinical significance
of these interventions and to determine the adverse health
consequences and subsequent health care avoided as a result
of the interventions.
Methods
Two field studies were conducted: a study of all pharmacies in
Australia to determine the baseline incidence rates for low-
frequency, high significance interventions, and a study of a
sample of pharmacies to collect data on the full range of non-
prescription interventions. Two studies were designed as
neither study alone presents the complete picture in relation
to pharmacy interventions. The results from both studies
were used in the calculation of intervention rates.
A pilot study undertaken in 28 pharmacies informed the
development of study materials, in particular the documen-
tation forms, and the research protocol. This pilot also pro-
vided an insight into the documentation behaviour of
pharmacy staff. Study materials and the protocol for the pilot
study and two field studies were approved by the Human
Research Ethics Committee of The University of Sydney (ref-
erence nos 6981 and 7320). The two studies were conducted
in the period from November 2004 to September 2005.
Pharmacy recruitment and participation
For the study of all community pharmacies in Australia, all
pharmacies with Pharmaceutical Benefits Scheme dispensing
approval were contacted by mail and invited to participate
(n = 4981). Of these, 1574 pharmacies (31.6%) consented to
157
Non-prescription medicines interventions
participate, and 934 (18.8%) completed the study. (The Phar-
maceutical Benefits Scheme is administered by the Australian
Government, who subsidise a wide range of medicines to
ensure affordable access for all Australians. Approximately
80% of prescriptions dispensed in Australia are subsidised
under the scheme.)
The study of a sample of pharmacies required 160 pharma-
cies. For the sample study three groups of pharmacies were
recruited on the basis of Quality Care Pharmacy Program
(QCPP) accreditation (the QCPP is an Australian national
program that aims to provide an industry-wide guarantee of
retail service quality and professional practice[33]). As part of
this programme pharmacies are required to observe profes-
sional standards of practice, including standards relating to
S2 and S3 medicines.[6] The groups selected for the sample
study were: Group 1, QCPP-accredited pharmacies previ-
ously found to be highly compliant with the relevant
standards (sample n = 60); Group 2, QCPP-accredited phar-
macies that were moderately-to-poorly compliant with the
relevant standards or had not been tested (sample n = 40);
and Group 3, non-QCPP-accredited pharmacies, registered
for accreditation but with no accreditation date booked
(sample n = 60). This met an objective of the study not
covered in this paper. A complete discussion of the sample-
size calculation is available in the full report.[32] Of the 1238
pharmacies contacted, 182 agreed to participate; however,
only 101 pharmacies completed the study.
Interventions
For the purposes of this study, an intervention was defined as
the promotion of the quality use of medicines by the identi-
fication and attempted resolution of an actual or potential
drug- or symptom-related problem arising from an over-the-
counter request. This did not include general counseling
or advice provided over the counter in relation to non-
prescription medicines, nor actions relating to the prevention
of diversion to illicit use of non-prescription medicines.
A literature review failed to provide authoritative guide-
lines for clinical intervention definitions or coding frames for
non-prescription interventions. A coding frame for determi-
nation of clinical significance, intervention characteristics
and action taken was developed empirically with reference to
previous limited literature. This coding frame included sec-
tions on outcome of the intervention, problem identified,
cause of the problem and actions taken. Further information
on this coding is available in the study report.[32] Definitions
of the outcomes of interventions can be found in Box 1.
In both studies, community pharmacy staff were asked to
record (on pre-printed forms) interventions related to phar-
macy (S2) and pharmacist-only (S3) medications, that
occurred in their pharmacy during an allocated 2-week study
period. For a list of examples of medications included in each
158 Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165
Kylie A. Williams et al.
Box 1 Classification of interventions according
to potential outcome
Potentially life-saving interventions
The patient was at substantial risk of death at the time
of the event; or
it is suspected that use, or continued use, of the medica-
tion(s) would have resulted in death.
Very significant interventions
or interventions that potentially averted emergency
medical attention or serious harm
The intervention averted emergency medical attention,
with or without hospitalisation.
This included:
prevention of disability, impairment, damage or dis-
ruption in the patients body function/structure,
physical activities or quality of life,
prevention of a birth defect,
prevention of serious drug toxicity or major adverse
event.
Significant interventions
or interventions that averted routine medical attention
The intervention resulted in an improvement in patient
care and/or optimisation of therapy.
This included:
prevention of a general practitioner visit,
decrease in length of hospital stay,
decrease in risk of moderate adverse events or
symptoms,
prevention of exacerbation of the condition.
Minor interventions
or interventions that averted minor symptoms
The intervention resulted in a minor improvement in
patient care and/or minor optimisation of therapy.
This included:
improvement in quality of life, mobility or comfort,
improvement in symptoms usually left untreated or
treated with non-prescription medicines.
Interventions that had no impact on the patient
Interventions that had no real impact on the patient or
the patients well being.
Interventions that were harmful to the patient
Interventions that have had, or may have had, a harmful
or negative impact on the patient or the patients well
being.
of these schedules, see Box 2. For each intervention, phar-
macy staff were required to record the age, sex and relevant
medical history of the consumer, and to outline the problem
they had identified and what was done to resolve it. For the
2011 The Authors. IJPP 2011
Kylie A. Williams et al. Non-prescription medicines interventions

