항히스타민 비교 자료

Ovid Technologies, Inc. Email Service
------------------------------
Results: The Journal of Allergy and Clinical Immunology
(C) Mosby-Year Book Inc. 2003. All Rights Reserved.

----------------------------------------------
Volume 111(4) April 2003 pp 770-776
----------------------------------------------

Sedation and performance impairment of diphenhydramine and second-generation
antihistamines: A meta-analysis
[Asthma, Rhinitis, Other Respiratory Diseases]
Bender, Bruce G. PhD; Berning, Shaun PharmD; Dudden, Rosalind MLS; Milgrom,
Henry MD; Tran, Zung Vu PhD
Denver, Colo
From National Jewish Medical and Research Center, Denver.
Received for publication July 1, 2002;
revised January 24, 2003;
accepted for publication January 27, 2003.
Reprint requests: Bruce G. Bender, PhD, National Jewish Medical and Research
Center, 1400 Jackson St, Denver, CO 80206.

----------------------------------------------

Outline

Abstract
Methods

Selection of studies
Inclusion criteria
Reported outcomes
Data abstraction
Analysis

Results

Diphenhydramine versus placebo
Diphenhydramine versus second-generation antihistamines
Second-generation antihistamines versus placebo
Additional factors potentially influencing outcomes

Discussion
References

Graphics

Table I. Data abstra...
Table I. Continued...
Fig. 1

----------------------------------------------

Abstract

Background: Antihistamines are among the most frequently used medications in the
United States. Despite dramatically higher cost, second-generation antihistamines
are replacing diphenhydramine because of the perception that they are not
constrained by its sedating effects.

Objective: We sought to examine, through meta-analytic procedures, the
collective evidence regarding the sedating and performance-impairing effects of
diphenhydramine relative to placebo and second-generation antihistamines.

Methods: A search that began with the MEDLINE database was limited to those
studies that included patients with atopic disease and control subjects, were
blinded and randomized clinical trials, objectively examined alertness and
psychomotor performance, reported means and variances, and were written in
English. Information was systematically abstracted from the resulting 18
articles, and effect size was calculated.

Results: Diphenhydramine impaired performance relative to placebo control and
second-generation antihistamines, including acrivastine, astemizole, cetirizine,
fexofenadine, loratadine, and terfenadine. However, results were quite varied,
the average sedating effect of diphenhydramine was modest, and in some instances
results of tests of performance in the diphenhydramine group showed less
sedation than in the control or second-generation antihistamine groups. A
significant (P

Conclusion: The absence of a consistent finding of diphenhydramine-induced
sedation is surprising given that most studies have been designed to increase
the probability of this outcome, including administering a 50-mg dose. On the
basis of this meta-analysis of performance-impairment trials, a clear and
consistent distinction between sedating and nonsedating antihistamines does not
exist.

----------------------------------------------

Abbreviations used ES: = Effect size; MeSH: = Medical Subject Headings

Allergic rhinitis affects 23.7 million Americans. 1 Because antihistamines are
the first choice for treatment of allergic conditions, they are among the most
frequently used medications in the United States, 2 with sales of prescription
antihistamines increasing 16% in 2001 alone. 3 Use of first-generation
antihistamines is widespread, largely because they are inexpensive and available
over the counter. Prescriptions for second-generation antihistamines, marketed
for their nonsedating benefits, have increased dramatically in recent years
despite their markedly higher cost.

Diphenhydramine is the most commonly used first-generation antihistamine. Its
capacity for sedation is so widely recognized that it is frequently the verum
control drug in studies seeking to show the relative nonsedating benefit of
second-generation antihistamines, including cetirizine, loratadine, terfenadine,
astemizole, fexofenadine, and acrivastine. 4 Studies of diphenhydramine largely
document some sedative side effect, but results are inconsistent, and some
studies find no sedation.

The purpose of this meta-analysis was to systematically search, compile, review,
analyze, and critique the existing diphenhydramine research literature and
perform statistical analysis on defined subsets of studies addressing sedation.
This project thus examined the collective evidence regarding the sedating and
performance-impairing effects of diphenhydramine relative to second-generation
antihistamines to provide physicians and consumers with a clearer understanding
of the relative risks associated with its use.

