Anda di halaman 1dari 12

Accepted: 28 March 2017

DOI: 10.1111/jan.13323

REVIEW

Is humidified better than non-humidified low-flow oxygen


therapy? A systematic review and meta-analysis

Zunjia Wen1,2,* | Wenting Wang1,* | Haiying Zhang2,* | Chao Wu2 | Jianping Ding1 |
Meifen Shen2

1
Nursing School of Soochow University, Su
Zhou, China Abstract
2
First Hospital Affiliated to Soochow Aims: To determine the effects of low-flow oxygen therapy with humidified or non-
University, Su Zhou, China
humidified oxygen in adult patients.
Correspondence Background: Although non-humidified oxygen in low-flow oxygen therapy is recom-
Meifen Shen, First Hospital Affiliated to
mended by many guidelines, humidifying oxygen regardless of oxygen flow has been
Soochow University, Su Zhou, Jiangsu
Province, China. routinely performed in China and Japan and further studies are needed to evaluate
Email: smf8165@126.com
the evidence.
Funding information Design: A systematic review and meta-analysis that comply with the recommenda-
This research received no specific grant from
tions of the Cochrane Collaboration were conducted.
any funding agency in the public, commercial
or not-for-profit sectors. Data sources: Studies (1980–2016) were identified by searching PUBMED,
EMBASE, Science Direct, Cochrane library, CNKI and Wanfang Database.
Methods: We performed a comprehensive, systematic meta-analysis of randomized
controlled trials on the efficacy of humidified and non-humidified low-flow oxygen
therapy. Summary risk ratios or weighted mean differences with 95% confidence
intervals were calculated using a fixed- or random-effects model.
Results: Twenty-seven randomized controlled trials with a total number of 8,876
patients were included. Non-humidified oxygen offers more benefits in reducing the
bacterial contamination of humidifier bottles, as shown by the mean operating time
for oxygen administration and the respiratory infections compared with humidified
oxygen therapy. No significant differences were found in dry nose, dry nose and
throat, nosebleed, chest discomfort, the smell of oxygen and SpO2 changes.
Conclusions: The routine humidification of oxygen in low-flow oxygen therapy is
not justifiable and non-humidified oxygen tends to be more beneficial. However,
considering that the quality of most included studies is poor, rigorously designed,
large-scale randomized controlled trials are still needed to identify the role of non-
humidified oxygen therapy.

KEYWORDS
care, humidification, meta-analysis, nurse, nursing, oxygen, review

1 | INTRODUCTION

Oxygen therapy is the administration of oxygen at higher concen-


trations than ambient air (20.9%) for combating hypoxia (Fulmer &
*These authors are equally contributed to this work. Snider, 1984). Previous studies (Heiring, Steensberg, Bjerager, &

2522 | © 2017 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/jan J Adv Nurs. 2017;73:2522–2533.
WEN ET AL. | 2523

Greisen, 2015; Ringbaek, Viskum, & Lange, 2002) have revealed


that long-term and low-flow oxygen administration in some chronic Why is this research or review needed?
diseases (such as chronic obstructive pulmonary disease) can not • Long-term and low-flow oxygen administration benefits
only improve hypoxia symptoms but also reduce the length of hos- many patients and is widely used in hospital and commu-
pital stay and prolong life expectancy. Nasal cannulas are the most nity settings.
widely used tool for oxygen administration in hospital and commu- • The related guidelines recommend non-humidified oxy-
nity; the oxygen delivered by nasal cannulas may or may not be gen if it is supplied to adults by nasal cannula at low
humidified by humidifier bottles. Notably, oxygen is routinely flow rate. However, in China and some other countries,
humidified clinically in China and Japan (Li, Wang, Zhong, et al., oxygen has been routinely humidified regardless of the
2010; Miyamoto, 2004) whenever low- or high-flow oxygen thera- oxygen flow rate.
pies are administered; however, in Europe and North America, oxy- • The evidence on the efficacy and effectiveness of oxy-
gen is not humidified in low-flow oxygen therapy (<5 L/min) gen humidification is lacking and warrants further investi-
(O’Driscoll, Howard, & Davison, 2008; American Thoracic Society, gation.
1995). The AARC clinical practice guideline (2007 revision and
update) has suggested that the humidification of oxygen is unnec- What are the key findings?
essary if it is supplied to adults by nasal cannula at flows ≤4 L/min
(Aarc 2007). Nonetheless, several guidelines support the humidifica- • Non-humidified oxygen seems to be more beneficial than
humidified oxygen in low-flow oxygen therapy in adult
tion of inhaled oxygen in long-term and low-flow oxygen therapy
populations.
(Magnussen et al., 2001; O’Reilly & Bailey, 2007). Therefore, the
efficacy and effectiveness of humidifying oxygen in low-flow oxy- • Some environmental and individual factors should be
considered when choosing whether to humidify oxygen.
gen therapy warrants further investigation.

How should the findings be used to influence


1.1 | Background policy/practice/research/education?

Considering the significance of this issue, the review authors found • Rigorously designed and high-quality large-scale random-
that many randomized controlled trials (RCTs) on the superiority of ized controlled trials are needed to identify the efficacy
humidified and non-humidified oxygen in low-flow oxygen therapy and effectiveness of non-humidified and humidified oxy-
have been performed, yet a meta-analysis synthesizing these data to gen therapy in children and adults.
provide robust evidence is lacking. Two previous Chinese meta-ana- • Incorporating this evidence into related nursing guideli-
lyses (Chu & Chen, 2016; Ren, Wang, Liang, & Zhou, 2016) have nes and translating it to clinical applications should be
concluded that non-humidified oxygen appears to be more beneficial promoted.
than humidified oxygen in low-flow oxygen therapy, but the included
data were all from Chinese papers and the sample sizes of the
included RCTs were small. Recently, several related RCTs that inves-
2.3 | Search methods
tigated this issue have been performed and published, thereby war-
ranting a larger sized meta-analysis. We planned, performed and reported this meta-analysis in compliance
with the PRISMA guideline (Liberati et al., 2009). Related articles that
have been either published in English or Chinese (1980 – 2016) were
2 | THE REVIEW
identified and selected by searching PUBMED, EMBASE, Science Direct,
Cochrane Central Register of Controlled Trials, China National Knowl-
2.1 | Aims
edge Infrastructure (CNKI) and Wanfang Database using the following
We performed this systematic review and meta-analysis of RCTs search terms: “oxygen inhalation therapy,” “oxygen inhalation ”, “oxygen
with the following aims: (1) to review the current evidence on the therapy,” “humidification”, “humidified,” “non-humidified,” “dried,” “low
oxygen humidification in low-flow oxygen therapy; (2) to compare flow,” and “low concentration.” We combined these terms in accordance
the efficacy of humidified and non-humidified oxygen in low-flow with the instructions of the database (Wen, Wang and Zhang). In addi-
oxygen therapy; and (3) to analyse and conclude the appropriateness tion, the reference lists of the retrieved studies and previous reviews
of oxygen humidification in low-flow oxygen therapy. and meta-analyses were reviewed and manually searched (Wen and
Wu) and we made no attempts to identify unpublished reports.

