dexamethasone and salbutamol in the treatment of acute ...has acute or recurrent wheezing. the lack...

PEDIATRICS Vol. 71 No. 1 January 1983 13

Dexamethasone and Salbutamol in theTreatment of Acute Wheezing in Infants

Asher Tal, MD, Chanan Bavilski, MD, David Yohai, MD,

Jacob E. Bearman, PhD, Rafael Gorodischer, MD, and

Shimon W. Moses, MD

From the Division of Pediatrics and Epidemiology Unit, Soroka University Hospital andFaculty of Health Sciences. Ben-Gurion University of the Negev, Beer-Sheba, Israel

ABSTRACT. Thirty-two infants, aged 1 to 12 months,hospitalized with acute wheezing, were studied. Theywere randomly divided into four treatment groups of

eight patients each. The treatments were intramusculardexamethasone or placebo (double-blind), and salbuta-mol (oral and inhaled), or none (open), in all four possiblecombinations. The study was carried out as a randomizedblock design with eight blocks of four infants each,

matched by age and clinical score. Average daily improve-ments, as reflected by changes in the clinical score andlength of hospital stay, was essentially the same for

infants treated with placebo, salbutamol alone, and dex-

amethasone alone. However, combined salbutamol-dex-amethasone treatment resulted in more than twice therate of improvement of the other treatments. The differ-ence was statistically highly significant (P < .01). Fur-

thermore, the response of this combined treatment wasobserved within 24 hours; none ofthe ten infants in whom

there was no significant improvement within 48 hoursand neither ofthe two patients who developed respiratoryfailure received the combined salbutamol-dexamethasone

treatment. A potentiating effect of corticosteroids on thefJ-adrenergic responsiveness is a possible explanation forthe advantage of this combined treatment in the manage-

ment of acute wheezing in infancy. Pediatrics 1983;

71:13-18; wheezing infant, corticosteroids, /32-adrenergicagent.

mucus production. It is not easy to distinguish

between bronchiolitis and asthma in an infant who

has acute or recurrent wheezing. The lack of clear-

cut diagnostic criteria is also evident from the dif-

ferent terminology used in various medical centers,

such as “spastic bronchitis,” “wheezy bronchitis,”

and “asthmatic bronchitis,” whereas others prefer

“the wheezy baby” ‘�P325� or “wheezing-associated

respiratory ifiness” (WARI).2’3 It is widely accepted

that wheezy bronchitis represents part of the spec-

trum of asthma,4 as it has been shown that infants

who have more than one wheezing episode will

frequently develop the typical picture of asthma in

later childhood. 1(p325),4,5

There are conflicting reports regarding the value

of corticosteroids or $-adrenergic agents in the

treatment of acute bronchiolitis or wheezy bronchi-

tis in infants less than 1 year of age.�’2

This study was designed to obtain information as

objective as possible on the effect of each of these

drugs, both alone and in combination, as well as

compared with placebo, on infants admitted to the

hospital with an attack of acute wheezing.

Acute wheezing in infancy is a frequent cause of

hospital admissions. Presenting as an acute ifiness,

it is usually accompanied by upper respiratory tract

symptoms, and characterized by cough, tachypnea,

retractions of the respiratory muscle, and increased

Received for publication Jan 20, 1982; accepted March 16, 1982.

Presented at the Israeli Society for Clinical Pediatrics, Herzlia,

Israel, 1981. This paper is part of the MD thesis of Ch. Bavilski,

Technion Medical School, Haifa, Israel.

Reprint requests to (A.T.) Department ofPediatrics “B,” Soroka

University Hospital, P0 Box 151, Beer-Sheba, Israel.

PEDIATRICS (ISSN 0031 4005). Copyright © 1983 by the

American Academy of Pediatrics.

