14819 April 15/97/97 CMAJ /Page 1 1 2 3
`
`First-line treatment of seasonal
`(ragweed) rhinoconjunctivitis
`A randomized management trial comparing
`a nasal steroid spray and a nonsedating
`antihistamine
`
`Elizabeth F. Juniper,* MCSP, MSc; Gordon H. Guyatt,*† MD, MSc;
`Penelope J. Ferrie,* BA; Lauren E. Griffith,* MSc
`
`Abstract
`
`Objective: To determine whether better health-related quality of life (HRQL) is
`achieved by initiating treatment of seasonal (ragweed) rhinoconjunctivitis (hay
`fever) with a nasal steroid (fluticasone) backed up by a nonsedating antihista-
`mine (terfenadine) or whether it is better to start with the antihistamine and add
`the nasal steroid when necessary.
`Design: Randomized, nonblind, parallel-group management study during the
`6 weeks of the ragweed pollen season in 1995.
`Patients: Sixty-one adults with ragweed pollen hay fever recruited from patients
`who had participated in previous clinical studies and from those who responded
`to notices in the local media.
`Setting: Southern Ontario.
`Interventions: Nasal steroid group: 200 µg of fluticasone nasal spray when needed
`(up to 400 µg/d) starting about 1 week before the ragweed pollen season and con-
`tinued throughout, with 1 to 2 tablets of terfenadine daily (maximum 120 mg/d) if
`needed. Antihistamine group: 1 60-mg tablet of terfenadine when needed (maxi-
`mum 120 mg/d) starting about 1 week before the ragweed pollen season and con-
`tinued throughout, with 200–400 µg/d of fluticasone nasal spray (maximum 400
`µg/d) if needed.
`Outcome measures: HRQL before, at the height of and toward the end of the rag-
`weed pollen season; HRQL was measured using the Rhinoconjunctivitis Quality
`of Life Questionnaire.
`Results: Overall, HRQL tended to be better in the group of patients whose first-line
`treatment was with fluticasone (p = 0.052), but the difference between the 2
`groups was small and not clinically important. Just over half (52% [16/31]) of
`the patients in the fluticasone group did not need additional help with terfen-
`adine, whereas only 13% (4/30) of those in the terfenadine group did not need
`additional help with fluticasone (p = 0.002).
`Conclusions: There is little difference in the therapeutic benefit between the 2 ap-
`proaches for the treatment of ragweed pollen hay fever. Therefore, the approach
`to treatment should be based on patient preference, convenience and cost. Re-
`gardless of the treatment, at least 50% of patients will need to take both types of
`medication in combination to control symptoms adequately.
`
`Résumé
`
`Objectif : Déterminer si l’on améliore la qualite de vie liée à la santé par un traite-
`ment initial de la rhinoconjonctivite (fièvre des foins) saisonnière (herbe à poux)
`aux stéroïdes par voie nasale (fluticasone) appuyé par un antihistaminique non
`sédatif (terfénadine), ou s’il est préférable de commencer par l’antihistaminique
`et d’ajouter les stéroïdes par voie nasale au besoin.
`Conception : Étude randomisée, non à l’insu, de traitement en groupe parallèle au
`cours des 6 semaines de la saison du pollen de l’herbe à poux en 1995.
`
`Evidence
`Études
`
`From the Department of
`*Clinical Epidemiology and
`Biostatistics and †Medicine,
`McMaster University Health
`Sciences Centre, Hamilton,
`Ont.
`
`This article has been peer
`reviewed.
`
`Can Med Assoc J 1997;156:1123-31
`
`Exhibit 1031
`IPR2017-00807
`ARGENTUM
`
`CAN MED ASSOC J • APR. 15, 1997; 156 (8)
`
`1123
`
`© 1997 Canadian Medical Association (text and abstract/résumé)
`
`000001
`
`

`

`14819 April 15/97 CMAJ /Page 1 1 2 4
`
`Juniper, Guyatt, Ferrie, et al
`
`Patients : Soixante et un adultes souffrant de fièvre des foins causée par le pollen
`de l’herbe à poux et recrutés parmi les patients qui avaient participé à des
`études cliniques antérieurs et parmi les personnes qui avaient répondu à des
`avis dans les médias locaux.
