Supplement: Tables and Figures Supplement Table 1. Baseline Characteristics by Study and Efavirenz Assignment Supplement Table 2. Baseline Psychoactive Medications by Efavirenz Assignment* Supplement Table 3. Suicidality Events by Country Supplement Table 4. Restricted to the United States: Suicidality Events by Study Supplement Table 5. Association Between Baseline Characteristics and Hazard of Suicidality (ITT Approach)* Supplement Table 6. Restricted to the United States: Association Between Baseline Characteristics and Hazard of Suicidality (ITT DSMB) Supplement Table 7. Restricted to the United States: Association Between Baseline Characteristics and Hazard of Suicidality (ITT Approach) Supplement Table 8. Cause of Death Classifications (ITT) Supplement Figure 1. Association between efavirenz and hazard of suicidality, according to baseline characteristics (ITT DSMB primary approach). Supplement Figure 2. Association between efavirenz and hazard of suicidality, according to baseline characteristics (ITT sensitivity approach). Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
14
Embed
Supplement: Tables and Figures - NATAP · Supplement: Tables and Figures . Supplement Table 1. Baseline Characteristics by Study and Efavirenz Assignment . Supplement Table 2. Baseline
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Supplement: Tables and Figures Supplement Table 1. Baseline Characteristics by Study and Efavirenz Assignment
Supplement Table 2. Baseline Psychoactive Medications by Efavirenz Assignment*
Supplement Table 3. Suicidality Events by Country
Supplement Table 4. Restricted to the United States: Suicidality Events by Study
Supplement Table 5. Association Between Baseline Characteristics and Hazard of Suicidality (ITT Approach)*
Supplement Table 6. Restricted to the United States: Association Between Baseline Characteristics and Hazard of Suicidality (ITT DSMB)
Supplement Table 7. Restricted to the United States: Association Between Baseline Characteristics and Hazard of Suicidality (ITT Approach)
Supplement Table 8. Cause of Death Classifications (ITT)
Supplement Figure 1. Association between efavirenz and hazard of suicidality, according to baseline characteristics (ITT DSMB
primary approach).
Supplement Figure 2. Association between efavirenz and hazard of suicidality, according to baseline characteristics (ITT sensitivity
approach).
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 1. Baseline Characteristics by Study and Efavirenz Assignment
EFV=efavirenz. *Prescription psychoactive (or antidepressant) medication ongoing within 30 days before study entry.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 2. Baseline Psychoactive Medications by Efavirenz Assignment* Generic name (Trade name) Efavirenz (n=3241) Efavirenz-free (n=2091)
Psychoactive Drug Category No. of participants on medication prestudy 422 (13%) 285 (14%) Antidepressant No. of participants on medication prestudy 325 (10%) 217 (10%)
Antipsychotic & Antidepressant No. of participants on medication prestudy 0 (0%) 1 (0%)
Amitriptyline & Chlordiazepoxide (Limbitrol (DS)) 0 (0%) 1 (0%) * Prescription medication ongoing within 30 days before study entry; prescribing indication was not collected
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 3. Suicidality Events by Country
A. Intention-to-Treat Data Safety Monitoring Board Approach Efavirenz-containing Efavirenz-free
Country n Events PYs IR (95% CI)* n Events PYs IR (95% CI)* Total n United States 2324 39 4345.9 8.97 (6.38, 12.27) 1627 13 3353.6 3.88 (2.06, 6.63) 3951 India 169 2 261.6 7.65 (0.93, 27.62) 86 0 131.8 0.00 (0.00, 27.98) 255 Brazil 155 0 224.6 0.00 (0.00, 16.42) 76 1 108.6 9.21 (0.23, 51.29) 231 South Africa 152 1 258.5 3.87 (0.10, 21.55) 78 0 134.4 0.00 (0.00, 27.45) 230 Malawi 147 0 250.7 0.00 (0.00, 14.71) 74 0 124.9 0.00 (0.00, 29.52) 221 Peru 86 1 142.8 7.00 (0.18, 39.00) 48 1 76.5 13.07 (0.33, 72.82) 134 Zimbabwe 73 0 127.5 0.00 (0.00, 28.92) 37 0 65.7 0.00 (0.00, 56.12) 110 Haiti 68 0 104.8 0.00 (0.00, 35.19) 32 0 51.6 0.00 (0.00, 71.50) 100 Thailand 67 4 100.6 39.78 (10.84, 101.85) 33 0 51.3 0.00 (0.00, 71.88) 100
B. Intention-to-Treat Approach Efavirenz-containing Efavirenz-free
