Simple behavioural interventions for nocturnal enuresis in children
Abstract
Background
Nocturnal enuresis (bedwetting) is a socially disruptive and stressful condition which affects around 15% to 20% of five year olds and up to 2% of adults. Although there is a high rate of spontaneous remission, the social, emotional and psychological costs can be great. Behavioural interventions for treating bedwetting are defined as interventions that require a behaviour or action by the child which promotes night dryness and includes strategies which reward that behaviour. Behavioural interventions are further divided into:
(a) simple behavioural interventions ‐ behaviours or actions that can be achieved by the child without great effort; and
(b) complex behavioural interventions ‐ multiple behavioural interventions which require greater effort by the child and parents to achieve, including enuresis alarm therapy.
This review focuses on simple behavioural interventions.
Simple behavioural interventions are often used as a first attempt to improve nocturnal enuresis and include reward systems such as star charts given for dry nights, lifting or waking the children at night to urinate, retention control training to enlarge bladder capacity (bladder training) and fluid restriction. Other treatments such as medications, complementary and miscellaneous interventions such as acupuncture, complex behavioural interventions and enuresis alarm therapy are considered elsewhere.
Objectives
To determine the effects of simple behavioural interventions in children with nocturnal enuresis.
The following comparisons were made:
1. simple behavioural interventions versus no active treatment;
2. any single type of simple behavioural intervention versus another behavioural method (another simple behavioural intervention, enuresis alarm therapy or complex behavioural interventions);
3. simple behavioural interventions versus drug treatment alone (including placebo drugs) or drug treatment in combination with other interventions.
Search methods
We searched the Cochrane Incontinence Group Specialised Trials Register, which contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, MEDLINE in process, and handsearching of journals and conference proceedings (searched 15 December 2011). The reference lists of relevant articles were also searched.
Selection criteria
All randomised or quasi‐randomised trials of simple behavioural interventions for treating nocturnal enuresis in children up to the age of 16. Studies which included children with daytime urinary incontinence or children with organic conditions were also included in this review if the focus of the study was on nocturnal enuresis. Trials focused solely on daytime wetting and trials of adults with nocturnal enuresis were excluded.
Data collection and analysis
Two reviewers independently assessed the quality of the eligible trials and extracted data. Differences between reviewers were settled by discussion with a third reviewer.
Main results
Sixteen trials met the inclusion criteria, involving 1643 children of whom 865 received a simple behavioural intervention. Within each comparison, outcomes were mostly addressed by single trials, precluding meta‐analysis. The only exception was bladder training versus enuresis alarm therapy which included two studies and demonstrated that alarm therapy was superior to bladder training.
In single small trials, rewards, lifting and waking and bladder training were each associated with significantly fewer wet nights, higher full response rates and lower relapse rates compared to controls. Simple behavioural interventions appeared to be less effective when compared with other known effective interventions (such as enuresis alarm therapy and drug therapies with imipramine and amitriptyline). However, the effect was not sustained at follow‐up after completion of treatment for the drug therapies. Based on one small trial, cognitive therapy also appeared to be more effective than rewards. When one simple behavioural therapy was compared with another, there did not appear to be one therapy that was more effective than another.
Authors' conclusions
Simple behavioural methods may be superior to no active treatment but appear to be inferior to enuresis alarm therapy and some drug therapy (such as imipramine and amitriptyline). Simple behavioural therapies could be tried as first line treatment before considering enuresis alarm therapy or drug therapy, which may be more demanding and have adverse effects, although evidence supporting their efficacy is lacking.
Plain language summary
Simple treatments for bedwetting in children
Bedwetting (nocturnal enuresis) is the involuntary loss of urine at night without an underlying organic disease as the cause. It can result in issues of psychosocial well‐being such as social problems, sibling teasing and lowered self esteem. It affects around 15% to 20% of five year olds and up to 2% of adults.
Simple behavioural strategies to help children gain control include star charts and other reward systems, fluid restriction, bladder training (including retention control training) and lifting or wakening. These are often used as a first attempt to control bedwetting and can be undertaken by families with less professional involvement.
The review found 16 trials which involved 1643 children. Most simple behavioural treatments were only studied in single small trials which makes the evidence less reliable. Simple treatments such as rewarding dry nights (e.g. with star charts), lifting and waking and bladder training appeared to be more effective than no treatment but they are not as effective when compared with other treatments known to work, such as enuresis alarm therapy and drug therapy. There does not appear to be one simple behavioural therapy that is more effective than another. On the other hand, simple treatments do not have any side effects or safety concerns. Therefore, simple methods could be tried as first line therapy before considering alarms or drugs for this common childhood condition.
Authors' conclusions
Background
This is one of seven reviews of interventions for bedwetting or nocturnal enuresis. The others focus on: desmopressin (Glazener 2004a), tricyclics and related drugs (Glazener 2004b), other drugs (Deshpande 2012), alarms (Glazener 2004d), complex behavioural and educational interventions (Glazener 2004e) and complementary and miscellaneous therapy (Huang 2011). This review is an update of previous updates by Glazener (Glazener 2002; Glazener 2004f) based on the original review by Lister‐Sharp and her colleagues at the NHS Centre for Reviews and Dissemination at the University of York, UK (Lister‐Sharp 1997).
With the promotion by the International Children's Continence Society (ICCS) of using standardised nomenclature in paediatric continence research (Neveus 2006) and the improved understanding about the pathophysiology of nocturnal enuresis, there has been an emphasis on classifying enuresis as either monosymptomatic or non‐monosymptomatic. The ICCS has also defined initial and long term treatment success. Unfortunately, the older studies have usually not used these definitions (or were unclear about their definition of initial success). There were no studies that included a two‐year follow‐up; therefore, long term success could not be assessed.
Description of the condition
Nocturnal enuresis is the intermittent involuntary loss of urine at night, in the absence of physical disease, at an age when a child could reasonably be expected to be dry (by consensus, at a developmental age of five years) (APA 1980; Neveus 2006; WHO 1992). Nocturnal enuresis has a significant negative impact on the child's psychosocial well‐being (Hagglof 1998; Von Gontard 2011a). Children with nocturnal enuresis usually do not have daytime bladder symptoms and are deemed to have monosymptomatic nocturnal enuresis (Neveus 2006). Between 10% to 28% of children who wet during sleep at night also have bladder problems during the day (including daytime wetting) (Forsythe 1974).These children are said to have non‐monosymptomatic nocturnal enuresis (Neveus 2006). A total of 6.6% of children have combined daytime urinary incontinence and nocturnal enuresis (Sureshkumar 2009).
Although daytime urinary incontinence is a significant problem it is usually considered separately to nocturnal enuresis, as the aetiologies underlying the two conditions are thought to be different (Neveus 2006). Physical causes such as structural abnormalities and functional disorders of the urinary tract, including overactive bladder syndrome or urinary tract infections, are more often found in children who also have daytime wetting (Jarvelin 1990). If daytime symptoms are present, investigations to identify physical (organic) causes such as urinary tract dysfunction, congenital malformation and neurogenic disorders are usually necessary (Djurhuus 1992), with treatment initially focused on addressing the daytime urinary symptoms (Hjalmas 2004; Neveus 2006). Daytime urinary incontinence is considered in a separate review (Sureshkumar 2003).
Prevalences and causes
Nocturnal enuresis is common in children. Estimating its prevalence is difficult because there is variation in diagnostic criteria with very few studies using ICCS (International Children's Continence Society) definitions (De Jonge 1973; Krantz 1994; Neveus 2006). The generally quoted prevalence rates are 15% to 20% of five year olds, 7% of seven year olds, 5% of ten year olds, 2% to 3% of teenagers and 0.5% to 2% of adults, with a spontaneous remission rate of 14% until adolescence (Blackwell 1989; Bower 1996; Forsythe 1974; Hirasing 1997; Rutter 1973; Yeung 2004). The prevalence of nocturnal enuresis is higher amongst children in residential care (Morgan 1970).
