Potentially relevant studies were identified by a computerized search of the Medline electronic database (PubMed, 1966–2008). Relevant text and keywords were: enuresis OR nocturnal enuresis OR monosymptomatic nocturnal enuresis, AND randomized controlled trial OR controlled trial OR meta-analysis. The search was limited to the English-language literature. The Cochrane reviews represent a fertile starting point for an evidence-based evaluation of the treatment of NE. All seven Cochrane NE reviews were based on a previous systematic review performed by Lister-Sharpe et al. at the Centre for Reviews and Dissemination [9]. Relevant trials were identified from a specialized register of controlled trials identified from Medline, CINAHL, the Cochrane Central Register of Controlled Trials, and hand searching of journals and relevant conference proceedings. Details regarding review methods, including identification of primary studies, quality assessments, and data extraction, are described in detail in each Cochrane review.

There exist some challenges in interpretation of studies. While there is no shortage of randomized, controlled trials (RCTs) in the NE literature, few are methodologically sound. Thus, the conclusions of these trials must be tempered by shortcomings, such as high dropout rates and poor rates of adherence to therapeutic regimen. Additionally, many trials are small and, frequently, outcomes may be represented by findings of single trials. Furthermore, especially when medical therapy is involved, the initial response rates may not be representative of durable, long-term response rates. Outcome measures may be different, as well. Some of the outcomes considered in the Cochrane reviews were: change in the mean number of wet nights per week during treatment; number of children failing to attain 14 consecutive dry nights during treatment; mean number of wet nights per week when participants were followed up after treatment cessation; and number of children failing during treatment and/or relapsing after treatment cessation.

The evidence

While multiple modalities may be employed in the treatment of NE, simple and complex behavioral interventions are often a component of the initial therapeutic attempt. The various behavioral techniques are summarized in Box 39.1. Additionally, educational interventions, including different methods of providing information or teaching for children and their parents, may be used to supplement behavioral therapy. In a Cochrane review (updated February 19, 2004) encompassing 13 trials, 387 out of 702 children received a simple behavioral intervention [11]. A meta-analysis could not be performed as each outcome in each comparison was reported in a single trial. Five trials compared simple interventions (including retention control training, rewards, lifting or waking) with no active treatment (control group). Star charts (with or without lifting or waking) were associated with significantly fewer wet nights and lower failure rates during treatment [13–15]. Additionally, fewer children relapsed after achieving success with reward therapy than controls. There was insufficient evidence to support retention control training alone.

Eight trials compared a simple behavioral intervention with another behavioral intervention, such as cognitive therapy, dry bed training (DBT) and alarms with or without imipramine. Evidence from single small trials revealed that children undergoing cognitive therapy had lower failure or relapse rates than those children using star charts. Additionally, children using alarms achieved fewer wet nights after the end of treatment than those practicing retention control training, or those who were randomly awakened. Five small trials included a placebo, desmopressin, imipramine or amitriptyline in one arm. There was insufficient evidence to compare desmopressin with retention control training; however, amitriptyline was associated with better initial results than waking, lifting, star chart and placebo in one small trial [16]. There was no significant association after treatment cessation. In another trial, imipramine with or without fluid deprivation and avoidance of punishment was better than the simple intervention alone [17]. Additionally, children using imipramine, with or without the addition of an alarm, had fewer absolute numbers of wet nights per week than during random wakening, while children receiving placebo had more wet nights per week [18].