MATERIALS AND METHODS

Selection of Patients

The study group included infants aged 1 to 12

months admitted with acute wheezing. The patients

studied included those who had wheezing associ-

ated with upper respiratory tract infection whose

illness had been diagnosed as bronchiolitis (first

attack of wheezing, with upper respiratory tract

symptoms, without atopic background), asthma (re-

current attacks of wheezing with evidence of atopic

background) or WARI (acute wheezing, not fitting

into the previous two definitions). Infants were

excluded if the cause ofwheezing was bronchopneu-

by guest on July 31, 2018www.aappublications.org/newsDownloaded from

14 TREATMENT OF ACUTE WHEEZING IN INFANTS

monia, congestive heart failure, foreign body aspi-

ration, cystic fibrosis, or postrespirator chronic lung

damage. Informed consent for the study was ob-

tamed from the parents, and patients were assigned,

on a random basis, to one of the treatment proto-

cols. The protocol for this study was approved by

the Helsinki Ethics Committee of the Soroka Uni-

versity Hospital.

Treatment Protocol

All infants were given oral or intravenous (IV)

fluids to maintain optimal hydration, and they were

given humidified oxygen when cyanosis and/or ir-

ritabiity were present. The treatment protocol is

shown in Fig 1. Dexamethasone (4 mg/mL) or

normal saline (as placebo) were prepared by the

hospital pharmacy in double-blind coded vials.

These agents were administered intramuscularly

(IM), 0.075 mL/kg, on admission, and then 0.025

mL/kg every eight hours during the next three days

(dexamethasone dose: 0.3 mg/kg on admission, and

then 0.1 mg/kg every eight hours). In addition, half

of the patients receiving each of these treatments

randomly assigned, were given salbutamol (Vento-

lin, Allen and Hanburys Ltd, England) or no addi-

tional treatment. Salbutamol was given by two

routes simultaneously: (1) inhalations: 0.5 mL (2.5

mg) of salbutamol respiratory solution with 2 mL

of water for injection, using a Hadson updraft neb-

ulizer and an electric compressor unit; inhalations

Intramuscular, Double-blind, Randomized

Dexamethasone Placebo

�r Randomized, OpenSalbutamol None

Stage 1

Stage 2

(oral + inhaled)

Salb,’,tamol*(oral + inhaled)

Fig 1 . Treatment protocol for infants hospitalized with

acute wheezing. Asterisk indicates salbutamol is given ifrelative therapeutic failure is present.

TABLE 1 . Clinical Scoring System

were given on admission and subsequently every six

hours, and (2) orally: salbutamol syrup, 0.15 mg/kg

every eight hours. If no improvement was observed

after 48 hours, a “relative therapeutic failure” was

registered, and salbutamol was administered (same

routes and doses) in those infants who were not

given this drug earlier (stage 2). Infants who devel-

oped respiratory failure’3 and needed assisted yen-

tilation were considered as a “complete therapeutic

failure.” They were transferred to the respiratory

intensive care unit, and the study protocol was

terminated, breaking the individual code to allow

for changes in therapy.

Clinical Evaluation

A clinical scoring system was established, based

on the one developed by Bierman and Pierson’4

(Table 1). The score included respiratory rate,

wheezing, cyanosis, and the use of accessory mus-

des. The maximum possible score was 12, which

indicated severe ifiness. The clinical score was ob-

tained on admission, three hours after the first IM

dose, and each morning (8 AM) until discharge. All

scores were performed by three of the investigators

(A.T., C.B., and D.Y.), and each patient was eval-

uated by the same physician during the entire pe-

riod. A controlled pilot scoring of 25 infants by

these physicians showed agreement with no differ-

ences greater than one point. For purposes of the

study, the rate of improvement was calculated as

the average decrease in clinical score per day. Du-

ration of hospitalization was the number of days

from admission until discharge.

Further Evaluation

On admission, a complete history was taken and

physical examination was performed. Anteroposte-

nor and lateral chest roentgenograms, CBC, serum

electrolytes, serum immunoglobulin (Ig)E levels (by

Phadebas IgE paper radioimmunosorbent test

[PRIST] method), and blood samples for titers of

antibodies against respiratory syncytial virus and

adenovirus were obtained. Measurements of site-

Score Respiratory Rate(breaths/mm)

Wheezing Cyanosis AccessoryRespiratory

MuscleUtilization

0 <30 None* None None

1 31-45 Terminal expiration with stetho-

scope onlyCircumoral on crying only +

-

2 46-60 Entire expiration and inspiration

with stethoscope onlyCircumoral at rest ++

3 >60 Expiration and inspiration without

stethoscopeGeneralized cyanosis at

rest++++

* If no wheezing is audible due to minimal air entry, score 3.