Box 2 Examples of pharmacy (S2) and sample of 20 interventions forms that were common across all
pharmacist-only (S3) medicines panels (to enable calculation of inter-panel agreement).
From the sample study, 888 intervention forms were
Pharmacy Pharmacist-only medicines
received, of which 469 were determined to be interventions
medicines (S2) (S3)
according to the operational definition. A random sample,
Diclofenac 12.5 mg tablets, Diclofenac 25 mg tablets, stratified across all significance coding categories, was
pack of 20 pack of 30 selected for review by a clinical panel (n = 189). Again this
Clotrimazole 1% cream for Clotrimazole 1% cream was determined by the maximum number of interventions
dermal use for vaginal use that could be assessed by each panel within available
Ranitidine 300 mg tablets, Pantoprazole 20 mg resources. Eighteen common interventions were used across
pack of 14 tablets, pack of 14 all panels.
Hydrocortisone 0.5% cream Hydrocortisone 1% cream
30 g for dermal use 30 g for dermal use
Evaluation by clinical panels
Loperamide 2 mg Levonorgestrel 1.5 mg
tablets, pack of 20 tablet for emergency Five clinical panels were used: three panels for the study of all
contraception pharmacies and two panels for the sample study. Each clinical
Azelastine 1 mg/mL Metoclopramide panel consisted of one general medical practitioner, one com-
nasal spray 5 mg+paracetamol munity or hospital pharmacist, and one clinical pharmacolo-
500 mg tablets for gist, recruited for their professional expertise. The panels
migraine-associated were required to examine the interventions to assess the fol-
nausea, pack of 8 lowing: clinical significance of the intervention, most likely
Budesonide 32 mg aqueous Orlistat 120 mg capsules and most serious potential adverse health consequences
nasal spray for weight control, avoided as a result of the intervention, and the probability of
pack of 42 (scale 01), likely duration of and level of health care needed
Flurbiprofen 8.75 mg Salbutamol 100 mg to treat the adverse health consequence. A standardised data-
lozenges, pack of 16 metered-dose inhaler collection form was developed and pilot-tested for use by the
Levocarbastine 0.5 mg/mL Glyceryl trinitrate 600 mg clinical panel (this form is available in the appendices of the
eye drops sublingual tablets, pack full report[32]).
Intervention forms were copied and sent to individual
of 100
panelists, who then assessed each form independently and
For a complete listing of S2 and S3 medicines, see National returned them for consensus matching. Consensus was iden-
Drugs and Poisons Schedule Committee, Standard for the tified if at least two of the three panelists agreed on the poten-
Uniform Scheduling of Drugs and Poisons No.24, National tial outcome avoided and the health care most needed to treat
Drugs and Poisons Scheduling Committee, Depart- the adverse health consequence. For the significance of the
ment of Health and Ageing: Canberra, July 2009. intervention and the probability and duration of the adverse
www.tga.gov.au/ndpsc/susdp.htm (accessed 6 July 2010). health consequence, the mean was used. Interventions that
did not achieve consensus were then discussed at a meeting of
study of all pharmacies, staff were asked to collect interven- the panelists in an attempt to reach agreement. Once the
tions that averted routine medical attention, emergency panels had assessed the interventions the adverse conse-
medical attention or serious harm, and potentially life-saving quences were coded with International Statistical Classifica-
interventions, in order to capture conservatively the highly tion of Diseases and Related Health Problems 9 (ICD-9)
significant and potentially life-saving interventions. In the codes by a medical coder.
sample study, pharmacies were asked to document all of their To enable comparisons between panels, 20 intervention
interventions for non-prescription medicines. forms in the study of all pharmacies and 18 intervention
Due to the numbers of intervention forms likely to be forms in the sample of pharmacies study, were common
received, only a sample of interventions could be sent to the across the respective panels. The kappa statistic was used to
clinical panels for assessment.A preliminary significance code determine correlations between panels (see Table 4, below).
was assigned to each intervention by the research team. This The determination of agreement across panels was difficult to
enabled appropriate sampling of interventions across signifi- assess. If data are recoded so that differences of 1 (on the
cance codes for the clinical panels. A random sample of 310 of severity scale) are coded to be the same, the Kappa sta-
these interventions were sent to the clinical panels. This was tistic improves substantially. This suggests that although
based on the maximum number of interventions that could exact agreement between panels was poor, there was good to
feasibly be assessed by each panel, and also took into account a excellent agreement on similar outcomes.

