This blog describes a tool, Guidelines for Use of Narrative Survey Instruments to Improve MDA for Lymphatic Filariasis Elimination.
In the elimination of lymphatic filariasis (LF), many communities have been able to stop mass drug administration (MDA) because transmission assessment surveys have found minimal LF transmission remaining. There are, however, areas where attaining consistently high coverage and compliance has been a challenge and transmission continues. In these areas, there may be a significant number of people – known as systematic noncompliers – who refuse treatment during multiple rounds, or, in the absence of directly observed treatments, health personnel may presume coverage and compliance to be greater than they are. In these areas, it seems inefficient to continue MDA rounds without understanding why people are taking the pills and why they are refusing. Without improvement to the MDA delivery, drugs and valuable resources may be wasted. Coverage surveys offer one way to understand the gap between coverage and compliance, but they may not give us the level of detail needed to fully understand why people are or are not complying.
Public health research that seeks to understand people’s health behaviors has traditionally focused on the individual and his or her knowledge, attitudes and practices, known as KAP. Does the individual understand how a disease is transmitted? Does the person know how the disease can be prevented? While this information is useful and informative, it may not provide the best explanation for why a person adopts particular health behaviors. Literature examining reasons why recipients take drugs offered during MDA for the elimination of LF suggests that few knowledge indicators explain recipients’ compliance with treatment. In fact, social factors – such as personally knowing the drug distributor, being aware of the MDA before it occurs, or knowing other people who have taken the LF drugs – have a strong influence on people’s decisions to comply with treatment. Through shifting the focus from the individual to the individual’s experiences within specific social contexts, we have a better chance of understanding compliance and non-compliance and hence identifying how best to intervene to improve compliance rates for different sections of the population.
We advocate for a novel approach to researching health behaviors that builds on social psychology theory. The approach asks individuals to relate an experience they had in a recent MDA situation. The related experience is a mini narrative that the individual is then asked to interpret using closed-form questions in the instrument. We have found the combination of narrative (i.e., qualitative data) and respondent interpretation (i.e,. quantitative data) to be extraordinarily powerful in understanding the factors associated with compliance in an LF MDA context.
Haitian school children gather to receive a two-drug treatment for lymphatic filariasis (LF), being administered by community members.
Narration combines description and explanation of personal experiences to describe, understand or explain situations, reactions and behaviors. People tell each other stories continuously: on the bus or at work, at social gatherings, on the phone, or in discussions with close family members. Narrative is the natural way people think about, talk about and store memories of experiences. Narrative is the mechanism by which people think of, make sense of, communicate and remember. It is therefore an ideal data source to understand people’s experiences with MDA, how people were mentally and physically oriented to their experiences, and their interpretations of their MDA experiences. Hence, narrative provides a highly informative window to an individual’s experience of being offered the LF drugs and of doing something with those drugs.
When we ask people to tell us of an experience, they bring elements of that experience into their working memories. It is highly unlikely that storytellers will recount a moment-by-moment story of the experience; such recounting could take longer than the original experience! Rather, individuals highlight certain actions, people, feelings, or things and keep others in the background. There are many possible synopsises of the experiences and the story may change slightly depending on the social context of the conversation.
Narrating an experience brings a particular synopsis into working memory to make the story more accessible and easier to relate. Research has shown that bodily functions are cued to re-enact the events described: narratives of direct experiences thus give access to a much deeper set of mental, physical and emotional details than the few lines of text or talk received in casual conversations. It may be useful to think of this effect as analogous with conducting an internet search: Typing the search terms into the search engine generates a series of results. Each preliminary result includes a summary of the source so users can decide whether to click and read the whole article or not. Users do not look at all the results (there are usually far too many for that). Users just look at the most salient. If needed it is possible to follow any number of links to see more details. Often each of these provides many opportunities for finding even more detail. It is the same with narratives: They provide first order, if selective, windows on individual experiences.
Internet search results for "Neglected Tropical Diseases" illustrate the order of memory: Headlines appear first, followed by brief excerpts. Users click on links to access the full text.
Stories of an experience take on different versions depending on the circumstances of the conversation (e.g., a job interview versus talking to friends over a meal). Social psychologists have shown the importance of situational factors – like time, place, what is going on and who is there – in determining the story versions that are told. For this reason, we advocate asking for very specifically cued or prompted narratives. We need to situate people in a specific context (as far as possible) and then ask them to relate their experience from that perspective.
The underlying logic of the narrative approach is that direct personal experience is the best source of information to understand what happens during an MDA. The instrument design does not assume that everyone experiences the MDA in the same way; it seeks to capture different perspectives and experiences. These different perspectives are the most reliable source of information to identify groups that are not receiving or not taking the drugs and, importantly, to identify factors associated with non-receipt or non-compliance that would then inform health systems teams to change their delivery procedures so as to achieve target levels (or greater) of drug receipt and ingestion.
In general, people experience things in different ways; two people will likely give different accounts of the same events. Think about asking students about their experience participating in a lecture. One student sitting on the front row may recount that this was the best lecture on public health she has heard in her university career thus far. Another student sitting by the door, constantly interrupted by late students, may report finding the lecture hard to follow and boring. Narratives about the same lecture may reflect as many experiences as there are students in the room. Each of these experiences are real and true, no matter how much they differ from each other.
Individuals interpret or make sense of events or objects in different ways. There are, thus, as many experiences of an MDA as there are people involved: the health teams involved in drug delivery or supporting the community through the MDA process, or the community members themselves. Everyone will have a different experience and will uniquely interpret his or her experience.
Linking qualitative and quantitative data collection methods expands and deepens the options available to understand what is being examined. Instruments developed using the approach we have tested are fully integrated: the qualitative (narrative) provides a bridge to the memory of the experience, which is then directly interpreted by the respondent using the interpretive questions (quantitative elements). Using this approach in Indonesia (in conjunction with District health teams) helped those teams improve their MDA processes and hence achieve significantly better compliance across diverse regions and social groups.
The integrated data sets that result from this approach can be analyzed in several different ways, thus providing researchers of different orientations with options to suit their needs. We have used traditional frequentist analytical methods to analyse the quantitative elements of the data with excellent results. But the best results we got came from quantitative analyses of both the narrative and quantitative elements of the data in fully integrated models. That's because – as described above – narrative provided a vitally important window to an individual's experience of being offered LF drugs and of doing something with those drugs. When analyzed approriately, narrative data provides a fuller picture of how a community experiences MDA, giving vital clues as to how it can be improved.
This blog is the first of a two-part series. A follow-up post will illustrate some of the advantages of using the narrative approach as evidenced by the results of two years of testing the approach in Indonesia.