`Contexte : Sud de l’Ontario.
`Interventions : Sujets traités aux stéroïdes par voie nasale : 200 µg de fluticasone
`en vaporisateur nasal au besoin (jusqu’à 400 µg/j) à compter d’environ 1 se-
`maine avant la saison du pollen de l’herbe à poux et pendant toute la saison, et
`1 ou 2 comprimés de terfénadine par jour (maximum : 120 mg/j) au besoin.
`Groupe traité aux antihistaminiques : un comprimé de 60 mg de terfénadine au
`besoin (maximum : 120 mg/j) à compter d’environ 1 semaine avant la saison du
`pollen de l’herbe à poux et pendant toute la saison, et de 200 à 400 µg/j de flu-
`ticasone en vaporisateur nasal au besoin (maximum : 400 µg/j).
`Mesures des résultats : Résultats relatifs à la qualité de vie liée à la santé avant la
`saison du pollen de l’herbe à poux, en plein coeur de la saison et vers la fin de
`celle-ci. Les résultats ont été mesurés au moyen du questionnaire sur la qualité
`de vie liée à la rhinoconjonctivite.
`Résultats : Dans l’ensemble, les patients traités d’abord à la fluticasone avaient ten-
`dance à avoir une meilleure qualité de vie (p = 0,052), mais l’écart entre les 2
`groupes était faible et sans importance sur le plan clinique. Un peu plus de la
`moitié (52 % [16/31]) des patients traités à la fluticasone n’ont pas eu besoin de ter-
`fénadine supplémentaire, tandis que 13 % (4/30) seulement de ceux qui ont été
`traités à terfénadine n’ont pas eu besoin de fluticasone supplémentaire (p = 0,002).
`Conclusions : Il y a peu de différence sur le plan des avantages thérapeutiques en-
`tre les 2 méthodes de traitement de la fièvre des foins causée par le pollen de
`l’herbe à poux. Il faudrait donc choisir le mode de traitement en fonction de la
`préférence du patient, de la commodité et du coût. Peu importe le traitement,
`au moins 50 % des patients devront prendre les 2 médicaments combinés pour
`bien contrôler les symptômes.
`
`At least 25% of adults report experiencing seasonal al-
`
`lergic rhinoconjunctivitis (hay fever),1 and despite ef-
`ficacious over-the-counter drugs about 20% of the
`population seek help from their primary care physician.2
`Hay fever not only produces troublesome symptoms, it also
`impairs normal daily activities and productivity.3–5
`A large number of clinical trials have demonstrated the
`individual efficacy and safety of fast-acting, nonsedating
`antihistamines and inhaled nasal steroids for the treat-
`ment of hay fever. A much smaller number of randomized
`trials have compared antihistamines with nasal steroids.6–11
`Although in most of the comparison studies the results
`tended to favour the latter, the artificial environment of
`the trials (regular and sustained daily use plus double-
`dummy techniques to achieve blinding) bears little resem-
`blance to how patients use these medications in real life.
`It is impossible to determine from all of these studies
`whether it is better to start treatment with an antihista-
`mine and add a nasal steroid for uncontrolled symptoms
`or whether the nasal steroid should be used first, with the
`antihistamine used as back-up.5 We therefore performed a
`management (effectiveness) study to determine whether
`adults with ragweed pollen hay fever would achieve better
`health-related quality of life (HRQL) by starting treat-
`
`1124
`
`CAN MED ASSOC J • 15 AVR. 1997; 156 (8)
`
`ment with fluticasone propionate nasal spray and adding
`terfenadine tablets when needed, or whether they would
`benefit more by starting treatment with terfenadine
`tablets and adding fluticasone nasal spray when needed.
`
`Methods
`
`Patient population
`
`We recruited 61 adults (aged 17–66 years) from south-
`ern Ontario who had either participated in previous clini-
`cal studies or had responded to notices in the local media.
