Full text of DOI:10.2165/00019053-200119030-00004

Pharmacoeconomics 2001; 19 (3): 255-265
1170-7690/01/0003-0255/$22.00/0
ORIGINAL RESEARCH ARTICLE
© Adis International Limited. All rights reserved.
Cost Effectiveness of Emedastine
versus Levocabastine in the Treatment
of Allergic Conjunctivitis in 7
European Countries
Carlos Gouveia Pinto,1 Antoine Lafuma,2 Francis Fagnani,2 Mark J.C. Nuijten3 and
Gilles Berdeaux4
1
2
3
4
Universade Técnica de Lisboa, Instituto Superior de Economia e Gestao, Lisboa, Portugal
Cemka, Bourg-la-Reine, France
Medtap Netherlands, Jisp, The Netherlands
Alcon France, Rueil-Malmaison, France
Abstract
Objective: To assess the cost effectiveness of emedastine, a new antihistamine,
versus levocabastine in the treatment of acute allergic conjunctivitis (AAC) in
Belgium, France, Germany, The Netherlands, Norway, Portugal and Sweden.
Design and setting: Randomised double-blind multicountry clinical trial fol-
lowed by economic modelling from the treatment provider perspective.
Patients: A total of 221 patients (109 emedastine, 112 levocabastine) with AAC
were included.
Methods: The clinical trial compared the efficacy and safety of emedastine 0.05%
and levocabastine 0.05%, both twice daily, for 42 days, using ocular redness,
itching, days without symptoms and clinical failure as outcome measures. The
cost of first-line treatment failure, including visits, drugs and laboratory exami-
nations, was established in each country from a panel of ophthalmologists and
general practitioners. Full sensitivity analyses were conducted.
Results: From day 7 to 42, patients treated with emedastine had less itching
(p < 0.001) and less redness (p < 0.001). The failure rate was 10% less (p < 0.02)
with emedastine and patients treated with emedastine had an incremental 8.5 days
(p < 0.01) without symptoms. Emedastine and levocabastine were equally well
tolerated. In all European countries, the cost of failure was lower with emedastine.
Emedastine was found to be economically dominant relative to levocabastine,
i.e. more effective and less expensive, in Belgium, Germany, Portugal and Swe-
den; in France, The Netherlands and Norway the incremental cost was low (less
than 1 euro per additional symptom-free day).
Conclusion: Through a model based on a randomised clinical trial and cost es-
timates of treatment failure derived from practitioner interviews, emedastine is a
cost-effective treatment of AAC.

256
Pinto et al.
Allergic conjunctivitis is one of the most com-
• it alters the quality of life/well-being of the pa-
tients[12,13]
mon ocular allergic disorders[1,2] with a prevalence
estimated from 5 to 22% of the population.[3-8] Acute
allergic conjunctivitis (AAC) is the most frequent
form, accounting for 50 to 90% of cases of allergic
conjunctivitis.[9] The onset of AAC is typically re-
lated to the presence of specific air-borne allergens
such as pollens, spores or moulds.[10]
• it is a cause of work or school days lost[14]
• although it is treated mainly on an outpatient
basis with low medical costs, its high prevalence
leads to high expenditures for sickness funds,
i.e. $US1.8 billion per year (1990 values) in the
US.[15]
The aim of this study was to compare the cost
(from a treatment provider perspective) of the treat-
ment of one episode of AAC by 2 antihistamines
(levocabastine and emedastine) in various European
Union (EU) countries: Belgium, France, Germany,
The Netherlands, Norway, Portugal and Sweden.
These drugs have been compared through a well-
controlled randomised double-blind clinical trial.
Characteristic signs and symptoms include bi-
lateral and periodic itching and occasionally burn-
ing, photophobia and foreign body sensation. Ocular
examination reveals mild conjunctival hyperaemia
and oedema, eyelid oedema and excessive lacrima-
tion. Concurrent rhinitis is common, although in a
subset of patients the ocular symptoms may occur
exclusively.
Traditional therapies include drugs aimed at con-
trolling the conjunctival secretion of histamine, the
primary mediator of the allergic reaction: oral anti-
histamines (H1 antagonists) as well as topical corti-
costeroids, mast cell stabilisers, nonsteroidal anti-
inflammatory drugs and antihistamines sometimes
given concurrently with a vasoconstrictor.[10] Di-
agnosis and treatment of AAC is performed either
by general practitioners (GPs) or by ophthalmolo-
gists, with some variations between countries.
Usually the different types of allergic conjunc-
tivitis do not threaten the patients’ vision, especi-
ally when diagnosed and treated by effective therapy.
Nevertheless, allergic conjunctivitis is an impor-
tant health problem because ofits morbidity,[11] since:
Patients and Methods
Experimental Design
Two sources of data were used: (i) failure rates
and time without symptoms were estimated from a
European clinical trial; and (ii) the standard aver-
age cost of a failure was obtained from a panel of
practitioners. A decision tree was constructed to
aggregate cost of first-line success and cost of fail-
ure.
