Selection Criteria for choosing articles from BNI and AMED, to include in Healthinfo4u site.

General criteria used during the project:

1. Select articles from 2001/2 using either the BNI/Amed heading or free text search as indicated on the master headings list.

2. From this selection then hand select those articles to remain in Healthinfo4u, according to criteria below.

3. Where a subject heading is available for a database, a search by that subject heading should be sufficient. If a subject heading does not adequately cover a subject, that should be indicated on the master headings list and a mixture of subjects and keywords may be necessary. If the results of a subject search number less than 5, consider using keywords to get a bigger result before extending your search to years previous to 2001. Also use keywords if your professional judgement suggests that more results should have been found. Never go back beyond 1999 even if fewer than 5 records found.

Exclusions:

1. Any articles referring to health service, hospital or ward/clinic administration

2. Any articles concerning training or organisation of nurses/other staff

3. All editorials, unless an overriding reason to include

4. Any articles that are controversial or clearly an opinion, and unsupported by any evidence base

5. Any articles so technical to be incomprehensible to the average lay reader

6. Any articles published before 2001 unless a search on 2001 output retrieves less than 5 articles

7. Any articles published earlier than 1999.

Inclusions from 2001 & 2002:

1. All articles that describe the clinical condition, its causes and treatment

2. All articles that contribute to the concept of improved information about living with, coping with a condition

3. All articles that include any patient information

4. All articles that are likely to be understandable by the lay public who can read newspaper editorials in the popular press

5. All articles that discuss in a balanced way new or experimental treatments provided they are well grounded.

6. Articles that are harder to understand if they contain valuable contribution to the knowledge base of chronic clinical conditions ( ie suitable for the Expert patient)

7. Articles that provide information that may be relevant to minority communities especially Ethnic minorities

Search Examples:

A. Examples of a basic subject search on Arthritis (simplified by the fact that the same subject heading is used in both BNI and AMED):

1. Type in arthritis/ and limit the publication year fields to 2001 onwards. Click Perform Search.
2. 45 results appear. Click the details button, and you will see that 8 results are from AMED and 37 from BNI.
3. Click Remove Duplicates. You will be asked for your field and database preferences. Select "No Field Preference" and ensure that BNI is the first database preference (this means that any duplicates would be removed from AMED rather than from BNI). Click Continue.
4. In this case, no duplicates were detected.
5. Move down the screen to view the results.
6. Customise Display allows you to select the fields you wish to view.
7. Select the references referring to the Selection Criteria.

B. Example of a subject search on Cancer (in BNI) or Neoplasms (in AMED):

1. cancer/
2. neoplasms/
3. 1 or 2 (limiting to 2001 onwards at this point)
4. Continue as points 3-7 above.

C. Examples of a subject search on Cervical Smears:

Refer to OVID's Quick Reference card - the Command Line Syntax section will give you the symbols for truncation, adjacency etc. Note also that you can limit your truncation by e.g. $3 so that only up to 3 letters will appear after symbol.
1. cervi$3 adj3 (neoplas$ or cancer$ or malignan$)
2. diagnosis/
3. smear$1 or cytolog$2
4. 2 or 3
5. 1 and 4 (limiting to 2001 onwards at this point)
6. Continue as points 3-7 above.

Tips:

When working out your search strategy, assume that you will have to search one database first and then the other, even though both are loaded. For example, carry out a suitable search for BNI first and then carry out a search for AMED, in each case using subjects and keywords as appropriate to that database. If the subject heading is the same for both, or if a keyword approach will be needed for both, this will not be necessary.

Further Information:

1. The results will be further scrutinised by experts before being added to the database
2. We expect circa 30 records for popular topics and as few as 5 for rarer headings to result.
3. It is estimated it will take approximately 7 minutes per topic to arrive at the results.
4. These instructions will be amended for the monthly update process we will subsequently undertake.