`The entry criteria were as follows: a diagnosis of seasonal
`allergic rhinoconjunctivitis; troublesome nasal symptoms
`requiring medication during the ragweed pollen season
`the previous year; positive skin-prick test result to ragweed
`pollen extract (wheal greater than 3 mm with 25 000
`Noon units); no perennial rhinoconjunctivitis (allergic or
`nonallergic) requiring treatment; no chronic nasal obstruc-
`tion, polyposis or sinusitis; no history of allergen injection
`therapy during the previous 12 months; and no history of a
`serious illness that might impair quality of life. Pregnant
`and nursing mothers were excluded, as were patients with
`other illnesses requiring treatment with antihistamines or
`
`000002
`
`

`

`14819 April 15/97 CMAJ /Page
`
`oral steroid therapy and those who could not communi-
`cate in English. All patients agreed to remain in the rag-
`weed pollen area (southern Ontario) for the duration of
`the study. Participants signed an informed consent form
`that had been approved by the Ethics Committee of the
`McMaster University Health Sciences Centre.
`
`Study design
`
`We used a randomized, nonblind study design to
`compare the 2 treatment regimens over a 6-week period
`that encompassed the ragweed-pollen season in 1995.
`Before the start of the season each patient underwent
`duplicate skin-prick tests with 10-fold serial dilutions of
`ragweed pollen extract (2.5 to 25 000 Noon units) and
`single dilutions of extracts of mixed grass pollen (preva-
`lent in the month before the ragweed season) and of the
`fungal spores Alternaria and Cladesporium (present dur-
`ing the first half of the ragweed season in southern On-
`tario). Sensitivity to the extract in each skin-prick test
`was estimated from the mean of 2 wheal diameters, mea-
`sured at right angles to each other. The estimated sensi-
`tivity to the ragweed pollen extract was determined from
`the mean wheal diameter of the 5 duplicate skin pricks.
`Participants were matched into pairs using the fol-
`lowing criteria in the following order: 1) severity of rag-
`weed pollen hay fever during the previous year; 2) skin
`sensitivity to the ragweed pollen extract; 3) skin sensitiv-
`ity to the fungal spore extracts; 4) skin sensitivity to the
`mixed grass pollen extract; and 5) sex. With the use of a
`random numbers table, 1 patient in each pair was ran-
`domly allocated to start treatment with the nasal steroid
`spray and the other to start with the antihistamine.
`
`Interventions
`
`We provided patients with enough medications for
`the whole ragweed pollen season and gave them both
`oral and written instructions on their optimal use. We
`told all patients that fluticasone nasal spray is a topical
`steroid that is slower acting than terfenadine but that
`nasal steroid sprays, if applied as soon as symptoms de-
`velop, can be used quite effectively as needed.12–14 We
`also told them that terfenadine is a fast-acting, nonsedat-
`ing antihistamine. Compliance with the recommended
`dosing was left entirely to the individual patient’s discre-
`tion. We asked patients to use only the medications we
`provided for their hay fever, not to give it to their friends
`and relatives and to contact us if they experienced any
`troublesome symptoms or adverse effects.
`Patients were told which treatment group they were in
`and provided with the medications only after all baseline
`values of the outcome measures had been recorded.
`
`Seasonal (ragweed) rhinoconjunctivitis
`
`Nasal steroid group
`
`Patients were told that the optimal approach to treat-
`ment was to start using 2 puffs (each puff 50 µg) of fluti-
`casone nasal spray in each nostril each morning (200
`µg/d) on Aug. 8, about 1 week before the start of the
`ragweed pollen season, and to continue with this dosage
`throughout the season. They were told that using the
`nasal spray only when needed might result in less effec-
`tive control of their symptoms. We recommended they
`increase the dose to 2 puffs in each nostril twice daily
`(maximum 400 µg/d) if their nasal symptoms became
`troublesome. If the symptoms continued to be trouble-
`some we advised patients to add terfenadine (60 mg)
`when needed, up to 120 mg/d, and to cut back on the
`terfenadine once the symptoms were controlled.
`
`Antihistamine group
`
`Patients in this group were told that the optimal ap-
`proach to treatment was to start using terfenadine on
`Aug. 8 and to take a 60-mg tablet every morning and
`evening (total 120 mg/d) throughout the ragweed pollen
`season. They were told that using less terfenadine might
`result in less effective control of their symptoms. We ad-
`vised patients to add fluticasone nasal spray when needed
`(1–2 puffs in each nostril, up to a maximum of 400 µg/d)
`if symptoms became troublesome once they were already
`taking the 120 mg of terfenadine daily and to cut back on
`the fluticasone once the symptoms were controlled.