Butler proposed the conceptual model that nocturnal enuresis occurs when there is poor arousal from sleep in response to a sensation of a full bladder with overactivity of bladder function or excessive overnight urine production, or both (Butler 2000). There are also many known risk factors for nocturnal enuresis including:
-
genetic influences (APA 1980; Bakwin 1971; Bakwin 1973; Eiberg 1995; Von Gontard 2001; Von Gontard 2011b);
-
delay in attaining bladder control (Jarvelin 1989; Koff 1995);
-
physiological factors (Djurhuus 1992; Norgaard 1993) such as constipation, faecal incontinence, daytime urinary incontinence, caffeine consumption (Blackwell 1989; Creighton 1990; Jura 2011; Sureshkumar 2009);
-
sleep apnoea and upper airway obstructive symptoms (Brooks 2003; Maizels 1993);
-
defective sleep arousal (Kawauchi 1998a; Yeung 2008);
-
psychological factors (Devlin 1991; Moffatt 1989; Rutter 1973; Shaffer 1977).
It is thought that these risk factors impact on sleep arousal, bladder function and overnight urine production.
Description of the intervention
Interventions used for treating nocturnal enuresis include pharmacological, behavioural (including alarm training), complementary and miscellaneous interventions. All of the following interventions are reviewed separately: alarms (Glazener 2004d), other complex behavioural and educational interventions (which require greater effort by the child and parents to achieve) such as dry bed training and full spectrum home training (Glazener 2004e), complementary and miscellaneous interventions such as acupuncture, homeopathy, chiropractic, spinal manipulation, functional magnetic stimulation, hypnosis and psychological therapies such as counselling and psychotherapy (Huang 2011), and pharmacological interventions such as desmopressin (Glazener 2004a), tricyclics and related drugs (Glazener 2004b) and drugs other than desmopressin and tricyclics (Glazener 2004c). Studies which compare more than one type of intervention are included in the reviews of each included intervention.
The current review is restricted to simple behavioural interventions which can be used by the child without great effort. These include fluid restriction, lifting, wakening, reward systems (e.g. star charts) and bladder training (e.g. retention control training). Other interventions were included only in this review if they were compared with simple interventions or used in combination with them.
How the intervention might work
1. Simple behavioural interventions
Behavioural interventions assume that the ability to remain dry at night is a learned response which can be achieved using psychological conditioning techniques if dryness has not arisen spontaneously. These include fluid restriction, lifting, wakening, reward systems (e.g. star charts) and bladder training (including retention control training).
Fluid restriction
Fluid restriction (deprivation) before bedtime is a method frequently used by parents (Shaffer 1977). Fluid restriction reduces the total overnight urine production which reduces the child's need to void overnight. However, fluid restriction (particularly during the day) may reduce the bladder volume (Sorotzkin 1984) and hence increase the chance of enuresis by reducing the ratio of bladder volume to overnight urine production. Even so, it might be useful to limit pre‐bedtime fluid intake, particularly drinks with diuretic properties (Novello 1987).
Lifting
In 'lifting', carers pick up the child, while still asleep, from the bed to allow them to urinate in an appropriate place, reducing the likelihood of urinating in the bed. Usually lifting is done without a password. It has been argued that this practice is counterproductive as the child is encouraged to urinate without waking and is denied the opportunity to learn to recognise the sensation of a full bladder which may trigger arousal (Butler 1994). On the other hand, some suggest that lifting is effective, precluding the need for professional help (Shaffer 1977).
Scheduled wakening
Scheduled wakening involves waking the child to urinate during the night (Warzak 1994), reducing the likelihood of urinating during sleep. A scheduled waking programme may be used with the child being woken progressively earlier after dry nights until the interval between going to bed and scheduled waking is one hour. Older individuals may use an alarm clock (Blackwell 1989), mobile phones or other forms of alarms to wake themselves. Lifting with a password is another form of waking where the child is woken 1.5 to 2 hours after falling asleep and asked a password to check whether they are awake before being taken to the toilet.
Star charts and other reward systems
Star charts and other reward systems are behavioural interventions which use positive reinforcement to encourage a desired behaviour. The child is rewarded for attaining a goal, such as remaining dry all night or getting up to go to the toilet. These schemes should be negotiated with the child and family. The aim is to positively reinforce dry nights and to reduce the negative emphasis on wet beds. These are often the first type of treatments proposed (Stewart 1975). However, rewards for dry nights may be problematic because children may feel a failure if they try but cannot attain the goal despite trying their best (Blackwell 1989).
Bladder training and retention control training
Retention control training (a form of bladder training) aims to increase the bladder capacity by using exercises such as delaying urination for extended periods of time during the day or drinking extra fluids (Warzak 1994). Stream interruption exercises (pelvic floor muscle training, sphincter control exercises) are other forms of bladder training that are sometimes tried (Novello 1987).
2. Other behavioural interventions (only included in this review if used in comparison arms)
Enuresis alarms
Enuresis alarms consist of an alarm system which is triggered by micturition (Glazener 2004d). The first enuresis alarms were bed‐based with the child sleeping on a pad or mat containing an electrical circuit. Urine coming into contact with the mat would cause an alarm to sound. The alarm is intended to train the child to recognise the sensation of having a full bladder and to either inhibit urination or waken to void (Forsythe 1989). There are now many different types of enuresis alarms, including body worn alarms with sensors placed in the child's underwear. Alarms are now either wired or wireless and stimuli range from noises (buzzer, various sounds, voice) to vibration and light.
Over‐learning with alarm training
Over‐learning is a reinforcement procedure which may be initiated after successful alarm treatment (e.g. after achievement of 14 consecutive dry nights). Extra drinks are given at bedtime to ensure that the bladder is full overnight causing the child to learn to wake to void. Alarm treatment is then continued until 14 consecutive dry nights are again achieved (Blackwell 1989; Young 1972).
Complex behavioural and educational interventions
Complex behavioural interventions include combinations of interventions and usually require greater effort by the child and parents or carers to achieve. These include Dry Bed Training and Full Spectrum Home Training (Glazener 2004e; Howe 1992).
Dry Bed Training
Dry Bed Training was initially developed in the early 1970s for use with people with learning disabilities (Azrin 1973). It aims to condition the child to wake to void. The original schedule involved an intensive training night, during which the person was woken every hour and taken to the toilet. If an accident occurred, 45 minutes of 'cleanliness training' (changing the bed) and 'positive practice' (practices of getting up and going to the toilet about nine times) was implemented. On subsequent nights, the individual was woken once and taken to the toilet, this nightly wakening occurring progressively earlier. Variants such as Modified Dry Bed Training forego the reprimands and positive practice elements (Butler 1988).
Full Spectrum Home Training
Full Spectrum Home Training (Houts 1983) combines enuresis alarm training, cleanliness training, retention control training and over‐learning procedures.
3. Other managements for enuresis (only included in this review if used in comparison arms)
Psychotherapy
Since psychotherapists consider enuresis to be a symptom rather than a condition in itself, the emphasis of treatment is on the "child's inner emotional disturbance" (Mishne 1993). Consequently, the intervention is aimed at addressing underlying psychological causative factors and modifying the environment which produced the symptom (Mishne 1993). Psychotherapy may be used in the management of children who have psychological problems in addition to enuresis, to address problems directly related to psychopathology (Warzak 1994).