Only costs contributing to the difference between
the 2 strategies were taken into account.[16] The cost
of success was estimated from the daily cost of
emedastine or levocabastine multiplied by the suc-
Table I. Sources of data related to resource utilisation costing
Resource
Belgium
France
Germany
The
Netherlands
Norway
Portugal
Sweden
Exchange rate (euro)
40.3399
Belgian
francs
6.55957 1.95583
2.20371
8.24
Norwegian
kroner
200.482 escudo 8.26 Swedish
kronor
French
francs
Deutschmarks Dutch
guilders
Private/public practice
Mainly
public
47 : 53
Mainly
public
1 : 1
12 : 88
60 : 40
36 : 64
Mainly public Unknown, fixed
Mainly public
arbitrarily at 1 : 1
Ratio of ophthalmologist to
general practitioner in allergic
conjunctivitis
7 : 93
15 : 85
Unknown, fixed
arbitrarily at 1 : 1 arbitrarily at 1 : 1
Unknown, fixed
Visits (references)
Drugs (references)
Laboratory examinations,
procedures (references)
27
38
27
28, 29
39
46
30, 31
40
30, 31
32
41, 42
34, 47-50
33
43
51
34, 35
44
34
36, 37
45
52
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

Cost Effectiveness in Allergic Conjunctivitis
257
1.0
0.9
Emedastine
Levocabastine
0.8
0.7
0.6
0.5
0
7
14
21
28
35
42
Time (days)
Fig. 1. Survival curves illustrating the instantaneous success rate incidence by treatment group. Failure is defined as patients whose
symptoms (redness and itching) were stable or worsened since the first visit.[60] χ2 analysis at day 42, p < 0.045; Wilcoxon life test,
p < 0.02.
cess rate. Therefore, the cost of success was inde-
pendent of the number of bottles required to treat
for 42 days. The cost of failure was estimated from
the cost of purchasing 1 bottle of emedastine/
levocabastine (mean time to failure was 8 days,
covered by 1 bottle of drug) plus the cost of failure,
multiplied by the failure rate.
The time horizon of the model (a decision tree)
was 42 days, i.e. the clinical trial duration. No dis-
counting was performed because of the timeframe
of the model.
Levocabastine ophthalmic suspension 0.05% or
emedastine ophthalmic solution 0.05% was given
twice a day in the affected eyes for 42 days.
The primary clinical efficacy end-points were
itching and ocular redness, and the secondary end-
points were chemosis, eyelid swelling, patient di-
ary data and physical global assessment. Itching
and redness were measured at days 0, 3, 7, 14, 30
and 42. Patients were asked to complete a diary
card 4 times a day. Itching and redness were self-
evaluated according to a score from 0 (no symp-
toms) to 9. Standardisation through a photo cata-
logue given to the patient was performed in order
to estimate the intensity of symptoms as reliably as
possible.
Clinical Efficacy and Effectiveness
Effectiveness was derived from a randomised,
multicentre, double-blind, parallel-group clinical
trial including 221 patients from Europe (UK, Italy,
France, Germany, Spain and Portugal), South Af-
rica and Australia, recruited by 18 investigators.[17]
Patients of either gender and any race who met
the following criteria could be included: scores
of 4 in itching (very frequently), and conjunctiva
hyperaemia scores of 2 (colour overall is more red
than pink, calibre of vessels has increased so that
vessels have lost their delicate web-like pattern) or
more in both eyes, a history of allergic conjuncti-
vitis for at least 1 allergy season and a positive skin
test (within the past 12 months) to at least 1 com-
mon pollen indigenous to the area at the time of the
study.
Two effectiveness measures were used from this
clinical trial: (i) failure, defined by the protocol as
‘no improvement or worsening in ocular redness,
combined with no improvement in itching’; and
(ii) time without symptoms estimated from the di-
ary card (at every recorded time with no redness
and no itching, 0.25 days were added to time with-
out symptoms, since the measures were performed
4 times a day).
The statistical analysis was conducted using SAS
software (release 6.12; SAS Institute, Cary, NC,
USA). The failure rate incidence was calculated at
each scheduled visit and the comparison between
emedastine and levocabastine was performed us-
ing the maximum likelihood ratio test applied to
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

258
Pinto et al.
survival curves. Missing values in diary cards were
estimated through the ‘last observation carried
forward’[18] technique. Comparison between the 2
groups were worked out using a nonparametric test.
All the tests were 2-sided, with alpha being fixed
at 5%.
the practitioners and the standard questionnaire are
available upon request.
Economic Analysis
All results were expressed in euro (EUR), taking
into account the fixed parity for the countries be-
longing to the euro zone[19] or the exchange rate (01
January 2000) for the others. The economic per-
spective was that of the treatment provider and fol-
lowed the guidelines of the countries participating
in this economic evaluation.[20-26]
Indirect costs were not taken into account, since
there were no national statistics documenting the
loss of productivity in AAC in all countries and
interviews were not considered as an appropriate
source for this type of data. Moreover, the number
of workdays lost has been found to be rather low
in AAC[14] and their associated cost could be close
to zero since patients return quickly to work. This
was confirmed by the French expert interviews and
preliminary data from an ongoing EU piggy-back
clinical trial (unpublished data).
Resource Utilisation
The number of days of treatment available from
a bottle was estimated from the following parame-
ters: size of the bottle (ml), size of the drop (μl),
expiry date, daily dosage. The size of the drops was
estimated from a standardised methodology taking
different sources of variations into account (the bot-
tle, the order of the drop, the measure itself) esti-
mated through a specific experimental design (bal-
anced block analysis of variance).