2011 The Authors. IJPP 2011

Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165 159
Non-prescription medicines interventions Kylie A. Williams et al.

Data weighting Table 1 Numbers of interventions

Study of all Sample

Results from the clinical panels were amalgamated with the
pharmacies study Total
corresponding intervention-form data and pharmacy demo-
graphic data. These data were subsequently weighted for Intervention forms received 6463 888
extrapolation to the Australian population for further clinical Intervention forms classified as 4917 469
intervention according to
and economic analysis. The purpose of the weighting was to
operational definitions
(1) correct for some differences in the characteristics of the Interventions assessed by clinical 310 189 499
pharmacies sampled in both studies compared to the known panels
national distribution of these characteristics and (2) weight Interventions assessed by clinical 87 64 151
the sample so that the numbers of interventions would reflect panels that related to S2 or S3
that expected in all Australian pharmacies over a period of 1 products only*
year. Thus the data were weighted for a number of variables, *The remainder of the interventions related to unscheduled medicines.
including distribution of pharmacies by state, pharmacy sales
turnover, data-collection periods and sampling of interven-
tions for evaluation by clinical panels. Further details of the Confidence intervals (CIs) were calculated according to the
data weighting can be obtained from the project report.[32] Poisson method using the conventional 95% level. The 95%
The pharmacies were considered to be representative of the Poisson confidence intervals were based on the actual counts
national distribution in terms of geographic accessibility and of interventions recorded in the study of all pharmacies and
QCPP accreditation. the sample study, and prorated to provide CIs for the extrapo-
lated annual counts and rates for Australia.
A summary of the numbers of interventions used in
Estimated cases avoided various stages of the analysis is provided in Table 1.
Healthcare and adverse health consequences avoided were
estimated as follows:
Results
Estimated cases avoided = estimated frequency of annual
From the study of all pharmacies, 6463 intervention forms
Australian interventions
probability of the adverse were received, and of these 4917 were initially categorised as
health consequence resulting had interventions that averted routine medical attention, emer-
the intervention not taken gency medical attention or serious harm, or were potentially
place ( calculated from the life-saving.
clinical panel assessments).

Intervention rates
Intervention rates were calculated as follows:
intervention rate = interventions per year
.
total units per year
This resulted in a number of interventions per unit
purchases rate.
Ideally the calculation of rates of intervention would be
based on actual sales from pharmacies; however, no such data
were available. Surrogate data, in the form of pharmacy pur-
chase data, were obtained from IMS Health (IMS Health is a
global company that provides pharmaceutical market intelli-
gence, information and analysis; www.imshealth.com). These
data are sourced from pharmaceutical manufacturers (repre-
senting over 98% coverage of the Australian market) and cor-
rected by inclusion of companies direct sales to pharmacies.
Data provided indicated that between June 2002 and May
2003, 65 838 526 S2 units and 20 043 924 S3 units were sold
(total 85 882 450 units).
Intervention rates
Weighted interventions across both studies allowed for the
calculation of a total rate of intervention per unit sales,
ranging from minor to potentially life-saving. Interventions
that had no effect were not used in the calculations.
The rate of professional intervention that currently occurs
in Australia for pharmacist-only (S3) and pharmacy (S2)
medicines is 5.66 per 1000 unit sales (95% CI 4.796.64).
Rates of intervention varied by clinical significance (Table 2).
Clinical panel evaluations
Totals of 310 and 189 interventions from the study of all phar-
macies and the sample study, respectively, were sent to the
clinical panels for evaluation. Examples of documented inter-
ventions and panel assessment of these are given in Box 3.
Table 3 outlines the panels ratings of the clinical signifi-
cance of those interventions and Table 4 presents the Kappa
analysis.