2.2 | Design and review methods


2.4 | Study selection
This systematic review and meta-analysis was conducted in concor-
dance with recommendations from the Cochrane Collaboration Study selection was made based on the first screening of identified
(Higgins & Green, 2011). titles or abstracts and on a second check-up of full-text articles
2524 | WEN ET AL.

(Wen, Wang and Zhang). Studies were considered to be eligible if 3 | RESULTS


the following criteria were met: (1) RCTs or quasi-RCTs design; (2)
adult population but not neonates or children; (3) studies were con- A total of 417 relevant publications were identified by compre-
ducted in the hospital regardless of confinement in regular, emer- hensive search and the abstracts of all citations were obtained.
gency, or intensive care units; (4) study patients have received low- We included 64 potentially related studies for further full-text
flow oxygen therapy (≤5 L/min); (5) containing the comparison review. After further screening and quality appraisal on those
groups of humidified and non-humidified oxygen administration; and studies, we included 27 RCTs for synthesized analysis
(6) reporting the relative outcome data (complications like dry nose (Figure 1).
and throat and others).

3.1 | Study characteristics


2.5 | Quality appraisal
The basic characteristics of 27 included studies are shown in
The Cochrane Collaboration’s “risk of bias” tool was adopted for
Table 1. In brief, for all the included studies, a total of 8,876 patients
evaluating the methodological quality and risk of bias of included
were involved, with 4,583 for non-humidified oxygen administration
RCTs. Seven specific domains were examined and measured in this
and 4,293 for humidified oxygen administration. Most studies were
tool, as follows: sequence generation, allocation concealment, blind-
conducted in China, with sample sizes ranging from 18 to 997. The
ing of participants and personnel, blinding of outcome assessment,
study settings also varied considerably. For oxygen administration,
incomplete outcome data, selective outcome reporting and “other”
most studies focused on the time point of longer than 24 h and they
issues. Every domain can be classified as “low risk of bias,” “high risk
chose sterile distilled water or sterile water for use as the humidifier
of bias,” or “unclear risk of bias” in accordance to the judgement cri-
fluid. The oxygen flow was all lower than 5 L/min. For outcome
teria (Cochrane Handbook for Systematic Reviews of Intervention.
measurements, some subjective indices such as the feeling of non-
Part 2: 8.5). Two reviewers (Wen and Wu) decided on the assess-
humidified nose and throat and chest discomfort were included for
ment and any dispute was solved by further discussion.
analysis. Some objective outcomes such as bacteria contaminations
of humidifier bottles and the mean operating time for oxygen admin-
2.6 | Data extraction istration were collected.
Figure 2 illustrates the risk of bias of the included RCTs. Even
The following data were extracted by two reviewers independently
if all 27 included RCTs mentioned randomization, only 12 pro-
(Wen and Wang): first author, year of publication, study design,
vided a detailed description of the methods used to produce the
patient population, oxygen administration (time, humidified fluid and
random sequence. Among them, some studies reported quasi-ran-
oxygen flow), main outcomes and study results. The studies selection
domization methods, such as assigning patients into non-humidi-
and data extraction were conducted by two authors independently
fied oxygen group or humidified group according to the odd or
and discussions resolved disagreements.
even property of admission date (Chen & Guo, 2015; Chen,
The main outcomes included: (1) dry nose; (2) dry throat; (3)
Huang, Peng, & Xiao, 2012). In general, sequentially numbered,
nosebleed; (4) chest discomfort; (5) the smell of oxygen during oxy-
opaque, sealed envelopes were assigned to each participant to
gen therapy; (6) bacteria contaminations of humidifier bottles; (7) the
avoid selective bias; only four RCTs (Andres et al., 1997; Franchini
mean operating time for oxygen administration; (8) respiratory infec-
et al., 2016; Liu & Chen, 2015; Zeng, Wang, Guo, & Huang,
tions; and (9) others (the cost of oxygen therapy and others).
2009) described adequate allocation concealments. Adequate blind-
ing, which is adopted for personnel, participants and outcome
2.7 | Data synthesis and analysis
assessment, is used to prevent against bias. Nevertheless, as a
All the extracted data were entered in the freeware program, Review simple and practical intervention, it seems to be impossible to
Manager (RevMan) Version 5.3. The data input was conducted and blind participants to oxygen administration. For most included
double-checked by two researchers (Wen and Zhang); the data syn- studies, no details on allocation concealment were reported, only
theses and interpretation were also performed by two authors (Wen two RCTs (Andres et al., 1997; Yang, Jing, & Han, 2014) have
and Shen) to ensure the accuracy of results. Binary outcomes were reported a blinding design on participants and outcome assess-
presented as Mantel–Haenszel-style odd ratios (ORs) with 95% con- ment. Most included studies reported completed outcome data
fidence intervals and continuous outcomes were reported as inverse (low risk of bias). Verification on selective reporting of outcomes
variance weighted mean differences (WMDs). A fixed-effect model are necessary because it may help evaluate the integrity of out-
was adopted in cases of homogeneity (p value of v2 test >.10 and come reporting and protect against bias; three studies (Tian,
I2<50%), whereas a random-effects model was used in cases of obvi- Huang, Yang, & Zuo, 2012; Zeng et al., 2009; Zhou, Zhou, Zhang,
ous heterogeneity (p value of v test >.10 and I >50%). Publication
2 2
& Huang, 2015) appeared to selectively report the outcomes,
bias was evaluated by using funnel plots and asymmetry was because the outcomes of their studies were rather simple and lim-
assessed by the Egger regression test (p value of <.1 was considered ited. No other significant biases were found. The overall risk of
to be significant for funnel plot asymmetry). bias is shown in Figure 3.
WEN ET AL. | 2525

Identification

Records identified through Additional records identified


database searching through other sources
(n = 468) (n = 18)

Records after duplicates removed


(n = 417)
Screening

Records screened Records excluded


(n = 417) (n = 353)

Full-text articles excluded, with


Full-text articles assessed reasons (n = 37):
for eligibility non-RCT (n = 9);
Eligibility

(n = 64) inappropriate intervention (n = 24);


information incomplete (n = 4)

Studies included in
qualitative synthesis
(n = 27)
Included

Studies included in
qualitative synthesis
(meta-analysis)
(n = 27)

