Skip to main content
. Author manuscript; available in PMC: 2019 May 1.
Published in final edited form as: Drug Alcohol Depend. 2018 Mar 10;186:147–153. doi: 10.1016/j.drugalcdep.2018.01.015

Table 2.

Multivariate models examining prescriber factors and relationship with practices for assessing and mitigating diversion (N=1,174)

Assesses all patients for diversion (logistic regression)
AOR
(95% CI)		 Frequency of office visits in early treatment (ordinal regression)
AOR
(95% CI)		 Always tests urine screen for buprenorphine (logistic regression)
AOR
(95% CI)		 Percentage of patients with medication counts (negative binomial regression)
IRR
(95% CI)
Magnitude of the diversion problem in the community 1.188**
(1.060, 1.332)		 1.202**
(1.060, 1.361)		 1.278*
(1.055, 1.550)		 1.186***
(1.081, 1.302)
Buprenorphine practice characteristics
Delivers buprenorphine treatment in individual medical practice (vs. no individual practice) 0.996
(0.759, 1.308)		 0.806
(−0.628, 1.034)		 0.953
(0.640, 1.418)		 0.887
(0.756, 1.041)
Payment typology
 Only accepts cash Reference Reference Reference Reference
 Accepts private insurance but not Medicaid 1.019
(0.729, 1.423)		 1.348
(0.963, 1.888)		 0.812
(0.450, 1.464)		 1.139
(0.934, 1.388)
 Accepts Medicaid 1.148
(0.776, 1.697)		 1.814**
(1.213, 2.711)		 0.955
(0.612, 1.489)		 0.992
(0.791, 1.244)
 All other types of payment 0.879
(0.367, 2.104)		 1.684*+
(1.009, 2.811)		 1.328
(0.448, 3.937)		 0.920
(0.556, 1.523)
Current number of buprenorphine patients 1.001
(0.998, 1.004)		 0.996
(0.993, 1.000)		 1.017***
(1.010, 1.024)		 1.003*+
(1.000, 1.011)
Years of prescribing buprenorphine 0.942**
(0.909, 0.976)		 0.992
(0.962, 1.024)		 0.894***
(0.861, 0.929)		 0.994
(0.977, 1.011)
Waivered to treat up to 100 patients (vs. 30 patients) 1.080
(0.742, 1.572)		 0.839
(0.632, 1.114)		 1.641**
(1.159, 2.322)		 0.917
(0.754, 1.115)
Prescriber characteristics
Medical specialty
 Addiction (psychiatry or medicine) 1.063
(0.701, 1.611)		 1.762**
(1.245, 2.493)		 1.077
(0.674, 1.721)		 0.816
(0.665, 1.003)
 Psychiatry 0.969
(0.711, 1.322)		 1.817***
(1.366, 2.418)		 0.551***
(0.404, 0.753)		 0.901
(0.732, 1.110)
All other specialties Reference Reference Reference Reference
Member of ASAM/AAAP (vs member of neither) 1.044
(0.713, 1.529)		 1.556*+
(1.042, 2.322)		 1.012
(0.680, 1.506)		 1.358***
(1.173, 1.573)
Age 0.991
(0.977, 1.006)		 1.000
(0.991, 1.009)		 0.992
(0.975, 1.009)		 1.003
(0.996, 1.011)
Female (vs. male) 1.081
(0.761, 1.535)		 1.267
(0.949, 1.692)		 1.270
(0.832, 1.941)		 0.937
(0.772, 1.138)
Non-white (vs. white) 0.541***
(0.401, 0.730)		 0.890
(0.693, 1.141)		 0.714*+
(0.510, 0.999)		 1.209
(0.953, 1.535)
Constant 7.248***
(3.085,
17.0
32)		 n/a 4.955**
(1.793, 13.690)		 18.495
(10.052,
34.030)
Threshold 1 n/a 0.171
(−0.578,
0.919)		 n/a n/a
Threshold 2 n/a 1.788
(1.105, 2.561)		 n/a n/a

Notes.

*

p<.05,

**

p<.01,

***

p<.001 (two-tailed tests).

+

p-value exceeds the false discovery rate (FDR) value for statistical significance when adjusting for multiple comparisons.

AOR = adjusted odds ratio. IRR = incidence rate ratio. CI = confidence interval. Pooled estimates were calculated from models estimated using 20 imputed datasets (n = 1,174). All models used robust standard errors to adjust for the clustering of prescribers within states. For the model of frequency of office visits, the thresholds represent estimated cutpoints on the underlying latent variable of office visits. Threshold 1 is the estimated cutpoint used to differentiate “monthly visits or less” responses from the other two categories when values of the independent variables are zero. Threshold 2 is the estimated cutpoint that differentiates the groups “monthly visits or less” and “every two weeks” from the “weekly or more” group when the independent variables are set at zero.