Pain trajectories and possible predictors of a favourable course of low back pain in patients consulting musculoskeletal physicians in The Netherlands

Table 4 Multivariable model

Predictor variables in multivariate model	Coefficient	OR	95% CI of OR
			Lower	Upper
Gender (female)	− 0.6271	0.5341	0.3629	0.7862
Time since complaints 1st started	0.0145	1.0146	0.9982	1.0312
Previous specialist visit	0.4623	1.5876	1.0645	2.3678
Previous visit pain clinic	0.1372	1.1471	0.9982	1.3162
Previous chiropractic treatment (treated without effect is reference)
Treated effective	0.2836	1.3279	0.4244	4.1546
Not treated	0.5804	1.7868	1.0872	2.9364
Work status (no work is reference)
Non-physical work	0.3539	1.4246	0.9218	2.2014
Physical work	0.5169	1.6768	0.9900	2.8399
Duration of current episode in years (non-linear spline variable)
Duration of current episode*	− 0.5197	0.5947	0.4298	0.8228
Spline variable duration of current episode*	1.9619	7.1128	1.9705	25.6750
Intercept	0.5462	1.7266	0.8101	3.6801

Multivariable model without treatment variables using baseline variables to predict a favourable outcome in patients presenting with high NRS for pain scores, including the Odds ratio’s (OR) and the 95% Confidence Intervals of the OR
*The Odds Ratio of the spline variables cannot be interpreted, but a longer duration of the current episode decreased the probability of improvement up to a period of four years and remained stable thereafter

ISSN: 2045-709X