Methods

Meta-analysis is an approach in which individual study outcomes are quantitatively
combined to arrive at overall conclusions regarding a clearly specified body of
research. The purposes of meta-analysis are to (1) increase statistical power
for primary end points and subgroups, (2) resolve uncertainty when studies
disagree, (3) improve estimates of treatment effectiveness, and (4) answer
questions not posed at the start of individual trials. 5 To date, no meta-analytic
review has specifically addressed the effects of diphenhydramine on psychomotor
performance compared with placebo and second-generation antihistamines.

Selection of studies

A literature search for relevant diphenhydramine studies was executed by using
the MEDLINE database. The source literature population of studies consisted of
controlled clinical trials in peer-review journals. A comprehensive search was
performed with a MEDLINE file dated from 1966 to identify studies for inclusion
in this meta-analysis. The term 'histamine h1 antagonists' and 34 specific
generic names of first- and second-generation antihistamines were used in the
MEDLINE database search, retrieving 25,894 citations. These were cross-indexed
with the research areas related to alertness and psychomotor performance. These
areas were examined to identify search terms in the Medical Subject Headings
(MeSH terms). Terms such as 'cognition disorders,' 'learning disorders,'
'psychomotor performance,' 'reaction time,' 'fatigue,' 'mental processes,'
'confusion,' and 'psychological tests' were exploded to include all conceptually
narrower terms in the MeSH hierarchy. The truncated text word for the concept of
sedation was used as well. This group of terms retrieved 1,112,254 citations.
The crossed retrieval was then limited to human studies and to English-language
studies, producing 2134 citations. These were limited to the MEDLINE publication
type 'randomized controlled trial,' resulting in a retrieval of 455 citations.
The same set was also limited to MeSH terms for supported research, yielding 404
citations. Combined, these sets produced a pool of 722 citations. Of these, only
125 contained the word 'diphenhydramine.' Abstracts were reviewed, and when the
information was insufficient to determine eligibility, the full text was
reviewed. Articles were rejected on the topics of psychiatric disorders and
therapies (39), cancer (12), anesthesia (17), caffeine (4), dentistry (2), or
ophthalmology (1). A final cause of rejection was the incomplete reporting of
means and variance (35), leaving 15 articles. Three articles, the indexing of
which did not match the extensive search criteria, were located in the
bibliographies of reviewed articles. Eighteen articles met all inclusion
criteria and were included in this meta-analytic study. 7-24 From these, 20
separate outcomes were evaluated (Table I).

----------------------------------------------
Table I. Data abstraction format for key variables
----------------------------------------------

----------------------------------------------
Table I. Continued
----------------------------------------------

Inclusion criteria

To qualify, studies included (1) atopic disease, even if patients were
asymptomatic; (2) blinded, randomized designs that included a placebo group; (3)
diphenhydramine as one of the drugs administered; (4) laboratory measures of
sedation or psychomotor perform-ance as outcomes; (5) appearance in an
English-language journal indexed between January 1966 and January 2002; and (6)
report of means and variances (SD or SE) allowing for the calculation of an
effect size (ES). Although blinded drug conditions might introduce difficulty if
encapsulation changes drug-absorption characteristics, the importance of
blinding subjects during evaluation of potential sedation necessitated this
inclusion criterion. All included studies from a single research group were
examined to ensure that each study's data were collected independently of each
other to avoid multiple publication biases. The most common causes of rejection
from this meta-analysis were (1) focus on a nonatopic disease, (2) inclusion of
anecdotal adverse reports without actual performance or sedation measures, and
(3) failure to report a measure of variance (ie, SD).

Reported outcomes

Studies included in this meta-analysis had to report at least one of the 5
categories of sedation-performance measures most commonly encountered in the
published literature:

1. Self-reported sleepiness. This measure included either a visual analog scale
(consisting of a 100-mm line on which participants selected a point between 2
extremes from 'very sleepy' to 'very awake' best representing their degree of
alertness) or the Stanford Sleepiness Scale (which allowed the subject to choose
one of 5 statements reflecting similar degrees of alertness). 6

2. Attention. Measures of attention included the CogScreen Dual Task Test, 10
Divided Attention Task,28 Critical Flicker Fusion, 11,20 and Continuous
Performance Test. 15

3. Memory. Measures of memory included the Automated Neuropsychological
Assessment Metrics Documentation Test, Running Memory Test, 10 Digit Symbol
Substitution Test, 14 Buschke Selective Reminding Test, 15 Serial Digit Learning
Test, 12,16 a computerized learning task, 21 and a classroom learning test. 7

4. Reaction time. Reaction time was measured by means of computerized calculation
of the time from presentation of a stimulus to the subject's response.

