Using evidence within HIA
Searching for existing evidence and conducting a review
A specific review of existing evidence is carried out as part of some HIAs. However the review may not be systematic, because there is a risk that the reviewer may (consciously or unconsiously) be looking for articles that support pre-conceived ideas or a particular viewpoint. This risk is present in all fields of scientific endeavour, not just HIA. Presenting one-sided evidence jeopardises a key value of HIA: the ethical use of evidence.
Key steps in undertaking an objective and methodological review are:
- Using a standard approach to guide the process
- Being systematic
- Focusing on a specific question
- Assessing the quality of the studies to be included in the review
- Analysing and combining findings
- Presenting clear conclusions and recommendations
Using a standard approach to guide the process
Using a standard approach to guide one's review can be highly useful, examples can be found at the Cochrane Collaboration or the National Health and Medical Research Council. Such approaches discuss issues such as creating a set of inclusion and exclusion criteria before starting the search, to reduce the temptation to exclude studies that may not agree with preconceived ideas.
To reduce the risk of presenting only one side of the evidence, it is important to be systematic. Developing a systematic search strategy is not always easy, and getting help from a person with library and searching skills is recommended. Deciding what electronic databases to search, the timeframe for the search, the keywords to be used, obtaining software to manage the abstracts, ordering and retrieving documents, are all issues where a trained person may help. An example is the "Systematic Review of Housing Improvement and Health", that contains the issues described above.
Full systematic reviews consume a lot of time and resources. Even when time is short, searches should be systematic and transparent – setting very tight inclusion and exclusion criteria help make it manageable.
Focusing on a specific question
There is an abundance of information that can be retrieved, and it is easy to become bogged down by setting a general question that touches on too many components. Setting a tightly focused question allows the reviewer to eliminate peripheral evidence, and helps keep the process manageable. Common ways of narrowing down a question are being very specific in defining the topic, or concentrating on specific target groups or geographical locations.
Assessing the quality of the studies included in the review
Assessing and combining the information in the articles identified during the literature search can be quite a challenge. However, quality appraisal of the studies is essential in order to determine the strength of the conclusions of each article or series of articles. Usually, a review presents different types of data, drawn from different study types that help to answer the research question. The most common way to deal with data from different study types is to present each piece of evidence separately and discuss the limitations of each. Methods for the quality assessment of studies are available at the Cochrane Collaboration or the National Health and Medical Research Council.
Analysing and combining findings
The methods used to analyse the findings should be identified. For example, some authors may use meta-analyses, and these should be discussed in depth. Others use methods such as content analysis, where concepts are identified and categorised on a variety of levels to identify key themes. Whatever method is used, it is always useful to base a new analysis on a proven methodological approach (the haphazard assembly of data should be avoided). Any gaps in available evidence should be discussed. Combining findings is discussed further in "Bringing together different types of evidence".
Presenting clear conclusions and recommendations
The development of recommendations is as important as the identification and assessment of impacts, therefore sufficient resources must be set aside for this task within the HIA. Recommendations should focus on the most important health impacts and on options to prevent or mitigate negative effects and enhance positive effects.
The choice of preventive, mitigation or enhancement options is likely to be a function of the social and financial significance of risks, their manageability, and the availability of resources. Decisions to actually enact these options will also be driven by the consensus obtained around them. Key to generating this consensus is early and regular communication with decision-makers and affected communities.