Find the latest information/resources on the Coronavirus (COVID-19)

Searching the Literature for Animal Testing Alternatives: A Tutorial

Tutorial for researching alternatives to animal research

Evaluate Your Search - Quiz!

Dr. X has tried running his literature search on a few databases, and needs to take a moment to evaluate his progress so far. Remember, he is studying the effect of caloric intake on the development of brain tumors in mice.  Below are some of the results of Dr. X's search. How can Dr. X narrow or broaden his search, if necessary?

Select the most correct option by clicking the radio button beside it. Each time you answer a question you will get immediate feedback.

Your answers will not be recorded or used by anyone else, so don't worry about making a mistake. Take the quiz as many times as you like.

Dr. X tried using his keyword lists in PubMed MEDLINE, ALTBIB, and RePORTER, but he did not find very many items that were relevant to his search. What should he do now?

Nothing. If there's no information out there, there's not much he can do about it.

Try Again! It is highly unlikely that Dr. X's search results would be so small, regardless of his study. Since Dr. X has tried three of the best databases for this kind of literature search, it is possible that there is something slightly wrong with his keywords or search question(s).

What other answer offers a strategy that would maximize Dr. X's results while minimizing his time outlay and frustration?

Try some other databases. Maybe he's just not looking in the right places.

Try Again! In general, searching multiple databases is the best way to expand your search results. However, Dr. X has already tried three of the major databases best suited for an animal testing alternatives search. It is likely at this point that there is something slightly wrong with his keywords and/or search strategythat is reducing his results.

What other answer offers a strategy that would maximize Dr. X's results while minimizing his time and frustration?

Contact the Health Sciences Library (by coming to the reference desk, calling or emailing the library, or using the Live Online Help service) to set up a consultation with an expert searcher.

Correct! Librarians at the Health Sciences Library are eager to consult with UNC-CH researchers on literature searches on any topic in the health sciences. We can help you construct keyword lists, develop search questions, suggest good databases to search, and even do the literature searches for you. There are some databases that we can search on a fee-based schedule that are not otherwise available to the campus community. Please contact the library by your favorite route (in person, by phone, or electronically) to set up a time to meet with a reference librarian.

Keep adding terms and tweaking his search strategy until he gets more results.

Try Again! Adding terms and tweaking the search strategy may help Dr. X get better results, especially if he uses the strategies explained earlier in this module. However, it is very easy to spend a lot of time wandering around in a database without finding the information that you need (and are sure is there!)

What other answer offers a strategy that would maximize Dr. X's results while minimizing his time outlay and frustration?

For the "reduction" part of his literature search, Dr. X wants to start by searching PubMed MEDLINE for studies that were similar to the protocol he is planning to use. Which of the following search questions is the best one for him to use?

(mice OR mouse) AND (cancer OR tumor) AND (brain OR cerebellum) AND (calories OR diet) - 56 hits

Try Again! Policy 12 of the US Department of Agriculture suggests that researchers should, at a minimum, look in more than one database, but there is no minimum number required by law.

"Mice"[MeSH] OR "Muridae"[MeSH] AND "Brain Neoplasms"[MeSH] AND "Diet"[MeSH] OR "Energy Intake"[MeSH] - 16984 hits

Correct! The most important aspect of database selection is the balance of subjects and formats searched. The number of databases is not as important as the breadth and depth of the literature search.

("Mice"[MeSH] OR "Muridae"[MeSH]) AND "Brain Neoplasms"[MeSH] AND ("Energy Intake"[MeSH] OR "Diet"[MeSH]) - 30 hits

Try Again! While Dr. X definitely does need to search one or more of the alternatives databases for specialized information on reduction, refinement, and replacement alternatives, he has no requirement to demonstrate results from scholarly and non-scholarly or hard and soft science fields. On the contrary, the balance of databases chosen (and the results found in them) should be Dr. X's main concern.

Any of the above - the MeSH database will automatically map terms and construct his search for him

Try Again! In order to get a sufficiently broad picture of the literature on animal testing alternatives, Dr. X needs to search more than one database.


After working on his keywords and search questions, Dr. X searched PubMed MEDLINE and found a good number of very relevant articles. What should he do now?

He's only gotten started. He has to search at least six other databases to satisfy the requirements in the law.

Try Again! Dr. X is interested, at this point, in casting a wide net for research alternatives in animal testing. He can adapt his literature searching as necessary if he is finding too many results, but the real focus is to find as much information relevant to his topic as possible.

Take a look at the search itself. Is anything missing or superfluous?

The search of PubMed MEDLINE is a start, but he should also look for articles and other publications from some databases in other fields, including the specialized alternatives databases and databases of research in progress.