ARTICLES 15

rial blood gases were obtained on admission, and

again three hours after the first IM administration

of the drug, and each morning for the next three

consecutive days. Additional examinations of site-

rial blood gases were performed according to the

clinical condition of the patient. In each case, values

for blood gases were determined ten minutes after

the infant was breathing room air, and were exam-

ined immediately. Blood pressure was measured on

admission and every second day.

Study Design

The study was carried out as a randomized block

design, with factorial arrangement of the treat-

ments within each block. The blocks were con-

structed by taking infants of the same age (�<6

months, >6 months) and with a comparable sever-

ity of illness. Infants with clinical score of 5 or less

were not admitted to the hospital. Those with

scores of 12 were admitted directly to the respira-

tory intensive care unit and, hence, not entered in

the study. Those entering the study were classified

as mildly (clinical score of 6), moderately (7 to 9),

or severely (10 or 11) ifi. As mentioned, the measure

of improvement was the mean decrease in clinical

score per day, for the first seven days of treatment,

or until discharge, whichever came first. (It should

be emphasized that all patients who were not des-

ignated “therapeutic failures” (“relative” or “com-

plete”) were discharged before seven days. For the

failures, average change in clinical score was cal-

culated to time of treatment failure).

The data were evaluated by analysis of variance

and t tests of three independent contrasts: main

effects of the two agents and their interaction.

RESULTS

Patients

During a 3-month period (October to December

1980), 32 patients were studied: 20 (62.5%) aged 1 to

6 months (mean 3.3 months) and 12 (37.5%) aged 6

to 11 months (mean 8.8 months). There were 20

(62.5%) boys and 12 (37.5%) girls, with a mean age

of 5.4 months. The infants were in eight blocks of

four patients each. Four blocks included patients

who were moderately ill; the remaining four blocks

included severely ifi patients. At the time of admis-

sion, 12 infants had their condition diagnosed as

bronchiolitis, 13 as asthma, and seven as WARI.

Past and Family History

Of the 32 infants ten (29%) had had previous

episodes of wheezing, and in 14 (43%) there was

positive family history of asthma or allergy. Only

six infants (18%) had been breast-fed for 2 months

or more.

Laboratory Findings

WBC count ranged between 6,800 to 19,0004tL

with 0% to 4% eosinophils in the peripheral blood.

Serum IgE was higher than predicted values’5 in

41% of the infants. In two infants significant rise in

antibody titers (from less than 1:4 to greater than

1:32 by complement fixation tests) against respira-

tory syncytial virus and adenovirus, respectively,

was found. Mean values of arterial blood Pco2,

Pco2, and pH obtained on admission are shown in

Table 2 and demonstrate that there were no signif-

icant differences between the four treatment groups

regarding the initial mean values of Po2, Pco2, and

pH. Changes in Po2 and Pco2 during the wheezy

episode were not significantly different between the

four treatment groups (Fig 2).

Radiographic Findings

Three main features were found: Hyperinflation

was found in 26 of the ifims taken on admission,

increased pulmonary markings were seen in 12

cases, and pulmonary subsegmental infiltrates were

seen in seven cases. Four films were interpreted as

normal by the pediatric radiologist.

The mean rate of improvement of each of the

four treatment groups is shown in Fig 3. The aver-

age initial clinical score was similar for all treatment

groups (about nine points). The rate of improve-

ment of the infants treated by the combination of

dexamethasone and salbutamol was more than

twice that of any of the other treatment groups.

There was no advantage to dexamethasone without

salbutamol (1.36 v 1.35), and there was no advan-

tage to salbutamol without dexamethasone (1.33 v

1.35). Dexamethasone had a significant effect when

TABLE 2. Arterial Blood Gas Values Obtained on Admission of Four Treatment Groupsof Patients*

Placebo Dexamethasone Salbutamol Dexamethasone andSalbutamol

Po2 (torr)Pco2 (ton)pH

68.60 ± 17.2038.00 ± 8.04

7.32 ± 0.06

64.10 ± 5.2238.80 ± 6.29

7.34 ± 0.04

79.50 ± 14.0037.60 ± 8.40

7.34 ± 0.06

75.30 ± 14.7035.30 ± 8.20

7.33 ± 0.06

* Values are mea ns ± SD.


p.---.

d----�-----

ds -

60

40

20

p - -

d-� .

ds�w00C’)

-J

0z-J

0

(N0U

(N0a.

