Analyses

Interview transcripts were imported and analyzed inductively in NVivo, using thematic analysis [29]. As a first step, SM read and reread the transcripts to become familiar with the data, noting impressions and initial ideas for codes. Then SM began assigning codes to sentences and extracts, distilling and condensing their meaning. SM formed a provisional codebook relevant to our study objective. As more transcripts were coded and after several read throughs, the codebook was refined and finalized. Initial codes included: alternative locations, reduced availability, travel bans, new control tactics, quarantine, immigrant women, elderly women, poverty, and childcare. Codes with similar meanings were then grouped together into themes. SM sent a sample of 4 transcripts to the research team, for co-authors to read and to discuss initial interpretations of the data and confirm codes and themes. Once the themes were refined, SM reached out to the participants and shared a summary of the findings, along with supporting quotes, in order to receive their feedback, ensure all quotes were accurately interpreted, and to validate the results. A total of 9 participants replied and validated the findings; 3 never replied; and 5 were no longer working at the organization and no alternate contact information was available. Findings were then interpreted deductively against a SDOH framework.