Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Heated humidification versus heat and moisture exchangers for ventilated adults and children

Authors' declarations of interest

Version published: 26 January 2004 Version history

https://doi.org/10.1002/14651858.CD004711

This is not the most recent version

view the current version 14 September 2017
Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

The objective of this review is to determine whether HHs or HMEs are the more effective in preventing complications in mechanically ventilated people.
If adequate data relating to age, types of artificial airway and duration of ventilation is available subgroup analyses will be conducted on these groups.

Background

When a person has an endotracheal tube (ETT) or tracheostomy inserted the normal humidifying function of the upper airway is bypassed. It is necessary, therefore, to provide humidification, since inhaling cold, dry gas leads to complications. A number of studies have highlighted the adverse effects of inadequate humidity on respiratory tract structure and function (Burton 1962; Mercke 1975; VanOostdam 1986), including inflammation, necrosis and squamous metaplasia of the airways (Hirsch 1975;Todd 1992). Ventilation with dry gases also causes deterioration in lung function (Fonkalsrud 1975; Tsuda 1977) which can lead to lung collapse. In addition, an increase in tenacity of sputum could lead to ETT occlusion (Oliver 1962; Marfatia 1975), a potentially life threatening event for any intubated patient. In the paediatric and neonatal populations, lack of humidification may have even greater ramifications. Water vapour and heat loss in the respiratory tract (which strongly influencefluid and heat balance) is more pronounced in infants than in adults(Racz 1971).

There are two main forms of humidification used at present, i.e. active or passive.Gases may be actively warmed and moistened by use of a heated humidifier (HH). This involves passing air over the surface of a heated water reservoir attached to the ventilator. The system may have a heated wire in the circuit to prevent the warmed air cooling and the moisture condensing as it moves from the reservoir to the patient (Carroll 1997). Alternatively, inhaled gases may be passively humidified with a heat and moisture exchanger (HME).Thisdevice conserves expired heat and moisture and returns them to the patient. In order to enhance the moisture conserving performance of the HME, some have a hygroscopic salt (calcium or lithium chloride) added which absorbs water vapour during expiration and releases it during inspiration. This type of HME is known as a hygroscopic condenser humidifier (HCH) (Wilkes 1998). HMEs and HCHs may have bacteria‐filtering properties in addition to the humidification element (HMEF or HCHF) (Kollef 1998).

Both types of humidification can lead to complications. Heated humidification may result in over‐hydration, which can cause lung damage (Tamer 1970;Noguchi 1973;Klein 1974; John 1980;Todd 1989). It can also result in bacterial contamination (Craven 1984) and increased inspiratory workload (Misset 1991). With HMEs,there is the potential for an increased incidence of tenacious sputum and, therefore, ETT occlusion (Cohen 1988; Martin 1990). These devices also increase ventilator circuit dead space that may lead to an increase in minute ventilation, carbon dioxide retention and increased work of breathing in patients on pressure support ventilation (Pelosi 1996; Iotti 1997).

Although humidification for mechanically ventilated patients is widely accepted as an essential practice (Kollef 1998), there is much debate over what constitutes an optimal level of humidity, and how this can best be achieved (Williams 1996). There is variation in practice. The American Association for Respiratory Care produced a clinical practice guideline on humidification during mechanical ventilation in 1992. This lists a number of contraindications to the use of HMEs, which include patients with thick, copious or bloody secretions and those with an expired tidal volume less than 70% of the delivered tidal volume (AARC 1992). According to Hess (Hess 2000), HMEs are underused in the US despite there being sufficient evidence to support their use for up to three days. In 1999, Edwards (Edwards 1999) stated the lack of review, in the paediatric population, of appropriate humidification techniques. A review by Cook and colleagues (Cook 1998) evaluated the influence of airway management on ventilator‐associated pneumonia (VAP).Five trials in this review compared HHs with HMEs. One of the conclusions was that HMEswere less likely to contribute to VAP. A survey, comparing use of HHs and HMEs, which was carried out in adult intensive care units across France and Canada revealed that HMEs were used more often in France whilst Canadian units were more inclined to use HHs (Ricard 2002).

Inadequate humidification of inhaled gases can cause considerable lung damage. We feel that there is a lack of strong evidence as to which method of humidification, active or passive, is more effective in preventing complications in mechanically ventilated people. This review, therefore,will attempt to systematically review all controlled trials that examine the use of HH versus HME.

Objectives

The objective of this review is to determine whether HHs or HMEs are the more effective in preventing complications in mechanically ventilated people.
If adequate data relating to age, types of artificial airway and duration of ventilation is available subgroup analyses will be conducted on these groups.

Methods

Criteria for considering studies for this review

Types of studies

The review will include randomized controlled trials (RCTs) comparing two or more types of humidification in mechanically ventilated adults and children.

Types of participants

Inclusion Criteria

Children (0 to 16 years) and adults (over 16 years) who are receiving invasive mechanical ventilation in any setting.

Exclusion Criteria

None

Types of interventions

Comparison of HH and HME
Any model of HH will be included
Any type of HME (eg hygroscopic; hydrophobic; combination) will be included

Types of outcome measures

Primary outcomes

Artificial airway occlusion

Secondary outcomes

Mortality (all cause and related to respiratory events)
Respiratory complications, as defined by study authors,
including: respiratory infection or
ventilator associated pneumonia, or hypoxaemia, or
increased PaCO2 or
aspiration from any cause,
‐ aspiration due to condensate in ventilator circuit
Work of breathing
Secretion clearance or inspissated mucus
Change in body temperature
Length of stay: intensive care unit
hospital
Supplemental humidification with nebulized or directly instilled saline
Cost of devices
Quality of Life measures

Search methods for identification of studies

We will search the Cochrane Anaesthesia Group's trials register; the current issue of the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (1966 to date), EMBASE (1980 to date) and CINAHL (1982 to date) for trials.
We will search the Cochrane Database of Systematic Reviews (CDSR) and the Database of Abstracts of Reviews of Effectiveness (DARE) to identify any relevant systematic reviews.