9

8

7

6

5

4

3

2

I 2 3 4 5�

TIME (days)I

�1 2 3 4 5

TIME (days)Fig 2. Average changes in Pco2 and Po2 in four treat-ment groups: p, placebo; d, dexamethasone; s, salbutamol;ds, dexamethasone and salbutamol.

Five infants who were treated with dexametha-

sone alone were considered as relative therapeutic

failures at the end of stage 1 of the treatment. A

definite clinical improvement was documented in

three of them when salbutamol was later adminis-


given in combination with salbutamol (P < .01).

At the end of stage 1 of the treatment protocol,

no clinical improvement was noted in ten (32%) of

the infants (relative therapeutic failure). Two ad-

ditional infants later developed respiratory failure

and needed assisted ventilation in the intensive care

unit (complete therapeutic failure). As shown in

Table 3, none of these 12 failures had been treated

by the combination of dexamethasone and salbu-

tamol.

DAYSFig 3. Average rate of improvement in four treatmentgroups. Abbreviations are defined in Fig 2 legend.

tered (stage 2). Such an improvement was seen in

only one of the infants in the placebo group who

“failed” during stage 1 and were given salbutamol

alone during stage 2. No statistical significance was

found in this relatively small group of infants.

One infant developed a remarkable tremor as a

side effect of salbutamol. No other side effects or

complications of the treatment were documented.

All infants were discharged after one to 21 days

(mean hospital stay 4.9 days). Average hospital stay

of the group given dexamethasone and salbutamol

was the shortest (2.5 days) but because of the large

-� -J variability, no statistical significance was noted.

DISCUSSION

The results of this study indicate a possible ad-

vantage of the combination of corticosteroids and

$-adrenergic agents in the treatment of acute

wheezing episodes in infancy. Although the number

of infants studied in each group was relatively small,

the rate of improvement in the group given dexa-

methasone and salbutamol was much greater than

in the other treatment groups. This difference was

highly significant (P < .01).

Some clinicians consider corticosteroids to be

important drugs in the management of acute wheez-

ing in infancy.9”6”7 This assumption is based on the

hypothesis that their anti-inflammatory action can

reduce bronchiolar swelling.’6 In fact, one study’8

indicated a beneficial effect of corticosteroids in

acute bronchiolitis in infants. Several other con-

trolled studies have shown that corticosteroids

alone offer little, if any, benefit in the treatment of

acute bronchiolitis.�


TABLE 3. Number of Treatment Failures for Four Treatment Groups of Infants

ARTICLES 17

Relat ive Failure* Comp lete Failuret Total

None Salbutamol None Salbutamol None Salbutamol

Placebo 3/8 3/8 1/8 0/8 4/8 3/8

Dexamethasone 4/8 0/8 1/8 0/8 5/8 0/8

* No improvement after 48 h.

t Respiratory failure.

The clinical impression exists that inhaled /32-

adrenergic drugs are often effective in treatment of

wheezy infants, but this was not confirmed by ob-

jective data.�’2 Using a whole-body plethysmo-

graph, a definite improvement in specific airway

conductance was found recently in 5/8 infants less

than 1 year of age, studied during an acute episode

of wheezing before and after inhaled salbutamol.’9

There is some evidence that corticosteroid ther-

apy may enhance the effect of f3-adrenergic agonist

on cAMP production.20’2’ These studies were per-

formed following the description ofthe /3-adrenergic

theory of the atopic abnormality in asthma by

Szentivanyi.22 This theory delineates the role of the

,&-adrenergic stimulus in maintaining bronchodila-

tion by means of high levels of cAMP. Corticos-

teroids exert a facilitating effect on the adrenergic

nervous system, and can cause an increase in cAMP

levels by decreasing the activity of phosphodiester-

ase in various tissues.23 Clark and Godfrey�4�2�

suggest that cortisol stimulation of the production

of adenylate-cyclase may, in part, explain the ap-

parent potentiation of /32-stimulant responsiveness

seen in asthmatic patients treated with steroids.

