Home‐based support for socially disadvantaged mothers
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
In keeping with the original review, this revised and updated review is designed to assess the impact, on maternal and infant health, and on infant development, of programmes offering additional home‐based support for women who have recently given birth and who are socially or economically disadvantaged, e.g. because they are poor, young, inner city, and/or single parents.
The further objective of this review is to update the earlier Cochrane review (Hodnett 1997), to account both for more recent publications and also for changes in methodological and editorial practice.
Background
Epidemiologic studies indicate that babies born to socio‐economically disadvantaged mothers are at higher risk of injury, abuse and neglect, health problems in infancy, and are less likely to utilise child health services regularly. Home visitation programs have long been advocated as a strategy for improving the health of disadvantaged children. For example, in Britain, home visitation by health visitors is considered to play a key role in child accident prevention by way of the advice provided on child development and home safety (Hodnett 1997) .
Over the last two decades, a number of home‐visiting programmes have been developed in the USA with the aim of improving parenting practices in disadvantaged families. These are increasingly being used outside the USA, in the UK and elsewhere. Such programmes typically involve structured visits by professional or trained lay people experienced in child health and development, after birth and in subsequent months. Programme evaluations and primary studies (albeit of varying quality) suggest a range of benefits from home‐visiting programmes to both infant and maternal health including child abuse and neglect (Hardy 1989; Olds 1986), numbers of emergency room visits, (Olds 1986) accidents (Olds 1986; Larson 1980b) and hospitalisations (Hardy 1989; Johnson 1993). There is some evidence for improvement of parental knowledge of child rearing and child development (Taylor 1988), home safety practices (Gielen 2002) and on immunisation rates (Barnes‐Boyd 2001). There is also evidence of effectiveness of home visiting programmes on maternal life‐course including the number of subsequent pregnancies, uptake of welfare and of employment (Olds 1993), and on maternal psychosocial health (Brown 1997; Seeley 1996).
A number of systematic reviews have been published during the last decade, synthesising evidence of the effectiveness of home‐based support programmes, more usually known as 'home visiting programmes' (e.g. Ciliska 2001; Elkan 2000; Guterman 1997; Heinicke 1988; Kendrick 2000; MacLoed 2000; MacMillan 2000; MacMillan 1994; Roberts 1996; Wekerle 1993). The strength of ongoing research and practice interest in the effectiveness of these programmes is testament to their widespread use. The published reviews include studies that consider a wide range of outcomes including parenting, maternal adjustment, parental attitudes and behaviour, child development, child abuse prevention, child abuse potential, maternal and infant physical and mental health, parental knowledge and perceptions, child abuse and neglect, child injury, out of home placement, uptake of health services, maternal psychological health and esteem, and breast feeding.
Whilst there are more studies evaluating home visiting programmes published regarding programmes offered in the developed world, there is also some evidence of increasing interest in early intervention programmes in the developing world (Bolam 1998; Cooper 2002; Hezekiah 1993; Jingheng 1994). Whilst these studies do not describe home visiting programmes as such, they reflect a concern for outcomes in early intervention such as infant care, immunisation levels, infant feeding practices, maternal mood, mother‐infant relationship, infant growth, access to pre and post‐natal care, disease prevention and breast feeding. These outcomes are similar, and in some instances identical, to outcomes measured in some home visiting programmes. Recent publications that describe home visiting programmes in the developing world, in Africa particularly, are evidence that such interventions are also being used outside the developed world (Brugha 1996; Ransjo‐Arvidson 1998). Evaluating the effectiveness of home‐visiting interventions is thus of potential relevance to both providers and consumers of health care, in all parts of the world.
This review will be the first update to the earlier Cochrane review 'Home‐based social support for socially‐disadvantaged mothers' (Hodnett 1997). To accommodate developments in editorial and methodological practice for systematic reviews since the original publication (completed within the Pregnancy and Childbirth Group), this protocol is being substantially rewritten.
For the purposes of this review the term 'Home visiting programmes' will be used throughout.
Objectives
In keeping with the original review, this revised and updated review is designed to assess the impact, on maternal and infant health, and on infant development, of programmes offering additional home‐based support for women who have recently given birth and who are socially or economically disadvantaged, e.g. because they are poor, young, inner city, and/or single parents.
The further objective of this review is to update the earlier Cochrane review (Hodnett 1997), to account both for more recent publications and also for changes in methodological and editorial practice.
Methods
Criteria for considering studies for this review
Types of studies
Trials will be eligible for inclusion in the review if the assignment of study participants to the intervention or control group is random or quasi‐random (for example, alternate record numbers).
Types of participants
Disadvantaged families including mothers with infants. For the purposes of this review, infants are defined as children of less than 12 months of age including but not exclusively, new‐borns. The term 'disadvantaged' includes low income parents, teenage parents, single parents (due to the fact that they may be living in circumstances of economic deprivation, and/or social isolation), educationally disadvantaged parents e.g. low literacy, and those in public housing whether located in urban or rural areas. Studies will be included where 70% or more of mothers in the study are in these categories. Studies that recruit 'all' mothers, for example in a particular time period of delivery, will be excluded, unless the study report specifically notes that 70% fulfil the criteria for disadvantage of this review, or data for disadvantaged participants are separately reported e.g. by using separate tables.
