The review evaluated the effectiveness of internet-based self-management interventions on health outcomes in children and adolescents with health conditions and found evidence for improvement in symptoms or disease control in obesity, recurrent pain, encopresis and particularly for asthma. The review was relatively well conducted, but the evidence was limited and so the reliability of the authors’ conclusions is unclear.

To evaluate the effectiveness of internet-based self-management interventions on health outcomes in children and adolescents with health conditions.

EBM Reviews, MEDLINE, EMBASE, CINAHL and PsycINFO were searched for peer-reviewed studies published in English between 1993 and 2008; search dates ranged from 1950 to January 2008. The bibliography of each retrieved article and review was handsearched. No attempt was made to locate or contact researchers for unpublished studies. Search terms were reported.

Randomised controlled trials (RCTs), quasi-randomised trials (quasi-RCTs), before-and-after design or feasibility RCT studies that evaluated internet-based or enabled self-management Interactive Health Communication Application (IHCA) interventions that targeted school-age children (six to 12 years) and/or adolescents (13 to 18 years) with subacute or chronic health conditions were eligible for inclusion. Eligible outcomes included: knowledge, social support, self-efficacy, emotions and health behaviours; health outcomes; quality of life (QOL); and resource utilisation. The review also assessed feasibility, compliance and usability. Studies were excluded if they involved other forms of IHCAs, including hand-held computers, video and telehealth programmes, CD-ROMs and video games. One study of acute hospitalised children was excluded.

Included studies were of children, adolescents and their parents. Most studies were of children with asthma; other individual studies were of children with recurrent pain, encopresis, traumatic brain injury or obesity. Ages ranged from six to 19 years. The proportion of males ranged from 36% to 79% (except for one study of female adolescents). Most studies were performed in North America. Some studies targeted specific ethnic groups (African-American, Taiwanese or mostly Caucasian). Interventions in the included studies were very varied: mean intervention duration was 36.5 weeks (range three weeks to two years); most were solely in-home and others were at school, clinic or combined clinic and home. All the interventions were psycho-educational and involved combinations of information and skills training to improve health outcomes. Specific therapeutic techniques included cognitive and/or behavioural therapy and educational modules. Controls either had standard education and medical care, had restricted access to general illness specific internet information plus standard medical care, were on waiting lists or had usual medical care. Timing of outcome measurement varied: pre- and post-intervention; up to one year; or at six-month intervals for two years.

Two independent researchers were involved in study selection. Agreement was by consensus.

Methodological quality was assessed by two reviewers independently. Any disagreements were resolved by consensus. The CONSORT checklist developed for trials of non-pharmacological treatments by Boutron et al. was used. The checklist had 23 items and scores were categorised as: poor quality (less than 8); fair quality (9 to 16); and good quality (more than 17). An internet link was provided to the checklist, which included randomisation and blinding.

Detailed data on the interventions were extracted. The authors’ main conclusions were extracted as either a statistically significant positive result (+) or no significant effect (0). One reviewer performed the data extraction, which was checked by a second reviewer. No attempt was made to contact authors for additional information.

A narrative synthesis was provided, due to the heterogeneity of the results.

Nine relevant studies were identified (n=1,415 children, adolescents and adults): eight RCTs (n=977, range 24 to 314) and one quasi-experimental study (n=438). Drop-out ranged from 0 to 63% in one control group. Four studies targeted young people alone. Five studies targeted young people and their parents, one such study was geared to the whole family (including siblings). All of the studies were in the fair quality range (mean score 13.8, range 10 to 16). Only three studies provided details of the randomisation procedure. Five studies used an intention to treat analysis.

All of the included studies measured improvements in health outcomes in terms of symptom management or disease control. Seven studies showed significant improvements in health outcomes. Four RCTs measured disease-related knowledge, which was significantly increased in two of three asthma RCTs, but not in an encompresis RCT. Two of six studies that measured participants’ QOL found a significant difference. Three of four asthma studies found a significant decrease in the use of health resources.

All of the studies reported on feasibility in terms of treatment efficiency, access and compliance: the most important factors identified were computer access (including a tendency for decline in website usage with time) and suitability of patient populations. Five RCTs reported on the usability of the interventions; one RCT found that the participants’ attitude towards computers at baseline positively correlated with greater website use for parents in the behaviour group. However, with respect to helpfulness, the therapists’ telephone calls were endorsed more frequently than the website on one RCT.

Four studies examined cost effectiveness, including labour costs, resource utilisation, health insurance and societal costs; no studies evaluated cost effectiveness from the participants’ perspective. Individual studies found significant savings with respect to: resource utilisation (therapist’s time); emergency room visits (savings of US dollar $907.10 per child versus $291.40); labour costs ($6.66 per treatment student); and cost-benefit ratio (adding an internet-based unit to an education programme improved ratio from 0.55 to 0.79 in one study and from 1.07 to 1.42 in another study of moderate or severe asthma).

There were the beginnings of an evidence base that internet-based self management interventions improved selected outcomes in certain childhood illnesses.

The review addressed a well-defined question in terms of participants, interventions, study design and relevant outcomes. Relevant databases were searched in English and unpublished studies were not eligible; therefore, the review was at risk of publication and language biases (publication bias was not assessed). Study quality was assessed using suitable criteria and the included studies identified were only of fair quality. Efforts were made to reduce error and bias throughout the review. Some relevant study details were reported, but minimal details of outcomes were reported and there were some errors; significant effects described did not correspond between the tables and text. A narrative synthesis was provided due to the heterogeneity of the results and wide variability of treatment programmes and health conditions. The review was relatively well conducted, but the authors’ conclusions were based on a relatively small number of studies of only fair quality of children with differing health conditions. Although the authors’ conclusions are conservative, their reliability is unclear.

Practice: The authors stated that a combination of enhanced self-efficacy with motivation and knowledge enabled users to change health behaviours, leading to changes in clinical outcomes.

Research: The authors identified a need for high-quality studies with rigorous study designs and adequate sample sizes. Studies were needed to be of children in both clinical and community environments and with a wide range of health conditions in order to elucidate the best types of internet interventions and modes of delivery. Studies should conduct cost-benefit analyses compared to standard approaches. A consensus should be developed not only for development and testing of internet interventions to ensure usability and feasibility but which also considers how to target relevant populations and keep their engagement. The authors encouraged uniformity of assessment intervals in order to compare different interventions and identified a need for research on how internet self-management differed from traditional self-management interventions.

Two reviewers were supported by Canadian Institute for Health Research Post-Doctoral Fellowships and one reviewer had a Summer Research Studentship from the Department of Anaesthesia and Pain Medicine, The Hospital for Sick Children, Toronto, Canada. There were no conflicts of interest.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.