`
`Eye symptoms
`
`We provided all patients with naphazoline eye drops
`and recommended that they use 1 drop in each eye
`when needed, up to 4 times per day. Patients who re-
`ported troublesome eye symptoms in previous years
`were also provided with sodium cromoglycate eye drops
`and advised to supplement the naphazoline eye drops
`with 1 drop of cromoglycate in each eye 4 times per day
`until the symptoms were controlled.
`
`Asthma
`
`Patients with asthma were instructed to continue
`taking their regular asthma medication throughout the
`study. If an inhaled β-agonist was required every day, we
`recommended 200 µg of beclomethasone dipropionate
`twice daily. If patients had already been prescribed an
`inhaled steroid and were needing their β-agonist daily,
`we recommended increasing the steroid dose to that
`recommended for an exacerbation by the physician
`treating their asthma.
`
`CAN MED ASSOC J • APR. 15, 1997; 156 (8)
`
`1125
`
`000003
`
`

`

`Juniper, Guyatt, Ferrie, et al
`
`Outcome measures
`
`Health-related quality of life
`
`Patients were seen 1 week before ragweed pollen was
`expected in the air (the first week of August), at the height
`of the ragweed pollen season (the first week of Septem-
`ber) and toward the end of the season (the third week of
`September). At each visit they were asked to complete the
`Rhinoconjunctivitis Quality of Life Questionnaire.3 This
`28-item disease-specific instrument is designed to mea-
`sure the 7 domains of functional impairment that are
`most important to patients with seasonal allergic rhino-
`conjunctivitis: sleep impairment, non-nasal symptoms
`(e.g., headache and fatigue), practical problems, nasal
`symptoms, eye symptoms, activity limitations and emo-
`tional function. Patients are asked to consider their expe-
`riences during the previous 7 days and to score their de-
`gree of impairment on a 7-point scale (0 = not bothered,
`6 = extremely bothered). The questionnaire has excellent
`reliability, responsiveness and construct validity and has
`been used successfully in a number of clinical trials over
`the last 6 years.3,11–14
`
`Medication use
`
`Patients were asked to return all used and unused flu-
`ticasone bottles and terfenadine packages at the final
`visit. We recorded the weight loss from each bottle of
`fluticasone and the number of terfenadine tablets used.
`To estimate the number of puffs of fluticasone used by
`each patient, we first estimated the mean weight loss per
`puff by weighing a bottle before and after 10 consecutive
`discharges into the air until the bottle was empty.
`In addition to estimating the actual amount of med-
`ication used by each patient, we calculated the number
`of bottles of fluticasone and packages of terfenadine each
`patient would have needed to provide the actual amount
`of medication used.
`
`Statistical analysis
`
`We examined differences between the treatment
`groups using a repeated measures analysis of variance,
`considering p values less than 0.05 (two-sided) as signifi-
`cant. Covariate analysis was used to adjust for differences
`between the 2 groups at baseline. All of the randomized
`subjects were included in the analysis (intention-to-treat
`analysis). The number of puffs of fluticasone used by each
`patient was based on a mean weight per puff of 0.0867 g,
`and the number of bottles of fluticasone that each patient
`needed was based on each bottle containing 170 puffs.
`Terfenadine (Seldane) can be purchased over the counter
`
`1126
`
`CAN MED ASSOC J • 15 AVR. 1997; 156 (8)
`
`in Canada in packages of 12, 24 and 36 tablets. After sur-
`veying about 10 pharmacies in the Hamilton area, we de-
`termined that the 36-tablet package had the highest sales
`during the ragweed pollen season. Therefore, we used
`this size to estimate the number of packages required by
`each patient.
