12.2.3. Health economic evidence review
Given the lack of published evidence assessing the cost-effectiveness of different interventions, including enuresis alarms, used in the treatment of bedwetting, the GDG identified this area as high priority for original economic analysis. Therefore, a cost-utility analysis was undertaken where costs and quality-adjusted life-years (QALYs) were considered from a UK National Health Service and Personal Social Services perspective. The time horizon for the analysis was 13 years, modelling patients from the time they entered at age 7 years until they reached age 20.
A summary of the analysis is provided below. The full report is presented in appendix G.
Summary of results
The results of the probabilistic sensitivity analysis are summarised in in terms of mean total costs and mean total QALYs and mean net benefit for each treatment sequence, where each mean is the average of 20,000 simulated estimates. The option with the greatest mean net benefit is the most cost-effective at a specified threshold (for example, £20,000). The percentage of simulations where each strategy was the most cost-effective gives an indication of the strength of evidence in favour of that strategy being cost-effective.
Basecase probabilistic sensitivity analysis results.
The results of the incremental analysis in the probabilistic analysis, excluding dominated and extendedly dominated strategies, are presented in .
Incremental analysis of basecase probabilistic results with dominated and extendedly dominated sequences removed.
The GDG considered that the differences between intervention sequences were relatively small and the probabilistic results indicated substantial uncertainty around the mean cost and benefit estimates. Small changes to the model inputs appears to result in substantial changes to the conclusions about modelled sequences’ relative and overall cost-effectiveness.
A series of sensitivity analyses were undertaken to test some of the assumptions feeding into the model and none of these affected the cost-effectiveness of the sequence alarm followed by combined alarm and desmopressin and then desmopressin alone compared to no treatment.
The economic analysis conducted and presented here represents the first undertaken to assess the cost-effectiveness of interventions used in the treatment of children with bedwetting. And although the analysis is directly applicable to decision making in the UK NHS, it has some potentially serious limitations, some of which may significantly impact the overall conclusions that can be drawn. The main limitations of the analysis are related to the fact that assumptions had to be made in the absence of evidence. Some of these key assumptions centre around:
treatment effectiveness being independent of age
health care resource use having been estimated by GDG
utility weights having been estimated by GDG
A full discussion of these can be found in appendix G.
12.2.6. Evidence to recommendations
Relative values of different outcomes
The GDG considered that sustained dryness was the outcome wished for by children and young people and their parents or carers. This was represented by the outcome of 14 consecutive dry nights to show initial success and indicate the effectiveness of the treatments being evaluated. The mean number of wet nights was also considered by the GDG in evaluating the effectiveness of treatments. Outcomes such as relapse and follow up rates were considered to evaluate sustained dryness.
Trade off between clinical benefit and harms
No evidence was identified of harms of alarm treatment.
Economic Considerations
Enuresis alarms were evaluated as part of original economic modelling undertaken for this guideline and were shown to be a likely cost-effective first line treatment option. The analysis showed that there was considerable uncertainty about which intervention was the most cost-effective first line option, and this was likely caused by the uncertainty around estimates of treatment effectiveness observed in the pairwise and network meta-analyses. The GDG considered that given the substantial uncertainty between interventions, it would be reasonable to recommend first line treatment with an alarm as it was consistently shown to be among less costly and still effective options.
As children and young people who have previously responded to an alarm are likely to respond to it again, it would be a good use of NHS resources to encourage children, young people and parents and carers to retain their alarm and reuse it before trying other options that have associated costs. The economic model assumed that prescribed alarms were given, not loaned, to patients and under this assumption, repeat use of alarms was considered cost-effective. Even if all alarms must be replaced at least once during treatment, they are still considered to be a cost-effective intervention.
Alarms are likely to be considered to be a cost-effective first line treatment regardless of age at initiation.
Quality of evidence (this includes clinical and economic)
The quality of evidence for the outcomes preferred by the GDG was generally low. The individual direct comparisons found in the evidence review were of underpowered studies with small sample sizes. Some studies did not give standard deviations and therefore mean difference and CI could not be calculated giving incomplete evidence.
The GDG considered that the available evidence on alarms compared to no treatment contained inadequate description of the study groups, mainly in terms of the patients’ age and the number of girls included in studies. One study compared supervised alarms to unsupervised alarms; the GDG considered that the type of supervision involved in the studies was not part of common clinical practice in England and Wales.
Other considerations
The GDG considered the direct evidence, the network meta-analysis and the health economic evidence in making their recommendations. They considered that the evidence from the direct comparisons indicated that alarms and desmopressin had similar effects on dryness (both complete dryness and reduced number of wet nights) when receiving treatment but children were more likely to have recurrence of bedwetting following use of desmopressin. In the study that examined monosymptomatic enuresis desmopressin had a faster response (described in Ng (2005) 113 as reduction in the number of wet nights, described in Wille (1986) 114 as the number of dry nights); however alarms had more continued success and children were less likely to experience a recurrence of bedwetting. The GDG considered that when children, young people and families and carers consult with healthcare professionals about bedwetting they are seeking a ‘cure’ from bedwetting. Pharmacological agents have an effect on bedwetting primarily while the child is taking them and do not affect the underlying pathophysiology; alarms develop a conditioned response of waking in response to full bladder which is more likely to continue after alarm treatment.