5. Eye-hand coordination. Included were measures of fine motor tracking, 11
symbol copying, 14 and fine motor speed. 12

6. Evoked brain potentials. The P300 event-related potential measure is viewed
as a direct, physiologic, central nervous system measure of sedation and is
obtained by means of electroencephalographic recording of brain wave activity.
This methodology was reported in 2 studies by one research group. 18,19

Data abstraction

The following information was abstracted from each published article and coded
and constituted the analytic data base: subject characteristics, including age,
sex, race, height, weight, and illness characteristics; study design; drug type,
dose, and delivery route; time lapse between dosing and psychometric measurement;
type of measure and its outcome (mean and SD); and funding source (industry vs
government;Table I). All data were independently abstracted by 2 investigators
(BGB, ZVT) to avoid intercoder bias. Every coded item was reviewed to ensure
accuracy and consistency, and disagreements were resolved by means of consensus.

Analysis

A protocol specifying the framework for this meta-analysis was written in
advance of data collection. The primary outcome in this study was the difference
in psychomotor performance between 2 treatment groups. For a given measure (eg,
self-reported sleepiness and attention), an ES was calculated as follows:
[DELTA]1 equals diphenhydramine minus placebo divided by pooled SDs (Fig 1, A);
[DELTA]2 equals diphenhydramine minus second-generation antihistamines divided
by pooled SDs (Fig 1, B); and [DELTA]3 equals second-generation antihistamines
minus placebo divided by pooled SDs (Fig 1, C). This standardized ES represents
the size of the average difference between 2 groups on one of the selected
outcome measures. 26 Pooled outcomes were calculated by assigning weights equal
to the inverse of the total variance for mean effects. The 95% CI is reported to
provide the reader the likely range of values associated with each ES. By using
this approach for calculating ES, scores greater than 0 indicated increased
sedation of the indicated drug. Thus an ES of 1.00 would indicate that the
average (or mean) subject in one group outperforms 84% of subjects in the other
group (ie, 1 SD above the mean is the 84th percentile).

----------------------------------------------
Fig. 1. Sedation and performance impairment.
----------------------------------------------

A random-effects model was used for all analyses because this approach collapses
into a fixed-effects model in the absence of significant ES heterogeneity. 25,26
For categoric variables, subgroup analyses were performed by using ANOVA-like
procedures for meta-analysis. The [alpha] level for a type I error was set at a
P value of less than .05. Bonferroni adjustments were not made because of the
increased likelihood of a type II error.

Results

Diphenhydramine versus placebo

Diphenhydramine altered performance on measures believed to reflect the effects
of sedation. When diphenhydramine was contrasted with placebo (Fig 1, A), the
mean ES across all 6 measures and all 18 studies was 0.36 (95% CI, 0.20-0.51;P =
.0001). Although this showed a moderate overall sedation effect of diphenhydramine,
results were distinctly variable across studies, and not all studies reported
diphenhydramine sedation. Significant measure-specific ESs were found for
attention (ES = 0.53; 95% CI, 0.36-0.71;P = .0001). The diphenhydramine-placebo
ES did not reach statistical significance for measures of memory, reaction time,
dexterity, or evoked brain potential measure, although it approached significance
for self-report (ES = 0.41; 95% CI, -0.07 to 0.90;P = .089) and memory (95% CI,
-0.01 to 0.33;P = .051). Because less than 5 ESs were available for pooling for
measures of reaction time, dexterity, and evoked brain potential, these results
were deemed insufficient to accurately assess whether the overall ES was
significant. Available data suggest a small ES in the direction of diphenhydramine-induced
impairment consistent with the overall findings (Fig 1, A).