A recent Cochrane review (updated March 20, 2008) evaluated studies of complex behavioral and educational interventions for NE [12]. Of 1174 children from 18 included trials, 746 received a complex behavioral or educational intervention. In eight of these trials, children undergoing a complex intervention were compared with those undergoing no treatment. Children allocated to DBT or full-spectrum home training (FSHT) with an alarm had fewer wet nights and were more likely to become dry during treatment (relative risk (RR) for failure 0.17, 95% confidence interval (CI) 0.11–0.28) and after treatment cessation (RR for failure of relapse after stopping DBT was 0.25, 95% CI 0.16–0.39) than those in control groups. However, if treatment was performed without an alarm, the difference between complex intervention and no treatment groups was not significantly different in the dry rate during treatment (RR for failure 0.82, 95% CI 0.66–1.02) and the rate of relapse after treatment cessation. In five trials, children receiving complex interventions were compared with those receiving other behavioral interventions. DBT alone, a three-step program of retention control training, wakening, and parental reassurance, and FSHT were compared with alarm alone, DBT with alarm, three-step program supplemented with counseling and parental education, and FSHT with pelvic floor muscle training. An alarm (RR for failure or relapse 1.7, 95% CI 1.06–2.73), DBT with alarm (RR for failure or relapse 2.81, 95% CI 1.80–4.38) or a three-step program with counseling and educational reinforcement (RR for failure or relapse 2.07, 95% CI 1.16–3.72) were significantly better than a complex intervention alone in the number of wet nights during treatment, failure rates during treatment, and combined failure and relapse rates after treatment cessation.

In four trials comparing supplemented complex intervention and other behavioral interventions, failure rates during treatment were marginally better after concomitant DBT with alarm compared to treatment with an alarm only (RR 0.6, 95% CI 0.38–0.94). There was not enough evidence to suggest that any method of providing instruction to children or parents regarding the implementation of complex behavioral interventions was better than another, except that live delivery of FSHT was better than filmed delivery in two small trials (RR for failure during treatment 0.36, 95% CI 0.15–0.90). In one trial, a three-step program with supplemental counseling and educational reinforcement seemed to be better than imipramine alone (RR for failure and relapse 0.27, 95% CI 0.16–0.43) [19]. In a second trial, children receiving various behavioral interventions had more wet nights than those on desmopressin (weighted mean difference (WMD) 1.67, 95% CI 0.35–2.99); however, there were not enough children to statistically assess failure rates [20].

The evidence

The most common enuresis alarms consist of a sensor on a bed pad or pajamas which is triggered by wetness, thereby waking the child when they begin to urinate. Complementary techniques such as overlearning or giving extra fluids at bedtime to stress the detrusor are often added after a child successfully becomes dry using an alarm. Alarm interventions have been evaluated in a recent Cochrane review (updated February 22, 2005) that encompassed 56 trials [21]. A total of 3257 children were included, of whom 2412 used an alarm.

The majority of the trials used a pad-and-buzzer type of alarm to wake the children when wetting occurred, although a few used an electric shock to the child’s skin. Seventeen trials compared an alarm with a no-treatment control group. In nine of these trials, standard alarm use was associated with over three fewer wet nights per week when compared to controls (WMD –3.34, 95% CI –4.14 to –2.55). In 13 trials, RR of failure was less in the alarm group in every trial (107 of 316 (34%) did not achieve 14 dry nights vs 250 of 260 (96%) in controls; RR 0.38, 95% CI 0.33–0.45). While approximately half of the children failed or relapsed after stopping alarm treatment (45 of 81, 55%), nearly all the children in the control group relapsed (80 of 81, 99%) (RR 0.56, 95% CI 0.46–0.68). There were no data regarding the number of wet nights after treatment cessation. Adding overlearning (RR 1.92, 95% CI 1.27–2.92) or DBT (RR 2.0, 95% CI 1.25–3.20) to alarm treatment contributed to lower relapse rates.

Alarm use was compared with behavioral intervention in eight trials. In one trial, alarm use was associated with fewer wet nights than stop-start training, both during treatment (WMD –2.25, 95% CI –4.2 to –0.3) and after treatment cessation (WMD –2.6, 95% CI –4.53 to –0.67) [22]. Likewise, alarm use was associated with fewer wet nights than DBT during treatment (RR for failure to achieve 14 dry nights was 0.22, 95% CI 0.09–0.53) and after treatment cessation (RR 0.59, 95% CI 0.37–0.95). While alarm use was associated with fewer wet nights, when compared to lifting, waking or rewards, the difference was not significant.