The Use of Adaptive Sampling to Reach Disadvantaged Populations for Immunization Programs and Assessments: A Systematic Review

Abstract

Vaccines prevent 4–5 million deaths every year, but inequities in vaccine coverage persist among key disadvantaged subpopulations. Under-immunized subpopulations (e.g., migrants, slum residents) may be consistently missed with conventional methods for estimating immunization coverage and assessing vaccination barriers. Adaptive sampling, such as respondent-driven sampling, may offer useful strategies for identifying and collecting data from these subpopulations that are often “hidden” or hard-to-reach. However, use of these adaptive sampling approaches in the field of global immunization has not been systematically documented. We searched PubMed, Scopus, and Embase databases to identify eligible studies published through November 2020 that used an adaptive sampling method to collect immunization-related data. From the eligible studies, we extracted relevant data on their objectives, setting and target population, and sampling methods. We categorized sampling methods and assessed their frequencies. Twenty-three studies met the inclusion criteria out of the 3069 articles screened for eligibility. Peer-driven sampling was the most frequently used adaptive sampling method (57%), followed by geospatial sampling (30%), venue-based sampling (17%), ethnographic mapping (9%), and compact segment sampling (9%). Sixty-one percent of studies were conducted in upper-middle-income or high-income countries. Data on immunization uptake were collected in 65% of studies, and data on knowledge and attitudes about immunizations were collected in 57% of studies. We found limited use of adaptive sampling methods in measuring immunization coverage and understanding determinants of vaccination uptake. The current under-utilization of adaptive sampling approaches leaves much room for improvement in how immunization programs calibrate their strategies to reach “hidden” subpopulations.