Participants and recruitment

The study population was defined by purposive sampling, because we wanted to include participants who were knowledgeable about the phenomenon we aimed to study: asthma attacks in children.^{32 33} For this, we included HCPs actively involved in the acute management of asthmatic children in the city of Esmeraldas and caregivers of asthmatic children participating in the cohort study. We included HCPs from the two principal hospitals in the city (Hospital Delfina Torres de Concha, HDTC and the IESS Hospital) or primary care centre with emergency care. The HCPs included paediatricians, general doctors working at the ED and respiratory therapists. Nurses were not included, as they are not involved in the management of children with asthma in this setting. All were contacted by telephone. We used the snowball technique, by asking participating HCPs to identify additional colleagues who met inclusion criteria. For caregivers, we wanted to include caregivers of children treated for an asthma attack at an ED. As there are no electronic records and registries of these children in Esmeraldas, we contacted by telephone the families that participated in the cohort study and invited them to participate. Again, we used the snowball technique to encourage participating caregivers to inform other carers. None of the potential participants refused to participate, though 20 of the 40 caregivers that agreed to participate in the focus group discussions (FGDs) were not able to do so for practical scheduling reasons.

General doctors were less experienced and younger, paediatricians were older and more experienced, while respiratory therapists were between these groups in age and experience (online supplemental table 1). Most caregivers were mothers of children with asthma except for two grandmothers and two fathers. Most were homemakers (16/20) and of Afro-Ecuadorian ethnicity (13/20) (online supplemental table 2).

bmjresp-2021-001066supp001.pdf