4. Drugs
Pharmacological interventions are used to reduce nocturnal enuresis by reducing overnight urine production (e.g. desmopressin), reducing bladder overactivity (e.g. oxybutynin) or by other mechanisms. Placebo drugs have been used to compare with other interventions to test their effectiveness. Drugs that have been compared with simple behavioural interventions include desmopressin (Glazener 2004a), tricyclic drugs (amitriptyline or imipramine) (Glazener 2004b), drugs other than desmopressin and tricyclics (e.g. oxybutynin) (Glazener 2004c) and placebo.
Why it is important to do this review
Nocturnal enuresis is common in children. Although nocturnal enuresis in itself is pathologically benign, it is socially and emotionally stigmatising and can affect the self esteem, peer relationships and educational opportunities of sufferers. It can be inconvenient and distressing to both the child and their family (Fitzwater 1992; Morison 1998; Redsell 2001; Theunis 2002; Warzak 1993). Although many different interventions have been tried for treating nocturnal enuresis, the relative effectiveness of each remains uncertain. Simple behavioural interventions are often the first interventions tried by parents or carers at home with minimal professional involvement. Therefore, it is important to identify effective interventions because successful treatment of nocturnal enuresis can result in improvement in health, self esteem and quality of life (Hagglof 1998).
Objectives
To determine the effects of simple behavioural interventions in children with nocturnal enuresis.
The following comparisons were made:
1. simple behavioural interventions versus no active treatment;
2. any single type of simple behavioural intervention versus another behavioural method (another simple behavioural intervention, enuresis alarm therapy or complex behavioural interventions);
3. simple behavioural interventions versus drug treatment alone (including placebo drugs) or drug treatment in combination with other interventions.
Methods
Criteria for considering studies for this review
Types of studies
Randomised or quasi‐randomised trials of simple behavioural interventions compared with an appropriate comparison group for the treatment of nocturnal enuresis in children.
Types of participants
Children (as defined by the trialists, usually up to age 16) with nocturnal enuresis.
Types of interventions
Any trial which used a simple behavioural intervention in at least one arm of the study for treating nocturnal enuresis.
Comparisons were made with no active treatment, other behavioural interventions and drugs, either alone or in combination.
Types of outcome measures
Primary outcomes
-
mean number of wet nights per week at the end of treatment;
-
number of children failing to attain 14 consecutive dry nights by the end of treatment.
Secondary outcomes
-
mean number of wet nights per week at follow‐up after treatment was stopped;
-
number of children failing to attain 14 consecutive dry nights during treatment or relapsing at follow‐up after treatment was stopped, or both;
-
adverse events.
Search methods for identification of studies
We did not impose any language or other limits on the searches.
Electronic searches
This review has drawn on the search strategy developed for the Incontinence Review Group. Relevant trials were identified from the Group's Specialised Register of controlled trials which is described, along with the group search strategy, under the Incontinence Group's module in The Cochrane Library. The register contains trials identified from MEDLINE, CINAHL, the Cochrane Central Register of Controlled Trials (CENTRAL) and handsearching of journals and conference proceedings. The Incontinence Group Trials Register was searched using the Group's own keyword system. The search terms used were:
({design.cct*} OR {design.rct*})
AND
({topic.urine.enuresis*})
(All searches were of the keyword field of Reference Manager 12, Thomson Reuters).
Date of the most recent search of the register for this review: 15 December 2011.
Most of the trials in the Incontinence Group Specialised Register are also contained in CENTRAL.
Searching other resources
The reference list of a previous systematic review of enuresis treatments was searched (Lister‐Sharp 1997). The reference lists of other relevant articles were also searched.
Data collection and analysis
The studies for this review were assessed using the methods described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
Selection of studies
The titles and where possible abstracts of all studies located by the searches were checked (PC,GN) to identify those likely to evaluate the effects of simple behavioural interventions for nocturnal enuresis. Full papers were then obtained and assessed to identify those which met the inclusion criteria. Studies were excluded if they were not randomised or quasi‐randomised trials for children with nocturnal enuresis. Excluded studies are listed with reasons for their exclusion.
Data extraction and management
The data were extracted using a standard form. Included trial data were processed as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Where appropriate, the results were converted to the mean and standard deviation of the number of wet nights per week, or the number of children failing to achieve full response during treatment, defined as 14 consecutive dry nights at the end of the treatment period, or the number who did not achieve full response during treatment plus those who relapsed after stopping active treatment at follow‐up (after completion of treatment). Where a mean value was reported with no standard deviation, we entered the data into 'Other Data Tables'.
Assessment of risk of bias in included studies
Standard risk of bias items according to the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) were assessed for each study. These include:
-
adequate sequence generation;
-
allocation concealment;
-
blinding of study participants;
-
blinding of study personnel;
-
blinding of outcome assessors;
-
incomplete outcome data addressed;
-
free of selective reporting;
-
free of other bias (e.g. whether children who had daytime wetting or who had physical (organic) causes of their enuresis were specifically excluded, whether there were baseline differences between groups or whether there was a washout period in cross‐over trials).
Measures of treatment effect
We intended, where possible, to calculate standardised effect sizes and 95% confidence intervals (CI): weighted mean differences (WMD) for pooled estimates or mean difference (MD) for comparing treatment effects where outcomes were continuous variables and risk ratios (RR) where they were dichotomous. A fixed‐effect model was to be used for calculation of summary estimates and 95% CI. However, we could not perform pooled estimates due to insufficient trials for each intervention.
Unit of analysis issues
The unit of allocation was individual patients. Cross‐over trials were only included if the design was evaluated to be appropriate for the question being investigated. Otherwise, data from the first arm of cross‐over trials was to be used like parallel group data.
Dealing with missing data
When intention‐to‐treat analysis was not mentioned in the paper, or if some participants were not analysed in the group to which they were randomised, we looked for sufficient information to restore them to the correct group. If participants could not be analysed in their allocated groups, we documented it in the 'Risk of bias' table and in the 'Risk of bias' section of the text.
Where there were missing outcome data in the included studies, the primary analysis used the number of participants with available data as the denominator (that is 'available case' analysis). We also attempted to contact the authors regarding missing data. Standard deviations were obtained from standard errors, confidence intervals, t values or P values that related to the differences between means in two groups, according to the Cochrane Handbook for Systematic Reviews of Interventions Section 7.7.3.3
Assessment of heterogeneity
Differences between trials were further investigated if statistically significant heterogeneity was found at the 10% significance level using the Chi2 test or assessment of the I2 statistic to quantify the degree of heterogeneity (Higgins 2011). We also applied visual inspection of the results. A fixed‐effect model was used if there was little or no heterogeneity and a random‐effects model was used when there was statistically significant heterogeneity.
Assessment of reporting biases
We planned to investigate the possibility of publication bias and related biases if any meta‐analyses were performed that involved 10 or more studies. However, this did not occur due to insufficient studies for meta‐analyses.
The methods that would be used for this are described below:
-
if there are 10 or more studies in an analysis, a funnel plot is produced;
-
we would perform a visual assessment of funnel plot asymmetry.
Statistical tests for funnel plot asymmetry are: for continuous outcomes using mean differences, the test proposed by Egger (Egger 1997); and for dichotomous outcomes using odds ratios or risk ratios, the tests proposed by either Harbord et al (Harbord 2006) or Peters et al (Peters 2006).
If evidence for funnel plot asymmetry was found, exploratory analyses would then be performed to investigate the possible causes (see Handbook section 10.4.4). These could include comparison of fixed‐effect and random‐effects estimates, comparison of treatment effects of small and large studies, or studies at low‐ and high‐risk of bias, or investigation of the effects of methods to correct for publication bias.
Data synthesis
Meta‐analysis would have been performed according to the recommendations of theCochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) had there been a sufficient number of trials for each comparison. We planned to use a fixed‐effect model. If there was substantial heterogeneity, assuming treatment effects differed between trials, we planned to use the random‐effects model to produce an overall summary estimate.
