Advising patients to increase fluid intake for treating acute respiratory infections
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To answer the questions:
(1) Does recommending increased fluid intake as a treatment for acute respiratory infections improve duration and severity of symptoms?
(2) Are there adverse effects from recommending increased fluids in patients with acute respiratory infections?
(3) Are any benefits or harm related to site (upper or lower respiratory tract), or different severity of illness?
Background
Acute respiratory infections form a large proportion of disease seen in primary care settings. Some studies estimate this as being up to 15% of primary care consultations (Fry 1993). Advice to increase fluid intake is a common treatment recommendation (Murtagh 1996; Evans 1998; Rosser 1998).
This advice is often non specific in terms of quantity of fluid recommended, but the usual implication is to drink more than normal. (There is debate over what is a normal healthy fluid intake) (Valtin 2002). The type of fluid is not usually specified, but is usually confined to oral fluids normally consumed by the patient. Sometimes specific fluids are recommended, such as fruit juice, soup, lemonade and tea (Kirkpatrick 1998; Schmitt 1999).
Benefits from fluids are attributed to: replacing increased insensible fluid loss from fever and from respiratory tract evaporation with tachypnoea (Shann 1985; Dhawan 1992); correcting dehydration from reduced intake (Gerigk 1996); reducing the viscosity of mucus (Middleton 1991; Rosser 1998); loosening nasal mucus (Saketkhoo 1978) and moistening the respiratory tract to maintain comfort (Middleton 1991; Evans 1998).
However, there are theoretical reasons for increased fluid intake to cause harm. Anti‐diuretic hormone (ADH) secretion is increased in lower respiratory tract infections of various aetiologies. Excessive ADH secretion has been reported in bronchitis, bronchiolitis and pneumonia. (Heim 1982; Dreyfuss 1988; Gozal 1990) The mechanism of increased ADH secretion might be due to a resetting of osmostat receptors (Dreyfuss 1988; Hill 1990) or a response to the perception of hypovolaemia by intrathoracic receptors (Gozal 1990; Steensel‐Moll 1990). Giving increased fluids (or even normal maintenance) then might theoretically lead to hyponatraemia and fluid overload (Dhawan 1992).
Therefore, a systematic examination of the evidence is needed to determine the benefit versus harm from increasing fluid intake. We also want to determine whether acute respiratory infections, in subgroups of severity of illness and site of infection (upper and lower respiratory tract), differ in response to increased fluid recommendations.
Objectives
To answer the questions:
(1) Does recommending increased fluid intake as a treatment for acute respiratory infections improve duration and severity of symptoms?
(2) Are there adverse effects from recommending increased fluids in patients with acute respiratory infections?
(3) Are any benefits or harm related to site (upper or lower respiratory tract), or different severity of illness?
Methods
Criteria for considering studies for this review
Types of studies
All randomised controlled parallel group trials that examine the effect of increasing fluid intake in acute respiratory infections.
Types of participants
Patients of all ages with acute respiratory infection presenting for treatment in a primary care setting.
'Acute respiratory infection' will be subdivided into upper and lower respiratory tract infection and will include the following clinical entities, as defined by the international classification of health problems in primary care (WONCA 1983):
Upper respiratory tract infection: (URTI)
This will include the ICHPPC defined conditions: 'acute upper respiratory tract infection' (cold, nasopharyngitis, pharyngitis, rhinitis), 'acute sinusitis', 'acute tonsillitis', 'acute laryngitis and tracheitis'. There must be an absence of abnormal chest signs to define these conditions.
Lower respiratory tract infection: (LRTI)
This will include the ICHPPC defined conditions 'acute bronchitis, bronchiolitis' (which includes tracheobronchitis) and 'pneumonia'.
Influenza will be defined according to two ICHPPC categories. 'Influenza, without pneumonia' will be included as an URTI. Influenzal pneumonia is included in ICHPPC category 'pneumonia' and will be considered as a LRTI.
For the purposes of this systematic review, we will exclude otitis media.