FIGURE 1 Flow diagram of study selection

Liu & Chen, 2015; Yuan, Zhu, & Zhang, 2013; Yue, He, &
3.2 | Main analysis
Chang, 2015; Zeng et al., 2009) reported the incidence of dry
3.2.1 | The incidence of dry nose nose and throat after oxygen administration; the summary OR
on the incidence of dry nose and throat was 0.93 (95% CI:
Four studies (Andres et al., 1997; Chen et al., 2012; Tian et al.,
0.78–1.10), with no evidence of heterogeneity (p = .35, I2 = 10%)
2012; Yang & Yu, 2015) reported the incidence of dry nose after
(Figure 4b).
oxygen administration; the summary OR on the incidence of dry
nose was 1.21 (95% CI: 0.76–1.93), with evidence of heterogeneity
(p = .07, I2 = 58%) (Figure 4a).
3.2.3 | The incidence of cough
Two studies (Andres et al., 1997; Franchini et al., 2016) reported the
3.2.2 | The incidence of dry nose and throat
incidence of cough after oxygen administration; the summary OR on
Nine studies (Cai, Dong, Lv, Tang, & Wu, 2013; Chen & Guo, the incidence of cough was 0.80 (95% CI: 0.42–1.52), with no evi-
2015; Han, 2011; Li, Wang, Wang, et al., 2010; Li et al., 2015; dence of heterogeneity (p = .17, I2 = 6%) (Figure 4c).
2526 | WEN ET AL.

T A B L E 1 The characteristics of included studies


Intervention

Time for Fluid for


Sample oxygen oxygen Oxygen
Study Country (T/C) Department administration humidification flow Outcomes
Andres et al. (1997) Canada 126/111 NA >2 days NA <4 L/min a, b, c, d, e, f
Campbell et al. (1988) USA 86/99 Respiratory department NA NA 5 L/min a, b, c, d, e
Franchini et al. (2016) Brazil 10/8 Basic health units/ >12 h NA NA f, g
Pulmonary outpatient clinic
Cai et al. (2013) China 51/29 Veteran ward >4 days SDW NA a, b, j, k
Zeng et al. (2009) China 30/30 Neurosurgery department >5 days SDW NA a, b
Chen and Guo (2015) China 328/352 Emergency department NA SDW NA a, b, c, d, i, k, l
Chen et al. (2012) China 235/245 Cadiovascular department >12 h SDW <4 L/min a, b, c, d, e, h
Han (2011) China 400/400 Respiratory department >24 h NA 4 L/min a, b, k
Haung et al. (2014) China 40/40 Blood purification centre NA SDW 1–4 L/min a, e, i, k
Huang et al. (2014) China 507/490 Cadiovascular department >12 h NA <4 L/min a, b, c, l, m
Li, Wang, Wang, et al. China 418/348 Neurosurgery/ >12 h NA < 4 L/min a, b, c, d, e, h
(2010), Li, Wang, respiratory department
Zhong, et al. (2010)
Li et al. (2011) China 130/130 NA >7 days SDW NA a, b, i, k
Li et al. (2015) China 173/168 NA >72 h SDW <4 L/min a, b, k
Liu and Chen (2015) China 174/174 The dialysis centre 3–4.5 h SDW NA a, b, e, i
Lou et al. (2016) China 101/99 Orthopaedic department NA SDW <4 L/min a, b, d, i, k, l
Ning et al. (2014) China 201/199 ICU >24 h SDW ≤4 L/min a, i, k, l
Tian et al. (2012) China 316/218 CCU ≥72 h SDW <4 L/min a, b, l
Wu (2016) China 90/90 Respiratory department 24–72 h SDW 1–4 L/min a, b, e, j, l, n, o
Xia et al. (2014) China 86/99 Emergency department NA NA 1–2 L/min a, b, c, d
Yang et al. (2014) China 165/65 NA >5 days SDW NA a, b, c, d, h, i
Yang et al. (2015) China 184/201 NA >24 h NA <4 L/min a, b, d, h, i, k, l, m
Yang and Yu (2015) China 70/70 Emergency department <48 h SDW <4 L/min a, b, k, l
Yuan et al., (2011) China 72/78 Respiratory department NA SDW ≤4 L/min a, b, d, i
Yuan et al. (2013) China 256/230 NA >24 h SDW NA a, b, d, e, i
Zheng et al. (2015) China 60/60 Cadiovascular department >24 h SDW <4 L/min a, b, d, e, g, h, i, l
Zhou et al. (2015) China 50/50 Neurosurgery department NA SDW 1–4 L/min k, m, p
Yue et al. (2015) China 224/210 Respiratory department >5 days SDW 2 L/min a, b, d, h, i

NA, not available; SDW, sterile distilled water; a, non-humidified nose; b, non-humidified throat and non-humidified mouth; c, headache; d, chest discom-
fort; e, nosebleed; f, cough; g, sleep problems; h, nausea; i, the smell of oxygen during oxygen therapy; j, SpO2; k, bacteria contaminations of humidifier
bottles; l, the mean operating time for oxygen administration; m, respiratory infections; n, heart rate; o, respiratory rate; p, the cost of oxygen therapy.

chest discomfort was 0.91 (95% CI: 0.53–1.55), with no evidence of


3.2.4 | The incidence of nosebleed
heterogeneity (p = .59, I2 = 0%) (Figure 4e).
Six studies (Andres et al., 1997; Chen et al., 2012; Liu & Chen,
2015; Wu, 2016; Yuan et al., 2013; Yue et al., 2015) reported the
3.2.6 | Bacteria contaminations of humidifier
incidence of nosebleed after oxygen administration; the summary
bottles
OR on the incidence of nosebleed was 1.34 (95% CI: 0.77–2.34),
with no evidence of heterogeneity (p = .82, I2 = 0%) (Figure 4d). Eight studies (Haung, Huang, & Chen, 2014; Li, Chen, & Jiang, 2011;
Li et al., 2015; Lou, Zhu, Liu, & Tian, 2016; Ning, Ouyang, Jiang, &
Yu, 2014; Yang, Huang, & Tian, 2015; Yang & Yu, 2015; Zhou et al.,
3.2.5 | The incidence of chest discomfort
2015) reported the bacterial contamination of humidifier bottles
Four studies (Andres et al., 1997; Chen et al., 2012; Yuan et al., after oxygen administration; the summary OR on the bacteria con-
2013; Yue et al., 2015) reported the incidence of chest discomfort tamination of humidifier bottles was 0.16 (95% CI: 0.06–0.43), with
after oxygen administration; the summary OR on the incidence of evidence of heterogeneity (p < .01, I2 = 90%) (Figure 5a).
WEN ET AL. | 2527

Blinding of participants and personnel (perfomance bias)


3.2.7 | The mean operating time for oxygen
administration (seconds)

Blinding of outcome assessment (detection bias)


Random sequence generation (selection bias)
Six studies (Chen & Guo, 2015; Huang, Tian, & Yang, 2014; Lou
et al., 2016; Tian et al., 2012; Yang et al., 2015; Zheng, Yao, Sun,

Incomplete outcome data (attrition bias)


Allocation concealment (selection bias)
Shen, & Zhao, 2015) reported the mean operating time for oxygen

Selective reporting (reporting bias)


administration; the summary WMD on mean operating time (sec-
onds) for oxygen administration was 35.84 (95% CI: 44.51 to
27.17), with evidence of heterogeneity (p < .01, I2 = 99%)
(Figure 5b).