Subgroup analysis and investigation of heterogeneity
If there was heterogeneity, we planned to investigate it using subgroup analysis restricted to primary outcomes and sensitivity analysis. However, subgroup analysis was unable to be performed in this review due to the limited number of trials. If there was evidence of statistical or clinical heterogeneity or other risk of bias, data could be analysed using random‐effects and fixed‐effect models.
Sensitivity analysis
Where appropriate, the possible effects of missing data could have been explored in a sensitivity analysis (e.g. assuming the 'worst' and 'best' case scenarios for all participants with missing outcome data).
Results
Description of studies
Results of the search
The flow of literature through the assessment process is shown in the PRISMA flowchart (Figure 1). The literature search produced a total of 451 abstracts and titles which were screened.Of these, 62 full text articles were assessed which identified 20 reports of 16 studies of simple behavioural interventions for treating nocturnal enuresis which were included in this review.
PRISMA study flow diagram
The current review is an update of a previously published review update (Glazener 2004f). There were three new trials (five reports of three studies) (Bryant 2003; Van Dommelen 2009; Van Hoeck 2007) included in this update. Two other studies are awaiting further information from authors (Rodriguez 1984; Unuvar 2005).
Included studies
Most interventions were addressed by either single trials, single arms in multi‐intervention trials or had data for a particular outcome in only a single trial. Some trials combined two simple behavioural interventions (waking or lifting and reward (Baker 1969; Fava 1981)) or simple behavioural interventions with placebo in one arm (fluid restriction, avoiding punishment and placebo (Bhatia 1990); waking and placebo (Fournier 1987); waking, reward and placebo (Mehrotra 1980)). Two studies used a combination of bladder training with enuresis alarm (Fielding 1980) or bladder volume alarm (Pretlow 1999) and compared them with enuresis alarm monotherapy. These studies were included as they assessed the additive effects of the bladder training to enuresis alarm training.
There were only a limited number of interventions where it was possible to combine the data (Bennett 1985; Van Hoeck 2007 for bladder training vs alarm and Ronen 1995; Van Dommelen 2009 for rewards vs control). However, due to the heterogeneity of the rewards and control interventions in the latter studies, these results need to be viewed with caution.
Simple behavioural interventions
These included:
-
waking or lifting (with or without password) (Baker 1969; El‐Anany 1999; Fournier 1987; Mehrotra 1980; Turner 1970; Van Dommelen 2009);
-
bladder training (Bennett 1985; Bryant 2003; Fielding 1980; Hamano 2000; Harris 1977; Pretlow 1999; Van Hoeck 2007);
-
fluid restriction and avoidance of punishment (Bhatia 1990);
-
star charts or other reward systems (Baker 1969; Fava 1981; Mehrotra 1980; Ronen 1995; Van Dommelen 2009).
Active comparison interventions
These included:
-
enuresis alarms (Baker 1969; Bennett 1985; Bryant 2003; Fielding 1980; Fournier 1987; Ronen 1995; Turner 1970; Van Hoeck 2007);
-
alarm triggered by ultrasound measurement of bladder volume (Pretlow 1999);
-
dry bed training (Bennett 1985);
-
cognitive therapy (Ronen 1995).
-
desmopressin (Hamano 2000);
-
the tricyclic drugs imipramine (Bhatia 1990; Fournier 1987;) and amitriptyline (Mehrotra 1980).
'Untreated' control groups
These were allocated to:
-
a waiting list (Baker 1969; Harris 1977; Ronen 1995) although one group were given star charts if they became dry while they were on the waiting list (Bennett 1985);
-
play therapy (Fava 1981);
-
a diary (Van Dommelen 2009);
-
placebo drug treatment (Bhatia 1990; Fournier 1987; Mehrotra 1980; Turner 1970; Van Hoeck 2007).
Excluded studies
Forty reports of 39 studies were excluded from this review. Twenty‐three studies did not include a simple behavioural intervention in one arm (Bak 2007; But 2006; Chertin 2007; Evans 2007; Jiang 1996; Jodorkovsky 2003; Li 1996; Li 2002; Ma 2007; Naitoh 2005; Ozden 2008; Reed 1994; Renjing 2004; Seabrook 2005; Tuygun 2007; Van Hoeck 2008; Van Kampen 2009; Vasconcelos 2006; Wang 1994; Wang 1999; Wang 2007; Xiao 1997; Yamanishi 1988); 13 were not randomised controlled trials (Bouchard 1981; Creer 1975; Dong 1998; Hagglund 1965; Jiang 1996; Kawauchi 1998b; Lebedev 1995; Li 1996; Paschalis 1972; Terakye 1997; Vasconcelos 2007; Wang 1994; Wang 1999) of which four also did not include a simple behaviour intervention; one was not about nocturnal enuresis (Kroll 2006); one focused on voiding dysfunction in children with no outcomes for nocturnal enuresis given (Ayan 2007) and two were conducted amongst adults with special needs (Barker 1979; Hanson 1988). One study was not able to be completed (Cohen 2006) and serious methodological flaws were identified in two other studies. In one, the groups were not comparable at baseline and children were switched between groups (Doleys 1977). In the second, recruitment to the random‐waking arm stopped during the trial (because of lack of success) and in the other arms of the trial children were switched between treatments according to initial response (McConaghy 1969). Therefore, these two studies were also excluded as the data were unreliable. All excluded studies are listed with reasons for exclusion in the Characteristics of excluded studies table.
Risk of bias in included studies
Most included trials either had high risks of bias or had insufficient information to judge the level of risk of bias. Trial sizes were generally small, with only one large trial with a sample size of 570 (Van Dommelen 2009): all the other trials had fewer than 150 participants and 10 trials had fewer than 100 participants (Figure 2; Figure 3).
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Random sequence generation (selection bias)
Of the 16 included trials, only one used adequate sequence generation (Bryant 2003). Twelve did not provide details of the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups. Three studies were at high risk of bias due to quasi‐randomised methods of sequence generation (alternation allocation (Pretlow 1999); assigned in chronological order (Ronen 1995); and sequence generated by rule on computer based on gender (Van Dommelen 2009)). Therefore, these had high risk of bias (Figure 2; Figure 3).
Allocation concealment (selection bias)
Of the 16 included studies, there was only one study that had adequate allocation concealment (Bryant 2003). Twelve did not describe the method used to conceal the allocation sequence in sufficient detail to determine whether the intervention allocations could have been foreseen in advance of, or during, enrolment. In three trials there was high risk of selection bias due to inadequate allocation concealment: in one trial, although allocation was by means of stratification based on age and sex, children who dropped out were replaced by the next child referred to the clinic (Turner 1970); one study was quasi‐randomised (alternation allocation) (Pretlow 1999) and one study assigned interventions in chronological order (Ronen 1995) (Figure 2; Figure 3).
Blinding
Blinding of participants and outcome assessors (who were usually the parents) was not possible for most of the simple behavioural interventions, although in studies involving drugs as a comparison intervention some blinding was achieved for the drug arms of the trial with the use of placebo drugs (Fournier 1987; Turner 1970; Van Hoeck 2007). There was only one study that had adequate blinding of study personnel (Fournier 1987). There was no blinding of the study personnel in four of the studies (Bennett 1985; Bhatia 1990; Bryant 2003; Ronen 1995) and in the other 11 studies, blinding of the study personnel was not mentioned (Figure 2; Figure 3).
