Hello Monali,

EPPI-Reviewer was originally designed to assist us in capturing detailed information about the studies in our in-house reviews and presenting that collected information in many different ways. In subsequent versions of the program we have kept this focus on data extraction as it is a major part of any systematic review.

One of the important elements of data extraction is the design and creation of your data extraction tool. When setting up your tool you must be sure that you are asking the correct questions that will help you collect the information from the studies that you require. When setting up these questions and answers in EPPI-Reviewer you can specify whether these codetypes are selectable and allow textual responses or are codetypes for collecting numeric outcome data. You will find this in the user manual under ‘Setting up the screening / coding tools’. You can also find descriptions of some of the data extraction tools we use on our own systematic reviews on the EPPI-Centre website (http://eppi.ioe.ac.uk).

If the data you are collecting is numeric please see the section called 'Meta-analysis' in the user manual.

EPPI-Reviewer can handle either single or double data extraction (similar to single or double screening) as it is all treated as a form of coding.

If you are carrying out double data extraction (two people data extracting the same study) a simple agreement or disagreement comparison as used in screening is usually not sufficient as data extraction tools are normally much more involved than screening tools. I have attached a document that I hope to incorporate into the user manual in the near future. It describes ‘Independent multi-user data extraction’ and how you would do this in EPPI-Reviewer. As well, I hope to create an instructional video in the next few weeks on data extraction in EPPI-Reviewer.

Once you have collected the detailed informaiton from your studies EPPI-Reviewer has a number of searching and reporting functions to help query and present that information. You will find details on this in the user manual under 'Analysis of your data'.

Best regards,

Jeff

Hello Monali,

When you are creating the child code you should set the code type as 'Selectable (show checkbox)'. This means that in the Document details window (where you carry out your coding) you will see a button labelled 'Info' next to the code. If you click on that button a window will open where you can enter textual information such as author or year.

Best regards,

Jeff

Hi Jeff,

A few papers that we are going to extract the data from are:

1) Acosta-Rojas, E.R., et al., Association Between Visual Impairment and Patient-Reported Visual Disability at Different Stages of Cataract Surgery. Ophthalmic Epidemiology, 2006. 13(5): p. 299-307.

2) Alió, J.L., D.P. Piñero, and A.B. Plaza-Puche, Visual outcomes and optical performance with a monofocal intraocular lens and a new-generation single-optic accommodating intraocular lens. Journal of Cataract & Refractive Surgery. 36(10): p. 1656-1664.

3) Arshinoff, S.A., Y.N.J. Strube, and R. Yagev, Simultaneous bilateral cataract surgery. Journal of Cataract & Refractive Surgery, 2003. 29(7): p. 1281-1291.

one papers compare 2 lens and other one is purely on Simultaneous bilateral cataract surgery.

My questions is: for patient baseline (before surgery) characteristics which are different from after surgery characteristics (outcomes). In data extraction there is only outcome, intervention, and outcome classification. How to input baseline (before surgery) characteristics fo 2 treatements of cataract surgery. (intervention)?

Monali

 Hi Jeff,

Could I please get a copy of the attachment you mention in this post about data extraction, and can you please tell me if there is an instructional video available on data extraction yet?

Thanks

Vivian Welch, Canada

Hi Jeff,

I wondered if you could also email me the document that describes ‘Independent multi-user data extraction’ and how I can do this in EPPI-Reviewer. We are hoping to use this approach in our systematic review.

THanks

Nada