2011 The Authors. IJPP 2011

160 Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165
Kylie A. Williams et al. Non-prescription medicines interventions

Table 2 Total intervention rates by significance (both studies)

Intervention rate per

1000 units*
Estimated
annual 95%
Recorded interventions Confidence
Significance of interventions interventions in Australia Rate interval

Interventions that averted emergency medical 101 101 324 1.18 0.961.43
attention or serious harm + potentially life-saving
interventions
Interventions that averted minor symptoms + 50 384 588 4.48 3.325.90
interventions that averted routine medical
attention
Total 151 485 912 5.66 4.796.64

*Annual unit purchases in Australia = 85 882 450.

Box 3 Examples of interventions

A number of case studies have been included to illustrate the range of intervention types and levels of significance recorded
in the census and sample study.
C3796: An adult male requested Product A (loratadine). On questioning, the pharmacist discovered that the patient had a
severe allergy to wasps and was stung an hour previously. The pharmacist noted that the patients face had severe swelling and
he had breathing difficulties. The pharmacist realised that stronger treatment was required, administered two 25 mg tablets
of promethazine hydrochloride and had his wife drive him to hospital immediately. The clinical panel felt that there was a
100% chance that anaphylaxis was prevented, which was life-saving.
C260: An adult male with a history of heart disease requested Product J (ibuprofen) for treatment of his gout from a phar-
macist as this was the product that a friend recommended. The pharmacist discovered that the patient was currently taking
warfarin, and so highlighted the contraindication and recommended a paracetamol/codeine combination instead. A review
of the patients medications subsequently took place, with the pharmacist giving advice on appropriate usage of other medi-
cations that the patient had at home. The clinical panel felt that there was a 1% chance that a gastrointestinal bleed was
avoided, with accident and emergency or, secondly, intensive care in hospital being the most likely treatments that would
have been needed
S11: An adult male requested a suspension for fever and pain containing ibuprofen. The pharmacist knew that the patient
was asthmatic and was taking several asthma medications, and recommended that he take a paracetamol mixture instead.
The clinical panel found that there was a 6% chance that the intervention avoided exacerbating the patients asthmas symp-
toms, with a visit to accident and emergency being the most likely level of health care needed if the intervention had not taken
place.
C147: An adult male asked for Product AB (combination gel) for a sore on his lip that was not healing. The pharmacist
examined it and felt that it needed to be seen by a doctor immediately, rather than being treated with an over-the-counter gel.
No product was sold. The patient returned to inform the pharmacist that the sore was cancerous. The clinical panel felt that
there was a 100% chance the pharmacist prevented the cancer from worsening.
C4905: An adult female requested Product P (sodium citrotartrate). On questioning, the pharmacy assistant discovered
that the patient had had urinary tract symptoms for a long time. The pharmacy assistant advised that sodium citrotartrate
was not an appropriate product for a urinary tract infection and that the patient needed to see a doctor immediately. The
clinical panel felt that there was 30% chance that a prolonged urinary tract infection was avoided by the intervention. The
most likely treatment would have been an urgent general practitioner visit.
S409: An adult female requested Product M (cetirizine). On questioning, the pharmacist discovered that this product had
not been working effectively and that the patient was a hay-fever sufferer who needed treatment daily. The pharmacist rec-
ommended Product N (fluticasone propionate) instead as a better treatment for daily use. The clinical panel felt that the
there was 97% chance that continuation of the symptoms of allergic rhinitis was avoided, with a regular general practitioner
visit being the most likely level of health care avoided.

2011 The Authors. IJPP 2011

Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165 161
Non-prescription medicines interventions Kylie A. Williams et al.

Table 3 Clinical significance of interventions assessed by clinical panel

Study of all pharmacies Sample of pharmacies

Clinical significance Frequency % Frequency %

Interventions that had no impact on the patient* 25 8.1 58 30.7

Interventions that averted minor symptoms 18 5.8 36 19.0
Interventions that averted routine medical attention 105 33.9 56 29.6
Interventions that averted emergency medical attention or serious harm 135 43.5 36 19.0
Potentially life-saving interventions 27 8.7 3 1.6
Total 310 100 189 100.0

*Interventions were pre-coded by the research team as having an impact to generate a sample of interventions to be given to the panels. On panel
assessment, some of these were considered to have no impact on the patient.