As the clinical trial patients were not followed-
up in case of failure, no primary data concerning
resource utilisation were available. Therefore, re-
source utilisation for the second-line treatment was
estimated through interviews with 10 GPs and 10
ophthalmologists (11 in Portugal), in each country,
except Norway (5 GPs and 5 ophthalmologists).
The interviews were standardised by the use of
a predetermined questionnaire. Interviewers were
trained before conducting interviews. The list of
Table I describes the sources used to value re-
source utilisation in monetary terms. Public prac-
tice was defined as a practitioner using national
tariffs: these estimates came mainly from national
tariff lists. The ratio of ophthalmologists versus GPs
practising in allergic conjunctivitis was taken from
Table II. Resource utilisation as prescribed/requested by ophthalmologists after failure of first-line topical antihistamine therapy (source:
practitioner interviews). The sum of percentages might exceed 100%, since a patient may receive more than 1 drug and some drugs may
include more than 1 entity
Resource (% of ophthalmologists
prescribing/requesting)
Belgium
France
Germany
The
Netherlands
Norway
Portugal
Sweden
Topical drugs
corticosteroids
mast cell stabilisers
antihistamines
antimicrobials
others
Systemic drugs
corticosteroids
135.0
71.0
33.0
13.0
0
18.0
20.0
0
117.5
30.5
14.9
35.0
28.1
9.0
106.0
36.0
64.0
6.0
0
0
1.5
0
96.4
44.0
32.4
10.0
0
10.0
3.4
0
100.0
60.0
20.0
0
118.2
95.5
4.5
0
18.2
0
161.5
55.0
61.5
35.0
0
10.0
20.0
0
0
20.0
20.0
0
73.0
0.5
9.1
0
antihistamines
20.0
38.0
2.32
1.78
0.54
72.5
14.2
1.31
1.26
0.05
1.5
1.0
1.49
1.41
0.08
3.4
20.0
10.0
1.81
1.80
0.01
9.1
0
1.13
1.13
–
20.0
10.1
2.26
1.82
0.44
Laboratory examinations, procedures
Visits (number)
ophthalmologists
other specialists
19.1
1.80
1.66
0.14
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

Cost Effectiveness in Allergic Conjunctivitis
259
Table III. Resource utilisation as prescribed/requested by general practitioners after failure of first-line topical antihistamine therapy (source:
practitioner interviews). The sum of percentages might exceed 100%, since a patient may receive more than 1 drug and some drugs may
include more than 1 entity
Resource (% of ophthalmologists
prescribing/requesting)
Belgium
France
Germany
The Netherlands Norway
Portugal
Sweden
Topical drugs
80.3
19.8
16.5
16.5
14.3
13.2
50.0
0
97.1
33.1
32.0
28.5
3.5
0
71.1
2.6
57.0
11.5
0
1.34
1.18
0.10
0.06
10.0
0
0
10.0
0
0
57.5
10.0
7.5
25.0
5.0
10.0
62.5
0
62.5
0
20.0
0
20.0
0
0
0
80.0
0
80.0
0
22.0
1.06
1.04
0.02
0
0
0
0
0
0
0
100
0
100
0
0
137.3
31.0
57.3
27.0
0
22.0
61.5
15.0
46.5
0
corticosteroids
mast cell stabilisers
antihistamines
antimicrobials
others
Systemic drugs
corticosteroids
antihistamines
others
Laboratory examination, procedures
Visits (number)
general practitioners
ophthalmologists
other specialists
102.5
0
50.0
0
82.5
20.0
27.5
1.61
1.41
0.09
0.11
17.0
2.32
1.68
0.48
0.16
9.3
2.0
1.58
1.56
0.01
0.01
1.03
1.01
0.02
0
1.87
1.21
0.47
0.19
IMS data.[53-57] These data were not available in
Sweden and Portugal, and they were fixed arbitrar-
were hospital-based and their activities were regu-
lated by both the national authorities and the global
budget approach. As the Dutch visit tariff was far
lower than the opportunity cost, specific surveys[58]
were conducted to estimate ‘real cost’.
Lastly, levocabastine was not available on the
Portuguese market. The cost of isospaglumic acid
(NAAGA) was taken as a proxy of the cost of levo-
cabastine, since it was the most often prescribed
drug in AAC.[59] From a medical point of view, this
was a conservative approach since isospaglumic
acid has never established superiority over levoca-
bastine.
:
ily at 1 1, since access to ophthalmologists was
unrestricted, like in Belgium, France and Germany.
The Netherlands presented a special case since most
of the ophthalmologists were hospital-based and
no statistics were available for the others. Sources
of resource utilisation cost data came mainly from
year 2000 data; a 2% yearly rate was applied to data
prior to year 2000, before converting costs to euros.
In all countries, prices of ‘listed-for-reimburse-
ment’ drugs[27,34,38-52] were controlled and refer-
enced by the national authority. The list of official
tariffs was used.
Except for ophthalmologists in The Netherlands
[costs regulated by COTG (Centraal Orgaan Tar-
ieven Gezondheidszorg)] and the private sector in
Portugal, the cost of visits is also regulated by the
state.[27,28,30,31,34,36,37,46-50] Some statistics estimat-
ing the extra cost linked to public activities were
available in France,[29] and the cost of a production
unit was increased by 10% in Germany to take into
account the 12% of private practice. In Portugal,
the cost of private visits was estimated using the
price recommended by the Portuguese Medical As-
sociation.[35] In The Netherlands, ophthalmologists
Results
Clinical Efficacy and Effectiveness
From day 7 to day 42, patients treated with eme-
dastine had a lower itching score (p < 0.0001) and
a lower redness score (p < 0.0001) than patients
treated with levocabastine.
