This assignment requires you to carry out additional work on the review of nutritional supplementation for older people that you have been working on throughout the course and to write a report of the results as detailed below:
1. Please download the following two articles:
- Beelen J, de Roos NM, de Groot LCPGM. A 12-week intervention with protein-enriched foods and drinks improved protein intake but not physical performance of older patients during the first 6 months after hospital release: a randomised controlled trial. British Journal of Nutrition 2017, 117, 1541-1549.
- Deutz NE, Matheson EM, Matarese LE, Luo M, Baggs GE, Nelson JL, Hegazi RA, Tappenden KA, Ziegler TR, on behalf of the NOURISH Study Group. Readmission and mortality in malnourished, older, hospitalized adults treated with a specialized oral nutritional supplement: a randomized clinical trial. Clinical Nutrition 2016, 35, 18-26.
2. Using the Cochrane risk of bias tool, assess the quality of conduct of these two studies, as reported, and construct a table providing your assessment together with reasons for your answers. For blinding of outcome assessment and incomplete outcome data, you should make separate assessments for mortality and weight change (use the categories in the template below).
3. Extract relevant data for these two studies for the outcomes:
- Mortality at six months
- Weight change (kg) at six months
If these time points are not available, you should use the data for the closest available time to six months, even if this is of short duration. You are not required to complete the full data extraction tables for these studies.
There may be more than one approach to extracting data and including in a meta-analysis. In some situations you may wish to refer to sections of the Cochrane Handbook for Systematic Reviews of Interventions in order to justify the approach you have used.
- Data may need to be extracted or estimated from various parts of a paper including the text, tables and figures
- If a study reports standard errors (SE or SEM), these may be converted to standard deviations (SD) using a formula (see Cochrane Handbook 7.7.3.2 or lecture slides)
- If no standard deviations are reported, these may need to be estimated from other studies in the review
- Most studies in this review only provided the mean change in weight from baseline to follow-up but some may provide mean weight at follow-up instead. If there are a mixture of studies including change scores and final scores it is acceptable to include these within the same meta-analysis but there are other approaches available. For more information see the
following sections of the Cochrane Handbook (7.7.3.1, 9.4.5.2, and the first three paragraphs of 16.1.3.2).
- A least squares mean is a mean that has been adjusted for covariates and is OK to include.
4. Add the results of the Beelen and Deutz studies to the two meta-analyses you set up in Revman during the meta-analysis practical. For each outcome, you should now have data for up to seven studies.
5. Assuming that these seven studies were the only ones found to be eligible for inclusion in the systematic review of nutritional supplementation for older people, write a short report (maximum 1000 words excluding tables and figures). The report should only include the items listed below; you should not include other aspects such as background to the topic, search strategies, characteristics of included studies or references. For each item you can see the proportion of the total mark.
a) Brief methods for the steps you carried out in 2-4 above with justification for the meta- analysis methods used.
b) Quality assessment for each of the two studies (Beelen and Deutz) comprising risk of bias tables. Give reasons for your answers.
c) Data extraction tables for weight change and mortality. You may include comments/footnotes if appropriate.
d) Forest plots for the updated meta-analyses.
e) Description of the results of the meta-analyses and conclusions. What is the overall evidence for the benefit or harm of this intervention? Comment on the reliability of the results.