Incomplete outcome data
Most studies (12) had high risk of attrition bias: two studies replaced drop‐outs with new study participants (Bennett 1985; Turner 1970). In one study the total numbers of participants did not add up (El‐Anany 1999) and the remaining nine studies had high attrition rates (which were often differential between interventions) and incomplete reporting of outcome data. In one study, the total number of participants at follow‐up was not mentioned (Harris 1977); therefore, we were unable to determine whether there was any attrition bias. Only three studies had a low risk of attrition bias as they reported the outcomes of all participants randomised (Fava 1981; Mehrotra 1980; Pretlow 1999) (Figure 2; Figure 3).
Five studies used intention‐to‐treat analysis (El‐Anany 1999; Fava 1981; Mehrotra 1980; Pretlow 1999; Van Hoeck 2007) and one study used an imputation method to deal with the high loss to follow up (Van Dommelen 2009). There was reporting bias in the remaining 10 studies where participants who dropped out were excluded in outcome analysis (Figure 2; Figure 3).
Selective reporting
Most studies did not state prespecified primary outcomes. In six trials there were high risks of reporting bias because of incomplete reporting of important and expected outcomes of interest (Bryant 2003; Fournier 1987; Hamano 2000; Ronen 1995; Turner 1970; Van Dommelen 2009). There was only one study where all prespecified and expected outcomes were reported (Mehrotra 1980).
Other potential sources of bias
As organic causes of wetting and the presence of daytime wetting can impact on treatment outcomes for nocturnal enuresis, we assessed whether these two characteristics were specifically excluded in included trials. Thirteen of the trials in the review specifically excluded children with organic causes for their bedwetting (Baker 1969; Bhatia 1990; Bryant 2003; El‐Anany 1999; Fielding 1980; Fournier 1987; Hamano 2000; Harris 1977; Pretlow 1999; Ronen 1995; Turner 1970; Van Dommelen 2009; Van Hoeck 2007). The remaining three trials did not mention whether organic causes were excluded. In addition, six trials specifically excluded children with daytime wetting (Bennett 1985; Hamano 2000; Harris 1977; Pretlow 1999; Van Dommelen 2009; Van Hoeck 2007). One further trial included children with day and night wetting in a parallel study but data from these participants were excluded from this review (Fielding 1980). Children with daytime wetting were included in two trials (Bhatia 1990; Bryant 2003) and the other seven trials did not mention whether daytime wetting was excluded.
Baseline wetting was measured in nine trials, although in two of these trials there was evidence that the groups were not comparable at baseline (Harris 1977; Pretlow 1999). In the remaining seven trials baseline comparison of measurements of wetting for individual interventions were not reported (Baker 1969; Bhatia 1990; Fava 1981; Mehrotra 1980; Turner 1970; Van Dommelen 2009; Van Hoeck 2007).
Effects of interventions
The 16 trials included 1643 children, of whom 845 received a simple behavioural intervention. However, these interventions were so disparate that only bladder training compared with alarm training could be combined for meta‐analysis.
1. Simple behavioural interventions compared with no active treatment (waiting list controls, play therapy)
See Comparisons 1.1. 1.2, 1.4, Other Data Tables 1.5
Six trials compared simple behavioural interventions with control management. The simple behavioural interventions were:
-
bladder training (Bennett 1985; Harris 1977);
-
rewards (including star charts) (Ronen 1995; Van Dommelen 2009);
-
lifting (with and without passwords) (Van Dommelen 2009);
-
rewards plus lifting or waking (Baker 1969; Fava 1981).
The control conditions included waiting list controls and play therapy. The interventions, control conditions and reported outcomes were so variable that they could not be combined.
Bladder training (compared with controls) was associated with fewer mean wet nights at the end of treatment (MD ‐1.90, 95% CI ‐3.67 to ‐0.13, Analysis 1.1.1) in one trial but there was no difference in number not achieving 14 dry nights at the end of treatment (RR 0.85, 95% CI 0.63 to 1.15, Analysis 1.2.1) (Bennett 1985). However, in this study some of the controls were given star charts if they achieved a dry night while on the waiting list and the star chart reward may have confounded the comparison (Bennett 1985). In Harris 1977 bladder training was also associated with lower absolute mean wet nights at the end of treatment compared with controls, but no test of significance could be applied as SDs were not given (Analysis 1.5.2).
Rewards were associated with significantly lower mean wet nights (MD ‐4.63, 95% CI ‐6.41 to ‐2.85, Analysis 1.1.2) (Ronen 1995) and number not achieving 14 dry nights at the end of treatment (RR 0.84, 95% CI 0.73 to 0.95, Analysis 1.2.3) (Ronen 1995; Van Dommelen 2009) compared with control, although the rewards and control interventions were quite heterogeneous between the two studies.
Lifting without a password (RR 0.79, 95% CI 0.68 to 0.92, Analysis 1.2.4) resulted in fewer children not achieving 14 dry nights at the end of treatment compared with controls. However, there was no statistically significant difference for lifting with a password compared with controls (RR 0.92, 95% CI 0.81 to 1.05, Analysis 1.2.5) (Van Dommelen 2009).
Rewards plus lifting resulted in fewer children not achieving 14 dry nights at the end of treatment (RR 0.22, 95% CI 0.06 to 0.78, Analysis 1.2.2), and also fewer failing or relapsing after treatment stopped (RR 0.22, 95% CI 0.06 to 0.78, Analysis 1.4.1) (Fava 1981), compared with controls. Rewards plus waking were also associated with fewer wet nights at the end of treatment compared with controls (Baker 1969), but no test of significance could be applied as SDs were not given (Analysis 1.5.1).
Overall, simple behavioural interventions appeared to be more effective than control management.
2. Simple behavioural interventions compared with other behavioural interventions
See Comparisons 2.1. 2.2, 2.3, 2.4, Other Data Tables 2.5
Eleven trials compared simple behavioural interventions with another behavioural intervention. The simple behavioural interventions studied were:
-
bladder training (either as monotherapy or in combination with other treatments such as alarm training) (Bennett 1985; Bryant 2003; Fielding 1980; Pretlow 1999; Van Hoeck 2007);
-
rewards (Ronen 1995);
-
waking or lifting (without a password) (El‐Anany 1999; Fournier 1987; Turner 1970; Van Dommelen 2009);
-
rewards plus waking (Baker 1969).
Comparison interventions were:
-
another simple behavioural intervention
-
lifting with a password (Van Dommelen 2009);
-
waking at the expected time of wetting (El‐Anany 1999) or waking using a bladder volume alarm (Pretlow 1999);
-
-
other behavioural interventions
-
enuresis alarm therapy (Baker 1969; Bennett 1985; Bryant 2003; Fielding 1980; Fournier 1987; Ronen 1995; Turner 1970; Van Hoeck 2007);
-
dry bed training (Bennett 1985);
-
-
combination therapies
-
bladder training plus enuresis alarm therapy (Bryant 2003);
-
enuresis alarm therapy plus imipramine (Fournier 1987);
-
-
cognitive therapy (Ronen 1995).
Bladder training was not as effective as enuresis alarm therapy based on fewer wet nights at the end of treatment with alarms (MD 2.25, 95% CI 0.30 to 4.20, Analysis 2.1.2) (Bennett 1985), fewer children not achieving 14 dry nights at the end of treatment with alarms (RR 2.73, 95% CI 1.75 to 4.26, Analysis 2.2.1)(Bennett 1985; Van Hoeck 2007) and fewer wet nights after completion of treatment with alarms (MD 2.60, 95% CI 0.67 to 4.53, Analysis 2.3.1)(Bennett 1985). The Bryant 2003 trial also reported fewer wet nights at the end of treatment with alarm therapy, but no test of significance could be applied as SDs were not given. Bladder training did not provide any additional benefit to alarm therapy: there was no statistically significant difference in numbers of children not achieving 14 dry nights at the end of treatment for bladder training combined with enuresis alarm therapy (RR 1.77, 95% CI 0.50 to 6.23, Analysis 2.2.2) (Fielding 1980) or bladder volume alarm (RR 0.89, 95% CI 0.43 to 1.83, Analysis 2.2.3) (Pretlow 1999) but the trials were small and the confidence intervals were wide. Therefore, alarm training does appear to be more effective than bladder training, with no apparent additional benefits for bladder training when added to enuresis or bladder volume alarm therapy.