Types of interventions
Home‐based maternal support by way of one or more postnatal home visits by either professional or specially trained lay providers. Studies that evaluate the standard service offered universally e.g. health visiting in the UK, will be excluded. Studies will be included where the intervention begins in the ante‐natal stage. However, where the intervention is limited to the ante‐natal period, the studies will be excluded
Types of outcome measures
Child outcomes: accidental or non‐accidental injury, failure to thrive, abuse and neglect, child immunizations, failure to keep scheduled clinic appointments, emergency ward visits, severe monilial diaper rash, hospital admissions, measures of infant development and developmental stimulation activities e.g. relevant domains of the HOME (Home Observation for Measurement of the Environment) inventory.
Maternal outcomes: repeat pregnancy within 18 months, school completion (in the case of adolescent mothers), breast feeding, maternal psychosocial health measures, realistic/unrealistic expectations of child development/behaviour e.g. AAPI (Adult‐Adolescent Parenting Inventory).
As the outcomes of interest are post‐natal, studies that include only birth outcomes will be excluded.
Rating scales. A wide range of instruments are available in this area of research. These instruments vary in quality and validity. For this analysis the minimum standards for the inclusion of data from outcome instruments will be i) that the psychometric properties of the instrument have been described in a peer‐reviewed journal; ii) that the instrument was either (a) a self‐report, or (b) completed by an independent rater or relative.
Search methods for identification of studies
The following search strategy will be used to search the Cochrane Controlled Trials Register, MEDLINE, EMBASE, CINAHL, PsychINFO, ASSIA, National Research Register, LILACS, and Sociofile. Grey literature will also be searched via ZETOC, Dissertation Abstracts A and SIGLE. Where necessary, the search terms will be modified to suit the requirements of particular databases.
((POSTNATAL or POST‐NATAL) OR (POST next NATAL))
((POSTPARTUM or POST‐PARTUM) OR (POST next PARTUM))
((POSTBIRTH or POST‐BIRTH) OR (POST next BIRTH))
((#1 or #2) or #3)
COMMUNITY‐HEALTH‐SERVICES*:ME
HOME‐NURSING*:ME
SOCIAL‐SUPPORT*:ME
(PARENT* near SUPPORT*)
(NURS* near HOME)
(HOME near VISIT*)
(HEALTH near VISIT*)
(DOMICIL* near NURS*)
(PUBLIC‐HEALTH near NURS*)
(FAMIL* near SUPPORT*)
(FAMIL* near INTERVENTION*)
(SOCIAL next SUPPORT*)
(HOME near SUPPORT*)
(COMMUNITY‐BASED or (COMMUNITY next BASED))
(((((((((((((#5 or #6) or #7) or #8) or #9) or #10) or #11) or #12) or #13) or #14) or #15) or #16) or #17) or #18)
MATERNAL‐WELFARE*:ME
MATERNAL‐BEHAVIOR*:ME
MOTHER‐CHILD‐RELATIONS*:ME
CHILD‐ABUSE*:ME
(MATERNAL near WEFARE)
(MATERNAL near HEALTH)
(MOTHER* near WELFARE)
(MOTHER* near HEALTH*)
(MATERNAL near (BEHAVIOR* or BEHAVIOUR*))
(CHILD* near ((ABUS* or INJUR*) or NEGLECT*))
(CHILD* near INJUR*)
(CHILD* near NEGLECT*)
(CHILD near MALTREAT*)
CHILD‐REARING*:ME
PARENTING*:ME
PARENTING
CHILD‐CARE*:ME
(CHILD* near (CARE or CARING))
CHILD‐DEVELOPMENT*:ME
(CHILD* next DEVELOP*)
(COGNITIVE* near DEVELOP*)
(SOCIAL near DEVELOP*)
(EMOTION* near DEVELOP*)
(LANGUAGE near DEVELOP*)
ATTACH*
FAILURE‐TO‐THRIVE
(FAILURE near THRIVE)
((((((((((((((((((((((((((#20 or #21) or #22) or #23) or #24) or #25) or #26) or #27) or #28) or #29) or #30) or #31) or #32) or #33) or #34) or #35) or #36) or #37) or #38) or #39) or #40) or #41) or #42) or #43) or #44) or #45) or #46)
POVERTY*:ME
POOR
(SOCIAL* next DISADVAN*)
DISADVANTAGED
(LOW next INCOME*)
(SOCIAL next PROBLEM*)
IMPOVER*
SOCIO‐ECONOMIC
INDIGEN*
(((((((#48 or #49) or #50) or #51) or #52) or #53) or #54) or #55) OR #56) (((#4 and #19) and #47) and #57)
In addition, the reference lists of published reviews will be screened for potentially relevant studies. All the studies from the original review will be reviewed to consider whether they meet the criteria for inclusion in the light of the new protocol.