Diphenhydramine versus second-generation antihistamines

Diphenhydramine also decreased performance scores in comparison with second-generation
antihistamines. Mean ESs were calculated for performance comparisons between
groups treated with diphenhydramine and one of the second-generation antihistamines,
including acrivastine, astemizole, cetirizine, fexofenadine, loratadine, or
terfenadine (Fig 1, B). Across measures and studies, the average ES of 0.31 was
significant (95% CI, 0.17-0.45;P = .0001) and again indicative of a moderate
diphenhydramine performance impairment. As with the diphenhydramine versus
placebo comparison, marked variability was observed, and not all studies
reported diphenhydramine-induced sedation. Significant mean ESs emerged on 3 of
the 6 individual measures, including self-report (ES = 0.45; 95% CI, 0.01-0.26;P
= .045), attention (ES = 0.35; 95% CI, 0.05-0.64;P = .029), and evoked brain
potential (ES = 0.53; 95% CI, 0.17-0.89;P = .018). The ES of 0.21 for memory
approached significance (95% CI, 0.03-0.42;P = .050). Although few studies
included reaction time or evoked brain potential measures, results from these
appeared consistent with the overall ES finding (Fig 1, B).

Second-generation antihistamines versus placebo

Placebo results were contrasted with those of the second-generation antihistamines,
including acrivastine, astemizole, cetirizine, fexofenadine, loratadine, or
terfen-adine (Fig 1, C). When all 6 performance measures were averaged, a small
but significant ES resulted (ES = 0.14; 95% CI, 0.01-0.26;P = .030), indicating
sedation in the groups treated with second-generation antihistamines relative to
placebo. The second-generation antihistamines responsible for positive ESs,
indicating self-reported sedation or impaired performance, were varied and
included acrivastine, astemizole, cetirizine, loratadine, and terfenadine. On
the individual measures, the ES was significant only for self-report (ES = 0.39;
95% CI, 0.05-0.74;P = .030), although the ES approached significance for
reaction time (ES = 0.22; 95% CI, -0.06 to 0.49;P = .10). Although fewer than 5
studies were available for measures of attention, dexterity, or evoked brain
potential measures, the results appeared consistent with an overall small
sedation effect.

Additional factors potentially influencing outcomes

Other potential mediating variables, including source of funding, medication
dose, days of treatment, presence or severity of allergic symptoms, and subject
age, race, and sex were coded but could not be analyzed because they were
infrequently reported or, in the case of medication dose, because there was
little variation across studies (eg, diphenhydramine was most commonly
administered at 50 mg, cetirizine at 10 mg, and loratadine at 10 mg).

Discussion

Diphenhydramine can cause sedation. On most measures included in this meta-analysis,
diphenhydramine produced evidence of sedation in comparison with placebo or
second-generation antihistamines. The overall ES of 0.36 represents a moderate
sedation factor. In some studies the ES was dramatically large. However, the
size and direction of results varied markedly. In some cases differences
indicated less sedation in the diphenhydramine group than in the comparison
group. Differences across studies in size, direction, and variance of specific
measures were not explained by differences in study drug, design, or subject
characteristics.

In this meta-analysis, limited to performance-impairment studies that included
diphenhydramine, no second-generation antihistamine produced performances
consistently superior to those of the placebo and diphenhydramine conditions,
including the 2 drugs no longer marketed in the United States, terfenadine and
astemizole. Second-generation antihistamines were not nonsedating, although
sedation effects were weaker than for diphenhydramine. The overall ES of 0.14,
indicating sedation effects in second-generation antihistamines versus placebo,
was small but significant. The ES of 0.39 for self-report measures was as large
as that seen when diphenhydramine was compared with placebo. The absence of a
consistent diphenhydramine sedation effect was unexpected given that most
studies were designed to enhance the probability of observing diphenhydramine
sedation. For example, in 15 of the 18 studies, subjects were provided 50 mg of
diphenhydramine, the maximum standard dose. Outcome measures reported and
included in this meta-analysis were taken at peak serum concentrations 2 to 4
hours after dosing. Additionally, dosing and evaluation in most cases were
conducted on the initial day of medication use before the tachyphylaxis to the
sedation effect could occur.

