Skip to content
INSIGHT Surveillance Toolkit

Data Collection Approaches

Qualitative

Qualitative research uses approaches and techniques designed to collect and analyze non-numerical data. Qualitative research explores subjective experiences, context, and perspectives including attitudes, knowledge, and behaviors based on an individual’s experience.

The most common qualitative research methods used in surveillance system assessments include:

In-Depth Interviews with Key Informants

In-depth interviews can be structured or semi-structured and are designed to understand the key informant’s perspective on the study topic. Key informants are not representative samples, but rather are people capable of providing information about your specific research question. Key informants are experts, and the job of the interviewer is to understand everything the key informant can share about the topic of study. Interviewers pose questions in a neutral manner, listen attentively to responses, and ask follow-up questions and probes based on those responses.

  • Who to Interview: People who know a lot about things you seek to understand and who are willing to talk to you.
  • Semi-Structured Interviews Versus Structured Interviews: Both require use of a written list of open-ended questions and topics that you want to cover (interview guide), but in semi-structured interviews, the respondent and the interviewer have discretion to follow new leads or explore new areas. In structured interviews, respondents answer nearly identical questions with fewer opportunities to probe for new information. Structured interviews might include questions with skip patterns, rank-order, pile sorting, and other approaches that will make things comparable but are still based on informant opinions. Semi-structured interviews are excellent for establishing rapport, exploring topics of interest, and probing for new areas of interest that you might not have anticipated

Sample questions for in-depth interviews:

  1. How does data flow from health facilities to the district?
  2. How has your syndromic surveillance system detected outbreaks in the past?
  3. What are the barriers and challenges to implementing event-based surveillance at the regional level?
  4. How could your surveillance system be improved?

Focus Group Discussions

Focus groups are a qualitative technique that brings together a diverse group of people (typically 6–10 participants) to discuss a topic in more detail. A facilitator guides the discussion using a semi-structured or structured discussion guide. Focus groups are a flexible approach to gathering in-depth information. They should include participants who represent the target audience for understanding your study question. When conducting focus groups, refer to Considerations when conducting stakeholder engagement for some key considerations in group dynamics.

Examples of when to use a focus group:

  • To gather community input about surveillance needs. (Community representatives, local organizations working in healthcare)
  • To understand community-based surveillance challenges and what is working well (Community representatives)
  • To discuss policies that may inhibit surveillance programming (Policy makers and managers)
  • To identify gaps in reporting at health facilities (health facility staff, data clerks)

Mapping Exercises

Mapping is a technique to visually represent concepts, relationships, and connections to help understand a topic of interest. Mapping exercises can use electronic tools or can be as simple as using paper and markers to design maps that reflect respondents understanding of a topic or perspective.

Data Flow Mapping can help to understand how information moves through a system. The objective of data flow mapping is to explore a person’s understanding of how something is transmitted from point A to point X, diagramming it on a piece of paper.

Data Flow Mapping Example: Mapping of data flow

Participatory Mapping can represent environments or spaces to look for patterns in how people navigate and interact in spaces or to identify areas to target for specific reasons.

Business Process Analyses (BPA)

A business process describes a set of activities and tasks that logically group together to accomplish a goal or produce something of value. In the context of public health, a business process will contribute a valuable service for the benefit of the local, regional, or national health organization, the public health community, or the target population.

In the context of this toolkit, BPA is conceptualized as the effort to understand an organization, its purpose and process workflows, taking into account its activities, the participants, and its information/data workflows. Generally, the outcomes of a BPA are:

  • Diagrams (e.g., process diagrams, data flow diagrams, system architecture diagrams)
  • Documentation of processes (e.g., textual descriptions)
  • Business rules and similar guidelines or procedures

For the specific case of mapping processes in healthcare, a framework has been proposed, that identifies five steps or phases1:

  1. Preparation, Planning, and Process Identification: This phase includes defining the scope of the assessment (which system(s) will be analyzed) and the relevant groups that will be involved in the project. As part of the preparation, there might be a need to account for capacity building in specific topics, such as the creation and interpretation of diagrams using Business Process Model and Notation (BPMN), which is a standardized graphical notation for specifying business processes.
  2. Data and Information Gathering: Information about the systems is obtained through different elicitation methods, including interviews, analysis of existing documentation and diagrams, descriptions of the system functionality, existing data models, etc.
  3. Process Map Generation: Once the information is gathered, drafts of maps illustrating the different processes and their interrelations, as well as the data flows inside and across different parts of the system(s), will be created with input from the different groups identified in phase 1. There might be a need to have maps that show the documented (and/or regulated) process, and one showing how it works in real practice, if adaptations or pragmatic modifications have been made.
  4. Analysis: The mapping will identify gaps, opportunities for improvement, and needs for change in the regulations surrounding the processes, systems, and data. This analysis will be reflected through annotations of the maps, as well as results from the reviews by subject matter experts.
  5. Taking it Forward: The process and data flow maps, will inform the next steps and plan for improvement of current systems and practices, as well as impact (sometimes) the need to reformulate existing rules and regulations.

Desk Review

A desk review involves the collection and review of existing information on a topic including documents, policies, reports, and both primary and secondary literature. The output is typically a report or summary of all information relevant to the identified topic. Depending on the topic reviewed, the output can help guide system design and implementation or provide external context for the surveillance system.

Quantitative

Quantitative research is the systematic collection, analysis, and interpretation of primarily numerical or categorical data. Statistical techniques are employed to draw objective and reliable conclusions.

The most common quantitative research methods used in surveillance system assessments include:

Survey

Surveys collect data directly from the target population via structured questionnaires with pre-generated response options. These options may include single or multiple-choice, Likert scales, or numeric ratings. The structured nature of survey questions and responses limits the opportunity to capture additional details or open-ended information from respondents. However, this structured approach facilitates statistical analysis and presentation of quantifiable results. Surveys typically do not allow for the depth of information or exploration found in qualitative methods; however, they are more efficient when collecting information from a large number of respondents. A larger number of respondents can potentially allow for:

  • A more representative picture of the topic of interest
  • More generalizable results
  • Disaggregation of data to look at sub-populations

Surveys can be administered to a sample or a census of the target population. The methods used for drawing a sample and the number of respondents sampled influence the results and conclusions that can be drawn from a survey. Surveys can be sent to participants online using secure web-based data capture tools, administered in person, or given via a video conferencing platform like Zoom.

Secondary Data Extraction or Collection

Secondary data extraction/collection is the analysis of surveillance data using case and population data to estimate values for each attribute. The population data could come from surveillance data or you may need to look at external sources, such as census data or Demographic and Health Surveys (DHS). For some attributes, linkages across systems may be required. Depending on the setting and data available, data linkage can be deterministic, using for instance national IDs, otherwise probabilistic methods will need to be used.

Mixed Methods

A mixed methods approach enriches the data obtained from the assessment and provides context to guide interpretation of results and recommended modifications to the system. Qualitative and quantitative data can be collected sequentially or simultaneously, and the data can be merged, connected, or embedded. We recommend a mixed method design to assess surveillance system performance and context to better understand the system’s strengths and existing gaps, as well as how to address them.

1 Antonacci, G., Lennox, L., Barlow, J. et al. Process mapping in healthcare: a systematic review. BMC Health Serv Res 21, 342 (2021). https://doi.org/10.1186/s12913-021-06254-1

Skip to toolbar