Table 1

Potential confounders for facility-level instrumental variables

Confounder category	Potential confounders	Covered	Data source
Geographic	Urban/rural	Yes.	Municipality data set and municipality level covariates in microdata (table 2).
Patient characteristics	Race, education, income, age, insurance status, health status/comorbid conditions, health behaviours	Insurance status is not likely an issue in the Norwegian healthcare system. Health behaviour is unobserved.	Microdata (table 2).
Facility characteristics	Procedure volume, facility volume, clinical services offered, departments, teaching status, profit status, trauma designation, delivery system type, practice type	Although the sizes of the facilities vary, the provided services and the organisation of the facility are part of the same nationwide healthcare system and should be similar. Profit status is likely not an issue in Norwegian healthcare system.	Municipality data set and municipality level covariates in microdata (table 2).
Treatment characteristics	Receipt of other treatments	We may adjust for contacts with health services for treatment and common prescriptions (eg, melatonin, antidepressants)	Microdata (table 2).