That few studies are available to evaluate the effect of 25 mg of diphenhydramine
is remarkable given that diphenhydramine is most commonly sold in 25-mg tablets,
and an undetermined number of adults take a 25-mg dose. Individual studies that
included diphenhydramine at 25 mg reveal inconsistent results, with some
detecting impaired performance 21,24 and others not. 7,16 If consumers are to
make an informed judgment about relative risk and benefits, then the histamine-suppressing
capacity and sedation effects of 25 mg of diphenhydramine must be concurrently
compared with those of second-generation antihistamines within a single study.
Equipotency, a basic principle followed in comparative studies of adverse
consequences of medications, calls for the use of clinically equivalent doses of
drugs under investigation. Given recent new evidence that untreated allergic
rhinitis can impose significant impairment on cognitive vigilance, speed, and
efficiency, 27 it will be of additional interest to assess performance
differences between antihistamines and within differing doses of the same
antihistamine relative to performance impairment caused by the illness itself.
Further benefits of antihistamine studies will be realized if they include
detailed information about the subjects studied, including an accounting of
whether any subjects dropped out of the study because they experienced sedation.
Despite their widespread use with children, very few studies have examined
sedation and performance impairment in children using first- or second-generation
antihistamines. Two studies have measured the consequence of administering 25 mg
of diphenhydramine, the maximum recommended dose, to children. One found that
computer-measured learning scores of children treated with diphenhydramine were
lower than those of children treated with loratadine, 22 whereas a second study
that followed children treated with the same 2 drugs through four 8-hour days in
a laboratory school found no differences in learning, reaction time, or
self-ratings of sedations. 7 There exists an inadequate number of pediatric
studies to establish a superiority of second-generation antihistamines in
children with rhinitis sufficient to justify their large expense over diphenhydramine.
The need clearly persists for further research with children.

Results from this meta-analysis provide guidance for the choice of measures for
future sedation studies. Self-report, reaction time, attention, and P300
measures produced the largest overall ESs, suggesting that these are the most
sensitive measures. However, very few reaction time and P300 studies were
represented in the analyses, and hence the choice of these measures warrants
less confidence at the present time than the use of self-report and attention.
Evoked potential measures provide physiologic evidence of a drug effect, but
little is known about whether the small changes detected in these studies,
typically presented as the percentage change from baseline, correspond to a
meaningful change in human performance. Whatever the measure used, investigators
must clearly state the measure of variance they are reporting (eg, SD vs SE) to
allow the comparison of ES across studies.

Although data are not plentiful, studies weigh toward more performance
disruption and sedation with diphenhydramine than placebo or so-called
nonsedating antihistamines. Given the importance of providing optimal care, as
well as maximizing safety, one can see why nonsedating antihistamines are widely
prescribed. Yet the information that we have is far from definitive, and the
sedating effects of allergic rhinitis have not been consistently separated from
the sedating effects of the drugs used to treat it. Diphenhydramine continues to
be the most extensively used over-the-counter medication for the treatment of
allergic rhinitis; an estimated 47% of persons with allergies take over-the-counter
products, most containing a first-generation antihistamine. 23 True experimental
studies that control for effects of dose, dosing regimen, and dose timing are
needed to resolve this dilemma.

References

1. Adams PF, Hendershot GE, Marano MA. Current estimates from the National
Health Interview Survey, 1996. Vital Health Stat 10 1999;200:1-212.

2. Marketletter. 1999:58.

3. Gebhart F. Steady, not stellar, growth marked the pharmaceutical market last
year. Drug Topics 2002;6:25-40.

4. Hindmarch I, Shamsi Z, Stanley N, Fairweather DB. A double-blind, placebo-controlled
investigation of the effects of fexofenadine, loratadine and promethazine on
cognitive and psychomotor function. Br J Clin Pharmacol 1999;42:200-6.

5. Kelley GA, Tran ZV. Effect sizes for continuous data: original metric versus
standardized difference. Prev Cardiol 2001;4:40-5.

6. Hoddes E, Zarcone V, Smythe H, Phillips R, Dement WC. Quantification of
sleepiness: a new approach. Psychophysiology 1973;10:431-6. Bibliographic Links

7. Bender BG, McCormick DR, Milgrom H. Children's school performance is not
impaired by short-term administration of diphenhydramine or loratadine. J
Pediatr 2001;138:656-60. Ovid Full Text Bibliographic Links

8. Berlinger WG, Goldberg MJ, Spector R, Chiang C-K, Ghoneim MM. Diphenhydramine:
kinetics and psychomotor effects in elderly women. Clin Pharmacol Ther
1982;32:387-91. Bibliographic Links

9. Carruthers SG, Shoeman DW, Hignite CE, Azarnoff DL. Correlation between
plasma diphenhydramine level and sedative and antihistamine effects. Clin
Pharmacol Ther 1978;23:375-82. Bibliographic Links