Data collection and analysis
Selection of Trials
Titles identified though searches of electronic databases will be reviewed to determine whether they meet the inclusion criteria. This process will also apply to titles identified from the references of previously published systematic reviews. Titles and abstracts will reviewed by two reviewers (EC and JP). Full copies of those which appear to meet the inclusion criteria, together with all the studies from the original review (Hodnett 1997) will be obtained and assessed by two reviewers (EC and JP). Uncertainties concerning the appropriateness of studies for inclusion in the review that cannot be resolved by consensus discussion, will be resolved through discussion with the editorial base.
Quality Assessment
Two reviewers (EC and JP) will independently assign each included study to quality categories according to their degree of concealment of allocation to intervention groups, as described in the Cochrane Collaboration Handbook (Clarke 2002). These are as follows:
(A) indicates adequate concealment of allocation (e.g. by sealed envelopes or telephone randomisation)
(B) indicates uncertainty about whether the allocation was adequately concealed (possibly where the method of allocation concealment is unknown)
(C) indicates that the allocation was definitely not adequately concealed (e.g. open random number lists or quasi‐randomisation methods).
The reviewers will also consider the degree to which those assessing outcomes are blind to the allocation status of participants. In an intervention such as a home‐visiting intervention, it is not possible to blind participants to allocation, and so this will not be assessed. We recognise that it is unusual for assessors to be blind in community based psychosocial interventions, as has been noted elsewhere (Barlow 2000). However, this criterion will be rated as 'met', 'unmet', or 'unclear' as described in the Cochrane Handbook (Clarke 2002).
Data Management
Data will be extracted independently by the two reviewers using a data extraction form. The data will be entered into RevMan 4.1 by one reviewer (JP) and checked for accuracy with reference to the primary studies by the second reviewer (EC).
Missing Data
In the first instance, authors will be contacted to supply data missing from included studies. Missing data and drop‐outs/attrition will be assessed for each included study, and the extent to which the results/conclusions of the review could be altered by the missing data will be assessed and discussed.
Where data necessary to meta‐analysis are not retrievable, the available data will be included in a textual discussion, and the study excluded from meta‐analysis.
Measurements of Treatment Effect
Where possible, data will be managed in a way that provides results and conclusions that are accessible to consumers and service providers who do not necessarily have a scientific or statistical educational background. For this reason, where there are dichotomous data, relative risk will be used to summarise the data rather than odds ratio, as the latter is conceptually more inaccessible. Continuous data will be reported as mean differences (see below: 'Data Synthesis').
Investigation/Assessment of Heterogeneity
Heterogeneity will be assessed using the Chi‐squared test of heterogeneity as well as visual inspection of the graph. Clinical heterogeneity will be discussed in the text of the review.
Data Synthesis
If studies are homogenous, a fixed effects model will be used in meta‐analysis. In the event of heterogeneity, a random effects meta‐analysis will be undertaken. Where there is clinical heterogeneity, as may be anticipated with a review such as this, that potentially includes similar interventions provided under varying circumstances and by varying providers, the statistical synthesis of results will use a random effects model.
Previous experience of reviews considering psycho‐social outcomes indicates that in many cases, there may be a number of different measures of the same outcome within a review (Barlow 2000). If this is the case, data will be reported as Standardized mean differences (effect sizes) in order to combine results across scales.
Subgroup analyses
The impact of certain study characteristics will be investigated by means of sub‐group analyses. Depending on the profile of included studies the following sub‐group analyses will be undertaken:
(i) the differential impact of studies which focus on minority ethnic groups (defined as minorities in the country in which the study was conducted);
(ii) the differential impact of studies which include a majority of teenage mothers;
(iii) the differential impact of studies which include a majority of mothers with learning disabilities;
Subgroup analyses will also be conducted in relation to differences in service delivery such as volunteer/professional home visitors, differences in the frequency and intensity of visiting, or services that begin ante‐natally. Subgroup analysis should however be treated with caution, and this will be considered in the textual discussion.
Assessment of Bias
If sufficient studies are found, funnel plots will be drawn to assess the presence of possible publication bias. Whilst funnel plot asymmetry may indicate publication bias, this is not inevitably the case, and possible explanations for any asymmetry found will be considered and discussed in the text of the review.
Sensitivity Analyses / Impact of study quality
Primary analyses will be based on available data from all included studies relevant to the comparisons and outcomes of interest. If there are studies where data appear to be skewed (where the Standard Deviation is greater than the mean), sensitivity analyses will be conducted to consider the effect of such studies on the overall result. If this is a problem, the possibility of including additional tables to report other data e.g. median figures, will be considered.
If there are studies of considerably lower quality than others e.g. where there are high levels of drop‐out among participants (defined as greater than 25%), sensitivity analyses will be considered.