Other bias
3.2.8 | The smell of oxygen
Four studies (Chen et al., 2012; Liu & Chen, 2015; Yuan et al., 2013;

Andres 1997 + + + + + + + Yue et al., 2015) reported the smell of oxygen after oxygen adminis-
tration; the summary OR on smell of oxygen was 1.35 (95% CI:
Cai 2013 ? ? ? + + + +
0.69–2.65), with no evidence of heterogeneity (p = .41, I2 = 0%)
Campbell 1988 ? ? ? + + + + (Figure 5c).

Chen 2012 + – ? – + + +

Chen 2015 + + ? – + + + 3.2.9 | SpO2

Franchini 2016 ? + ? ? + + + Two studies (Cai et al., 2013; Wu, 2016) reported SpO2 after oxygen
administration; the summary WMD on SpO2 was 0.60 (95% CI:
Han 2011 ? – ? – + + +
3.32–2.21), with evidence of heterogeneity (p < .01 I2 = 97%)
Huang J 2014 ? – ? – + + + (Figure 5d).

Huang S 2014 ? – ? – + + +

Li 2010 ? – ? – + + + 3.2.10 | Respiratory infections

Li 2011 ? – ? – + + + Four studies (Huang et al., 2014; Yang et al., 2015; Yue et al., 2015;
Zhou et al., 2015) reported the respiratory infection after oxygen
Li 2015 ? – ? – + + +
administration; the summary OR on respiratory infection was 0.39
Liu 2015 ? – ? – + + + (95% CI: 0.21–0.73), with no evidence of heterogeneity (p = .14,

Lou 2016 ? – ? – + + + I2 = 45%) (Figure 5e).

Ning 2014 + – ? – + + +

+ – – + + +
3.3 | Subgroup and sensitivity analyses
Tian 2012 ?

– ? – + + + No subgroup analyses were performed in this study. Sensitivity anal-


Wu 2016 ?
yses, which investigate the influence of a single study on the overall
Xia 2014 + – ? – + + +
risk estimate by abandoning one study in each turn, suggested that
Yang 2014 + ? + + + + + the overall risk estimates do not substantially change by any single
study.
Yang B 2015 ? – ? – + + +

Yang T 2015 + – ? – + + +

Yuan 2011 + – ? – + + + 4 | DISCUSSION

Yuan 2013 ? – ? – + + +
Findings from our meta-analysis indicate that non-humidified and
Yue 2015 + – ? – + + + humidified oxygen in low-flow oxygen administration provide no

+ + – + – + effect differences on dry nose, dry nose and throat, nosebleed, chest
Zeng 2009 ?
discomfort, smell of oxygen and SpO2 after oxygen administration, but
Zheng 2015 ? – ? – + + + non-humidified oxygen provide more beneficial effects on the reduc-
Zhou 2015 ? – ? – + – + tion of the bacterial contamination of humidifier bottles, shortening
the mean operating time for oxygen administration and decreasing
FIGURE 2 The risk of bias summary for included studies respiratory infections. Therefore, the use of non-humidified oxygen
2528 | WEN ET AL.

Random sequence generation (selection bias)


Allocation concealment (selection bias)
Blinding of participants and personnel (performance bias)
Blinding of outcome assessment (detection bias)
Incomplete outcome data (attrition bias)
Selective reporting (reporting bias)
Other bias

0% 25% 50% 75% 100%


F I G U R E 3 The overall risk of bias
Low risk of bias Unclear risk of bias High risk of bias
graph for included studies

seems to be better than humidified oxygen in low-flow oxygen ther- Our review revealed that non-humidified oxygen offers more
apy. However, considering the quality of most included RCTs is poor, advantages in reducing bacteria contaminations of humidifier bottles,
our results should be interpreted with caution. the mean nursing operating time and respiratory infection without
In the past decades, humidification of the inhaled oxygen in low- increasing the incidence of other complications. Respiratory infection
flow oxygen therapy was considered unnecessary given that oxygen leads the list of hospital-acquired infections. While bacteria contami-
is not readily soluble in water (Miyamoto, 2004). The routinely used nations of humidifier bottles are closely related to respiratory infec-
cold bubble humidification for oxygen administration may not actu- tion (Kobayashi, Yamazaki, & Maesaki, 2006), the wet condition of
ally work (Vargas & Esquinas, 2016). A recent critical study (Fran- humidifier bottles with water inside may be more adaptable to bac-
chini et al., 2016) demonstrated that the cold bubble humidification teria proliferation. A survey (Li, Wang, Zhong, et al., 2010) reported
does not humidify inspired oxygen at all and offers no benefits in that the bacteria contamination rate of humidifier bottles with water
the promotion of mucociliary clearance, mucus hydration and pul- inside after 24-h oxygen therapy reached 57.5%, yet the non-humi-
monary function. Therefore, many related guidelines (Aarc 2007; dified oxygen group is only 44.2%, with the top three bacteria being
O’Driscoll et al., 2008; O’Driscoll et al., 1995) have recommended Staphylococcus aureus, Escherichia coli and Bacillus. The contamination
the direct use of non-humidified oxygen in low-flow oxygen therapy. of humidifier bottles possibly increases the risk for respiratory infec-
This may explain why the recent studies on this issue are quite few tion with the inhalation of contaminated oxygen. Ideally, the non-
in European and American areas; however, in China and Japan, humidified oxygen requires no addition or change of sterile water,
humidification of oxygen in clinic is still routinely performed whether which can significantly reduce the operating time for oxygen ther-
the oxygen flow is low or high (Li, Wang, Zhong, et al., 2010; Miya- apy, thereby relieving the nursing workload and decreasing the medi-
moto, 2004). Many Chinese medical researchers have investigated cal expenditure. Our results on the bacterial contamination of
this issue in the past few years as this situation warrants changes. humidifier bottles and the mean nursing operating time are signifi-
Patients receiving low-flow oxygen therapy by nasal cannula cantly heterogeneous. These outcomes should be interpreted care-
actually only inhale 2.4–19% of provided oxygen because the upper fully as the possible causes of the heterogeneity may be the
respiratory tract can provide 75% heat and moisture to make the difference on the nursing intervention and operating time calcula-
inhaled air comfortable for us (Branson, 1998). Additionally, the rela- tion. Some of the included RCTs did not indicate the accurate oxy-
tive humidity of inhaled air is more important than absolute humid- gen flow in the inhalation therapy, possibly contributing to the
ity; the discomfort caused by non-humidified oxygen to patients is results’ heterogeneity.
far lower than the indoor air brings (Sottiaux, 2006). This may Although our results support the non-humidified oxygen applica-
explain the results that non-humidified and humidified oxygen pro- tion in the low-flow oxygen therapy, several details should be taken
duce no significant difference on the complications, such as the inci- into consideration. No significant change in airway humidity was
dence of non-humidified nose, non-humidified throat, nosebleed and observed when oxygen was given pre-nasally without humidification
chest discomfort. However, these outcomes are rather subjective as (Dellweg, Wenze, Hoehn, Bourgund, & Haidl, 2013). The room air
the measurements on these outcomes are not accurate enough. Sev- humidity may act as an important factor. A Chinese survey in Xin-
eral studies (Campbell, Baker, & Crites-Silver, 1988; Ning et al., jiang, northwest area with dry climate of China found that non-humi-
2014; Xia, Gu, Zhang, Liu, & Lv, 2014; Yang et al., 2014; Zheng dified oxygen leads to more complications when compared with the
et al., 2015) conducted classified evaluation on the severity of these humidified oxygen in low-flow oxygen therapy (Yang et al., 2014).
symptoms, but due to the criteria requirements, we did not include Hence, we should take air humidity into account when performing
these data in the meta-analysis. Besides, non-humidified and humidi- oxygen therapy. Besides, more studies are needed to address other
fied oxygen inhalation exerted similar effects on SpO2, indicating possible effects achieved with humidified oxygen administration,
similar therapeutic values. However, we only included two RCTs on such as the clarification of the effect of humidified oxygen on the
this outcome. Further studies are needed to elucidate the differ- recovery of patients with chest disease. Also, as the relative risks/
ences in efficacy. benefits of high vs. low oxygen therapy and humidified vs. non-
WEN ET AL. | 2529

Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio


Study or Subgroup Events Total Events Total Weight M-H, Random, 95% CI M-H, Random, 95% CI
Andres 1997 52 126 32 111 27.8% 1.73 [1.01, 2.98]
Chen 2012 10 235 8 245 15.7% 1.32 [0.51, 3.40]
Tian 2012 221 316 211 281 35.2% 0.77 [0.54, 1.11]
Yang T 2015 52 70 46 70 21.3% 1.51 [0.73, 3.12]

Total (95% CI) 747 707 100.0% 1.21 [0.76, 1.93]


Total events 335 297
Heterogeneity: Tau2 = 0.13; Chi2 = 7.12, df = 3 (P = 0.07); I2 = 58% 0.1 0.2 0.5 1 2 5 10
Test for overall effect: Z = 0.81 (P = 0.42)
Favours [nonhumidified] Favours [humidified]
(a) The forest plot for the incidence of dry nose
Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Cai 2013 4 51 2 29 0.9% 1.15 [0.20, 6.69]
Chen 2015 138 328 144 352 29.7% 1.05 [0.77, 1.42]
Han 2011 71 400 103 400 31.3% 0.62 [0.44, 0.87]
Li 2010 47 418 30 348 10.7% 1.34 [0.83, 2.17]
Li 2015 120 173 121 168 13.9% 0.88 [0.55, 1.40]
Liu 2015 7 174 8 174 2.8% 0.87 [0.31, 2.45]
Yuan 2013 14 256 11 230 4.0% 1.15 [0.51, 2.59]
Yue 2015 14 224 12 210 4.3% 1.10 [0.50, 2.44]
Zeng 2009 12 30 11 30 2.4% 1.15 [0.41, 3.26]

Total (95% CI) 2054 1941 100.0% 0.93 [0.78, 1.10]


Total events 427 442
Heterogeneity: Chi2 = 8.90, df = 8 (P = 0.35); I2 = 10%
0.2 0.5 1 2 5
Test for overall effect: Z = 0.85 (P = 0.39)
Favours [nonhumidified] Favours [humidified]
(b) The forest plot for the incidence of dry nose and throat
Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio
Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Andres 1997 101 126 95 111 94.7% 0.68 [0.34, 1.35]
Franchini 2016 5 10 2 8 5.3% 3.00 [0.40, 22.71]

Total (95% CI) 136 119 100.0% 0.80 [0.42, 1.52]


Total events 106 97
Heterogeneity: Chi2 = 1.85, df = 1 (P = 0.17); I2 = 46%
0.01 0.1 1 10 100
Test for overall effect: Z = 0.68 (P = 0.50)
Favours [nonhumidified] Favours [humidified]
(c) The forest plot for the incidence of cough

Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio


Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Andres 1997 22 126 12 111 48.8% 1.75 [0.82, 3.71]
Chen 2012 1 235 1 245 4.5% 1.04 [0.06, 16.77]
Liu 2015 2 174 1 174 4.6% 2.01 [0.18, 22.39]
Wu 2016 2 90 1 90 4.5% 2.02 [0.18, 22.71]
Yuan 2013 1 256 1 230 4.9% 0.90 [0.06, 14.44]
Yue 2015 5 224 7 210 32.7% 0.66 [0.21, 2.12]

Total (95% CI) 1105 1060 100.0% 1.34 [0.77, 2.34]


Total events 33 23
Heterogeneity: Chi2 = 2.21, df = 5 (P = 0.82); I2 = 0%
0.05 0.2 1 5 20
Test for overall effect: Z = 1.04 (P = 0.30)
Favours [nonhumidified] Favours [humidified]
(d) The forest plot for the incidence of nosebleed

Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio


Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Andres 1997 29 126 26 111 75.0% 0.98 [0.53, 1.79]
Chen 2012 3 235 2 245 6.8% 1.57 [0.26, 9.49]
Yuan 2013 1 256 1 230 3.7% 0.90 [0.06, 14.44]
Yue 2015 1 224 4 210 14.5% 0.23 [0.03, 2.08]

Total (95% CI) 841 796 100.0% 0.91 [0.53, 1.55]


Total events 34 33
Heterogeneity: Chi2 = 1.90, df = 3 (P = 0.59); I2 = 0%
0.01 0.1 1 10 100
Test for overall effect: Z = 0.36 (P = 0.72)
Favours [nonhumidified] Favours [humidified]
(e) The forest plot for the incidence of chest discomfort

FIGURE 4 The forest plots of different outcomes for included studies


2530 | WEN ET AL.

Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio


Study or Subgroup Events Total Events Total Weight M-H, Random, 95% CI M-H, Random, 95% CI
Huang J 2014 1 40 10 40 9.4% 0.08 [0.01, 0.63]
Li 2011 2 130 37 130 12.1% 0.04 [0.01, 0.17]
Li 2015 3 173 95 168 13.3% 0.01 [0.00, 0.04]
Lou 2016 2 101 20 99 12.0% 0.08 [0.02, 0.35]
Ning 2014 53 201 77 199 15.8% 0.57 [0.37, 0.87]
Yang B 2015 41 184 67 201 15.8% 0.57 [0.36, 0.90]
Yang T 2015 20 70 20 70 15.0% 1.00 [0.48, 2.08]
Zhou 2015 0 50 4 50 6.7% 0.10 [0.01, 1.95]

Total (95% CI) 949 957 100.0% 0.16 [0.06, 0.43]


Total events 122 330
Heterogeneity: Tau2 = 1.58; Chi2 = 68.44, df = 7 (P < 0.00001); I2 = 90%
0.005 0.1 1 10 200
Test for overall effect: Z = 3.62 (P = 0.0003)
Favours [nonhumidified] Favours [humidified]
(a) The forest plot for bacteria contaminations of humidifier bottles

Nonhumidified oxygen Humidified oxygen Mean Difference Mean Difference


Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI
Chen 2015 15.64 3.32 328 58.47 8.11 352 17.1% –42.83 [–43.75, –41.91]
Huang S 2014 49.5 18.02 507 77.19 17.48 490 17.0% –27.69 [–29.89, –25.49]
Lou 2016 47.2 13.06 101 105.8 15.01 99 16.6% –58.60 [–62.50, –54.70]
Tian 2012 49.5 18.02 316 77.19 17.48 281 16.8% –27.69 [–30.54, –24.84]
Yang B 2015 48.76 17 184 78.18 17.84 201 16.7% –29.42 [–32.90, –25.94]
Zheng 2015 49.32 16.42 60 77.94 17.08 60 15.8% –28.62 [–34.61, –22.63]

Total (95% CI) 1496 1483 100.0% –35.84 [–44.51, –27.17]


Heterogeneity: Tau2 = 114.12; Chi2 = 363.41, df = 5 (P < 0.00001); I2 = 99%
–50 –25 0 25 50
Test for overall effect: Z = 8.10 (P < 0.00001)
Favours [nonhumidified] Favours [humidified]
(b) The forest plot for mean operation time for oxygen administration

Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio


Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Chen 2012 2 235 3 245 19.8% 0.69 [0.11, 4.18]
Liu 2015 1 174 3 174 20.3% 0.33 [0.03, 3.20]
Yuan 2013 3 256 2 230 14.1% 1.35 [0.22, 8.16]
Yue 2015 15 224 7 210 45.8% 2.08 [0.83, 5.21]

Total (95% CI) 889 859 100.0% 1.35 [0.69, 2.65]


Total events 21 15
Heterogeneity: Chi2 = 2.86, df = 3 (P = 0.41); I2 = 0%
Test for overall effect: Z = 0.87 (P = 0.39) 0.01 0.1 1 10 100
(c) The forest plot for the smell of oxygen Favours [nonhumidified] Favours [humidified]

Nonhumidified oxygen Humidified oxygen Mean Difference Mean Difference


Study or Subgroup Mean SD Total Mean SD Total Weight IV, Random, 95% CI IV, Random, 95% CI
Cai 2013 98.06 1.03 51 97.29 1.15 29 50.5% 0.77 [0.26, 1.28]
Wu 2016 90 3 90 92 2 90 49.5% –2.00 [–2.74, –1.26]

Total (95% CI) 141 119 100.0% –0.60 [–3.32, 2.11]


Heterogeneity: Tau2 = 3.73; Chi2 = 36.39, df = 1 (P < 0.00001); I2 = 97%
Test for overall effect: Z = 0.43 (P = 0.66) –10 –5 0 5 10
(d) The forest plot for the SpO2 Favours [nonhumidified] Favours [humidified]

Nonhumidified oxygen Humidified oxygen Odds Ratio Odds Ratio


Study or Subgroup Events Total Events Total Weight M-H, Fixed, 95% CI M-H, Fixed, 95% CI
Huang S 2014 8 507 9 490 25.3% 0.86 [0.33, 2.24]
Yang B 2015 2 184 3 201 8.0% 0.73 [0.12, 4.39]
Yue 2015 3 224 13 210 37.1% 0.21 [0.06, 0.73]
Zhou 2015 2 50 11 50 29.6% 0.15 [0.03, 0.71]

Total (95% CI) 965 951 100.0% 0.39 [0.21, 0.73]


Total events 15 36
Heterogeneity: Chi2 = 5.47, df = 3 (P = 0.14); I2 = 45%
Test for overall effect: Z = 2.99 (P = 0.003) 0.01 0.1 1 10 100
Favours [nonhumidified] Favours [humidified]
(e) The forest plot for the incidence of respiratory infection

