Enrichment sampling for a multi-site patient survey using electronic health records and census data

Technology-assisted adaptive recruitment strategy for a large nation-wide COVID-19 vaccine immunogenicity study in Brunei

A national study was conducted in Brunei to assess and compare the immunogenicity of the various brands of COVID-19 vaccines administered to the population as part of the National COVID-19 Vaccination Programme. Most of the population have had received at least 2 doses of BBIBP-CorV, AZD1222 or MRNA-1273 vaccines. Neutralising antibodies against SARS-CoV-2 induced by these vaccines will be analysed to infer population-level immune protection against COVID-19. During the 5-week recruitment period, 24,260 eligible individuals were invited to the study via SMS, out of which 2,712 participants were enrolled into the study. This paper describes the novel adaptive strategy used to recruit the study participants. Digital technology was leveraged to perform targeted online recruitment to circumvent the limitations of traditional recruitment methods. Technology also enabled stratified random selection of these eligible individuals who were stratified based on age, gender and vaccine brand. Data was extracted from the electronic health records, the national mobile health application and a third-party survey platform and integrated into a dedicated research platform called EVYDResearch. The instant availability and access to up-to-date data on EVYDResearch enabled the study team to meet weekly and adopt an adaptive recruitment strategy informed by behavioural science, where interventions could be quickly implemented to improve response rates. Some examples of these include incorporating nudge messaging into SMS invitations, involving the Minister of Health to make press announcements on this study, media coverage, setting up an enquiries hotline and reaching out to foreign language speaking expatriates of a local multinational company to participate in this study. Data integration from various data sources, real time information sharing and a strong teamwork led to good outcomes adaptable to the progress of recruitment, compared to the more time-consuming and static traditional recruitment methods.

Survey design and analysis considerations when utilizing misclassified sampling strata

BACKGROUND

A large multi-center survey was conducted to understand patients' perspectives on biobank study participation with particular focus on racial and ethnic minorities. In order to enrich the study sample with racial and ethnic minorities, disproportionate stratified sampling was implemented with strata defined by electronic health records (EHR) that are known to be inaccurate. We investigate the effect of sampling strata misclassification in complex survey design.

METHODS

Under non-differential and differential misclassification in the sampling strata, we compare the validity and precision of three simple and common analysis approaches for settings in which the primary exposure is used to define the sampling strata. We also compare the precision gains/losses observed from using a disproportionate stratified sampling scheme compared to using a simple random sample under varying degrees of strata misclassification.

RESULTS

Disproportionate stratified sampling can result in more efficient parameter estimates of the rare subgroups (race/ethnic minorities) in the sampling strata compared to simple random sampling. When sampling strata misclassification is non-differential with respect to the outcome, a design-agnostic analysis was preferred over model-based and design-based analyses. All methods yielded unbiased parameter estimates but standard error estimates were lowest from the design-agnostic analysis. However, when misclassification is differential, only the design-based method produced valid parameter estimates of the variables included in the sampling strata.

CONCLUSIONS

In complex survey design, when the interest is in making inference on rare subgroups, we recommend implementing disproportionate stratified sampling over simple random sampling even if the sampling strata are misclassified. If the misclassification is non-differential, we recommend a design-agnostic analysis. However, if the misclassification is differential, we recommend using design-based analyses.