`With sufficient statistical power, even the most trivial
`differences between the treatment groups can reach statis-
`tical significance. To interpret HRQL data that reaches
`statistical significance, it is important to know what mag-
`nitude of change or difference can be considered clinically
`important. The minimal important difference (MID) is
`defined as “the smallest difference in score in the domain
`of interest which patients perceive as beneficial and would
`mandate, in the absence of troublesome side effects or
`excessive cost, a change in the patient’s management.”15
`Using a standardized “anchor-based” method16 we have
`determined that the MID for the Rhinoconjunctivitis
`Quality of Life Questionnaire is about 0.5.17 The sample
`size for our study was determined on the basis of the
`MID, the pooled variance3,12–14 and error rates of α = 0.05
`(two-sided) and β = 0.1.
`
`Results
`
`The profile of the study is summarized in Fig. 1. The
`demographic characteristics and allergy history of the 61
`patients are shown in Table 1. Complete data sets were
`provided by 60 of the patients; the remaining patient, in
`the nasal steroid group, experienced nausea using the flu-
`ticasone and asked to be changed to beclomethasone. In
`keeping with the management study philosophy, this was
`permitted, but the patient failed to keep the final appoint-
`ment.
`Although the patients were carefully matched, those in
`the fluticasone group appeared to have slightly better
`HRQL than those in the terfenadine group before the rag-
`weed pollen season (Fig. 2). Even though this difference
`was small (0.24 for overall quality of life, where MID = 0.5)
`and not statistically significant (p > 0.05) and was probably
`due to residual symptoms induced by grass pollen and fun-
`gal spores, we investigated the treatment effect after doing
`a covariate adjustment for baseline differences.
`For overall rhinoconjunctivitis-specific quality of life
`and for each of the 7 domains covered by the question-
`naire, both groups of patients experienced a deterioration
`in HRQL between the beginning and the height of the
`ragweed season which resolved toward the end of the sea-
`son (Fig. 2, Table 2) (p < 0.001). However, the deteriora-
`tion in HRQL was small, and only in the eye-symptom
`domain could it be considered clinically important.
`At the height of the ragweed pollen season the patients
`whose first-line treatment was with fluticasone tended to
`
`000004
`
`

`

`14819 April 15/97 CMAJ /Page
`
`Seasonal (ragweed) rhinoconjunctivitis
`
`Table 1: Characteristics of patients with seasonal (ragweed) rhino-
`conjunctivitis entered into study comparing fluticasone nasal spray
`with terfenadine tablets as first-line treatment
`
`Treatment group;
`no. of patients*
`Fluticasone
`Terfenadine
`n = 31
`n = 30
`15/16
`16/14
`41.0 (11.4)
`45.7 (10.5)
`
`3.77 (1.23)
`
`3.77 (1.26)
`
`11
`12
`7
`
`10
`11
`9
`15
`23
`
`7
`
`2
`
`14
`12
`5
`
`16
`
`69
`
`13
`25
`
`8
`
`0
`
`Characteristic
`Sex (male/female)
`Mean age (and SD†), yr
`Mean diameter (and SD) of wheals after
`duplicate skin-prick test with 5
`concentrations of ragweed pollen
`extract
`Severity of hay fever symptoms
`the previous year
`Mild
`Moderate
`Severe
`Medications taken for hay fever
`the previous year
`Antihistamine alone
`Nasal steroid alone
`Antihistamine + nasal steroid
`Skin sensitivity to fungal spores
`Skin sensitivity to grass pollen
`No previous experience in clinical
`studies
`Nasal steroid used within 6 weeks
`before randomization
`
`have better HRQL than those whose first-line treatment
`was with terfenadine (Table 2). For overall HRQL, this
`difference was on the borderline of statistical significance
`(p = 0.052); however, the mean difference in scores, after
`we adjusted for differences at baseline, was only 0.11
`(MID = 0.5) and therefore of little clinical importance.
`Similar trends were seen for all domains except the eye-
`symptom domain, for which there was no evidence of any
`difference between the 2 groups. For the nasal-symptom
`domain the difference between the 2 groups was statisti-
`cally significant (p = 0.005), but the difference in scores
`was only 0.21 and still of little clinical importance.