The MEDLINE MeSH terms and search strategy are included below:

#1 ARTIFICIAL RESPIRATION/ all subheadings
#2 ventilat$
#3 artificia$ near3 (respir$ or airway$)
#4 TRACHEOSTOMY/ all subheadings
#5 tracheostomy
#6 #1 or #2 or #3 or #4 or# 5
#7 HUMIDITY/ all subheadings
#8 HEAT/ therapeutic use
#9 humidifi$
#10 (heat and moisture and exchanger$)
#11 hme
#12 (artificial or swedish) near3 nose
#13 #7 or #8 or #9 or #10 or #11 or #12
#14 #6 and #13

Methodology Terms

Modified version of the first two phases of the Cochrane RCT search strategy.
We will combine the population terms and intervention terms as outlined above with the methodology terms. Where duplicate publication occurs, we will use the publication with the most data for the review.

We will make all reasonable efforts to contact recognized experts in the field, to obtain any unpublished data. We will also utilize professional networks to obtain any unpublished data.
We will check reference lists of relevant papers for potentially relevant studies. Non‐English language articles will be eligible for selection to reduce the risk of publication bias and we will make efforts to have them translated.

Data collection and analysis

Selection of Studies

The above search strategy will identify a set of potentially relevant citations. Two reviewers (MK and, CL or DT)will independently examine each retrieved citation, and those thought to fulfil the selection criteria will be retrieved in full. Where a judgement cannot be made based on the citation, or when consensus cannot be reached, we will obtain the full article. Two reviewers (MK and, CL or DT) will then compare each article against the selection criteria independently, to determine which ones will be selected for data extraction. If differences exist they will be resolved either by consensus or by referral to another member of the team. We will devise a standard data extraction form and will pilot the form on a mixed set of articles. Two reviewers (MK and DT) or (DG and CL) will extract the data from each study independently and each pair will meet to compare the data. If differences exist they will be resolved either by consensus or by referral to another member of the team. If data is missing or further information is required, we will make reasonable attempts to contact the authors to obtain the required information. We will keep records of all articles (identified from the search strategy, excluded and included).

Data analysis

We will present event rates for binary outcomes as relative risk (RR) and will calculate the 95% confidence interval (CI).

For continuous outcomes we will calculate a weighted mean difference (WMD) and 95% confidence interval (CI). We will use fixed effects models to combine data if there is negligible statistical heterogeneity. In case of non negligible statistical heterogeneity we will employ random effects models.

Heterogeneity

We will explore potential for clinical heterogeneity. If clinical heterogeneity appears to be negligible, we intend to combine the effects. If clinical heterogeneity is not negligible, data will be tabulated and no meta‐analysis will be performed. With regards to statistical heterogeneity, we will use I2 statistics to test for heterogeneity.

Addressing publication bias

Where possible, we will enter data from all included studies into a funnel graph (trial effect against standard error) in an attempt to investigate the likelihood of overt publication bias (Egger 1997).

Sub‐groups analyses

If adequate data relating to age, duration of ventilation and types of artificial airway is available, sub‐group analyses will be conducted on these groups.

  • A‐priori sub‐group analysis by age is planned for the following age categories

    • preterm (birth to 40 weeks post‐conceptual age) and neonatal infants (birth to 28 days);

    • infants and children (28 days to 16 years);

    • adults (over 16 years)

  • A‐priori sub‐group analysis by duration of mechanical ventilation is planned for the following categories

    • short term ventilation (defined as less than six hours);

    • medium term ventilation (defined as greater than six hours but less than 48 hours);

    • long term ventilation (defined as greater than 48 hours)

  • Further, a‐priori sub‐group analysis is planned for ETT versus tracheostomy.

Sensitivity Analyses

We will perform a sensitivity analysis on general study quality. Analysis will be undertaken on the basis of adequate allocation concealment vs. not adequate concealment.

Methodological Quality of Studies

The quality of the studies to be included will be assessed independently by two reviewers (MK and DT) or (DG and CL), without blinding to authorship or journal of publication. Differences in the reviewers' allocation of studies into quality categories will be resolved either by consensus or by referral to another member of the team.

The quality of each trial will be assessed according to the following criteria by each reviewer working independently :

  1. method of randomization;

  2. extent of allocation concealment;

  3. clear inclusion and exclusion criteria;

  4. baseline comparability of treatment groups for important variables;

  5. use of intention to treat analysis, where appropriate, (participants analysed according to the group to which they were initially allocated, regardless of whether or not they withdrew, fully adhered to treatment, or crossed over and received the other treatment);

  6. extent of loss to follow up;

  7. blinded outcome assessment;

If differences are identified they will be resolved either by consensus or by referral to another member of the team.

Data extracted will include:

  1. country and setting where study was performed;

  2. inclusion and exclusion criteria;

  3. details of intervention;

  4. outcomes measured;

  5. duration of study;

  6. numbers enrolled and completing in each group;

  7. baseline characteristics of each group;

  8. results per group;

  9. mechanical ventilation variables such as cuffed or uncuffed ETT, FiO2 etc.