When reward treatment was compared with cognitive therapy, there was no statistically significant difference in mean wet nights at the end of treatment (MD 0.77, 95% CI ‐0.29 to 1.83, Analysis 2.1.3). However, the numbers not achieving 14 dry nights at the end of treatment (RR 2.80, 95% CI 1.24 to 6.30, Analysis 2.2.6) and the number failed or relapsed after completion of treatment was lower (better) for cognitive therapy compared to rewards alone (RR 3.43, 95% CI 1.11 to 10.59, Analysis 2.4.3) (Ronen 1995) suggesting that cognitive therapy may be superior to rewards.
When reward treatment was compared with enuresis alarm therapy (Ronen 1995), no statistically significant difference in mean wet nights at the end of treatment was demonstrated (MD 0.70, 95% CI ‐0.65 to 2.05, Analysis 2.1.4). When reward plus waking was compared with enuresis alarm therapy, there appeared to be no difference in the total mean wet nights at the completion of treatment (Analysis 2.5.2) (Baker 1969) although no test of significance could be applied as SDs were not given. When reward treatment was compared with lifting with a password (RR 1.07, 95% CI 0.92 to 1.25, Analysis 2.2.12) or lifting without a password (RR 0.93, 95% CI 0.78 to 1.10, Analysis 2.2.11) there was no statistically significant difference in numbers not achieving 14 dry nights (Van Dommelen 2009).
When random waking was compared with enuresis alarm therapy using a continuous signal (conventional) alarm, there was no difference in mean wet nights (MD 0.33, 95% CI ‐1.23 to 1.89, Analysis 2.1.5) (Turner 1970) and Analysis 2.5.3 (Fournier 1987) where no test of significance could be applied as SDs were not given), and number not achieving 14 dry nights at the end of treatment (RR 1.17, 95% CI 0.88 to 1.55, Analysis 2.2.8) (Turner 1970). Similarly, when random waking was compared with enuresis alarm therapy using a twin signal alarm, there was no difference in mean wet nights (MD ‐0.35, 95% CI ‐1.84 to 1.14, Analysis 2.1.6) or number not achieving 14 dry nights at the end of treatment (RR 1.08, 95% CI 0.85 to 1.37, Analysis 2.2.9) (Turner 1970). When waking at set times was compared with waking linked to expected time of wetting there was no difference in the number not achieving 14 dry nights at the end of treatment (RR 0.79, 95% CI 0.50 to 1.22, Analysis 2.2.7) or the number failed or relapsed after completion of treatment (RR 1.06, 95% CI 0.79 to 1.44, Analysis 2.4.4) (El‐Anany 1999). When lifting without a password was compared with lifting with a password there was no difference in number not achieving 14 dry nights at the end of treatment (RR 0.86, 95% CI 0.74 to 1.02, Analysis 2.2.10) (Van Dommelen 2009).
When bladder training was compared with dry bed training (Bennett 1985), there were no statistically significant differences in mean wet nights at the end of treatment (MD 1.85, 95% CI 0.00 to 3.70, Analysis 2.1.1), numbers not achieving 14 dry nights (RR 1.67, 95% CI 0.85 to 3.26, Analysis 2.2.4) or mean wet nights after completion of treatment (MD 1.10, 95% CI ‐1.22 to 3.42, Analysis 2.3.2).
3. Simple behavioral interventions compared with drug interventions
See Comparison 3.1. 3.2, 3.3, 3.4, Other Data Tables 3.5
Six trials compared a simple behavioural intervention with a drug or placebo. The trials were all small, with at most 60 children per arm. The simple behavioural interventions studied were:
-
bladder training either alone or in combination with oxybutynin or placebo (Hamano 2000; Van Hoeck 2007);
-
waking (Fournier 1987; Turner 1970);
-
rewards plus waking plus placebo (Mehrotra 1980);
-
fluid restriction and avoidance of punishment (Bhatia 1990).
Comparison drug interventions were:
-
desmopressin (Hamano 2000);
-
imipramine (Bhatia 1990; Fournier 1987);
-
amitriptyline (Mehrotra 1980);
-
oxybutynin (Van Hoeck 2007);
-
placebo (Bhatia 1990; Fournier 1987; Turner 1970; Van Hoeck 2007).
When bladder training plus placebo drug was compared with a placebo drug alone (Van Hoeck 2007), there was no difference in number not achieving 14 dry nights at the end of treatment (RR 0.97, 95% CI 0.88 to 1.06, Analysis 3.2.6). When bladder training plus placebo was compared with oxybutynin (Van Hoeck 2007), there was no difference in number not achieving 14 dry nights at the end of treatment (RR 1.11, 95% CI 0.95 to 1.30, Analysis 3.2.5). When bladder training plus placebo was compared with bladder training plus oxybutynin (Van Hoeck 2007), there was no difference in number not achieving 14 dry nights at the end of treatment (RR 1.07, 95% CI 0.93 to 1.23, Analysis 3.2.7).
When bladder training was compared with desmopressin (Hamano 2000), there was no difference in mean wet nights (MD ‐0.10, 95% CI ‐0.40 to 0.20, Analysis 3.1.2) and number not achieving 14 dry nights (RR 1.25, 95% CI 0.97 to 1.62, Analysis 3.2.4) at the end of treatment. Nor was there a difference after treatment stopped, largely because most children who were cured while using desmopressin relapsed once the drug was stopped (number failed or relapsed after completion of treatment (RR 0.92, 95% CI 0.81 to 1.05, Analysis 3.4.5)).
When waking was compared with placebo (Turner 1970), there was no difference in mean wet nights (MD ‐0.99, 95% CI ‐2.54 to 0.56, Analysis 3.1.1), and number not achieving 14 dry nights (RR 1.22, 95% CI 0.91 to 1.64, Analysis 3.2.1) at the end of treatment. In another study (Fournier 1987) the numbers were too small to determine whether waking was superior to placebo (Analysis 3.5.1) and no test of significance could be applied as SDs were not given.
When waking plus reward plus placebo drug was compared with amitriptyline (Mehrotra 1980), amitriptyline appeared to be more effective with lower numbers not achieving 14 dry nights at the end of treatment (RR 2.83, 95% CI 1.42 to 5.67, Analysis 3.2.2). However, there was no significant difference in the number who failed or relapsed after completion of treatment (RR 1.50, 95% CI 0.90 to 2.49, Analysis 3.4.1). When waking plus reward plus placebo was compared with waking plus reward plus amitriptyline, the addition of amitriptyline was just superior to the behavioural therapy alone, with a lower number not achieving 14 dry nights at the end of treatment (RR 1.55, 95% CI 1.00 to 2.39, Analysis 3.2.3). However, the number who failed or relapsed after completion of treatment (RR 1.00, 95% CI 0.70 to 1.43, Analysis 3.4.2) was not statistically different. This suggests that the amitriptyline may be effective during treatment but has no sustained effect after stopping treatment .
When random waking was compared with imipramine, or imipramine combined with alarm therapy, although the mean number of wet nights was lower for imipramine or imipramine combined with alarm therapy (Analysis 3.5.2 and Analysis 3.5.3, respectively), the numbers were small and no test of significance could be applied as SDs were not given (Fournier 1987).