`Table 3 shows the amount of medication used by the
`2 groups. Of the 31 patients in the fluticasone group 16
`(52%) never needed to use any terfenadine, whereas
`only 4 (13%) of the 30 patients in the terfenadine group
`never used fluticasone (p = 0.002). Although we in-
`structed patients not to use more than 2 terfenadine
`tablets per day, the mean use in the terfenadine group
`was 2.07 tablets per day, which suggested that a number
`of patients ignored this instruction.
`
`Discussion
`
`The patients whose first-line treatment of seasonal al-
`lergic rhinoconjunctivitis was with fluticasone nasal spray
`
`*Unless otherwise stated.
`†SD = standard deviation.
`
`Patients with seasonal (ragweed)
`rhinoconjunctivitis meeting inclusion criteria
`(n = 61)
`
`Skin-prick tests using
`• serial dilutions of ragweed pollen extract
`• single dilution of mixed grass pollen extract
`• single dilutions of Alternaria and
`Cladesporium fungal spore extracts
`
`R
`
`Fluticasone nasal spray (200 µg/d,
`up to 400 µg/d if needed), with
`addition of terfenadine tablets
`(60–120 mg/d) if needed
`(n = 31)
`
`HRQL measured
`• before the start of the ragweed
`pollen season (baseline)
`• at the height of the season
`• toward the end of the season
`
`Completed study and provided com-
`plete data sets (n = 30)
`(1 patient did not attend final visit)
`
`Terfenadine tablets (120 mg/d),
`with addition of fluticasone nasal
`spray (100–200 µg/d,
`maximum 400 µg/d) if needed
`(n = 30)
`
`HRQL measured
`• before the start of the rag-
`weed pollen season (baseline)
`• at the height of the season
`• toward the end of the season
`
`Completed study and provided
`complete data sets (n = 30)
`
`Fig. 1: Trial profile. See Methods for inclusion criteria. R = randomization, HRQL = health-related quality of life.
`
`CAN MED ASSOC J • APR. 15, 1997; 156 (8)
`
`1127
`
`000005
`
`

`

`Juniper, Guyatt, Ferrie, et al
`
`Overall quality of life
`
`Nasal symptoms
`
`Practical problems
`
`Non-nasal symptoms
`
`Activity limitations
`
`Emotional function
`
`Sleep impairment
`
`Eye symptoms
`
`Fig. 2: Mean scores for HRQL for patients with seasonal (ragweed) rhinoconjunctivitis measured before, at the height of and to-
`ward the end of the ragweed pollen season. Solid lines represent patients whose first-line treatment was fluticasone nasal spray,
`with terfenadine tablets as back-up; broken lines represent patients whose first-line treatment was terfenadine tablets, with flu-
`ticasone nasal spray as back-up. Scores range from 0 (not bothered) to 6 (extremely bothered).
`
`1128
`
`CAN MED ASSOC J • 15 AVR. 1997; 156 (8)
`
`000006
`
`

`

`Seasonal (ragweed) rhinoconjunctivitis
`
`highly controlled conditions does the intervention have a
`biological effect?) to the management or effectiveness
`study (what is the effect on clinically important outcomes
`in real life?). There are a large number of explanatory
`studies of inhaled nasal steroids and nonsedating antihist-
`amines, none of which has taken into account how pa-
`tients use these medications outside the artificial environ-
`ment of the explanatory clinical trial. We wanted our
`study to be as close to real life as possible and to provide
`practical information for clinicians. Therefore, we chose a
`design as close to the management end of the continuum
`as possible. That was why we omitted a placebo group,
`which would have required a double-dummy design, in-
`troduced artificiality into the study, interfered with patient
`
`Table 3: Medication use during study period
`
`Treatment group
`
`Fluticasone
`
`Terfenadine
`
`04
`
`26
`
`16
`0
`15
`
`5.27 (1.62)
`
`2.61 (1.97)
`
`4
`17
`9
`35
`
`07
`
`24
`55
`
`0.13 (0.28)
`
`2.07 (0.43)
`
`006
`
`24
`84
`
`16
`13
`
`20
`
`17
`
`Medication use
`No. of patients using
`Fluticasone alone
`Terfenadine alone
`Both
`Fluticasone
`Mean no. of puffs daily per patient
`(and SD)
`No. of patients using
`0 bottles
`1 bottle
`2 bottles
`Total no. of bottles used
`
`Terfenadine
`Mean no. of tablets daily per patient
`(and SD)
`No. of patients using
`0 packages
`1 package*
`2 packages
`3 packages
`Total no. of packages used
`
`*One package = 36 tablets.