Logsdon et al24 showed in vitro the stimulation of

leukocyte adenylate-cyclase by hydrocortisone and

isoproterenol in leukocytes of asthmatic and non-

asthmatic subjects. Recently, various autonomic

nervous system abnormalities were described in

allergic subjects, including hyporesponsiveness to

/3-adrenergic stimulation and hyperresponsiveness

to cholinergic and a-adrenergic stimuli.25’26 Frazer

et al27 found autoantibodies to $2-adrenergic recep-

tors in the serum of three patients with allergic

respiratory disease. These abnormalities were

found also in “preallergic” subjects (asymptomatic

with IgE-sensitized mast cells).25’26 It is possible that

/3-adrenergic or cholinergic receptors are involved

in the process of acute wheezing in infancy. This

could be related to the reported relative ineffective-

ness of theophylline in young infants with bron-

chiolitis, as shown in a retrospective study.28

The potentiating effect of corticosteroids on the

/3-adrenergic responsiveness is a feasible explana-

tion for the advantage of the combination of dexa-

methasone and salbutamol in the treatment of

acute wheezing in infants as shown in our study.

Average daily improvement, as shown by changes

in clinical score divided by the duration of hospi-

talization, was more than twice the rate of improve-

ment with steroids alone, f32-adrenergic drug alone,

or placebo. Furthermore, the response to this com-

bined regimen was observed within 24 hours; the

ten infants in whom there was no significant im-

provement within 48 hours and the two additional

patients who later developed respiratory failure had

not been given the combined treatment.

We found no correlation between the response to

treatment and demographic or clinical factors or

laboratory findings (such as age, sex, atopic back-

ground, serum IgE levels, or severity of illness).

There were no differences in the course of the

disease or response to treatment between infants

whose condition was diagnosed as asthma versus

bronchiolitis or WARI. There seems to be little

value in differentiating between bronchiolitis,

asthma, and WAR! in young wheezy infants, and

so it seems appropriate to use the terminology of

Godfrey,’ ie, wheezy baby syndrome.

In conclusion, it is evident from this pilot study

on a small number of infants that the combination

of dexamethasone and salbutamol was effective and

safe in the treatment of infants with acute wheezing.

These data justify a more extensive study to cor-

roborate the results before the use of such a com-

bined therapy can routinely be recommended.

ACKNOWLEDGMENTS

This study was supported by a research grant from the

Faculty of Health Sciences, Ben-Gurion University of theNegev, Beer-Sheba, Israel.

We thank Professor J. Bar-Ziv for reviewing all thechest radiographs, Professor S. Uretsky for critical reviewof the manuscript, Rimona Rothem and the hospitalpharmacy staff for their assistance, and Agneta Skibin forperforming the serum IgE tests.

REFERENCES

1. Godfrey 5: Clark JJH, Godfrey S (eds): Asthma. London,

Chapman and Hall, 1977, pp 288, 3252. Henderson FW, Clyde WA, Collier AM: The etiology and

epidemiologic spectrum ofbronchiolitis in pediatric practice.

J Pediatr 1979;95:183

3. Parrott RH: Editorial commentary. J Pediatr 1979;95:1904. Tabachnik, E, Levison H: Infantile bronchial asthma. J

Allergy Clin Immunol 1981;67:339



5. Williams HE, McNicol KN: Prevalence, natural history and

relationship of wheezy bronchitis and asthma in childhood:An epidemiological survey. Br Med J 1969;4:321

6. Leer JA, Green JL, Heimlich EM, et al: Corticosteroid

treatment in bronchiolitis: A controlled collaborative study

in 297 infants and children. Am J Dir Child 1969;117:4957. Dabbous IA, Tkachyk JS, Stamm 8,1: A double blind study

on the effect ofcorticosteroids in the treatment of bronchiol-

itis. Pediatrics 1966;37:477

8. Danus D: Bronquiitis: Analisis de 30 casos tratados conMeticorten. Pediatria 1965;8:113

9. Milner AD: Response to bronchodilator drugs in the firstfive years of life. Eur J Clin Pharmacol 1980;18:117

10. Radford M: Effect of salbutamol in children with wheezybronchitis. Arch Dis Child 1975;50:535

11. Rutter N, Milner AD, Huller EJ: Effect of bronchodilatorson respiratory resistance in infants and young children withbronchiolitis and wheezy bronchitis. Arch Dir Child1975;50:719

