Skip to main content
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Addiction. 2020 Jul 13;116(2):305–318. doi: 10.1111/add.15112

Table 3.

Marginal effect (Absolute least square (LS) mean difference) of CBT compared to HL on percent drinking days and drinks per drinking day averaged over the entire study and across three study phases. Sensitivity analyses for these results

Absolute least squares mean differences Sensitivity analyses*
Percent Drinking Days
  Estimate (std. error) Estimate (std. error)
Study
Period
CBT HL LS mean diff
(95%CI)
p-value CBT HL LS mean diff
(95%CI)
p-value
Overall 7.58 (0.678) 10.26 (0.695) 2.68 (0.77, 4.58) 0.0059 10.20 (0.876) 13.63 (0.896) −3.42(−5.88, −0.97) 0.0063
Intervention (baseline-6wks) 21.14 (0.978) 25.33 (1.001) 5.20 (2.45, 7.94) <0.001 22.40 (1.025) 27.62 (1.049) −5.22(−8.10, −2.34) 0.0004
Follow up (7-30 wks) 5.35 (0.748) 8.11 (0.767) 2.76 (0.65, 4.86) 0.0102 7.46 (0.897) 10.81 (0.917) −3.35(−5.87, −0.83) 0.0091
Maintenance (31-46 wks) 3.64 (0.696) 5.72 (0.712) 2.08 (0.13, 4.04) 0.037 9.17 (1.167) 11.55 (1.193) −2.38(−5.62, 0.86) 0.1496
Drinks per Drinking Day
  Estimate (std. error) Estimate (std. error)
Study
Period
CBT HL LS mean difld>
(95%CI)
p-value CBTld> HLd> LS mean diff
(95%CI)
p-value
Overall 1.15 (0.091) 1.69 (0.093) 0.55 (0.29, 0.80) <.0001 1.47 (0.116) 2.00 (0.119) −0.53(−0.85, −0.20) 0.0015
Intervention (baseline-6wks) 2.69 (0.128) 3.47 (0.131) 0.78 (0.42, 1.14) <.0001 2.98 (0.136) 3.75 (0.139) −0.77(−1.15, −0.39) <0.0001
Follow up (7-30 wks) 0.83 (0.103) 1.51 (0.106) 0.67 (0.38, 0.96) <.0001 1.16 (0.126) 1.79 (0.129) −0.64(−0.99, −0.28) 0.0004
Maintenance (31-46 wks) 0.66(0.096) 0.98 (0.098) 0.31 (0.05, 0.58) 0.0218 1.29 (0.141) 1.55 (0.144) −0.25(−0.65, 0.14) 0.2082
*

Sensitivity analyses assumed that those who were lost-to-follow-up had returned to baseline drinking levels