Figure 1 describes time to failure by treatment
group. The mean time to failure was 8 days across
all groups of patients pooled together. Fewer pa-
tients treated with emedastine (p < 0.02) had treat-
ment failure (21.10%) compared with those treated
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

260
Pinto et al.
with levocabastine (33.04%). The median time with-
out symptoms (redness and itching) was 1.75 days
with levocabastine and 10.25 days with emedastine:
this difference was statistically significant (p < 0.01)
[patients treated with emedastine had 8.5 days more
without symptoms].
in Germany, The Netherlands and Norway prescri-
bed 1 product on average. Practitioners in Belgium
and Sweden asked their patients to re-visit more
often than those in France and Portugal. Belgian
ophthalmologists and German GPs requested the
most laboratory examinations.
Topical corticosteroids were mostly prescribed
by ophthalmologists, whereas systemic antihista-
mines were mostly used by GPs. The exception was
France, where ophthalmologists coprescribed an oral
antihistamine with a topical compound in 72.5% of
patients. No major differences between GPs and
ophthalmologists were found either in the average
number of visits or in the requests for laboratory
examination.
Table IV describes the standard cost of failure
of the first-line antihistamine in each country for
GPs and for ophthalmologists. Outside GP visits in
Sweden, visit costs represented 1.5 to 3 times the
cost of drugs, except in Portugal where this ratio
was closer to 18 for ophthalmologists. The cost
of an ophthalmologist visit is about 50% or more
higher than that of a GP visit, except in Germany
where it was 50% lower. An especially large differ-
ence was observed for Portuguese ophthalmologists
(EUR96.85) and Portuguese GPs (EUR37.97).
Table V describes the difference in costs of the
2 strategies according to the treatment provider per-
spective. The cost of successful emedastine treat-
ment was higher than that of levocabastine: the extra
cost of 1 emedastine bottle is not fully balanced by
the 3 (10 in France) additional therapy days avail-
The safety of the 2 drugs was found to be sim-
ilar, and both compounds were well tolerated.
Economic Analysis
The size of 1 bottle of levocabastine is 4ml (ex-
cept in France, 3ml), the size of 1 drop is 39.6μl
(internal data) and the mean number of drops per
bottle is 101 (75 in France). With 2 drops a day in
each eye (the dosage used in the clinical study), the
number of days of treatment available from a bottle
of levocabastine was 25 (18 days in France). Be-
cause the emedastine bottle is bigger (5ml) and the
drop size is smaller (34μl), the number of days of
treatment per bottle of emedastine at the same daily
dosage was limited by the expiry date to 28. To
summarise, because the drop size of emedastine
was smaller and because its bottle size was bigger,
3 additional days were available per bottle.
As well as this daily cost, tables II and III de-
scribe the resource utilisation of ophthalmologists
and GPs, respectively, when facing failure of the
first-line antihistamine therapy. Some differences
were found between countries and between GPs
and ophthalmologists. Prescriptions in France and
Sweden were usually for 2 products (true for both
GPs and ophthalmologists), whereas practitioners
Table IV. Cost per patient (in euro; 2000 values) of failure of first-line topical antihistamine treatment of allergic conjunctivitis
Resource
Ophthalmologist
drugs
visit
procedure
Total
Belgium
France
Germany
The Netherlands Norway
Portugala
Sweden
18.59
37.70
6.37
21.13
47.96
9.01
14.21
40.43
0.17
20.54
64.60
4.98
25.88
36.90
5.00
5.25
91.60
0.00
30.66
49.98
4.81
62.66
78.10
54.81
90.12
67.78
96.85
85.45
General practitioner
drugs
visit
procedure
Total
12.89
35.85
1.28
16.35
28.00
16.39
60.74
32.58
53.56
4.81
16.70
27.56
1.60
27.97
14.21
2.51
9.88
28.09
0.00
36.50
20.50
0.96
50.02
90.95
45.86
44.69
37.97
57.96
a
Cost of isospaglumic acid (NAAGA) taken as a proxy of the cost of levocabastine, since the latter is not available on the market.
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

Cost Effectiveness in Allergic Conjunctivitis
261
Table V. Cost per patient (in euro; 2000 values) of 42 days of treatment started either with levocabastine or with emedastine. The economic
perspective is that of the treatment provider
Cost
Belgium
France
Germany
The
Netherlands
Norway
Portugala
Sweden
Levocabastine
cost of success
cost of failure
Total
14.48
22.05
36.53
8.59
24.48
33.07
17.82
27.28
45.10
17.00
20.45
37.45
16.58
20.00
36.58
12.87
25.45
38.32
17.94
28.12
46.06
Emedastine
cost of success
cost of failure
Total
Incremental cost (emedastine –0.02
– levocabastine)
21.79
14.72
36.51
23.18
16.95
40.13
0.83/DFS
24.12
17.85
41.97
–3.13
24.57
13.67
38.24
0.09/DFS
23.55
13.31
36.86
0.03/DFS
16.57
16.44
33.01
–5.31
25.01
18.48
43.49
–2.57
Conclusion
Emedastine
dominant
Cost of isospaglumic acid (NAAGA) taken as a proxy of the one of levocabastine since the latter is not available on the market.