12. Lenney W, Milner AD: At what age do bronchodilatorswork? Arch Dir Child 1978;53:532

13. Newth CJL: Recognition and management of respiratory

failure. Pediatr Clin North Am 1979; 26:617

14. Bierman CW, Pierson WE: The pharmacologic management

of status asthmaticus in children. Pediatrics 1974;54:24515. Kjellman NIM, Johansson SGO, Roth A: Serum IgE levels

in healthy children quantified by a Sandwich technique(PRIST*). Clin Allergy 1976;6:51

16. Ellis EF: Therapy of acute bronchiolitis. Pediatr Res1977;11:263

17. Leffert F: The management of acute severe asthma. J Pe-diatr 1980;96:1

18. Oski FA, Salitsky 5, Barness LS: Steroid therapy in bron-chiolitis: A double blind study, abstracted. Am J Dis Child

1961;102:75919. Tal A, Bar-Yishay E, Godfrey 5: Use of whole body plethys-

mograph to study the response of wheezy infants to treat-ment, abstracted. Prog Respir Res 1981;17:229

20. Van Arsdel PP Jr, Paul GH: Drug therapy in the manage-ment of asthma. Ann Intern Med 1977;87:68

21. Townley RG, Reeb R, Fitzgibbons T, et al: The effect ofcorticosteroids on the beta-adrenergic receptors in bronchial

smooth muscle. J Allergy 1970;45:11822. Szentivanyi A: The beta adrenergic theory of the atopic

abnormality in bronchial asthma. J Allergy 1968;42:20323. Sahn SA: Corticosteroids in chronic bronchitis and pulmo-

nary emphysema. Chest 1978;73:324. Logsdon PJ, Middleton E, Caffey RG: Stimulation of leu-

kocyte adenylate-cyclase by hydrocortisone and isoprotere-nol in asthmatic and non-asthmatic subjects. JAllergy ClinImmunol 1972;50:45

25. Shelhamer JH, Metcalfe DD, Smith U, et al: Abnormal

beta adrenergic responsiveness in allergic subjects: Analysisof isoproterenol-induced cardiovascular and plasma cyclicadenosine monophosphate responses. J Allergy Clin Im-munol 1980;66:52

26. Smith U, Shelhamer JH, Kaliner M: Cholinergic nervous

system and immediate hypersensitivity: II. An analysis of

pupillary responses. J Allergy Clin Immunol 1980;66:37427. Frazer CM, Venter JC, Kaliner M: Autonomic abnormalities

and autoantibodies to beta-adrenergic receptors. N Engi J

Med 1981;305:1165

28. Brooks U, Cropp GJA: Theophylline therapy in bronchiol-itis: A retrospective study. Am J Dis Child 1981;135:934


1983;71;13Pediatrics Shimon W. Moses

Asher Tal, Chanan Bavilski, David Yohai, Jacob E. Bearman, Rafael Gorodischer andDexamethasone and Salbutamol in the Treatment of Acute Wheezing in Infants

ServicesUpdated Information &

http://pediatrics.aappublications.org/content/71/1/13including high resolution figures, can be found at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtmlentirety can be found online at: Information about reproducing this article in parts (figures, tables) or in its

Reprintshttp://www.aappublications.org/site/misc/reprints.xhtmlInformation about ordering reprints can be found online:


http://http://pediatrics.aappublications.org/content/71/1/13

http://www.aappublications.org/site/misc/Permissions.xhtml

http://www.aappublications.org/site/misc/reprints.xhtml

1983;71;13Pediatrics Shimon W. Moses

Asher Tal, Chanan Bavilski, David Yohai, Jacob E. Bearman, Rafael Gorodischer andDexamethasone and Salbutamol in the Treatment of Acute Wheezing in Infants

http://pediatrics.aappublications.org/content/71/1/13the World Wide Web at:

The online version of this article, along with updated information and services, is located on

Copyright © 1983 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397. American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007.been published continuously since 1948. Pediatrics is owned, published, and trademarked by the Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has


http://pediatrics.aappublications.org/content/71/1/13

dexamethasone and salbutamol in the treatment of acute ...has acute or recurrent wheezing. the lack...

Documents