Try Again! Dr. X has made a good decision to omit the details of his experiment from his search. He is interested in alternative methods and approaches to the use of mice in laboratory research, not just in cancer studies. At this point it is important to cast a wide net for all relevant publications. If he needs to restrict his results later, Dr. X has a number of options.

Take a look at the search itself. Is anything missing or superfluous?

A cursory search of one of the alternatives databases so that he gets some "soft science" results.

Correct! The National Institutes of Health Library suggests that one of the red flags for IACUCs and other investigatory bodies could be the omission of enrichment and replacement keywords from a search for animal testing alternatives. Remember, the search for alternatives has an end goal of improving the general welfare of animals used in laboratory research at every step of the process.

Write up his results and move on to his laboratory work.

Try Again! The terms for the test animal/subject are particularly imporant to a search for alternatives, as they can help narrow a potentially overwhelming body of literature down to relevant results for Dr. X's research. Without the description of the animal, a search of a database such as PubMed MEDLINE will return a tremendous number of articles that mention the concept "pain."

Take a look at the search itself. Is anything missing or superfluous?

Dr. X wants to see what kinds of results he gets in some databases other than PubMed MEDLINE for the search he did in Question #2. What should he do?

Check the help pages in each database (as available) to find out what database idiosyncracies may exist.

Try Again! Dr. X could depend on the MeSH mapping tool in PubMed to get him relevant records. However, some of these terms (for instance, "calories") do not map to any term. Thus, Dr. X will only find articles that have the world "calories" as a word in the title or abstract. There is a subject heading ("Energy Intake") that describes the concept in which Dr. X is interested, but the mapping tool did not locate it.

Go back to his keywords list and select ones that are appropriate for each individual database.

Try Again! In this search question, Dr. X did a good job by using relevant MeSH keywords, but he wasn't careful enough in constructing his search question itself. In general, it is a good idea to "nest" related terms (that are connected by an OR) with parentheses. Otherwise, the database works from the beginning of the search question to the end. This question literally asks PubMed MEDLINE to find articles about energy intake OR mice with brain cancers and diet. Clearly Dr. X meant to find articles about (mice) and (brain cancers) and (diet or energy intake).

Rewrite his search question to fit the requirements of each database.

Correct! In this search question Dr. X has selected appropriate MeSH keywords, and has constructed a well-formed question using Boolean operators and parentheses to nest related terms. The fairly short list of results contains articles that are extremely relevant to his research, and that may be helpful to him as he is evaluating the demand for his experiment.

All of the above

Try Again! While the PubMed MEDLINE search system is capable of automatically mapping keywords to MeSH equivalents, it does not always do a reliable job. Some terms may not map properly (or at all) even when a clear synonym exists in the hierarchy. The system is also not capable of parsing Boolean "logic" in anything other than a linear fashion - which may not be the logic you used when constructing the question. It is always better to construct the best search question possible yourself without relying on the fallibilities of the automatic system to fill in for you.


Dr. X used the following search question in several non-PubMed databases: "(mice OR murid*) AND (pain OR stress OR anesthesia OR euthanasia OR biopsy OR blood collection)". What complaint might the IACUC have with this search question?

No complaints. Even though it doesn't mention diet or cancer specifically, the real need is to find alternatives to using mice.

Almost! It is very important to read the help pages before tackling a new database, particularly if you have never used it before. Not all databases have the same options available for searching, and making an innocent mistake in your search question can cause hours of frustration.

What else should Dr. X do before searching other databases?

There aren't enough details about the experiment itself to narrow Dr. X's search to relevant articles.

Almost! As you already know from searching PubMed MEDLINE, the keywords you use in a database can radically change the results you get from the database. Some databases, including MEDLINE and AGRICOLA, use controlled vocabularies. It is best to take words from those vocabularies to use as your keywords. Others use free-text searching, so a range of keywords may be necessary to ensure that all aspects of a concept are represented.

What else should Dr. X do before searching other databases?

Dr. X only used terms for painful aspects of his research protocol, and none for the replacement of animals or the enrichment of their environment.

Almost! The same search question may not work from one database to another, so it is very important to modify the question to fit the database that you are searching. Not all databases are able to use Boolean logic and/or truncation searching. Some databases use different symbols for the same function (example: * vs. ? for truncation searching).

What else should Dr. X do before searching other databases?

By specifying "mice AND murid" as search terms Dr. X has unnecessarily restricted his search. He should drop these to widen his search.

Correct! No two databases are the same - each has different content, different indexing/controlled vocabulary words, and different rules for searching. Some databases may not support Boolean searching or truncation. Others may automatically search for similar words. You may be able to specify certain areas of the document to search, or you may only be able to enter one or two words that could be anywhere in the record. Tailoring each search to each database will not only get you better results from your search, it will end up saving you time and stress through efficiency.

Support Research, Teaching, & Learning - Give to the HSL