Assessing the completeness and comparability of outcomes in systematic reviews addressing food availability




Poster session 3


Tuesday 18 September 2018 - 12:30 to 14:00

All authors in correct order:

Visser M1, Kredo T2, Durao S2, Saldanha I3
1 Centre for Evidence-Based Health Care, Division of Epidemiology and Biostatistics, Faculty of Medicine and Health Sciences, Stellenbosch University, South Africa
2 Cochrane South Africa, South African Medical Research Council, South Africa
3 School of Public Health, Brown University, USA
Presenting author and contact person

Presenting author:

Solange Durao

Contact person:

Abstract text
Systematic reviews (SRs) provide evidence to inform healthcare decision-making. All outcomes important for patients, clinicians and policy-makers should be prespecified in SRs to avoid reporting bias. Furthermore, completeness of outcome specification may influence the utility of a SR for decision-makers. A framework to assess outcome completeness has been reported, but has not been used in SRs of nutrition interventions.

To assess the completeness of outcome prespecification in SRs of interventions to improve food availability in low- and middle-income countries, and to assess comparability of outcome domains across reviews.

We searched the literature to identify SRs of interventions addressing food security (February 2015); and selected reviews of interventions to increase food availability, such as agricultural, fortification, or food provision. We sought the original protocols of eligible reviews, or used the Methods section of the SR for data extraction if the original protocol was not available. We identified key outcome domains reported in at least 20% of included SRs. We will assess the completeness of each outcome domain, in duplicate, according to five elements:
1) domain/title;
2) specific measurement;
3) specific metric;
4) method of aggregation; and
5) time-points.
We will use descriptive statistics to assess the extent of outcome completeness, based on how many outcome elements are specified, and comparability in how individual elements are described across reviews reporting the same outcome domains.

Our search retrieved 2591 records. Of these, we selected 57 SRs, 22 of which addressed interventions to increase food availability. We identified key outcome domains in these SRs as anthropometry (14/22; 63.6%); biochemical (16/22; 72.7%); developmental and cognitive (9/22; 40.9%), and dietary intake (5/22; 22.7%). We will present the assessment of completeness of outcome prespecification, as well as the comparability of outcome elements in each domain across these SRs.

This study will give an indication of the usefulness of outcome reporting in SRs addressing food availability interventions, specifically for end users who need to compare outcomes across reviews to make policy decisions.

Healthcare consumer involvement:
Improved outcome reporting relevant to consumers.


Relevance to patients and consumers: 

This study addresses the issue of how well researcher/s specify the measurements or outcomes they set out to measure before a study is conducted. This process is important since it ensures that study results are well described for individual studies and are relevant to consumers. It also applies to authors of systematic reviews (SRs). This analysis looks at how well authors pre-specify the outcomes of SRs summarizing the effects of interventions to improve the availability of food in low- and middle-income countries, such as agriculture, food fortification and food provision. Are these outcomes comparable across the different SRs ?