(2004; 98 pages)
3.3 Study of documents
In the field of drug use and distribution there are many secondary data sources that researchers can tap. These include:
• published studies
• reports of agencies involved in the implementation of health care (baseline surveys, health surveys, monitoring reports and evaluations)
• sales and consumption figures
All these data sources can be used to:
• get a global overview of drug consumption, prescription and distribution
• get specific information on drug use in commonly occurring diseases
• identify drug use problems.
How to do document reviews
Before conducting fieldwork on drug use practices, we should ask: What do we know about the problem in the local settings? Have studies on the problem been done in other countries? What reasons for the problem were identified? What sub-problems have been described? What population groups are affected? What can you learn from drug sales statistics in your country? Have studies been done on community members’ perception of the problem in your country? What data are still lacking?
There are several websites which help you to find published - and sometimes unpublished - information (see box). Increasingly, journals offer their articles online. Articles can be downloaded free of charge, see for example http://bmj.bmjjournals.com.
BOX 4. INFORMATION SEARCHES
A computerized literature search using key words in Medline (which has abstracts of medical articles) and Popline (which covers family planning and related health issues). A site where you can search these and other health-related databases free of charge is: http://igm.nlm.nih.gov/
The INRUD Drug Use Bibliography, an annotated bibliography of published and unpublished articles, books, reports and other documents related to drug use, with a special focus on developing countries, contains over 4,000 entries and is updated regularly. See: www.inrud.org
Getting hold of unpublished reports is usually more difficult. If possible, it is often best to visit a few well-functioning documentation centres of health-related organizations and institutions. UNICEF national offices can be a good source of information. Also, when conducting key informant interviews ask for any relevant reports or data. It is best to ask health workers in the research area what data they or others have collected, and if they can be accessed. It is very important to do this, in order not to bother people in the area with questions and surveys that have already been undertaken by others.
A survey of prescriptions can give a very accurate picture of physician prescribing practices. To study prescriptions you need to request permission from health care institutions or pharmacies. However, they may not be systematically collected, for example, patients keep them. In that case, interviews with consumers can include a question on prescriptions kept at home, or you can conduct exit interviews at health facilities. Prescriptions are a good starting point for discussion on (non) compliance.
Sales statistics provide useful information and can be obtained from IMS Health affiliates in each country, though this may be expensive. Sales figures can be used to describe national drug consumption patterns; one can, for example, identify the 10 most commonly used drugs. A drawback is that the agencies collating sales figures often ask researchers to pay for these data; only outdated figures are available free of charge. Sales statistics reflect private sector drug consumption. Procurement data from the ministry of health can provide a good indication of volumes of drug use in the public sector. Ministries of health may also have collated data on medicine provision by type of health facility and region.
Strengths and weaknesses of document reviews
The strengths of using document reviews are:
• they are a cost-efficient way of doing research
• they avoid duplication of efforts.
The weaknesses of document reviews are:
• consent is often needed from the “owner” of the data
• it is sometimes difficult to assess the reliability and accuracy of the data
• data are often outdated
• data may be costly.