10. Kay, GG, Berman B, Mockoviak SH, Morris CE, Reeves D, Starbuck V, et al.
Initial and steady-state effects of diphenhydramine and loratadine on sedation,
cognition, mood, and psychomotor performance. Arch Intern Med 1997;157:2350-6.
Bibliographic Links

11. Mattila MJ, Mattila M, Konno K. Acute and subacute actions on human
performance and interactions with diazepam of temelastine (SK & F93944) and
diphenhydramine. Eur J Clin Pharmacol 1986;31:291-8. Bibliographic Links

12. Philpot EE, Brooker AE, Biegalski CS. Effects of sedating and nonsedating
antihistamines on flying performance. Mil Med 1993;158:654-60. Bibliographic
Links

13. Ramaekers JG, O'Hanlon JF. Acrivastine, terfenadine and diphenhydramine
effects on driving performance as a function of dose and time after dosing. Eur
J Clin Pharmacol 1994;47:261-6. Bibliographic Links

14. Roth T, Roehrs T, Koshorek G, Sicklesteel J, Zorick F. Sedative effects of
antihistamines. J Allergy Clin Immunol 1987;80:94-8. Bibliographic Links

15. Sands L, Katz IR, DiFilippo S, D'Angelo K, Boyce A, Cooper T. Identification
of drug-related cognitive impairment in older individuals. Am J Geriatr
Psychiatry 1997;5:156-66.

16. Scavone JM, Greenblatt DJ, Harmatz JS, Engelhardt N, Shader RI. Pharmacokinetics
and pharmacodynamics of diphenhydramine 25 mg in young and elderly volunteers. J
Clin Pharmacol 1998;38:603-9. Ovid Full Text Bibliographic Links

17. Schweitzer PK, Muehlbach MJ, Walsh JK. Sleepiness and performance during
three-day administration of cetirizine or diphenhydramine. J Allergy Clin
Immunol 1994;94:716-24. Bibliographic Links

18. Simons FER, Fraser TG, Reggin JD, Simons KJ. Individual differences in
central nervous system response to antihistamines (H1-receptor antagonists). Ann
Allergy Asthma Immunol 1995;75:507-14.

19. Simons FER, Fraser TG, Reggin JD, Simons KJ. Comparison of the central
nervous system effects produced by six H1-receptor antagonists. Clin Exp Allergy
1996;26:1092-7. Ovid Full Text Bibliographic Links

20. Tharion WJ, McMenemy DJ, Rauch TM. Antihistamine effects on the central
nervous system, cognitive performance and subjective states. Neuropsychobiology
1994;29:97-104. Bibliographic Links

21. Vuurman EFPM, van Veggel LMA, Sanders RL, Muntjewerff ND, O'Hanlon JF.
Effects of Semprex-D and diphenhydramine on learning in young adults with
seasonal allergic rhinitis. Ann Allergy Asthma Immunol 1996;76:247-52.

22. Vuurman EFPM, van Veggel LMA, Uiterwijk MMC, Leutner D, O'Hanlon JF.
Seasonal allergic rhinitis and antihistamine effects on children's learning. Ann
Allergy 1993;71:121-6. Bibliographic Links

23. Weiler JM, Bloomfield JR, Woodworth GG, Grant AR, Layton TA, Brown TL, et
al. Effects of fexofenadine, diphenhydramine, and alcohol on driving performance.
Ann Intern Med 2000;132:354-63. Ovid Full Text Bibliographic Links

24. Witek TJ Jr, Canestrari DA, Miller RD, Yang JY, Riker DK. Characterization
of daytime sleepiness and psychomotor performance following H1 receptor
antagonists. Ann Allergy Asthma Immunol 1995;74:419-26. Bibliographic Links

25. NRC Committee on Applied and Theoretical Statistics. Combining information:
statistical issues and opportunities for research. Washington (DC): Academy
Press; 1992.

26. Hedges LV, Olkin I. Statistical methods for meta-analysis. Orlando (FL):
Academic Press; 1985.

27. Wilken JA, Berkowitz R, Kane R. Decrements in vigilance and cognitive
functioning associated with ragweed-induced allergic rhinitis. Ann Allergy
Asthma 2002;89:372-80.

Key words: Allergic rhinitis; antihistamines; diphenhydramine; meta-analysis;
sedation

----------------------------------------------
Accession Number: 00004483-200304000-00019
----------------------------------------------