`
`tended to experience better HQRL during the ragweed
`pollen season than those who started with terfenadine.
`However, the differences in mean scores between the 2
`groups at the height of the season were small and of little
`clinical importance. Because there was little difference in
`the therapeutic benefit, even for eye symptoms, between
`the 2 regimens, other factors such as patient preference
`(patient perception of efficacy, patient preference for topi-
`cal or systemic preparations, and side effects), conve-
`nience and cost should be considered when making treat-
`ment recommendations. With regard to convenience and
`cost, it is noteworthy that 52% of the patients who started
`with fluticasone never needed additional terfenadine,
`whereas only 13% of those who started with terfenadine
`managed without additional fluticasone.
`Even at the height of the ragweed pollen season pa-
`tients experienced minimal impairment of HRQL (Fig.
`2). We recognize that there was no placebo control group.
`Nevertheless, all of the patients had moderate to severe
`sensitivity to ragweed pollen and a history of troublesome
`symptoms the previous year. Therefore, these results
`strongly suggest that both treatment regimens were effec-
`tive. The regimens had 3 important features that we be-
`lieve contributed to the success. First, the back-up med-
`ication was added when the first medication, on its own,
`was insufficient to control symptoms. Patients frequently
`change medication at the height of the season, complain-
`ing that “nothing works” and not realizing that 2 different
`types of treatment in combination may be necessary. Sec-
`ond, the patients were advised to start taking their med-
`ications either just before pollen was expected in the air or
`immediately after they experienced their first symptoms.
`It is much easier to keep symptoms under control from
`the beginning than to try and bring severe symptoms un-
`der control. Third, the patients were given written in-
`structions on the use of the medications.
`In designing clinical trials, there is a continuum from
`the explanatory or efficacy study (under optimum and
`
`Table 2: Differences in mean scores for health-related quality of life (HRQL)* between the treatment
`groups (after adjustment for differences at baseline)
`
`Difference in mean scores†
`
`p value§
`
`Sleep impairment
`Non-nasal symptoms
`Practical problems
`Nasal symptoms
`Eye symptoms
`Activity limitations
`Emotional function
`
`−0.03
`0.17
`0.15
`0.21
`−0.27
`0.20
`0.23
`
`0.15
`0.09
`0.39
`0.31
`0.04
`0.33
`0.05
`
`0.054
`0.274
`0.015
`0.005
`0.614
`0.037
`0.165
`
`0.0037
`0.0052
`0.0001
`0.0035
`0.0001
`0.0002
`0.0003
`
`*HRQL determined using Rhinoconjunctivitis Quality of Life Questionnaire3 scores (0 = not bothered, 6 = extremely bothered).
`†Minimal important difference = 0.5.
`‡Positive differences indicate better HRQL for patients in the fluticasone group.
`§Repeated measures analysis with covariate adjustment for differences in baseline score.
`
`CAN MED ASSOC J • APR. 15, 1997; 156 (8)
`
`1129
`
`000007
`
`

`

`Juniper, Guyatt, Ferrie, et al
`
`dosing choices and probably altered the results. Our study
`design allowed many opportunities for patients to make
`their own decisions. For instance, patients were free to
`read and respond to the package inserts for both flutica-
`sone and terfenadine. If they asked for additional informa-
`tion on allergen avoidance, it was provided. We advised
`patients on the best way to use the medications; they were
`free to follow or ignore this advise.