Emedastine
cost effective dominant
Emedastine Emedastine
Emedastine Emedastine Emedastine
cost effective cost effective dominant
dominant
a
DFS = day free of symptoms.
able in a bottle of emedastine. Because emedastine
was more effective than levocabastine, the costs of
failure of the former were lower. Consequently, 42
days of treatment started with emedastine costs that
were less than those with levocabastine, except for
in France, The Netherlands and Norway where less
than EUR1 had to be paid per additional symptom-
free day. Since emedastine provided more days with-
out symptoms, this model suggests that emedastine
economically dominates levocabastine in the treat-
ment of AAC in Belgium, Germany, Portugal and
Sweden, and reaches a low cost-effectiveness ratio
in France, The Netherlands and Norway, although
no official thresholds have been published in AAC.
Discussion
Levocabastine was chosen as the comparator
since this was one of the most potent and well doc-
umentedtopical compoundsinthetreatmentofAAC.
One study[61] has demonstrated the equivalence of
levocabastine and lodoxamide in the treatment of
AAC, whereas a comparison between lodoxamide
and isospaglumic acid showed the superiority of
lodoxamide on ocular symptoms,[62,63] indirectly
suggesting a better efficacy for levocabastine over
isospaglumic acid. Individual studies have demon-
strated the equivalence (a trend in favour of levo-
cabastine without statistical significance)[64,65] or
the superiority[66-68] of levocabastine over cromo-
glycates. This was confirmed by a review of the
clinical trials performed by Janssens[69] and Noble
and McTavish,[70] supporting the better efficacy of
levocabastine, although the efficacy of sodium cro-
moglycate 2% is still debatable since this low con-
centration is associated with a mediocre level of
efficacy,[62,66]even against placebo. Lastly, azelast-
ine did not show any superiority over levocabast-
ine.[71]
Sensitivity Analysis
Sensitivity analyses were performed on the par-
ameters associated with a high level of uncertainty:
the ratio of private/public paying patients, the ratio
of ophthalmologists/GPs in the treatment of AAC
and extreme values of the first-line failure cost.
The results of the main sensitivity analyses are
presented in table VI and figure 2. In Germany and
Portugal, emedastine was always less expensive than
levocabastine. In the other countries, the additional
cost was always lower than EUR1.5 per symptom-
free day.
Our results were based on modelling since no
economic parameters were collected during the cli-
nical trial. However, it has to be emphasised that
the primary objective of the clinical trial was to
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

262
Pinto et al.
evaluate redness and itching, and, therefore, sev-
eral compulsory visits were scheduled in the pro-
tocol to determine the onset of action. As most of
the cost would have been protocol driven in this
trial,[72] no resource utilisation data were collected.
The cost of failure after first-line treatment was
estimated from a small sample of ophthalmologists
and GPs since we did not expect large variability.
This was confirmed by our findings.
No costs related to adverse effects were inclu-
ded in this model, since the tolerability of the 2
drugs was found to be similar, and, therefore, they
would not have contributed to the cost difference
between the 2 strategies.
Indirect costs were not taken into account. This
was due to the lack of national statistics. Neverthe-
less, the number of days of work missed is expected
to be low for a nonsevere and acute disease; con-
sequently, indirect costs should have very few con-
sequences for the results of our study. Indeed, since
the full activity of antihistamines is reached after 3
or 5 days of treatment, patients could make up for
this lost time when returning to work. Even in case
of failure, patients could go back to the practitioner
within 3 days and obtain topical corticosteroids,
which are efficacious very quickly. Lastly, not in-
cluding indirect costs leads to an underestimation
of the true benefits of emedastine.
We took efficacy measures from our clinical trial
as a proxy of effectiveness since, at that time, no
national data were available. This is in accordance
with most of the international health economics
guidelines. In fact, if withdrawals were accounted
for in clinical trials, and if these were integrated
when calculating the costs of failure, efficacy would
have been close to effectiveness. However, the dos-
ages of the 2 drugs were fixed by the protocol and
might, therefore, have been far from usual practice.
Lastly, concomitant medications were fixed by the
protocol and their use might differ in usual prac-
tice.
Therefore, despite the methodological difficul-
ties, the sensitivity analyses we performed showed
the robustness of our findings, including the ex-
treme sensitivity analysis based on the minimal and
maximal costs reported during the practitioner
interviews. Emedastine dominated levocabastine
except in France, The Netherlands and Norway,
where the incremental cost was lower than EUR1
per symptom-free day.
France had the highest incremental cost-effec-
tiveness ratio, because the price of levocabastine in
France is very low, 50% less than in the other coun-
tries. Overall, drug prices in France are lower than
in other EU countries, a fact that is even more pro-
nounced with ophthalmological drugs.[73] There-
fore, for a new drug to be more cost effective, its
incremental efficacy must be very high, and it is
possible for ceiling effects to jeopardise the valua-
tion of innovation. Nevertheless, the incremental
cost to be paid per day without symptoms is rather
low. On a French budget approach, knowing that
about 300 000 episodes of AAC are treated with
levocabastine per year,[53] an additional EUR2.1 mil-
lion would have to be allocated for the treatment of
AAC if all patients were switched to emedastine,
which represents less than 2.5% of the drug costs
for this disease.