Country n Events PYs IR (95% CI)* n Events PYs IR (95% CI)* Total n United States 2324 48 5736.2 8.37 (6.17, 11.09) 1627 15 4040.0 3.71 (2.08, 6.12) 3951 India 169 5 568.0 8.80 (2.86, 20.54) 86 2 286.8 6.97 (0.84, 25.19) 255 Brazil 155 1 512.9 1.95 (0.05, 10.86) 76 1 251.9 3.97 (0.10, 22.12) 231 South Africa 152 1 503.7 1.99 (0.05, 11.06) 78 1 256.2 3.90 (0.10, 21.75) 230 Malawi 147 0 479.3 0.00 (0.00, 7.70) 74 0 236.3 0.00 (0.00, 15.61) 221 Peru 86 4 293.6 13.63 (3.71, 34.89) 48 1 160.4 6.23 (0.16, 34.73) 134 Zimbabwe 73 0 258.0 0.00 (0.00, 14.30) 37 0 130.0 0.00 (0.00, 28.38) 110 Haiti 68 0 223.2 0.00 (0.00, 16.52) 32 0 112.1 0.00 (0.00, 32.91) 100 Thailand 67 4 212.5 18.82 (5.13, 48.20) 33 0 112.1 0.00 (0.00, 32.90) 100
IR=incidence rate per 1000 PYs. PYs=person-years. Countries are ordered by sample size. *Exact Poisson 95% confidence interval for crude incidence rate per 1000 PYs
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 4. Restricted to the United States: Suicidality Events by Study
A. Intention-to-Treat Data Safety Monitoring Board Approach Efavirenz-containing Efavirenz-free
HIV-1 RNA categories ≥100,000 vs. <100,000 copies/mL 1 5330 0.73 (0.45, 1.17) 0.190 0.85 (0.51, 1.42) 0.53 History of AIDS Yes vs. No 1 5332 0.93 (0.50, 1.72) 0.81 1.33 (0.68, 2.62) 0.40 Injection drug history Yes vs. No 1 5332 3.23 (1.82, 5.72) <0.001 2.70 (1.50, 4.88) <0.001 Psychiatric history or psychoactive Rx
Yes vs. No 1 5332 3.72 (2.35, 5.88) <0.001 3.75 (2.33, 6.03) <0.001
BMI categories Underweight (< 18.5) vs. Normal (<25) 3 5280 1.38 (0.62, 3.05) 0.29 . Overweight (<30) vs. Normal (<25) 0.68 (0.39, 1.17) Obese (30+) vs. Normal (<25) 0.68 (0.32, 1.43)
Weight categories <60 vs. ≥80 kg 2 5321 1.24 (0.67, 2.29) 0.39 1.48 (0.77, 2.86) 0.158 60-<80 vs. ≥80 kg 0.85 (0.50, 1.44) 0.84 (0.49, 1.44)
BMI=body mass index. DF=degrees of freedom. HR=hazard ratio. ITT=intention-to-treat. *All on-study follow-up was included (ITT approach); Cox model Wald CI and p-values are presented, each model was stratified by study. †In multivariable analysis, race or ethnic group was omitted because analysis of this covariate was restricted to the United States, and BMI category was omitted due to collinearity with body weight and missing height observations; otherwise covariates were included in this multivariable model.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 6. Restricted to the United States: Association Between Baseline Characteristics and Hazard of Suicidality (ITT DSMB)*
HIV-1 RNA categories ≥100,000 vs. <100,000 copies/mL 1 3950 0.95 (0.52, 1.73) 0.87 0.98 (0.50, 1.93) 0.95 History of AIDS Yes vs. No 1 3951 1.13 (0.57, 2.25) 0.73 1.70 (0.75, 3.88) 0.20 Injection drug history Yes vs. No 1 3951 2.81 (1.44, 5.47) 0.002 2.30 (1.16, 4.59) 0.018 Psychiatric history or psychoactive Rx
Yes vs. No 1 3951 4.22 (2.28, 7.82) <0.001 3.51 (1.85, 6.64) <0.001
BMI categories Underweight (< 18.5) vs. Normal (<25) 3 3899 1.74 (0.61, 4.94) 0.177 . Overweight (<30) vs. Normal (<25) 0.64 (0.33, 1.22) Obese (30+) vs. Normal (<25) 0.53 (0.21, 1.38)
Weight categories <60 vs. ≥80 kg 2 3940 2.02 (0.93, 4.40) 0.20 2.70 (1.14, 6.41) 0.072 60-<80 vs. ≥80 kg 1.23 (0.65, 2.33) 1.28 (0.67, 2.45)
BMI=body mass index. DF=degrees of freedom. DSMB=data safety monitoring board. HR=hazard ratio. ITT=intention-to-treat. *Cox model Wald CI and p-values are presented, each model was stratified by study. †Multivariable analysis was restricted to participants of white, black, or Hispanic race or ethnic group with complete covariate information; BMI was omitted due to collinearity with body weight and missing height observations.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 7. Restricted to the United States: Association Between Baseline Characteristics and Hazard of Suicidality (ITT Approach)*
HIV-1 RNA categories ≥100,000 vs. <100,000 copies/mL 1 3950 0.79 (0.45, 1.38) 0.41 0.80 (0.43, 1.49) 0.48 History of AIDS Yes vs. No 1 3951 0.87 (0.44, 1.71) 0.69 1.21 (0.54, 2.72) 0.65 Injection drug history Yes vs. No 1 3951 3.29 (1.84, 5.89) <0.001 2.81 (1.53, 5.14) <0.001 Psychiatric history or psychoactive Rx
Yes vs. No 1 3951 4.14 (2.39, 7.19) <0.001 3.47 (1.96, 6.13) <0.001
BMI categories Underweight (< 18.5) vs. Normal (<25)
3 3899 1.41 (0.50, 3.97) 0.21 .
