This well-conducted review assessed the effectiveness and costs of paediatric home care (PHC), including whether and how cost-effectiveness differs between different groups of children. The authors' noted that the evidence base for PHC was weak, and that the review did not provide any clear messages for health services. The authors' made a number of recommendations for future research.

Principally, to systematically review the effectiveness and costs of paediatric home care (PHC), including whether and how cost-effectiveness differs between different groups of children. Other stated objectives were to determine the rate of growth of the evidence base, and what recommendations could be made for further research.

The authors listed 20 electronic databases, publications lists and current research registers that they searched for studies published since 1985. The search dates for each database and the search terms used were reported in full. Additional studies were identified through reference lists, handsearches, personal contact with researchers, and forward citation searching.

Study designs of evaluations included in the review

The predefined criteria were: randomised or pseudo-randomised trials; studies with a health economics element; non-randomised controlled trials of PHC compared with other models.

Specific interventions included in the review

Home care for several conditions usually associated with primary and secondary care (see Participants Included in the Review) were included. The specific therapies implemented at home were intravenous therapies, parenteral and enteral feeding, oxygen therapy, dialysis and nebuliser therapy.

Participants included in the review

The predefined criteria were children aged under 18 years with serious acute or chronic illness. The main conditions and categories identified for the review were: low birth weight (less than 1,500 g) or medically fragile babies; asthma or diabetes; technology-dependent children; and children with mental health problems.

Outcomes assessed in the review

The outcome areas predefined for randomised controlled trials (RCTs) were: mortality; service use; clinical, physical and psychological outcomes; costs; impact on family, social life and education; knowledge of the condition.

The outcome areas predefined for economic studies were: costs tothe health service, family and other agencies; analysis of costs and benefits.

The outcome areas predefined for non-randomised controlledstudies were: clinical outcomes; costs; impact.

How were decisions on the relevance of primary studies made?

Two reviewers independently screened all citations, abstracts, and retrieved papers, working to agreement. Any unresolved disagreements were referred to a third reviewer.

Established methodological quality tools for the evaluation of randomised and non-randomised trials were used. The authors adapted an existing economic evaluation tool to assess studies with an economic component. The quality ratings were not used to exclude studies. One reviewer assessed the quality of the included studies, while a second reviewer checked the assessment.

Data from RCTs and economic studies were extracted directly into computer databases developed for the review. Data from non-randomised studies were entered directly into evidence tables. A second reviewer checked the data before analysis. The data fields included: publication details; details of the intervention or model of care; study details; demographics; quality criteria; mortality; other clinical, social and economic outcomes; and knowledge of the condition.

How were the studies combined?

The analysis was predominantly descriptive. Only two RCTs had a limited number of comparable outcome measures that allowed pooled standard mean differences (SMDs) to be calculated.

How were differences between studies investigated?

Differences between the studies were considered in a narrative summary.

Ten RCTs, 16 economic papers and 14 non-randomised controlled trials were included.

Very low birth weight babies: there was limited reporting of the clinical or developmental outcomes of earlier discharge, accompanied by home care. Physical development was measured in two RCTs using the Bayley Psychomotor Index, and a pooled analysis showed a weak trend towards PHC (pooled SMD 0.238, 95% confidence interval, CI: -0.097, 0.573, P=0.164). Mental development was measured in the same two RCTs using the Bayley Mental Developmental Index, and a pooled analysis showed a strong trend towards PHC (pooled SMD 0.327, 95% CI: -0.009, 0.663, P=0.056). The impact on the family members was rarely reported.

Diabetes and asthma: whether PHC for children with diabetes or asthma affects clinical or social outcomes for children, their families or the health service, remained unsure.

Technology-dependent children: studies of home intravenous therapy, parenteral and enteral nutrition, oxygen therapy, dialysis and nebuliser therapy were identified. Controlled studies were rare, as were studies that measured clinical outcomes and the impact on the families or children's quality of life.

Children with mental health problems: apart from parents' satisfaction with services, few other effects were reported.

Generic paediatric home care: no major clinical effects were evident at early follow-up in the single identified study. Very partial follow-up after 5 years suggested that psychological adjustment may be improved by PHC. Family satisfaction with services was higher with home care, although no direct impact on the children's mothers or on the family was detected.

Very low birth weight babies: PHC may be cheaper than the alternative; three studies showed savings of 12 to 26% over hospital care, but the costing methods used were weak.

Diabetes and asthma: early discharge with home care after diagnosis may reduce the parents' costs, largely by reducing the children's initial length of hospital stay.

Technology-dependent children: PHC for technology-dependent children may be cheaper for the health service.

Children with mental health problems: health service use after home care may be lower, with reductions in health service costs. Admission to residential care may also be lower, with reductions in social care costs.

The authors noted that the evidence base for PHC was weak, as were the methods. Common methodological weaknesses included sample sizes, timing of data collection, objectivity, long-term follow-up, accurate description of PHC models, impact beyond the hospital, and the ages of children researched. Narrow ranges of children and parents (in terms of socioeconomic status, ethnicity and geographical location) were included in the studies, and the children's views were largely absent. The results of the review did not provide any clear messages for health services.

The authors adopted a rigorous protocol for a wide-ranging review. Comprehensive literature searches were unlikely to have omitted many relevant publications. A relevance assessment was performed in parallel, and the data extraction was checked independently before analysis. It was not mentioned whether the quality assessment was also performed in parallel, although this element was not as crucial to the synthesis as in reviews where quality scores are used to exclude studies. The decision to present the results in a narrative, rather than a quantitative synthesis, was dictated by the wide dissimilarities in focus, outcome reporting and quality of the studies. Only two RCTs had some outcome measures that were sufficiently comparable to be pooled, although the authors did not report testing the clinical or statistical heterogeneity of the pooled studies. The authors' conclusions appear to follow from the available evidence.

Practice: The authors did not state any implications for practice.

Research: The authors proposed research topics, including controlled trials, systematic reviews and economic evaluations, in the following areas. Generic and specific PHC: research into generic PHC for very dependent children and their families, across several sites.

Home intravenous therapy: a national survey of current practice; systematic reviews of outcomes using case series; multicentre controlled studies of home versus hospital care for paediatric home intravenous therapy.

Home parenteral and enteral nutrition: a systematic review (updated in the case of parenteral nutrition) using case series.

Home dialysis: non-randomised controlled trials, empirical evaluation, and economic modelling that includes costs falling to other agencies and families.

Diabetes and asthma: high-quality trials of models of home care that explore which children and families would benefit the most.

Fragile babies: research to identify what support the most fragile babies and their families need and, if provided, what benefits it delivers at what cost.

Home oxygen therapy: a national survey to establish current practices and numbers, to ensure adequate sample sizes for subsequent multiple site evaluations; rigorous, well-designed, non-randomised controlled trial research on the effectiveness of different models of care for oxygen-dependent children, the impact that home oxygen therapy has on children and their families, and the ways in which services can enhance positive outcomes.

Nebulisers: research to identify whether children with asthma should have nebulisers at home, rather than using different modes of drug administration, with studies of different age groups; the clinical safety of home nebuliser use for children with cystic fibrosis, concentrating on infection rates; and an evaluation of services or training programmes that enable families to use nebuliser equipment effectively and safely.

Mental health problems: a multicentre RCT of home care controlling for different treatment regimes.

NHS R&D Health Technology Assessment (HTA) Programme, project number 98/05/03.

Parker G, Bhakta P, Lovett C, Olsen R, Paisley S, Turner D. Paediatric home care: a systematic review of randomized trials on costs and effectiveness. J Health Serv Res Policy 2006;11:110-23.

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.