Table VI. Sensitivity analyses of incremental cost (a negative figure represents savings with emedastine) in euros (2000 values) of 42 days
of treatment started either with levocabastine or with emedastine according to different scenarios. The economic perspective is that of the
treatment provider
Scenario
Belgium
France
Germany
The
Netherlands
Norway
Portugala Sweden
Patients treated by general practitioner 0.09/DFS
0.96/DFS
0.70/DFS
1.47/DFS
0.02/DFS
–5.72
–1.41
–0.24
0.14/DFS
–4.09
0.37/DFS
–3.56
0.08/DFS
–2.05
0.34/DFS
–0.65
–1.79
–8.82
–4.43
–9.85
–0.93
–4.21
0.21/DFS
–8.01
Patients treated by ophthalmologist
Minimal cost of failure
–0.82
0.30/DFS
–2.54
Maximal cost of failure
–10.36
a
Cost of isospaglumic acid (NAAGA) taken as a proxy of the cost of levocabastine since the latter is not available on the market.
DFS = day free of symptoms.
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

Cost Effectiveness in Allergic Conjunctivitis
263
2
0
1.47/DFS
0.83/DFS
0.37/DFS
0.09/DFS
0.30/DFS
0.34/DFS
0.21/DFS
0.02/DFS
−0.02
0.03/DFS
−0.24
−3.13
−0.65
−2
−2.57
−2.54
−3.56
−4
−6
−4.43
−5.31
−8.01
−8
−10
−12
−9.85
−10.36
Belgium
France
Germany
The Netherlands
Country
Norway
Portugal
Sweden
Fig. 2. Incremental costs and cost-effectiveness ratios of emedastine versus levocabastine. Confidence limits were calculated
according to the extreme scenario. Negative values reflect savings from emedastine. Positive values reflect the incremental cost to
be paid per incremental day free of symptoms (DFS).
From a macroeconomic point of view, the cost-
shifting generating the savings observed in several
countries is likely to occur since it is directly linked
to the incremental efficacy of emedastine and, there-
fore, does not require any adaptations of the organ-
isation of the healthcare system: emedastine avoids
some visits only because of its better efficacy (cost-
shifting within and only within the outpatient en-
velope).
Finally, besides the economic consequences, ef-
ficacious treatment is the way to avoid the short
and long term complications of allergic conjunc-
tivitis, which were not taken into account in our
short term modelling approach. Moreover, symp-
toms of AAC are linked with deterioration of qual-
ity of life.[12,13] The isotropic hypothesis suggests
that increasing the number of symptom-free days
will have positive consequences for quality of life.
France, Germany, The Netherlands, Norway, Por-
tugal and Sweden.
Acknowledgements
The clinical trial was supported by an unrestricted grant
from Alcon Laboratories Inc., Fort Worth, Texas, USA, and
was conducted according to the local laws of each participat-
ing country. The economic analysis was supported by an
unrestricted grant from Alcon France SA, Rueil-Malmaison,
France.
References
1. Allansmith MR. Immunology of the external ocular tissues. J
Am Optom Assoc 1990; 61: S16-S22
2. Friedlaender MH. Allergic and multisystem disease. In: Fried-
lander MH, editor. Allergy and immunology of the eye. 2nd
ed. New York (NY): Raven Press, 1993: 259-60
3. Weeke ER. Epidemiology of hay fever and perennial allergic
rhinitis. Monogr Allergy 1987; 21: 1-20
4. Smith JL. Epidemiology and natural history of asthma, allergic
rhinitis and atopic dermatitis. In: Middleton Jr E, Reed CE,
Elli E, et al. editors. Allergy: principle and practice. Vol. 2. St
Louis (MO): CV Mosby, 1988: 891-929
5. Croner S, Kjellman MH. Development of atopic disease in re-
lation to family history and cord blood IgE levels. Pediatr
Allergy Immunol 1990; 1: 14-20
Conclusion
6. Naclerio R. Allergic conjunctivitis. N Engl J Med 1991; 325:
860-9
On the basis of the results of a randomised clin-
ical trial, and resources and costs associated with
treatment failure estimated through clinician inter-
views, our economic model found that emedastine
is a cost-effective treatment of AAC in Belgium,
7. Dodge R, Burrows B. The prevalence and incidence of asthma
and asthma-like symptoms in a general population sample.
Am Rev Respir Dis 1980; 122: 567-75
8. Dart JKG, Buckley RJ, Monnickendan M, et al. Perennial al-
lergic conjunctivitis: definition, clinical characteristics and
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

264
Pinto et al.
prevalence. A comparison with seasonal allergic conjunctivi-
tis. Trans Ophthalmol Soc U K 1986; 105 (5): 513-20
9. Friedlander MH, Okumuto M, Kelley L. Diagnosis of allergic
conjunctivitis. Arch Ophthalmol 1984; 102: 1198-9
10. Abelson MB, Weintraub D. Levocabastine eye drops: a new
approach for the treatment of allergic conjunctivitis. Eur J
Ophthalmol 1994; 4 (2) : 91-101
11. Knight A. The role of levocabastine in the treatment of allergic
rhinoconjunctivitis. Br J Clin Pract 1994; 48 (3): 139-43
12. Juniper EF, Thompson AK, Ferrie PJ, et al. Validation of the
standardized version of the Rhino-conjunctivitis Quality of
Life Questionnaire. J Allergy Clin Immunol 1999; 104: 364-9
13. De Graaf-in’t Veld T, Koenders S, Garrelds IM, et al. Relation-
ship between nasal hyperreactivity, quality of life, and nasal
symptoms in perennial rhinitis. J Allergy Clin Immunol 1996;