Table 4 Kappa analysis

Actual data Recoded data

Expected % Actual % Kappa Expected % Actual % Kappa

agreement* agreement statistic agreement* agreement statistic

Study of all pharmacies

Panel 1 versus Panel 2 33 35 0.0152 46 85 0.7248
Panel 1 versus Panel 3 28 30 0.0311 43 90 0.8246
Panel 2 versus Panel 3 28 35 0.0972 30 80 0.7163
Sample of pharmacies study
Panel 1 versus Panel 2 19 33 0.1787 24 67 0.561

*Expected agreement if left to chance.

Kappa >0.75, excellent agreement; Kappa 0.40.75, fair to good agreement; Kappa <0.4, poor agreement.
Recoded so that differences in values that differ by 1 (severity) are coded to be the same.

Table 5 Health care avoided from total interventions (both studies) The most common adverse health consequences avoided
Estimated (by ICD-9 code) were exacerbations of an existing condition
annual Estimated (e.g. hypertension or asthma) and adverse drug effects
interventions annual (Table 6). This result was not unexpected given the large pro-
Health care avoided Frequency in Australia* cases avoided portion of consumers involved in the interventions who had
Urgent GP 122 229 995 84 650 at least one pre-existing condition (Table 7).
Next regular GP 57 133 837 35 783
Accident and emergency 140 172 584 30 808
Self-care 36 39 135 6074 Discussion
Standard ward in hospital 21 13 134 1161
Intensive care in hospital 15 13 668 76
Major findings
Other 15 55 431 13 106
Total 406 657 784 171 658 It is estimated that Australian pharmacies perform 485 912
Not reported 93 149 926 5806 interventions per annum in the process of dealing with non-
Total 499 807 710 177 464
prescription medicines, with 101 324 per annum averting
*These are weighted frequencies that were calculated using a systematic emergency medical attention, serious harm or being poten-
weightings procedure. GP, general practitioner. tially life-saving. This equates to an estimated annual Austra-
lian intervention rate of 5.66 interventions per 1000 unit sales
for pharmacist-only (S3) and pharmacy (S2) medicines.
When the estimated annual health care avoided is consid-
ered, the main impact of the interventions was the avoidance
Study limitations
of urgent general practitioner (GP) visits, followed by avoi-
dance of regular GP visits and avoidance of accident and A major obstacle encountered in this work was pharmacy
emergency treatment (Table 5). The additional impact in recruitment. The reasons for the reluctance to participate in
the hospital system includes avoidance of admission to a the study must inevitably be multi-factorial. Commu-
standard ward, followed by intensive care. nity pharmacists and their staff serve the community in a

2011 The Authors. IJPP 2011

162 Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165
Kylie A. Williams et al. Non-prescription medicines interventions

Table 6 Most common potential adverse health consequences avoided

Estimated annual
Recorded interventions Estimated annual
ICD-9 code Description* frequency in Australia cases avoided

Study of all pharmacies

493.9 Asthma, unspecified 17 17 256 4996
533.9 Peptic ulcer, unspecified as acute or chronic, without 18 17 982 4876
mention of haemorrhage
401.9 Essential hypertension, unspecified 11 10 840 2555
995.2 Unspecified adverse effect of drug, medicinal and biological 27 9776 2462
substance
965.4 Poisoning by aromatic analgesics, not elsewhere classified 46 51 103 372
239.9 Neoplasms of unspecified nature, site unspecified* 17 7311 228
578.9 Haemorrhage of gastrointestinal tract, unspecified 12 4559 197
Sample of pharmacies study
995.2 Unspecified adverse effect of drug, medicinal and biological 9 71 707 21 850
substance
372.3 Conjunctivitis, unspecified 6 18 500 17 883
401.9 Essential hypertension, unspecified 28 124 017 13 640
54.9 Herpes simplex without mention of complication 3 2033 1964
998.1 Haemorrhage or haematoma complicating a procedure 4 30 151 1508
493.9 Asthma, unspecified 5 15 619 1195
333.99 Other and unspecified extra pyramidal diseases and 5 2199 44
abnormal movement disorders

*These descriptions come directly from the International Statistical Classification of Diseases and Related Health Problems 9 (ICD-9).
These are weighted frequencies that were calculated using a systematic weightings procedure.

