Application of data mining techniques and logistic regression to model drug use transition to injection: a case study in drug use treatment centers in Kermanshah Province, Iran

Table 3 Logistic regression model

Variables	Odds ratio (OR)	95% confidence interval for OR		Wald	P-value
Occupational status
Unemployed	–	–	–	–	–
Self-employed	0.548	0.134	2.232	0.705	0.401
Employed	0.303	0.075	1.218	2.831	0.092
Housewife	0.090	0.017	0.477	8.006	0.005
Education
Illiterate	–	–	–	–	–
Elementary	0.202	0.044	0.926	4.239	0.040
Leadership	0.438	0.183	1.050	3.421	0.064
High School	0.919	0.459	1.838	0.057	0.811
College education	0.905	0.467	1.754	0.087	0.767
The first used drug
Opioids	–	–	–	–	–
Others	0.417	0.275	0.633	16.909	< 0.001
Number of years of drug use	1.047	1.021	1.075	12.390	< 0.001
Motivation for starting the drug use
Pleasure	–	–	–	–	–
Drug use of friends	0.592	0.271	1.293	1.729	.0.188
Curiosity	1.592	0.725	3.498	1.341	0.247
Use as a pain reliever	2.207	0.929	5.244	3.215	0.073
Emotional distress and mental	0.000	0.00		0.000	0.996
Others	0.760	0.287	2.014	0.304	0.581
Unemployment	0.492	0.167	1.454	1.647	0.199
Family history of drug use
yes	1. 403	0.938	2.099	2.712	0.100
History of taking hashish
yes	0.446	0.299	0.665	15.694	< 0.001
History of taking heroin
yes	0.107	0.070	0.164	106.375	< 0.001
History of taking cocaine
yes	0.165	0.101	0.270	51.905	< 0.001
History of taking hallucinogens
yes	0.059	0.024	0.146	37.579	< 0.001
History of taking crystal
yes	0.674	0.451	1.008	3.684	0.055
History of taking methadone
yes	0.487	0.252	0.941	4.591	0.032
History of suicide
yes	0.353	0.220	0.566	18.650	< 0.001
History of prison
yes	0.594	0.394	0.897	6.143	0.013
Number of referrals to drug treatment centers
1	–	–	–	–	–
2	0.291	0.101	0.835	5.271	0.022 0.032 0.001
3	0.556	0.326	0.950	4.608
4	0.381	0.217	0.670	11.235
> 5	0.604	0.384	0.949	4.780	0.029

ISSN: 1747-597X