98 (3): 508-13
14. US Department of Health and Human Services (DHHS). Asthma
and allergies: an optimistic future. Bethesda (MD): DHHS,
1980. DHHS publication no. NIH80-388.
15. McMenamin P. Cots of hay fever in the United States in 1990.
Ann Allergy 1994; 73: 35-9
16. Drummond M, Stoddard G, Torrance G. Methods for economic
evaluation of health care programmes. Oxford: Oxford Uni-
versity Press, 1989
17. Verin Ph, Easty LD, Secchi A, et al. Clinical evaluation of twice
daily emedastine 0.05% eye drops (Emadine® eye drops) ver-
sus levocabastine 0.05% eye drops in patients with allergic
conjunctivitis. Am J Ophthalmol. In press
28. Nomenclature générale des actes professionnels. Nomenclature
des actes professionnels. Paris: Union des caisses nationales
de Sécurité Sociale, 2000
29. Dumesnil S, Grandfils N, Le Fur P, et al. Santé, soins et protec-
tion sociale en 1997, Enquête sur la santé et la protection
sociale 1997, N°1255. Paris: CREDES éditeur, 1999 Feb
30. Broglie MG, Pranschke-Schade S, Schade H-J, et al. Gebühren
– Handbuch 2000. Kommentar für Ärzte. EBM. GOÄ. Wies-
baden: Inter Medical Kommunikations GmbH, 2000
31. Grunddaten zur Vertragsärzlichen Versorgung in der Bundes-
republik Deutschland. Köln: Kassenärztlichen Vereinigun-
gen, 2000
32. LINH (Landelijk Informatie Netwerk Huisatsenzorg) jaarrapport
Kompas. Utrecht: Nederlands Instituut voor Onderzoek van
de Gezondheidszorg (Nivel), 1998
33. Data on file, NMA, Norwegian Medical Association (Oslo), 2000
34. Portaria no. 348-B/98, Diário da República, Número 138/98, 2.°
Suplemento, I-Série B, de 18 de Junho 1988, Lisboa
35. Codigo de nomenclatura e valor relativo de actos medicos. Lisboa:
Ordem dos medicos, 1998
36. Landstingförbundets årliga sammanställning av patientavgifter
i öppen hälso-och sjukvård ån 2000. Stockholm: The Federa-
tion of Swedish County Council, 2000
37. Socialstyrelsen. Socialstyrelsons meddelandeblad nr 16/99: om
högkostnadsskydd vid köp av läkemedel m.m. från och med
den 1 juni 1999. Stockholm: The National Insurance Board
Statistics, 1999
38. Associations Pharmaceutiques Belges. Tarif des spécialités
pharmaceutiquesn n°2. Bruxelles: La Société Royale, 2000 Apr
39. Edition du Vidal. Paris: Vidal, 2000
40. Lauer taxe. Fürth: Pharma Daig & Lauer GmbH, Fischer Soft-
ware GmbH, 2000 May
18. Pocock SJ. Clinical trials: a pragmatic approach. New York (NY):
John Wiley and Sons, 1983
19. Council Regulation (EC) No. 2866/98 of 31 December 1998 on
the conversion rates between the euro and the currencies of
the Member States adopting the euro. Official Journal of the
European Communities 1999 Dec 31: L359/1-L359/2
20. Alves Da Silva E, Gouveia PC, Sampaio C, et al. Methodolog-
ical guidelines for economic evaluation studies of medicines.
