Table 3.

Clinical predictors of long COVID (imputed data).

Variable	Categories	Variance inflation factor (VIF)	Model 1 PR (95% CI)
Migration background
	Dutch origin		Reference
	African Surinamese origin		1.40(1.09–1.81)
	South Asian Surinamese origin		1.53 (1.07–2.19)
	Moroccan origin		1.34 (1.01–1.77)
	Turkish origin		1.44 (1.03–2.00)
	Other origin		1.08 (0.85–1.37)
	Unknown origin	1.05	1.13 (0.88–1.43)
Age (years)		1.18	0.98 (0.99–1.02)
Sex	Male		Reference
	Female	1.05	1.26 (1.08–1.48)
Smoking	Never smoked		Reference
	Current smoker		1.06 (0.89–1.25)
	Past smoker	1.03	0.96 (0.73–1.27)
Alcohol consumption	No		Reference
	Yes	1.02	0.67 (0.42–1.06)
Vaccinated against COVID-19	No		Reference
	Yes	1.05	0.75 (0.54–1.08)
Number of chronic health conditions	None		Reference
	1 condition		0.92 (0.75–1.14)
	2 conditions		0.85 (0.67–1.09)
	3 conditions		1.02 (0.79–1.31)
	4+ conditions	1.18	0.99 (0.76–1.29)
Number of days admitted in hospital		1.03	1.01 (1.01–1.02)
Rehospitalization	No		Reference
	Yes	1.50	1.07 (0.75–1.53)
Admitted to the ICU	No		Reference
	Yes	1.78	1.41 (1.13–1.74)
Received oxygen	No		Reference
	Yes	1.49	1.70 (1.32–2.20)
Received antibiotics	No		Reference
	Yes	1.52	1.01 (0.84–1.22)
Received corticosteroids	No		Reference
	Yes	1.37	0.64 (0.54–0.76)
Received remdesivir	No		Reference
	Yes	1.09	1.23 (0.97–1.57)

PR = prevalence ratios with 95% confidence interval obtained via robust Poisson regression models.

Variance inflation factor (VIF) analysis to identify predictors that are correlated. VIF scores greater than five indicate high correlation between predictors.

Model 1: multivariate robust Poisson regressions.

Vaccinated against COVID-19 refers to receipt of any time of COVID-19 before admission irrespective of number of doses at admission.

Received corticosteroids: hydrocortisone, prednisone, dexamethasone, methyldprednisone.

Received antibiotic medications applies to all forms of antibiotics.

Bold values inside the table signify statistically significant findings.