Table 7 Medical history of consumers varied from 0 to 45 in the study of all pharmacies, and from 0
Study of all pharmacies* Sample study* to 55 in the sample study over a 2-week period.
The calculation of intervention rates in this study may have
Medical history Frequency % Frequency %
been limited by the lack of actual sales data from participating
Hypertension 1276 19.7 188 21.2 pharmacies and potential under-reporting or over-reporting
Other 892 13.8 142 16.0
of interventions. The evidence available suggests that, on
Asthma 595 9.2 65 7.3
balance, there has been under-reporting in this study. This is
Heartburn/ulcer 505 7.8 40 4.5
Arthritis 490 7.6 45 5.1 supported by data from the pilot study in which the docu-
Heart disease 474 7.3 53 6.0 mentation forms were tested, where documentation rates
Diabetes 325 5 47 5.3 decreased in the second week of recording from those in the
Pregnancy 309 4.8 26 2.9 first week, and pharmacists reported that they felt that only
*Pre-existing conditions are not mutually exclusive, so row and column 5080% of interventions had been documented. However,
totals do not apply. the counter-balance to the under-reporting could be that par-
ticipating pharmacies may be biased towards high performers
compared to non-participants.
professional manner and simultaneously operate commer-
Although the use of clinical panels is a useful and accepted
cially in a retail environment. Documentation of activities,
technique to assess averted negative clinical outcomes and/or
unless a by-product of their normal activities, is not routine
identify clinical benefits, it may be challengeable on the
practice in pharmacies. However, it is surprising that a study
grounds of its subjective approach to determining clinical sig-
of this focus and importance, which could affect the profes-
nificance and potential outcomes. However, in a large study
sional and commercial future of the profession of pharmacy
such as this it is not logistically possible to follow-up all con-
in Australia, did not attract more participation. International
sumers to document outcomes.
studies have also reported low participation rates, at around
5% of pharmacies nationally.[28,29,31]
The samples were broadly representative on a number of
Discussion of findings
factors and were adjusted by weighting. In addition, reporting
of interventions indicated a spectrum of pharmacy behav- A Swedish study, the only study which provided a compari-
iour: the number of interventions reported by pharmacies son, reported an intervention rate of 1.2 DRPs per 1000

