Critical appraisal is the careful analysis of studies to determine their relative value.
EXAMPLE:
Researching your systematic review, you find one rigorous randomized controlled trial (RCT) which concludes that exercise alone had a positive impact on heart disease, and three badly conducted studies saying exercise had no impact. Since you have not appraised the articles, you conclude in your review that the evidence shows exercise has no impact on heart disease.
This systematic review would be misleading and inaccurate.
To be unbiased in your systematic review, assess the bias of the studies you find.
Bias means anything that skews the study results away from the purely scientific. It can refer both to active personal bias (on the part of any person involved), and to passive elements of the study itself that could determine the outcomes in other ways than intended.
The major types of bias, according to the Cochrane Handbook, are:
Selection bias: Bias in how participants of the study are selected or self-select.
Performance bias: Bias in how different groups are treated or perform.
Detection bias: Bias across groups in how the researchers assess and determine the study outcomes and in how participants affect these outcomes.
Attrition bias: Bias across groups in participant withdrawals from the study.
Reporting bias: Bias in how the information is conveyed in the final document, especially in terms of which outcomes are included and which are not.
Source: “Books – Spines Up,” CCAC North Library 
For more information, see Chapter 7 of the Cochrane Handbook. Most of these types apply only to the studies analyzed in your systematic review, but it is also important to avoid reporting bias in your final review.
How to Read a Paper
Finding and Evaluating Evidence
Making Sense of Critical Appraisal
Facebook
Instagram
Twitter
YouTube