`There is no doubt that some patients like to keep their
`symptoms as well controlled as possible and take their
`medication regularly and prophylactically throughout the
`entire pollen season. Others prefer to tolerate some mild
`impairments in order to keep medication use to a mini-
`mum. It was this observation, made a number of years
`ago, that led us to compare the regular use of beclometha-
`sone dipropionate nasal spray with its use as needed for
`seasonal allergic rhinoconjunctivitis.12,13 Although explana-
`tory studies suggested that prophylactic, regular use of
`nasal steroids should provide optimum symptom control,
`such a regimen is unacceptable to patients who like to
`keep medication use to a minimum and who know that
`the condition will resolve spontaneously at the end of the
`pollen season. Our randomized trials showed that there
`was minimal impairment of HRQL at the height of the
`pollen season in the both the regular and the as-needed
`treatment groups.12,13 When we examined patient satisfac-
`tion, most of the patients in the group instructed to use
`the medication as needed were very satisfied with the level
`of symptom control.13 It was on the basis of those find-
`ings, the results from another management study of nasal
`steroid use for seasonal allergic rhinoconjunctivitis14 and
`the recognition that some patients want to minimize their
`medication use that we decided, in the present study, to
`tell patients how to use nasal steroids effectively on an as-
`needed basis.
`Although we tried to replicate real life as much as pos-
`sible, there were 3 problems that we could not overcome
`and that may have affected the results. First, our patients
`were volunteers. Although they represented both sexes
`and a wide range of age, academic achievement and so-
`cioeconomic backgrounds, they were all interested in the
`management of their condition. Second, our patients
`were provided with hay fever medication before the rag-
`weed pollen season. In real life some patients become se-
`verely symptomatic and limited before they buy medica-
`tions or seek help. Third, none of our patients paid for
`their medications.
`We did not compare costs in the 2 treatment groups
`because they differ greatly across national health care sys-
`tems. Instead, we calculated the number of bottles of fluti-
`casone and packages of terfenadine patients needed so
`that direct costs in any given country can be determined.
`In some countries both nasal steroid sprays and nonsedat-
`
`1130
`
`CAN MED ASSOC J • 15 AVR. 1997; 156 (8)
`
`ing antihistamines are available over the counter, and the
`costs are borne entirely by the patient. In other countries,
`both are available only by prescription, and the cost of the
`drugs, the dispensing fees and the physician visits (possi-
`bly 2 or more) are carried by the health care provider.
`Elsewhere it is a mixture, often with the costs of the drugs
`being borne by different payers. In addition, we did not
`attempt to calculate indirect costs. However, because
`there was little evidence of any difference in HRQL be-
`tween the 2 treatment groups, indirect costs (e.g., loss of
`earnings) would have probably been similar.
`Our primary aim was to compare drug types, but for
`study purposes we had to select a representative of each
`class. We selected fluticasone and terfenadine because
`they are used extensively for hay fever and we believe
`both are acceptable representatives of nasal steroids and
`nonsedating antihistamines. However, caution should be
`exercised when extrapolating these results to other nasal
`steroid sprays and nonsedating antihistamines.
`With regard to convenience and cost, our results
`favour starting treatment with fluticasone because over
`half the patients in the fluticasone group did not need to
`add the back-up medication. However, a limitation of
`any clinical trial is that it only provides mean data about
`a group of patients. Some patients respond better to and
`prefer using a topical nasal steroid spray, whereas others
`prefer a systemic nonsedating antihistamine. Only by
`trying each approach in an individual patient is it possi-
`ble to determine which will be more beneficial. What-
`ever the final choice, at least 50% of patients are likely
`to require 2 different types of medication in combina-
`tion to achieve optimal HRQL.
`
`This study was supported by a grant from Glaxo Wellcome Inc.
`
`References
`
`5.
`
`3.
`
`1. Richards S, Thornhill D, Roberts H, Harries U. How many people think they
`have hay fever, and what they do about it. Br J Gen Pract 1992;42:284-6.
`2. Royal College of General Practitioners Office of Population Censuses and
`Surveys. Morbidity statistics from general practice: third national study (1981–82).
`London: Government Statistics Service; 1986.
`Juniper EF, Guyatt GH. Development and testing of a new measure of health
`status for clinical trials in rhinoconjunctivitis. Clin Exp Allergy 1991;21:77-83.
`4. Bousquet J, Bullinger M, Fayol C, Marquis P, Valentin B, Burtin B. Assess-
`ment of quality of life in patients with perennial allergic rhinitis with the
`French version of the SF-36 health status questionnaire. J Allergy Clin Im-
`munol 1994;94:182-8.
`Internationa