2011 The Authors. IJPP 2011

Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165 163
Non-prescription medicines interventions
customers.[29] However, this per customer rate is not directly
comparable to a rate calculated per unit sold.
Prescription intervention rates have been more widely
reported in the literature,and an Australian study[14] identified
a rate of 8.7 interventions per 1000 prescription items dis-
pensed. Given that prescription intervention rates might be
expected to be higher due to greater risk associated with these
drugs, the non-prescription intervention rate of 5.7 per 1000
unit sales identified in this study seems a reasonable estimate.
Reassuringly, interventions that averted minor symptoms
or routine medical attention were performed more commonly
(4.48 per 1000 units) than interventions that averted emer-
gency medical attention or where potentially life-saving (1.18
per 100 units).However,it is evident from these data that there
is a group of consumers who, without the intervention of
pharmacy staff, would not be using non-prescription medi-
cines safely and appropriately. One of the consequences of this
inappropriate use, as determined by the clinical panel, would
be increased utilisation of other sectors of the healthcare
system. Estimates suggest that each year pharmacy staff pre-
vent 84 650 urgent visits to a GP, 35 783 routine GP visits and
30 808 visits to hospital accident and emergency departments.
For the interventions reviewed by the clinical panel,
the most common estimated adverse health consequences
avoided were exacerbations of hypertension, asthma and
peptic ulcers, and unspecified adverse effects of drugs, which
suggests that the high-risk groups were patients with those
underlying diseases. Although the rank for interventions
from the sample of pharmacies study varied from that in
the study of all pharmacies, the same groups of patients
appear to be at risk. Significant proportions of our samples
reported having an underlying chronic condition, especially
hypertension and asthma. Thus it appears that consumers
with chronic conditions are more at risk when it comes to
self-medicating with non-prescription medicines.
These findings may therefore have implications for both
regulators involved in the decision-making processes regard-
ing descheduling of medications, and for pharmacists and
their staff in identifying at-risk consumers.
In Australia, medicines are increasingly being descheduled
from S2 (pharmacy medicines) to the unscheduled category.
These products can be sold in any retail outlet (e.g. supermar-
kets) without any professional advice. The results from this
study have suggested that some consumers are incorrectly
self-selecting products, and that in the absence of pharmacy
advice this may lead to adverse consequences. Further inves-
tigation of the incidence and outcomes of inappropriate use
of unscheduled medicines should be undertaken. In addition,
there is little surveillance of medicines once they are avail-
able without prescription. When considering a medicine for
unscheduled use, generally the only evidence for safe and
effective use that is considered is usually that gathered from
when the agent was restricted to prescription-only availabil-
164 Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165
Kylie A. Williams et al.
ity (S4). Thus a post-descheduling (S4 to S3 and S3 to S2)
pharmaco-vigilance system should be developed and imple-
mented on a national basis to assist in the analysis and man-
agement of risk. The definitions and processes associated
with implementing such a system would vary by country and
scheduling system, and warrants further investigation.
In the past, pharmacists may have approached the issue of
risk management of non-prescription medicines from a
more product-based point of view. However, the findings
from this study have suggested that when taking a particular
non-prescription medicine certain groups of consumers are
more at risk than others. Thus, any risk assessment should
perhaps be based on the patient rather than the drug.
Conclusion
This study has demonstrated the way in which community
pharmacy encourages appropriate use of non-prescription
medicines and prevents harm through intervening at the
point of sale or supply. Despite the limitations of this study it
is evident that these interventions help avoid adverse events
and further costs to the healthcare system.
Declarations
Conflict of interest
The Author(s) declare(s) that they have no conflicts of
interest to disclose.
Funding
This work was supported by the Australian Govern-
ment Department of Health and Ageing as part of the Third
Community Pharmacy Agreement.
Acknowledgements
This research was funded by the Australian Government
Department of Health and Ageing as part of the Third Com-
munity Pharmacy Agreement. The research team would like
to acknowledge and thank the following for their contribu-
tion: the Expert Advisory Group (EAG) who provided expert
guidance for this project; the Pharmacy Guild of Australia as
managers of the Third Community Pharmacy Agreement
Research and Development Grants (CPA R&D Grants)
Program; IMS Health Australia Pty Ltd for their generous
provision of pharmacy purchase data; Professor Beth Reid
(Professor and Head of Health Information Management,
School of Health Information Management, The University
of Sydney) for her assistance with the development of the
medical coding procedure; and all the pharmacy staff who
helped collect intervention data for these studies.
2011 The Authors. IJPP 2011
Kylie A. Williams et al. Non-prescription medicines interventions

References management in acute care government 23. Westerlund T et al. Evaluation of a