The evidence of continued success associated with alarms from the evidence review and from professional experience of the GDG was pivotal in the GDG decision to recommend alarms as the choice for first line treatment if the child, young person and family and carers could use alarm. The GDG considered that the first line use of an alarm as a non-pharmacological intervention in children and young people was also perceived as an advantage. The patient carer members specifically highlighted concern about using pharmacological interventions in young children if alternatives are available.
Children with special needs
There was no evidence that one type of alarm was better than another. The GDG considered that if different alarms were available children and families should be given choice. The evidence also indicated that alarms have been used successfully as treatment in children and young people with hearing problems and childrenand young people with behavioural problems. The GDG considered it important that these children and young people do not lose out on a potentially good treatment modaility and where possible, and with the needs of the child, young person and family and carers considered, alarms should be considered as treatment.
Children with and without daytime symptoms
The GDG were interested in whether there was any difference in response to alarm among population groups with different wetting patterns i.e. population who did or did not have daytme wetting. The evidence indicated that bedwetting in children and young people who also had daytime wetting did respond to an alarm and there was no difference in the number achieving 14 consecutive dry nights when alarm was compared with enuresis and desmopressin. The GDG considered it important that children with daytime wetting did have access to alarm treatment and made a recommendation to ensure that alarm is considered as an option for these children and young people. No specific evidence was found regarding treatment of secondary enuresis but the professional experience of the GDG was that these children and young people do respond similarly to children and young people with primary enuresis.
Children and young people who are very infrequent bedwetters will not wet often enough to have the conditioned responses by which an alarm works.
Assessment at 4 weeks
The GDG discussed the lack of evidence for when a child or young person should be assessed after starting treatment. From clinical experience the GDG discussed the benefits of following up early at 4 weeks or less to encourage the patient and report on progress with the treatment. The GDG made a consensus decision on assessment at 4 weeks after starting treatment. In younger children it may be advisable to stop at this stage as child may respond when older and proceeding with treatment for longer at this stage may engender negativity in the child, young person and family and carers about the alarm.
Continue alarm until minimum of 2 weeks uninterrupted dryness has been achieved
The GDG discussed the lack of evidence for how long the alarm should be used. The GDG discussed from clinical experience that to ensure continuing success it was important the child or young person continued to use the alarm until 14 consecutive dry nights was achieved to reduce the chance of experiencing a recurrence of bedwetting after treatment.
Review at 3 months
The GDG considered that it can take several weeks for an alarm to have an effect and it is important to inform and encourage the family or carers to use it for some weeks to get an effect. However a child not getting any response from an alarm should stop the alarm. The GDG considered it not appropriate to continue beyond 3 months if there was not continuing improvement. The use of an alarm can be difficult for a family and an assessment of motivation and ability to continue its use should be made.
Addition of reward systems
The evidence supported the addition of reward systems to alarms and this finding is consistent with psychological theory. The GDG considered that as part of the instructions in how to use the alarm it might be useful to suggest to the family and help them to consider how they will approach this.
Use of alarm in children between 5 and 7
While the GDG considered that children between 5 and 7 years may not require treatment those that do, and are appropriately motivated and mature enough to cope with an alarm, should not be denied use of an alarm by virtue of age alone.
Combination of alarm and pharmacological treatments
The direct evidence indicated that the combination of alarms with desmopressin were similarly effective to alarms alone in the number of wet nights at end of treatment and drop out rates for children with MNE. However relapse rates were inconclusive for children and young people with bedwetting and possible daytime symptoms. The GDG did not consider the evidence supportive of using combination treatment as first line, although the evidence indicated this may be better for children with severe wetting and may also be helpful for children and young people with behavioural difficulties. The evidence comparing alarms to imipramine (two small studies) had contradictory findings (for number of wet nights at the end of treatment). Alarms had fewer wet nights at follow up compared to imipramine. The addition of imipramine to an alarm was not supported by clinical evidence.
12.2.8. Supporting recommendations -evidence to recommendations
Economic considerations
No economic evidence was identified to support these recommendations, however the GDG felt that informing children and their parents/carers of the aims of treatment, how to use the device and how and when they might observe results may encourage adherence to a treatment that may produce only a gradual response. Additionally, offering children and their parents/carers advice on what to do if they experience a recurrence following success may reduce re-presentation at clinics.
Quality of evidence (this includes clinical and economic)
No evidence was identified.
Other considerations
The GDG considered that while alarms may have a sustained effect on dryness an alarm requires considerable effort and perseverance from both child, young person and family, including siblings and extended family. The GDG considered that an important part of considering an alarm was assessing whether the child, young person and family have the necessary motivation, time and energy to use an alarm. Contectual factors such as a new baby in the house, might make an alarm a less attractive first line treatment. The GDG were particularly concerned that in situations where family members are already finding it difficult to cope with bedwetting and where parents or carers may be expressing anger to the child, the introduction of an alarm might result in a more punitive approach to the child.
The GDG considered it important that child, young person and parents or carers were properly informed about how an alarm works and that it may take some weeks for it to have an effect. The GDG also discussed from clinical experience the importance of recording the time child waked and how wet they were as this can be a sign of commitment and allows for positive feedback during follow up clinics.
The GDG considered that the use of an alarm can be difficult for a child, young person and parent or carers to master and that families may need considerable advice and support and access to expertise when starting to use an alarm. The GDG used their experience both as professional and patient members to develop recommendations around the likely information children, young people and parents and carers need when using an alarm. This covered how to use the alarm, what to expect, the signs of response and information about dealing with problems with the alarm and how to return it. Offering and agreeing appropriate support was considered by the GDG to be vital in helping families have confidence in their use of alarm.