Lisbon: Ministry of Health, National Institute of Pharmacy
and Medicines (Infarmed), 1998
41. Z-Index, May 2000. Den Haag: Uitgave van Z-Index, 2000
42. Van der Kuy A, editor. Farmacotherapeutisch Kompas. Amstel-
veen: College voor Zorgverzekeringen, 1999
43. Felleskatalogen. 42nd ed. Oslo: Elanders Publishing, 2000
44. Simposium terapêutico. Enciclopedia de especialidades farma-
cêuticas Portuguesas. Lisbon: Ed Simposium, 1999
45. LINFO. Farmaceutiska Specialiteter i Sverige (FASS) 2000.
Stockholm: Läkemedelsinformation AB, 2000
21. Collège des économistes de la santé. Rapport d’une commission
d’experts présidée par le Pr Emile Lévy, sous l’égide du collège
des économistes de la santé. Recommandations de bonnes
pratiques des méthodes d’évaluation économique des stratégies
thérapeutiques. Paris: Société Française d’Evaluation des Soins
et des Technologies, 1997
22. German recommendations for health care economic evaluation
studies. Revised version of the Hannover consensus. Hanno-
ver consensus group. Med Klin 2000; 95 (1): 52-5
23. Belgium Society for Pharmacoepidemiology. A proposal for
methodological guidelines for economic evaluation of phar-
maceuticals. Antwerpen: Belgian Society of Pharmacoepide-
miology – Pharmaco-Economic Committee (ESPE-PEC), 1995
24. Belgium Society for Pharmacoepidemiology. Perspectives for
pharmaco-economic evaluation in Belgium. Antwerpen: Bel-
gian Society of Pharmacoepidemiology – Pharmaco-Economic
Committee (ESPE-PEC), 1998
25. Health Insurance Council. Dutch guidelines for pharmaco-
economics research. Amstelveen: Zickenfondstaad, 1999. Pub-
lication no. CURE/cuo.228/2
26. Norwegian guidelines for pharmaco-economic analysis in
connection with application for reimbursement. Oslo: The
Norwegian Medicines Control Authority Department of Phar-
macoeconomics, 1999 Nov
27. Data on file, Institut National d’Assurance Maladie-Invalidité
(Bruxelles), 2000
46. Nomenclature générale des actes professionnels. Nomenclature
des actes de biologie médicale. Paris: Union des caisses na-
tionales de Sécurité Sociale, 2000
47. Tarieflijst Instellingen. Utrecht: CTG (Centraal Tarieven Ge-
zondheidszorg), 1999
48. Utrecht: CTG (Centraal Tarieven Gezondheidszorg), 2000. TB-
Number 5600-1000-00-1
49. Utrecht: CTG (Centraal Tarieven Gezondheidszorg), 2000. TB-
Number 5600-1500-00-1
50. Utrecht: CTG (Centraal Tarieven Gezondheidszorg), 2000. TB-
Number 5600-1900-00-1
51. Data on file, Rikstrygdeverket (Oslo), 2000
52. Hälso-och sjukvårdens ansvarsnämnd (HSAN). Stockholm:
Avgiftshandboken, 2000
53. Data on file, IMS Data (Paris), 2000
54. Data on file, IMS Data (Bruxelles), 2000
55. Data on file, IMS Data (Frankfurt), 2000
56. Data on file, IMS Data (Amsterdam), 2000
57. Data on file, IMS Data (Oslo), 2000
58. Oostenbrink JB, Koopmanschap MA, Rutten FFH. Handleiding
voor kostenonderzoek: methoden en richtlijnprijzen voor
economische evaluaties in de gezondheidszorg. Amstelveen:
College voor Zorgverzekeringen, 2000
59. Data on file, IMS Data (Lisboa), 1999
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)

Cost Effectiveness in Allergic Conjunctivitis
265
60. Study C-95-54. Internal report. Fort Worth (TX): Alcon Labo-
ratories, 1997. (Data on file)
61. Richard C, Trinquand C, Bloch-Michel E. Comparison of top-
ical 0.05% levocabastine and 0.1% lodoxamide in patients
with allergic conjunctivitis. Study Group. Eur J Ophthalmol
1998; 8 (4): 207-16
62. Denis D, Bloch-Michel E, Verin P, et al. Treatment of common
ocular allergic disorders; a comparison of lodoxamide and
NAAGA. Br J Ophthalmol 1998; 82 (10): 1135-8
63. Gunduz K, Ucakhan O, Budak K, et al. Efficacy of lodoxamide
0.1% versus N-acetyl aspartyl glutamic acid 6% ophthalmic
solutions in patients with vernal keratoconjunctivitis. Oph-
thalmic Res 1996; 28 (2): 80-7
68. Davies BH, Mullins J. Topical levocabastine is more effective
than sodium cromoglycate for the prophylaxis and treatment
of seasonal allergic conjunctivitis. Allergy 1993; 48 (7): 519-24
69. Janssens MM. Levocabastine: a new topical approach for the
treatment of paediatric allergic rhinoconjunctivitis. Rhinol
Suppl 1992; 13: 39-49
70. Noble S, McTavish D. Levocabastine: an update of its pharma-
cology, clinical efficacy and tolerability in the topical treat-
ment of allergic rhinitis and conjunctivitis. Drugs 1995; 50
(6): 1032-49
71. Sabbah A, Marzetto M. Azelastine eye drops in the treatment
of seasonal allergic conjunctivitis or rhinoconjunctivitis in
young children. Curr Med Res Opin 1998; 14 (3): 161-70
72. Drummond MF, Davies L. Economic analysis alongside clini-
cal trials: revisiting the methodological issues. Int J Technol
Assess Health Care 1991; 7: 567-73
64. Ciprandi G, Cerqueti PM, Sacca S, et al. Levocabastine versus
cromolyn sodium in the treatment of pollen-induced conjunc-
tivitis. Ann Allergy 1990; 65 (2): 156-8
65. Frostad AB, Olsen AK. A comparison of topical levocabastine
and sodium cromoglycate in the treatment of pollen-provoked
allergic conjunctivitis. Clin Exp Allergy 1993; 23 (5): 406-9
66. Azevedo M, Castel-Branco MG, Oliveira JF, et al. Double-blind
comparison of levocabastine eye drops with sodium cromo-
glycate and placebo in the treatment of seasonal allergic con-
junctivitis. Clin Exp Allergy 1991; 21 (6): 689-94
67. Odelram H, Bjorksten B, Klercker T af, et al. Topical levo-
cabastine versus sodium cromoglycate in allergic conjuncti-
vitis. Allergy 1989; 44 (6): 432-6
73. Le secteur de l’ophthalmologie pharmaceutique. Revue de
l’ophtalmologie Française 1999; 117: 52-5
Correspondence and offprints: Dr Gilles Berdeaux, Alcon
France, Rue Henri Sainte-Claire Deville 4, F-92563 Rueil-
Malmaison Cedex, France.
E-mail: gilles.berdeaux@alconlabs.com
© Adis International Limited. All rights reserved.
Pharmacoeconomics 2001; 19 (3)