Table 1 Key model assumptions

Risk group assumptions

 % of women in ‘high risk’ group (propensity for concurrent partners)

25%

 % of men in ‘high risk’ group (propensity for concurrent partners)

35%

Personality assumptions

 Decrease in odds of being high risk for each SD increase in conscientiousness score

0.67

[40, 79]

 Decrease in drinks per drinking day, for each SD increase in conscientiousness score

0.32

[41]

Inequitable gender norm assumptions (men only)

 OR for association between ‘low risk’ group and inequitable gender norms, not modifiable by interventions

0.50

[80]

 Effects of age, education, race, and urban/rural location on inequitable gender norms

Table S1

2016 DHS^a

 Effect of inequitable gender norms on incidence of concurrency

2.50

[15]

    Relative rate of endorsing inequitable gender norms after

     Gender-transformative interventions at individual level

0.50

Vague prior

     Gender-transformative interventions at community level

0.50

Vague prior

 Annual probability of reverting to pre-intervention gender norms

0.50

Vague prior

Alcohol assumptions

    Effects of age, sex, education, employment, race, urban/rural, marriage on

     Daily probability of alcohol consumption

 Table S2

2016 DHS^a

     Number of drinks per drinking day

Table S3

2016 DHS^a

 Increase in drinks per drinking day, comparing men who always endorse inequitable gender norms to those who never do

6.25

[15]

 Ratio of reported drinking frequency (% days) to true drinking frequency

0.65

Calibrated to

 Ratio of reported drinks per drinking day to true drinking volume

"

alcohol sales

 Probability of confounding between risk group (propensity for concurrent partners) and number of drinks per drinking day

0.50

Vague prior

    Relative rate of drinking (per day) after versus before intervention

     Single session of alcohol counselling

0.80

[23]

     Multiple sessions of alcohol counselling

0.50

[23]

 Annual probability of reverting to pre-intervention drinking levels

0.50

[81,82,83]

Casual sex assumptions

 Annual rate of entry into casual sex state: single high-risk males aged 17

0.08

[8]

 Annual rate of entry into casual sex state: single high-risk females aged 17

0.15

[3]

 Relative rate of entry into casual sex state: single low-risk males

0.30

[8]

 Relative rate of entry into casual sex state: single low-risk females

0.15

[3]

 Relative rate of entry into casual sex state: male binge drinkers

1.38

[2, 6, 8, 12]

 Relative rate of entry into casual sex state: female binge drinkers

2.15

[3,4,5, 12]

 Relative rate of entry into casual sex state per 0.1 decrease in inequitable gender norm score (men only)

0.88

[14, 17]

Condom use assumptions

 Reduction in odds of condom use, per day of binge drinking/week (OR)

0.83

[1, 7]

 Relative rate of condom use, comparing men who always endorse inequitable gender norms to those who never do

0.63

[10, 19]