FIGURE 5 The forest plots of different outcomes for included studies


WEN ET AL. | 2531

humidified oxygen therapy in child and adult remain different (Nath, design (randomization, allocation, blinding, outcomes evaluation and
Ponnusamy, Willis, Bissett, & Clarke, 2010), more individual factors interpretation) to provide more insights and evidence into this issue.
should be considered.
Considering that only three fair-quality RCTs (Andres et al., 1997;
CONFLICT OF INTEREST
Franchini et al., 2016; Yang et al., 2014) were included, sensitivity
analysis of those better quality studies should be performed. Those No conflict of interest has been declared by the authors.
studies that used sound designs for the method of randomization and
blinding are likely to have less risk of bias. Andres et al. (1997) recom-
AUTHOR CONTRIBUTIONS
mended that accepting non-humidification as the standard method for
delivering low-flow oxygen therapy may decrease the patient’s risk of All authors have agreed on the final version and meet at least one
developing nosocomial pneumonia, reduce expenditures for oxygen of the following criteria [recommended by the ICMJE (http://www.ic
therapy and eliminate one source of biomedical waste. However, the mje.org/recommendations/)]:
study of Franchini et al. (2016) showed that unheated bubble humidi-
fication does not humidify inspired oxygen to prevent deterioration of • substantial contributions to conception and design, acquisition of
data or analysis and interpretation of data;
mucociliary clearance, mucus hydration and pulmonary function; this
may support the finding that humidification performed no better than • drafting the article or revising it critically for important intellectual
content.
non-humidification, which is consistent with Andres’s results (Andres
et al., 1997). Yang et al. (2014) found that the dryness of oropharyn-
geal condition is related to the humidification, but the study was con- REFERENCES
ducted in the dry area of China and the data are limited for our
American Thoracic Society (1995) Standards for the diagnosis and care
synthesized analysis; these results should be interpreted with caution. of patients with chronic obstructive pulmonary disease. American
Similarly, Andres et al. (1997) also found that the mean symptom Thoracic Society. American Journal of Respiratory and Critical Care
scores for nasal dryness in humidification group were significantly Medicine, 152(5 Pt 2), S77–S121.
AARC (2007) AARC clinical practice guideline. Oxygen therapy in the
lower than the non-humidification group. Therefore, as the effect of
home or alternate site health care facility–2007 revision & update.
cold bubble humidification works remain conflicting and future studies Respiratory Care, 52(8), 1063–1068.
should focus more on this problem. Andres, D., Thurston, N., Brant, R., Flemons, W., Fofonoff, D., Ruttimann,
Several limitations in this review should be noted. First, as most A., . . . Neil, C. (1997). Randomized double-blind trial of the effects of
included RCTs were conducted in China, population differences may humidified compared with nonhumidified low flow oxygen therapy
on the symptoms of patients. Paediatrics & Child Health.
exist. Second, the quality of included studies is disturbing as most
Branson, R. D. (1998). The effects of inadequate humidity. Respiratory
studies do not use a blind design on the allocation concealment and Care Clinics of North America, 4(2), 199–214.
participants. For some studies with a blind design, the authors only Cai, X., Dong, M., Lv, H., Tang, X. Y., & Wu, W. (2013). Research on con-
state that treatment allocation is concealed without providing details tamination situation of oxygen humidification bottles with and with-
out humidification fluid. Nursing and Rehabilitation Journal, 12(12),
on how this was performed. The lack of blinding design on partici-
1122–1124.
pants and personnel may lead to significant bias. Future studies Campbell, E. J., Baker, M. D., & Crites-Silver, P. (1988). Subjective effects
should focus more on study design. Third, due to data limitation, we of humidification of oxygen for delivery by nasal cannula. A prospec-
did not perform subgroup analysis and funnel plot. Some of our tive study. Chest, 93(2), 289–293.
Chen, Q., & Guo, Y. (2015). Non-humidification nasal catheter oxygen
results exhibited evident heterogeneity, with the possible causes of
inhalation applicated in emergency patients. Public Medical Forum
the heterogeneity difficult to identify. Finally, targeted on adult pop- Magazine, 19(7), 884–886.
ulation, our findings are unlikely to be relevant to neonates or chil- Chen, X., Huang, G., Peng, D., & Xiao, J. (2012). The effect of humilified
dren, a separate analysis to establish benefits and harms for these and nonhumilified low-flow nasal oxygen on relieving respiratory
symptoms. Practical Clinical Medicine, 13(3), 119–120.
specific population groups is needed.
Chu, S., & Chen, X. (2016). A systematic review of the efficacy of contin-
uing low-flow non-humidified nasal cannula oxygen therapy. Chinese
Nursing Management, 16(1), 88–93.
5 | CONCLUSION Dellweg, D., Wenze, M., Hoehn, E., Bourgund, O., & Haidl, P. (2013).
Humidification of inspired oxygen is increased with pre-nasal cannula,
compared to intranasal cannula. Respiratory Care, 58(8), 1323–1328.
Non-humidified oxygen appears to be of greater benefit than humid- Franchini, M. L., Athanazio, R., Amato-Lourenco, L. F., Carreirao-Neto,
ified oxygen in low-flow oxygen therapy. However, some environ- W., Saldiva, P. H., Lorenzi-Filho, G., . . . Nakagawa, N. K. (2016). Oxy-
mental and individual factors should be considered when considering gen with cold bubble humidification is no better than dry oxygen in
the necessity of humidifying oxygen. Meanwhile, considering that preventing mucus dehydration, decreased mucociliary clearance and
decline in pulmonary function. Chest, 150(2), 407–414.
our results are based on poor quality and heterogeneous trial evi-
Fulmer, J. D., & Snider, G. L. (1984). American College of Chest Physi-
dence, the results should be interpreted with caution despite our sig- cians/National Heart, Lung and Blood Institute National Conference
nificant results. Future studies should focus more on the study on Oxygen Therapy. Heart and Lung, 13(5), 550–562.
2532 | WEN ET AL.

