The Scottish Government

« Previous | Contents | Next »

Listen

Methodology

This is a summary of our methodology (see also the Discussion section in this report for thoughts on the search process, its benefits, challenges and limitations). The searching was restricted to between 1990 and summer 2005. We identified the key resources to search, devised a search strategy and conducted the comprehensive, international literature search. Identified references were imported or entered into a computer-based library software package, Reference Manager.

The bulk of the searching focused on major electronic databases. Librarian support at the University of Bath advised us to focus on PsychInfo, Web of Science, EMBASE and PubMed (Medline). We employed a dual strategy, to ensure maximum coverage, combining use of 'free-text' terms with pre-determined MeSH headings (where appropriate). CINAHL and SIGLE were also searched in this way. Computer-based searches, or 'eyeballing' of reference lists and abstracts, were also undertaken. These included: Addiction Abstracts (from 2001), the Cochrane Collaboration, the Campbell Collaboration, ETOH, the National Research Register, Project CORK, Drug Misuse Information Scotland, Caredata, and publications lists of the UK Drug Strategy Directorate Publication List, the CDMR (Centre for Drug Misuse Research in Glasgow) and Bancroft and colleagues. Additionally, the Scottish Executive's Substance Misuse Research Team provided a reference list from its database, together with a collation of further thoughts and suggestions from the Advisory Group. E-mails were sent to individuals suggested by the Advisory Group, including to the DAT Association for Scotland, and to the ENCARE network (European network for children affected by risky environments within the family). In addition, one of the research team ( LT) was testing the search strategy for another project, a systematic review, and anything new to arise from this was included. The research team also subscribe to various journals and news alert services (for example, Daily Dose, NSPCC, Drink and Drug News) and were therefore aware of any newly released, relevant, literature.

It is important to highlight that the core aim of the review was to scope the literature base on the topic of parental substance misuse, rather than undertake a systematic or comprehensive in-depth literature review of all that we identified. The amount of literature identified would thus influence how we proceeded to summarise themes and gaps within that literature. In this case, given the time available to us for the review, and that our comprehensive literature search identified several thousand articles, it was not possible for us to review the literature in depth. After discussion with the Advisory Group, it was agreed that we would scope at a broader level, as opposed to using strict quality assessment criteria that would have meant we read a very small amount of the literature in more detail. We therefore employed the following three level strategy:

Scoping Level One: To quantitatively scope the database in relation to the study objectives, we divided a printed, paper copy of the database contents (titles and abstracts) between the four of us and used a checklist to tally the references, corresponding to the key objectives of the study and areas of importance agreed following discussion with the study Advisory Group (for example, impact, domestic violence, risk, resilience). Several thousand references were screened in this way (Table One). This process allowed us to screen for relevance, remove duplicates and references that were not relevant to the objectives of the review, and to identify areas for more detailed review.

Scoping Level Two: Given the large number of relevant references that remained in the database following level one, and the time available to us, we felt that it would be most useful to review in more detail a selection of key areas identified in the study objectives, with the agreement of the study's Advisory Group. We decided at this stage not to look at Risks and Impact, as the literature search in both areas had resulted in a large number of references. The research team felt judged that it was unlikely that many news areas for future research would therefore be identified on these subjects. We thus selected seven areas for further review, as it seemed likely that reviewing these areas would lead to both more information on effective practice, and priorities for future work. A selection of abstracts (or all abstracts if there was a limited number to choose from) were selected and reviewed, and a short report was written on each topic. The seven areas were:

1. Child Protection and Domestic Violence.
2. Resilience.
3. Pregnancy, Motherhood and FASD (fetal alcohol spectrum disorder).
4. Service needs.
5. Fathers ¹.
6. Parental mental health.
7. Children's views.

Scoping Level Three: In discussion with the Advisory Group, two areas were identified as warranting more detailed review. The two areas were Prevalence and Interventions. This entailed reviewing a larger number of abstracts as well as selected full articles and reports.

In order to ensure a good spread of literature in the selection of abstracts, articles and reports for more detailed review, we attempted to use the following criteria when identifying abstracts and literature for review under scoping levels two and three: Scottish work; not all USA work (though this was very hard, given the limited amount of non- US literature); more recent work; reviews, systematic reviews or meta-analyses; a range of study designs and methodologies ( RCT, intervention, evaluation and qualitative); and a split between alcohol and drugs. However, given the time available and the amount of literature identified, a lesser emphasis was placed on quality analysis and an in-depth review of the data and a greater emphasis placed on continuing to scope the literature. The themes and gaps that we have identified, and the recommendations that we make, allow for detailed literature reviews on particular issues to be conducted, based on this broader scoping review. This would also allow for confirmation of the findings that we have reviewed and the conclusions that we have made from scoping such a large amount of literature.

Table One - Quantitative Summary of Database References

« Previous | Contents | Next »