(2004; 98 pages)
3.6 Observation techniques
Direct observation produces more reliable information than interviews on actual behaviour. The researcher can see which medicines are being sold, prescribed or used, or are available in medicine cabinets; what information on their use the seller or doctor provides; whether drugs are purchased on prescription, etc. The most common observational techniques to study drug use are:
• simulated client visits
• inventories of community drug outlets and medicine cabinets.
How to conduct simulated client visits
Observational research is often done to describe drug distribution patterns in informal drug outlets and pharmacies. However it is difficult to carry out because drug sellers may not want to have an onlooker present. They may feel that it will disrupt or even endanger their business, particularly if some transactions are illegal. If they have a researcher present, they may do business in a more ‘correct’ way than they normally would, and they may feel that their professional competence is being tested.
This problem can be solved by conducting simulated client visits. A researcher or an assistant can pose as a patient or client. The advantage is obvious; it gives an unbiased picture of normal procedures. Simulated client methods are often used to measure the quality of prescribing and dispensing drugs in health facilities, pharmacies and drug shops. It involves a researcher posing as a client and going to a health centre or pharmacy with a complaint. The objective is to determine how a sample of providers react to the complaint; what treatments they recommend, and what information they give.
Posing as a patient or client, however, reduces the amount of information that can be gathered. The researcher can only observe his or her own visit, and in pharmacies at most a few drug purchases while waiting in turn. Furthermore it is difficult to pose many questions and make notes, without revealing one’s true identity. To counter this restriction, researchers can combine unobtrusive observation with interviewing. This approach was used by Wolffers (1987), who had assistants visit 28 pharmacies to buy tetracycline over the counter, and then let other assistants interview the pharmacy personnel about tetracycline a few days later. You may also consider using different kinds of ‘clients’, representing the diversity in wealth, gender, age and ethnicity of people visiting the facility.
The method is somewhat controversial. One can question how ethical it is to conduct such visits without asking health workers and/or drug sellers for informed consent. The ethical issue can be resolved by asking the respondents or their professional organization for consent before conducting the visits and without giving details on when the visit will take place, to avoid bias. In the fieldwork it is recommended to use the simulated client visit to observe what information and advice drug sellers give. If a drug is bought during a visit, funding and accountability measures are needed.
An advantage of the simulated client visit is that it is a rapid method. A typical transaction does not last more than five minutes. It can also be used to evaluate the effects of rational drug use interventions, such as to measure information provided on drugs before and after an intervention. Ross-Degnan and colleagues (1996) tested the effects of a face-to-face education outreach intervention in Kenyan and Indonesian pharmacies.
Using trained surrogate patients posing as mothers of a child under five with diarrhoea, they measured sales of oral rehydration salts; sales of antidiarrhoeal drugs; and history taking and advice to continue fluids and foods. Sales of oral rehydration salts in intervention pharmacies increased by an average of 30% in Kenya, and 21% in Indonesia, compared to controls. Discussion of dehydration during pharmacy visits increased significantly in Kenya.
When planning to use simulated client visits as a method, you need to take decisions on how to sample the drug outlets and how many observations to do per outlet (see 6.2). It is important to consider the usual opening hours and the volume of transactions each day. For example, on market day pharmacies may be very busy and minimal advice given.
The guidelines provided in box 7 will help you to conduct effective simulated client visits.
BOX 7. GROUND RULES FOR CONDUCTING GOOD SIMULATED CLIENT VISITS
Record the results of the simulated client visits systematically. It is helpful to make a simple form to be filled in immediately after the visit. However, as with non-formal interviewing, the researcher has to be alert for the unexpected. The analysis and interpretation of the data depends on the extent to which the observations are structured. In some cases the analysis is quantitative, for example, when reporting in how many cases prescription drugs were sold over-the-counter. In other cases the observation is less structured, for example, focusing on the communication during the drug transaction. The researcher then has to categorize and analyse the findings in much the same way as with semi-structured interviews (see Chapter 6 for more information on analysis).
Strengths and weaknesses of simulated client visits
The strengths of simulated client visits are:
• they can provide more reliable information than interviews
• drug use and distribution in its natural context can be observed
• if done well, this method gives information on what drug sellers really do
• a representative sample of pharmacies/health centres can be observed.
• results can be generalized
• results can be quantified
• they can be used to evaluate effects of training of pharmacy sellers and health workers
• they can be used as a participatory method. You can ask people living in the communities that you are studying to act as surrogate clients, and collect data.
The weaknesses of simulated client visits are:
• data are sometimes hard to interpret
• it is difficult to do a lifelike simulation, especially if you are playing a type of client you are not so familiar with (female students acting as mothers, for example)
• the observation period is short (the time needed to buy the drug, or consult a health worker)
• it is difficult to probe on why advice is given
• the depth of information collected is limited
• the findings need to be complemented by interviews.
Inventory of community drug outlets and medicine cabinets
Inventories of medicine outlets and medicine cabinets are a second useful observation tool. The main aim of this tool is to describe the types of medicines commonly used in the community. The assumption is that commonly used medicines are those that people store in their medicine cabinets, and that storekeepers sell in community drug outlets. This tool can provide information about medicines used to promote health (such as vitamins and tonics) as well as illness-related medicine use. And, it can be used to conduct surveys of drug prices.1 The form overleaf can serve as an aid in collecting the information on medicine cabinets.
1 See WHO/HAI (2003), Medicine prices: a new approach to measurement.
You need to explain to the respondents how the data will be used, as you do with the other tools. It is also important that you ask for their consent.
To conduct the inventory of medicines in community stores, ask a key informant to guide you around the community and introduce you to storekeepers. Explain that you are interested in common drug use practices and why. Say that for this purpose you would like to know what medicines are sold in the general shops and what they are used for. You can make an inventory of the types of medicines, and what they are used for. You can combine this method with (semi-) structured interviews. At the household level the inventories can be used as the basis for semi-structured interviews about drug use practices and perceptions.
Strengths and weaknesses of medicine inventories
The strengths of medicine inventories are:
• they give an accurate picture of medicines commonly used in the community
• the medicines inventories are a good starting point for discussion on how medicines are used, and why they are used
• they can be a good participatory method - community health workers can help in making the inventories
• they can be used to collect data on pricing of drugs in different outlets.
The weaknesses of medicine inventories are:
• they can be time-consuming if households/shops have large medicine stocks, unless a selected list of medicines is surveyed
• sometimes people/sellers do not have the package information - it is hard to validate the types of medicines kept.