Overweight (<30) vs. Normal (<25) 0.64 (0.36, 1.16) Obese (30+) vs. Normal (<25) 0.53 (0.22, 1.26)
Weight categories <60 vs. ≥80 kg 2 3940 1.35 (0.65, 2.79) 0.62 1.72 (0.77, 3.86) 0.33 60-<80 vs. ≥80 kg 0.97 (0.56, 1.68) 0.99 (0.56, 1.73)
BMI=body mass index. DF=degrees of freedom. HR=hazard ratio. ITT=intention-to-treat. *Cox model Wald CI and p-values are presented, each model was stratified by study. †Multivariable analysis was restricted to participants of white, black, or Hispanic race or ethnic group with complete covariate information; BMI was omitted due to collinearity with body weight and missing height observations.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Table 8. Cause of Death Classifications (ITT)
United States* Multinational† All Participants (Total)
5798.86 PY 4056.76 PY 3074.83 PY 1552.82 PY 8873.68 PY 5609.57 PY Cause of death
Freq (%) Incidence
Freq (%) Incidence
Freq (%) Incidence
Freq (%) Incidence
Freq (%) Incidence
Freq (%) Incidence
Suicide 6 (0.26) 1.03 1
(0.06) 0.25 2 (0.22) 0.65 0 0 8
(0.25) 0.90 1 (0.05) 0.18
Substance Abuse
3 (0.13) 0.52 2
(0.12) 0.49 0 0 0 0 3 (0.09) 0.34 2
(0.10) 0.36
Homicide 0 0 0 0 2 (0.22) 0.65 0 0 2
(0.06) 0.23 0 0
Accident 5 (0.22) 0.86 3
(0.18) 0.74 2 (0.22) 0.65 0 0 7
(0.22) 0.79 3 (0.14) 0.53
Unknown 3 (0.13) 0.52 1
(0.06) 0.25 5 (0.55) 1.63 0 0 8
(0.25) 0.90 1 (0.05) 0.18
Other cause‡
30 (1.29) 5.17 21
(1.29) 5.18 26 (2.84) 8.46 15
(3.23) 9.66 56 (1.73) 6.31 36
(1.72) 6.42
All causes, total
47 (2.02) 8.11 28
(1.72) 6.90 37 (4.03) 12.03 15
(3.23) 9.66 84 (2.59) 9.47 43
(2.06) 7.67
ITT= intention-to-treat. PY=person-years. Incidence is presented per 1000 PY. Death frequency and incidence are summarized using the intention-to-treat approach to account for all deaths reported during study follow-up. *United States: Suicide method: gunshot (n=3), hanging (1), overdose (n=2: 1 efavirenz, 1 efavirenz-free), method not reported (1). Substance abuse: substance not reported (2), cocaine (1), heroin (1), amphetamine (1). Accident type: motor vehicle (n=6: 4 efavirenz, 2 efavirenz-free), asphyxia (1 efavirenz-free), fire (1 efavirenz). †Multinational (n): Brazil (231), Haiti (100), India (255), Malawi (221), Peru (134), South Africa (230), Thailand (100), Zimbabwe (110). Suicide method: hanging (1), method not reported (1). Homicide: assassinated by gunshot (1), stabbed during a street quarrel (1). Accident type: motor vehicle (1), earthquake (1). ‡ Other cause: e.g., infection, cancer, organ failure.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Figure 1. Association between efavirenz and hazard of suicidality, according to baseline characteristics (ITT DSMB primary approach). DSMB=data safety monitoring board. IR=crude incidence rate. ITT=intention-to-treat. PY=person-years. Rx=medication. Univariate estimated hazard ratios were quantified from a Cox model with a Wald CI and p-value, stratified by study, and are plotted on a logarithm scale. The continuous age variable presents the efavirenz association when age is held fixed at 25, 35, and 45 years old. *The first p-value tests the main effect of efavirenz; subsequent p-values test for an interaction between efavirenz and the baseline characteristic. †Race or ethnic group was limited to white, black and Hispanic participants enrolled in the United States.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014
Supplement Figure 2. Association between efavirenz and hazard of suicidality, according to baseline characteristics (ITT sensitivity approach). IR=crude incidence rate. ITT=intention-to-treat. PY=person-years. Rx=medication. Univariate estimated hazard ratios were quantified from a Cox model with a Wald CI and p-value, stratified by study, and are plotted on a logarithm scale. The continuous age variable presents the efavirenz association when age is held fixed at 25, 35, and 45 years old. *The first p-value tests the main effect of efavirenz; subsequent p-values test for an interaction between efavirenz and the baseline characteristic. †Race or ethnic group was limited to white, black and Hispanic participants enrolled in the United States.
Downloaded From: http://annals.org/ by Jules Levin on 06/30/2014