When fluid restriction and avoidance of punishment was compared to imipramine (either alone or in combination with fluid restriction and avoidance of punishment), the addition of imipramine appears to be superior with lower numbers of those who failed or relapsed after completion of treatment (RR 2.00, 95% CI 1.12 to 3.57, Analysis 3.4.3 and RR 8.00, 95% CI 2.11 to 30.34, Analysis 3.4.4, respectively) (Bhatia 1990). Therefore, the tricyclic antidepressants imipramine and amitriptyline alone or in combination with simple behavioural interventions appeared to be superior to simple behavioural interventions alone.
Adverse effects
There were no reported adverse effects directly caused by simple behavioural interventions. Eight trials mentioned adverse events but only five specifically mentioned adverse events for individual interventions (Baker 1969; El‐Anany 1999; Hamano 2000; Mehrotra 1980; Van Hoeck 2008). The main adverse events, leading to the high dropout rate, were: failure of the treatment because it was too demanding of the children or their families (Pretlow 1999); because it led to family strife (Pretlow 1999; Turner 1970); or emotional problems (Baker 1969); or simply because it did not work (Bennett 1985; El‐Anany 1999; Fournier 1987; McConaghy 1969; Pretlow 1999). Adverse events caused by the comparison intervention were mentioned (for example, headaches with oxybutynin (Van Hoeck 2007), nasal discomfort with desmopressin (Fournier 1987; Hamano 2000) and drowsiness with amitriptyline (Mehrotra 1980)). The remaining eight trials failed to report whether adverse events occurred or not (Bennett 1985; Bhatia 1990; Bryant 2003; Fava 1981; Fielding 1980; Harris 1977; Ronen 1995; Van Dommelen 2009).
Discussion
Summary of main results
Comparison with no active treatment
Simple behavioural interventions generally appeared to be more effective than no active treatment. Rewards for dry nights with or without lifting or waking appeared to be more effective (Baker 1969; Fava 1981; Ronen 1995; Van Dommelen 2009), but each finding was based on single, small trials. Bladder training (retention control training) also appeared to be more effective than no active treatment, although the studies were too small and of low methodological quality to provide useful information (Baker 1969; Bennett 1985; Harris 1977). These interventions are widely used by families as first line treatment. The role of lifting is more controversial because of concerns that it encourages the child to pass urine whilst asleep, thus preventing their learning to wake to the sensation of a full bladder, even if there may be more instances of dry nights. In one study, lifting without a password (but not lifting with a password) was superior to no active treatment (Van Dommelen 2009). One explanation for the failure of lifting with a password (compared with the success of lifting without a password) may be related to increased tiredness in the child caused by their trying to remember the password which may affect their treatment response.
Comparison with other behavioural interventions
Enuresis alarm therapy was more effective than bladder training (retention control training) for treating nocturnal enuresis (Bennett 1985; Bryant 2003; Van Hoeck 2007), with no additional benefit when bladder training was combined with alarm therapy (Fielding 1980; Pretlow 1999). Cognitive therapy also appeared to be more effective than rewards in one small trial (Ronen 1995). In the other studies of behavioural interventions, no single behavioural intervention was more effective than any other behavioural intervention, with no differences between alarm therapy and rewards, lifting or waking or combination of rewards and waking, or between rewards and various forms of lifting or waking or between bladder training and dry bed training.
Comparison with drugs
Four of the six trials which included a drug arm were conducted over 20 years ago (Bhatia 1990; Fournier 1987; Mehrotra 1980; Turner 1970) before the recognition of the effectiveness of desmopressin (Glazener 2004a), tricyclic drugs (Glazener 2004b) and enuresis alarms (Glazener 2004d) for treating nocturnal enuresis.
The tricyclics amitriptyline and imipramine were compared with simple behavioural therapies and appeared to be more effective than simple behavioural therapies such as waking, lifting, rewards, fluid restriction or avoidance of punishment, either alone or in combination (Bhatia 1990; Fournier 1987; Mehrotra 1980).
There was no difference demonstrated between bladder training and desmopressin (Hamano 2000), oxybutynin or placebo, either alone or in combination with bladder training (Van Hoeck 2007).
Although the tricyclics appeared to be more effective than some simple behavioural therapies, the potential adverse effects of the tricyclics needs to be considered. The main problems with many of the simple behavioural interventions were the high dropout and non‐compliance rates as families found themdifficult and found the waking at night to attend to the child disruptive and stressful.
Overall completeness and applicability of evidence
This review included 16 trials comparing various simple behavioural strategies for treating nocturnal enuresis, involving 1643 participants. Simple behavioural therapies are generally more effective than no treatment but inferior to proven effective treatments such as alarm training and tricyclics. As most trials are small and of poor methodological quality, care needs to be taken with interpretation of the results.
Quality of the evidence
The methodological quality of the 16 included trials was low. Of the nine trials which reported means of continuous outcomes, five failed to provide SDs (Baker 1969; Bryant 2003; Fielding 1980; Fournier 1987; Harris 1977). The sample sizes were small for most studies. The only combination of interventions that could be meta‐analysed was bladder training versus enuresis alarm therapy. In the other comparisons outcomes were reported for single trials only, precluding meta‐analysis. The confidence intervals are wide and this is likely to obscure or overestimate treatment effects. Only ten of the 16 studies gave information about the follow‐up results after the intervention was finished (Bennett 1985; Bhatia 1990; Bryant 2003; El‐Anany 1999; Fava 1981; Fielding 1980; Hamano 2000; Mehrotra 1980; Ronen 1995; Van Dommelen 2009).
Most included trials either had high risks of bias or had insufficient information to judge the level of risk. Thirteen of the 16 trials did not provide information about the method used to generate the allocation sequence, blinding was usually not possible for simple behavioural therapies, most studies had high attrition rates, with 10 of the studies not using intention‐to‐treat analysis or imputation methods to deal with the high loss to follow up.
Other potential biases included not providing adequate baseline or background information about the participants, inclusion of children with organic causes of wetting or daytime urinary incontinence (who may respond differently to the interventions) and failure to report details of adverse reactions.
Potential biases in the review process
Many of the studies did not provide adequate detail to allow the reviewers to assess the adequacy of the quality of the study. Therefore, subjective judgement was sometimes made by the reviewers based on the scant information provided.
As there was no clear definition of what is a "simple behavioural intervention" for treating nocturnal enuresis, the reviewers had to make a subjective judgement. It could be argued that some of the included interventions should be classified as a complex behavioural intervention.
Agreements and disagreements with other studies or reviews
The results of this study are consistent with the other reviews of nocturnal enuresis, showing superiority of proven effective treatments such as enuresis alarm therapy and tricyclic antidepressants over simple behavioural therapies. However, there was no difference found between desmopressin and simple behavioural therapies: this may have been caused by the inclusion of only one small, low quality study of desmopressin versus bladder training (a simple behavioural therapy which did not appear to be effective when compared with other active treatments).
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Comparison 1 Simple behavioural intervention vs control, Outcome 1 Mean wet nights on treatment.
Comparison 1 Simple behavioural intervention vs control, Outcome 2 Number not achieving 14 consecutive dry nights (failed) on treatment.
Comparison 1 Simple behavioural intervention vs control, Outcome 4 Number failed or relapsed after completion of treatment.
| Study | Behavioural | Control |
| Waking + reward vs control (wait list) | ||
| Baker 1969 | Mean wet nights per week 3.1; N = 10 | Mean wet nights per week 5.9; N = 10 |
| Bladder training vs control (wait list) | ||
| Harris 1977 | Mean wet nights per week 2.6; N = 9 | Mean wet nights per week 5; N = 9 |
Comparison 1 Simple behavioural intervention vs control, Outcome 5 Mean wet nights on treatment (no SDs).
Comparison 2 Simple behavioural intervention vs another behavioural intervention, Outcome 1 Mean wet nights on treatment.