Patients with underlying medical conditions are excluded, as their fluid requirements may differ from the normal population. Patients with central nervous system (CNS) infection are excluded as this alters their fluid management (Brown 1994). Patients with diarrhoea are excluded, as discussion of their fluid requirements has been covered in a previous systematic review (Hahn 2003).
Types of interventions
Treatment with, or recommendation for, increased oral fluid intake.
Types of outcome measures
Symptoms
‐ severity and duration, however measured in the studies.
‐ including but not restricted to: fever, mucus production, nasal congestion, sore throat, cough, headache.
Complications
‐ symptoms of dehydration (nausea, vomiting, postural dizziness)
‐ prolonged illness
‐ symptoms of water overload and hyponatraemia (behavioural disturbance, headache, confusion, convulsions, coma)
Health Service Utilization
‐ including ‐requirement for hospital admission
‐visits to primary care doctor
Search methods for identification of studies
See: Cochrane Acute Respiratory Infections Group search strategy.
Electronic searches of the following databases will be conducted ‐ the latest issue of the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (1966 to date), EMBASE (1974 to date), Current Contents (1966 to date) and Web of Science/ Science Citation Index will be searched.
The following search strategy will be used:
(exp Respiratory Tract Infections/ OR respiratory infection*OR upper respiratory tract infection* OR URTI)
AND (exp Fluid Therapy/ OR fluid therapy OR exp Water‐Electrolyte Balance/ OR water electrolyte balance OR fluid balance OR exp water/ OR exp drinking/ OR exp drinking behaviour/ OR drink* adj (fluid* or water) OR exp Infusions, Parenteral/ OR parenteral infusion* OR exp thirst/ OR thirst* OR exp water deprivation/ OR water intake OR fluid intake OR rehydration OR exp Rehydration Solutions/ OR rehydration solution* OR oral rehydration therapy OR (give fluid*) OR (give NEAR fluid*)
To identify randomized controlled trials, this search strategy will be combined with the Cochrane Highly Sensitive Search Strategy phases one and two as published in appendix 5c of the Cochrane Reviewers' Handbook (Clarke 2003). There will be no constraints based on language or publication status when searching for trials.
Data collection and analysis
Abstracts from the initial search will be read to identify studies that meet the inclusion criteria. Full text articles will be retrieved and reviewed to determine eligibility. These studies will be assessed independently by at least two reviewers. Differences of opinion among the reviewers will be resolved by discussion.
Data extraction
Data will be extracted from the studies independently by two reviewers using a standard form.
Differences in extraction will be compared and resolved by discussion.
Quality Assessment
Study quality assessment will be done using a modification of the method outlined in the Cochrane Reviewers' Handbook and that published in the literature (Chalmers 1990).
1. Method of treatment assignment
(3) adequate, blinded randomisation technique described
(2) randomised + double blind stated but method not described, or method suspect e.g. envelopes
(1) randomisation stated but method not described + investigator not blinded
(0) randomisation not mentioned
2. Control of selection bias after treatment assignment
(3) intention to treat analysis and full follow up
(2) intention to treat analysis and < 15% loss to follow up
(1) analysis by treatment received only or no mention of withdrawals
(0) analysis by treatment received and no mention of withdrawals or >15 % withdrawals/ loss to follow up/post randomisation exclusions.
3. Blinding
(3) Blinding of patient, care giver and investigator
(2) Blinding of investigator assessing outcome or (patient and care giver)
(1) Blinding impossible or impossible to judge if attempted
(0) Blinding not done when it could have been
4. Outcome assessment
(2) All patients had standardised assessment
(1) No standardised assessment or not mentioned
Data analysis
Data analysis will be on an intention to treat basis. Both fixed and random effects models will be used. However, only the random effects model will be used if significant heterogeneity is found. Continuous data will be analysed using weighted mean differences and 95% confidence intervals. Dichotomous data will be expressed as odds ratios with 95% confidence intervals.
If data of sufficient quality is obtained, subgroup analysis will be performed on the basis of:
‐ upper and lower respiratory tract infection (as already defined).
‐ age groupings: infants under two years old, children ‐ as defined in the studies; adults older than 18 years old.
‐ severity of illness: as measured in the studies, and also determined on the basis of health service utilisation i.e. requirement for admission