1. Blenkinsopp A, Bradley C. Over the
counter drugs: patients, society and the
increase in self medication. Br Med J
1996; 312: 629632.
2. Soller RW. Evolution of self-care with
over-the-counter medications. Clin
Ther 1998; 20(Suppl. 3): C134C140.
3. Charupatanapong N. Perceived likeli-
hood of risks in self-medication prac-
tices. J Soc Adm Pharm 1994; 11: 1828.
4. Brass EP. Changing the status of drugs
from prescription to over-the-counter
availability. N Engl J Med 2001; 345:
810816.
5. International Pharmaceutical Federa-
tion. Standards for Quality of Pharmacy
Services. The Hague: International
Pharmaceutical Federation, 1997:
www.fip.org/statements (accessed 6
July 2010).
6. Pharmaceutical Society of Australia.
Standards for the Provision of Pharmacy
Medicines and Pharmacist Only
Medicines in Community Pharmacy.
Canberra: Pharmaceutical Society of
Australia, 2005.
7. Pharmaceutical Society of Australia.
Competency Standards for Pharmacists
in Australia. Canberra: Pharmaceutical
Society of Australia, 2003.
8. NSW Health, Pharmaceutical Services
Branch. Guide to Poisons and Thera-
peutic Goods Legislation for Pharma-
cists. January 2009. www.health.
nsw.gov.au/resources/publichealth/
pharmaceutical/pdf/poisons_
pharmacists.pdf (accessed 6 July 2010).
9. Gilbert A et al. A review of pharmaceu-
tical scheduling processes in six coun-
tries and the effect on consumer access
to medicines. Int J Pharm Prac 2006; 14:
95104.
10. Galbally R. National Competition
Review of Drugs, Poisons and Controlled
Substances Legislation: Final Report Part
A. Woden: Commonwealth of Austra-
lia, 2000: www.tga.gov.au/docs/html/
rdpdfr.htm (accessed 6 July 2010).
11. Dooley MJ et al. A prospective multi-
centre study of pharmacist initiated
changes to drug therapy and patient
funded hospitals. Br J Clin Pharmacol
2004; 57: 513521.
12. Parthasarathi G et al. Assessment of
drug-related problems and clinical
pharmacists interventions in an Indian
teaching hospital. J Pharm Pract Res
2003; 33: 272274.
13. Leemans L et al. Frequency and trends
of interventions of prescriptions in
Flemish community pharmacies.
Pharm World Sci 2003; 25: 6569.
14. Benrimoj SI et al. Clinical intervention
rates in community pharmacy: a ran-
domised trial of the effect of education
and a professional allowance. Int J
Pharm Prac 2003; 11: 7180.
15. Norris P. Purchasing restricted medi-
cines in New Zealand pharmacies:
results from a mystery shopper study.
Pharm World Sci 2002; 24: 149153.
16. Rutter PM et al. Evaluation of com-
munity pharmacists recommenda-
tions to standardized patient scenarios.
Ann Pharmacother 2004; 38: 1080
1085.
17. Alte D et al. Evaluation of consultation
in community pharmacies with
mystery shoppers. Ann Pharmacother
2007; 41: 10231030.
18. Benrimoj SI et al. Monitoring quality
standards in the provision of non-
prescription medicines from Austra-
lian community pharmacies: results of
a national programme. Qual Saf Health
Care 2007; 16: 354358.
19. Benrimoj SI et al. A system for moni-
toring quality standards in the provi-
sion of non-prescription medicines
from Australian community pharma-
cies. Pharm World Sci 2008; 30: 147
153.
20. Taylor JG et al. Assisting consumers
with their OTC choices. CPJ 2006; 139:
3845.
21. Sclar DA et al. Pharmacy consultation
and over-the-counter medication pur-
chasing outcomes. J Clin Pharm Ther
1996; 21: 177184.
22. Guignard AP et al. Economic impact of
pharmacists interventions with non-
steroidal antiinflammatory drugs. Ann
Pharmacother 2003; 37: 332338.
model for counseling patients with dys-
pepsia in Swedish community pharma-
cies. Am J Health Syst Pharm 2003; 60:
13361341.
24. Viktil KK et al. Comparison of drug-
related problems in different patient
groups. Ann Pharmacother 2004; 38:
942948.
25. Ernst ME et al. Drug-related problems
and quality of life in arthritis and low
back pain sufferers. Value Health 2003;
6: 5158.
26. Vinks THAM et al. Identification of
potential drug-related problems in the
elderly: the role of the community
pharmacist. Pharm World Sci 2006; 28:
3338.
27. Alderman CP, Farmer C. A brief analy-
sis of clinical pharmacy interventions
undertaken in an Australian teaching
hospital. J Qual Clin Prac 2001; 21:
99103.
28. Dreyer AC et al. Cost savings effected
by community pharmacists interven-
tions in the supply of medicines in the
Republic of South Africa. J Res Pharm
Econ 1999; 9: 120.
29. Westerlund LOT et al. Nonprescription
drug-related problems and pharmacy
interventions. Ann Pharmacother 2001;
35: 13431349.
30. Gray S et al. Evaluating the potential
impact of community pharmacy inter-
ventions on patient care in Northum-
berland. Qual Prim Care 2004; 12:
4751.
31. Hammerlein A et al. Survey of drug-
related problems identified by commu-
nity pharmacies. Ann Pharmacother
2007; 41: 18251832.
32. Benrimoj SI et al. A Cost-Benefit Analy-
sis of Pharmacist-Only (S3) and Phar-
macy Medicines (S2) and Risk-Based
Evaluation of the Standards. Final
Report Appendices. Sydney: The
University of Sydney. June 2005. www.
guild.org.au/research/project_display.
asp?id=246 (accessed 6 July 2010).
33. The Pharmacy Guild of Australia.
Quality Care Pharmacy Program.
www.guild.org.au/qcpp (accessed 6
July 2010).

2011 The Authors. IJPP 2011

Royal Pharmaceutical Society 2011 International Journal of Pharmacy Practice, 19, pp. 156165 165