Comparison 2 Simple behavioural intervention vs another behavioural intervention, Outcome 2 Number not achieving 14 consecutive dry nights (failed) on treatment.
Comparison 2 Simple behavioural intervention vs another behavioural intervention, Outcome 3 Mean wet nights after completion of treatment.
Comparison 2 Simple behavioural intervention vs another behavioural intervention, Outcome 4 Number failed or relapsed after completion of treatment.
| Study | Behavioural int. | Other behavioural |
| Bladder training + alarm vs alarm alone | ||
| Fielding 1980 | Mean wet nights per week 1.5; N = 16 | Mean wet nights per week 0.6; N = 17 |
| Waking + reward vs alarm | ||
| Baker 1969 | Mean wet nights 3.1; N = 10 | Mean wet nights 1.8; N = 10 |
| Waking vs alarm | ||
| Fournier 1987 | Mean wet nights per week 3.3; N = 8 | Mean wet nights per week 2.5; N = 8 |
| Bladder training vs alarm | ||
| Bryant 2003 | Mean wet nights per week 4.0; N = 37 | Mean wet nights per week 3.0; N = 34 |
| Bladder training vs bladder training + alarm | ||
| Bryant 2003 | Mean wet nights per week 4.0; N = 37 | Mean wet nights per week 4.3; N = 45 |
Comparison 2 Simple behavioural intervention vs another behavioural intervention, Outcome 5 Mean wet nights on treatment (no SDs).
Comparison 3 Simple behavioural intervention vs drug intervention, Outcome 1 Mean wet nights on treatment.
Comparison 3 Simple behavioural intervention vs drug intervention, Outcome 2 Number not achieving 14 consecutive dry nights (failed) on treatment.
Comparison 3 Simple behavioural intervention vs drug intervention, Outcome 4 Number failed or relapsed after completion of treatment.
| Study | Behavioural | Drug |
| Waking vs placebo | ||
| Fournier 1987 | Mean wet nights per week 3.3; N = 8 | Mean wet nights per week 5; N = 8 |
| Waking vs imipramine | ||
| Fournier 1987 | Mean wet nights per week 3.3; N = 8 | Mean wet nights per week 1.9; N = 8 |
| Waking vs alarm + imipramine | ||
| Fournier 1987 | Mean wet nights per week 3.3; N = 8 | Mean wet nights per week 1; N = 8 |
Comparison 3 Simple behavioural intervention vs drug intervention, Outcome 5 Mean wet nights on treatment (no SDs).
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Mean wet nights on treatment Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 1.1 Bladder training vs control (wait list + reward) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Reward vs control (wait list) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Number not achieving 14 consecutive dry nights (failed) on treatment Show forest plot | 4 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2.1 Bladder training vs control (wait list + reward) | 1 | 21 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.85 [0.63, 1.15] |
| 2.2 Reward + lifting vs control (play) | 1 | 20 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.22 [0.06, 0.78] |
| 2.3 Reward vs control (wait list) | 2 | 325 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.84 [0.73, 0.95] |
| 2.4 Lifting (without password) vs control | 1 | 287 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.79 [0.68, 0.92] |
| 2.5 Lifting (with password) vs control | 1 | 284 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.92 [0.81, 1.05] |
| 3 Mean wet nights after completion of treatment | 0 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 4 Number failed or relapsed after completion of treatment Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.1 Reward + lifting vs control (play) | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 5 Mean wet nights on treatment (no SDs) Show forest plot | Other data | No numeric data | ||
| 5.1 Waking + reward vs control (wait list) | Other data | No numeric data | ||
| 5.2 Bladder training vs control (wait list) | Other data | No numeric data | ||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Mean wet nights on treatment Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 1.1 Bladder training vs Dry Bed Training (complex intervention) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Bladder training vs alarm | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.3 Reward vs cognitive therapy | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.4 Reward vs alarm | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.5 Waking vs alarm (continuous signal) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.6 Waking vs alarm (twin signal) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Number not achieving 14 consecutive dry nights (failed) on treatment Show forest plot | 8 | Risk Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2.1 Bladder training vs alarm | 2 | 80 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.73 [1.75, 4.26] |
| 2.2 Bladder training + alarm vs alarm alone | 1 | 33 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.77 [0.50, 6.23] |
| 2.3 Bladder training + bladder volume alarm vs alarm alone | 1 | 40 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.89 [0.43, 1.83] |
| 2.4 Bladder training vs Dry Bed Training (complex intervention) | 1 | 22 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.67 [0.85, 3.26] |
| 2.5 Reward vs alarm | 1 | 39 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.9 [0.99, 3.66] |
| 2.6 Reward vs cognitive therapy | 1 | 40 | Risk Ratio (M‐H, Fixed, 95% CI) | 2.8 [1.24, 6.30] |
| 2.7 Waking linked to wetting vs waking at set time | 1 | 125 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.79 [0.50, 1.22] |
| 2.8 Waking vs alarm (continuous signal) | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.17 [0.88, 1.55] |
| 2.9 Waking vs alarm (twin signal) | 1 | 30 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.08 [0.85, 1.37] |
| 2.10 Lifting (without password) vs lifting with password | 1 | 283 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.86 [0.74, 1.02] |
| 2.11 Lifting (without password) vs reward | 1 | 286 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.93 [0.78, 1.10] |
| 2.12 Lifting (with password) vs reward | 1 | 283 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.07 [0.92, 1.25] |
| 3 Mean wet nights after completion of treatment Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3.1 Bladder training vs alarm | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.2 Bladder training vs Dry Bed Training (complex intervention with alarm) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4 Number failed or relapsed after completion of treatment Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.1 Bladder training + alarm vs alarm alone | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.2 Reward vs alarm | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.3 Reward vs cognitive therapy | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.4 Waking linked to wetting vs waking at set time | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 5 Mean wet nights on treatment (no SDs) Show forest plot | Other data | No numeric data | ||
| 5.1 Bladder training + alarm vs alarm alone | Other data | No numeric data | ||
| 5.2 Waking + reward vs alarm | Other data | No numeric data | ||
| 5.3 Waking vs alarm | Other data | No numeric data | ||
| 5.4 Bladder training vs alarm | Other data | No numeric data | ||
| 5.5 Bladder training vs bladder training + alarm | Other data | No numeric data | ||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Mean wet nights on treatment Show forest plot | 2 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 1.1 Waking vs placebo | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Bladder training vs desmopressin | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Number not achieving 14 consecutive dry nights (failed) on treatment Show forest plot | 4 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 2.1 Waking vs placebo | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.2 Waking + reward + placebo vs amitriptyline | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.3 Waking + reward + placebo vs waking + reward + amitriptyline | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.4 Bladder training vs desmopressin | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.5 Bladder training + placebo vs oxybutynin | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.6 Bladder training + placebo vs placebo | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.7 Bladder training + placebo vs Bladder training + oxybutynin | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 3 Mean wet nights after completion of treatment | 0 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 4 Number failed or relapsed after completion of treatment Show forest plot | 3 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 4.1 Waking + reward + placebo vs amitriptyline | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.2 Waking + reward + placebo vs waking + reward + amitriptyline | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.3 Fluid restriction and avoidance of punishment vs imipramine | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.4 Fluid restriction and avoidance of punishment vs fluid restriction and avoidance of punishment + imipramine | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4.5 Bladder training vs desmopressin | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 5 Mean wet nights on treatment (no SDs) Show forest plot | Other data | No numeric data | ||
| 5.1 Waking vs placebo | Other data | No numeric data | ||
| 5.2 Waking vs imipramine | Other data | No numeric data | ||
| 5.3 Waking vs alarm + imipramine | Other data | No numeric data | ||
