. Author manuscript; available in PMC: 2018 Aug 30.

Published in final edited form as: N Engl J Med. 2018 Aug 30;379(9):834–845. doi: 10.1056/NEJMoa1716677

Table 1.

Univariable and multivariable mixed-effects logistic regression model results for risk factors associated with the reporting of one or more cVDPV2 cases in a province after OPV2 withdrawal (01 May 2016). Data are for all provinces in Nigeria, Pakistan, Syrian Arab Republic and Democratic Republic of the Congo as of 08 August 2017

Fixed Effect	Data source	Univariable Odds Ratio (95% Credible Interval)	Multivariable Odds Ratio (95% Credible Interval)
Routine immunisation coverage+	Non-polio AFP cases 12-23 months old* or P3 coverage²⁹,‡	1.69 (1.06 – 3.05)	2.59 (1.26 – 6.33)
Serotype-2 population immunity in the first half of 2016+	Non-polio AFP cases <36 months old	2.83 (1.28 – 6.80)	4.65 (1.71 – 15.28)
Number of tOPV campaigns during 6 months prior to OPV2 withdrawal	Vaccination campaign calendar (Polio Information System)	1.05 (0.44 – 2.18)	-
Time (days) since the last tOPV campaign prior to April 2016	Vaccination campaign calendar (Polio Information System)	1.01 (0.81 – 1.27)	-
Population size (log10)	Worldpop³⁰	0.96 (0.10 – 9.10)	20 (0.49 – 1243)
Population density (log10 people / km²)	Worldpop³⁰ and WHO geodata	0.18 (0.02 – 0.95)	-

⁺

OR for a 10% absolute decrease

Pakistan and Syrian Arab Republic

^‡

Nigeria and DRC