Han, C. (2011). Randomised controlled trials of humidified and nonhu- Ren, Y., Wang, C., Liang, Y., & Zhou, H. (2016). Meta-analysis of oxygen
midified oxygen in the Low-flow nasal cannula oxygen in patients inhalation therapy with humidifying and non-humidifying oxygen. Chi-
with respiratory symptoms. Chinese Journal of Medicinal Guide, 13(7), nese Journal of Practical Nursing, 32(7), 553–556.
1257–1258. Ringbaek, T. J., Viskum, K., & Lange, P. (2002). Does long-term oxygen
Haung, J., Huang, X., & Chen, C. (2014). Low-flow oxygen administration therapy reduce hospitalisation in hypoxaemic chronic obstructive pul-
with wet and dry oxygen in special care unit. Journal of Qiqihar Medi- monary disease? European Respiratory Journal, 20(1), 38–42.
cal College, 35(8), 1215–1216. Sottiaux, T. M. (2006). Consequences of under- and over-humidification.
Heiring, C., Steensberg, J., Bjerager, M., & Greisen, G. (2015). A random- Respiratory Care Clinics of North America, 12(2), 233–252.
ized trial of low-flow oxygen versus nasal continuous positive airway Tian, F., Huang, S., Yang, B., & Zuo, H. (2012). Continuous low-to-mid flow
pressure in preterm infants. Neonatology, 108(4), 259–265. non-humidification oxygen inhalation therapy through nasal tube in
Higgins, J. P., & Green, S. (2011). Cochrane handbook for systematic patients with cardiovascular disease. Journal of Nursing, 19(12), 59–62.
reviews of interventions (Version 5.1.0). USA: The Cochrane collobra- Vargas, M., & Esquinas, A. M. (2016). Unheated or no humidification bub-
tion. ble for long-term nasal low-flow oxygen: A matter of nasal mucosa
Huang, S., Tian, F., & Yang, B. (2014). The effect of low flow nasal can- response or disease progression. Chest, 150(3), 750.
nula oxygen therapy with nonhumidified oxygen in patients with car- Wu, A. (2016). The effect of dry and humidified low-flow dual lumen
diovascular disesase. International Journal of Nursing, 33(2), 484–486. nasal cannula oxygen administration in COPD patients. Medical Fron-
Kobayashi, N., Yamazaki, T., & Maesaki, S. (2006). Bacteriological moni- tier, 6(9), 84–85.
toring of water reservoirs in oxygen humidifiers: Safety of prolonged Xia, L., Gu, M., Zhang, K., Liu, Y., & Lv, S. (2014). Effect of low-flow, dry
and multipatient use of prefilled disposable oxygen humidifier bottles. oxygen therapy on the comfort experience of patients with acute
Infection Control and Hospital Epidemiology, 27(3), 320–322. exacerbations of chronic obstructive pulmonary disease. Journal of
Li, X., Chen, G., & Jiang, W. (2011). Clinical application study of using Nursing Science, 29(15), 50–51.
low flow oxygen with disposable oxygen humidification bottle with- Yang, B., Huang, S., & Tian, F. (2015). The effects of nonhumidified low
out humidification. The Journal of Medical Theory and Practice, 24(17), flow nasal cannula oxygen administration. Chinese General Practice
2047–2048. Nursing, 13(31), 3126–3129.
Li, Z., Liu, R., You, Y., Dong, F., Guo, S., & Li, Z. (2015). Clinical research Yang, Y., Jing, C., & Han, X. (2014). The effects of humidified and nonhu-
on non -wetting low -to -moderate flow oxygen to prevent pul- midified oxygen administration by nasal cannula in Urumqi, Xinjiang.
monary infections in patients with cardiovascular diseases. Chinese Today Nurse, 3(6), 141–142.
Journal of Nosocomiology, 25(13), 3008–3010. Yang, T., & Yu, C. (2015). The application of nohumidified low-flow oxygen
Li, L., Wang, W., Wang, G. M., Liu, Z., Lin, Y., Wu, S., & Chen, Y. (2010). in emergency department. Guide of China Medicine, 13(9), 153–154.
A comparison study on the respiratory symptoms between patients Yuan, Q., Liu, M., Huang, S., & Chen, Z. (2011). Effect of Non-humidified
receiving low-to-mid flow humidified and dry oxygen by nasal can- Low-flow Oxygen Therapy by Nasal Cannula. Journal of Nursing
nula. Chinese Journal of Nursing, 45(1), 31–32. (China), 5(18), 37–39.
Li, L., Wang, G., Zhong, L., Lin, C., Chen, Y., Lin, Y., . . . Wang, W. (2010). Yuan, Z., Zhu, L., & Zhang, Q. (2013). Clinical observation on the effect
Survey on bacterial contamination of clinical oxygen humidifier bot- of nonhumidified low flow oxygen administration by nasal cannula.
tles in hospital. Chinese Journal of Disinfection, 27(4), 437–438. Today Nurse, 12(24), 125–126.
Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche, P. C., Ioanni- Yue, X., He, J., & Chang, Y. (2015). Clinical effect of conventional wet
dis, J. P., . . . Moher, D. (2009). The PRISMA statement for reporting low -flow oxygen and dried low flow oxygen. Nursing Journal of Chi-
systematic reviews and meta-analyses of studies that evaluate health nese People’s Liberation Army, 32(5), 25–28.
care interventions: Explanation and elaboration. PLoS Medicine, 6(7), Zeng, X., Wang, R., Guo, N., & Huang, L. (2009). Study of two oxygen
e1000100. method on the dry nose feeling in patients undergone low flow oxy-
Liu, Y., & Chen, C. (2015). Clinical application of dry oxygen in the low- gen therapy. Chinese General Nursing, 7(6), 1513–1514.
flow nasal cannula oxygen administration in dialysis patients. Chinese Zheng, L., Yao, F., Sun, J., Shen, L., & Zhao, L. (2015). The effects of non-
Journal of Medical Device, 28(10), 97–98. humidified low flow nasal oxygen administration in cardiovascular
Lou, L., Zhu, L., Liu, Q., & Tian, Y. (2016). The effect of low-flow oxygen patient. Nursing Practice and Research, 12(6), 40–41.
without humidified oxygen humidifier in orthopedic surgery patients. Zhou, X., Zhou, R., Zhang, W., & Huang, L. (2015). The effect of dry oxy-
Journal of Clinical Medicine Practice, 20(10), 175–176. gen administration by nasal cannula on reducing lung infection. Jour-
Magnussen, H., Goeckenjan, G., Kohler, D., Matthys, H., Morr, H., Worth, nal of Qilu Nursing, 21(5), 90–91.
H., & Wuthe, H.; Deutsche Gesellschaft fur Pneumologie Wis-
senschaft Sektion: Klinische Pneumologie, F. (2001) [Guidelines to
long-term oxygen therapy]. Pneumologie, 55(10), 454–464.
Miyamoto, K. (2004). Is it necessary to humidify inhaled low-flow oxygen SUPPORTING INFORMATION
or low-concentration oxygen? Nihon Kokyuki Gakkai Zasshi, 42(2),
138–144. Additional Supporting Information may be found online in the
Nath, P., Ponnusamy, V., Willis, K., Bissett, L., & Clarke, P. (2010). Current
supporting information tab for this article.
practices of high and low flow oxygen therapy and humidification in
UK neonatal units. Pediatrics International, 52(6), 893–894.
Ning, X., Ouyang, R., Jiang, H., & Yu, Y. (2014). Application of non-humi-
dified low flow nasal oxygen inhalation in ICU patients. Modern Clini-
How to cite this article: Wen Z, Wang W, Zhang H, Wu C,
cal Nursing, 13(8), 34–36.
Ding J, Shen M. Is humidified better than non-humidified
O’Driscoll, B. R., Howard, L. S., & Davison, A. G.; British Thoracic Society.
(2008). BTS guideline for emergency oxygen use in adult patients. low-flow oxygen therapy? A systematic review and meta-
Thorax, 63(Suppl 6), vi1–vi68. analysis. J Adv Nurs. 2017;73:2522–2533. https://doi.org/
O’Reilly, P., & Bailey, W. (2007). Long-term continuous oxygen treatment in 10.1111/jan.13323
chronic obstructive pulmonary disease: Proper use, benefits and unre-
solved issues. Current Opinion in Pulmonary Medicine, 13(2), 120–124.
WEN ET AL. | 2533

The Journal of Advanced Nursing (JAN) is an international, peer-reviewed, scientific journal. JAN contributes to the advancement of evidence-
based nursing, midwifery and health care by disseminating high quality research and scholarship of contemporary relevance and with potential to
advance knowledge for practice, education, management or policy. JAN publishes research reviews, original research reports and methodological
and theoretical papers.

For further information, please visit JAN on the Wiley Online Library website: www.wileyonlinelibrary.com/journal/jan

Reasons to publish your work in JAN:

• High-impact forum: the world’s most cited nursing journal, with an Impact Factor of 1·998 – ranked 12/114 in the 2016 ISI Journal Citation
Reports © (Nursing (Social Science)).
• Most read nursing journal in the world: over 3 million articles downloaded online per year and accessible in over 10,000 libraries worldwide
(including over 3,500 in developing countries with free or low cost access).
• Fast and easy online submission: online submission at http://mc.manuscriptcentral.com/jan.
• Positive publishing experience: rapid double-blind peer review with constructive feedback.
• Rapid online publication in five weeks: average time from final manuscript arriving in production to online publication.
• Online Open: the option to pay to make your article freely and openly accessible to non-subscribers upon publication on Wiley Online Library,
as well as the option to deposit the article in your own or your funding agency’s preferred archive (e.g. PubMed).