Alcohol Consumption Advice in Sexual Health Clinics
Alcohol Consumption Advice in Sexual Health Clinics
The trial was a two-arm, parallel group, assessor-blind, randomised controlled trial. Ethical approval was obtained from West London Research Ethics Committee 3 (10/H0706/29), and the study protocol was registered with Controlled Clinical Trials (ISRCTN 99963322) prior to data collection. The detailed study protocol has been published elsewhere, and key features are described here.
Study participants were recruited in three sexual health clinics in central and west London. To take part in the study, potential participants had to be aged 19 years or above, be drinking excessively according to the Modified-Single Alcohol Screening Question (M-SASQ), and be willing to provide written informed consent. According to the M-SASQ, men who drink more than eight standard drinks on one occasion once a month or more, and women who drink more than six standard drinks on one occasion once a month or more are considered to be drinking excessively. We excluded any person who was unable to communicate in English sufficiently to complete baseline questionnaires, anyone who did not have an address or contact telephone number, and anyone who believed they may not be contactable again 6 months later.
Clinic staff gave all those attending the service a postcard with information about the study. Those who agreed to meet a researcher were given information about the study and asked to provide written informed consent. Following assessment of eligibility, baseline data were collected using a computer-assisted self-completion questionnaire. Participants were then randomised via a remote telephone service provided by an independent Clinical Trials Unit using permuted blocks, stratified by site. Block size was randomly assigned between four and six, with an equal allocation probability between arms. Researchers who collected follow-up data at 6 months were blinded to allocation status. Participants who completed the follow-up interview were offered a £15 honorarium in recognition of their time and any inconvenience related to their involvement in the study.
The SHEAR study had two treatment conditions. Those randomised to control treatment received a general health information leaflet with advice about smoking, alcohol, diet and exercise. In the brief advice group, participants were given feedback from the treating clinician (lasting 2–3 min) which consisted of information about the possible health consequences of excessive alcohol consumption, written information about alcohol and health and an offer of an appointment with an Alcohol Health Worker (AHW). Brief intervention with the AHW lasted up to 30 min and used the 'FRAMES' approach which combines active listening and feedback about risks associated with excessive alcohol consumption and emphasises personal responsibility for change. For any participant who was drinking at a harmful or dependent level, the AHW had the option of arranging a follow-up appointment or referring them on to local alcohol services for individual alcohol counselling, detoxification, or other treatments. If the participant was unable to attend an appointment on the day they were seen, then the AHW offered them an appointment at a later date or the option of telephone-based information and advice.
All clinicians who treated study patients received training on delivering brief advice prior to the start of the study. In addition to this, the lead researcher RS spoke to front-line clinicians on the days when recruitment was taking place. She provided support and advice to clinicians, gave feedback on their performance, and checked that brief advice was being delivered in accordance with the trial protocol. All AHWs who took part in the study were experienced practitioners who had undertaken specific training in counselling people who misuse alcohol. They each received regular clinical supervision during the trial and were encouraged to discuss work with trial participants along with other patients they saw during these sessions.
Clinicians delivering brief advice and AHWs were asked to complete a treatment proforma for each person they saw. These proforma were based on ones we used in a previous trial, and required clinicians to indicate whether they had delivered each of the four components of brief advice and AHWs to record the number and length of session(s), interventions delivered during the session(s) and further information of referrals that were subsequently made. Proforma were examined at the end of the study to examine treatment fidelity.
All outcomes were measured 6 months after randomisation and assessed behaviour in the 3 months prior to the date of the assessment. The primary outcome was mean weekly alcohol consumption (measured using the Form 90), and the main secondary outcome was the proportion of participants who reported any unprotected sex during the previous 3 months. Other secondary outcomes included mean units of alcohol consumed per drinking day, percentage days abstinent (both measured using the Form 90), and whether the participant was drinking excessively (defined as more than eight UK units/64 g of alcohol on one occasion for men, and more than six UK units/48 g for women). Sexual behaviour outcomes included total number of sexual partners, number of unprotected sexual partners, any incidence of regretted sex, any incidence of unprotected sex after drinking alcohol or while drunk, how long they knew their last sexual partner before they had sex with them, unplanned pregnancy and any new diagnosis of a sexually transmitted infection. These were assessed using questions derived from a previous study. Finally, we collected data on health-related quality of life (measured using the EuroQol-5 Dimensions scale; EQ-5D), and health and social care resource use during the past 6 months measured using a modified version of the Adult Service Use Schedule (AD-SUS). The cost of the brief advice was directly calculated from salaries using a microcosting approach, and national UK unit costs for the year 2010–2011 were applied to medication, hospital contacts and community health and social services.
The initial sample size calculation was based on identifying differences in mean weekly alcohol consumption as found in our previous trial of brief advice in an emergency department and suggested a minimum of 160 per arm. However, in the first few months of the trial the rate of recruitment was higher than expected and the sample size was therefore increased to provide additional power to test the primary and main secondary outcomes. The final sample size was based on a practical size of 380 per arm (760 in total). If the intervention reduced the proportion of participants who had unprotected sex from 65% to 50%, the power to detect a significant difference would be above 90%, assuming 25% drop out, and a clustering design effect of 1.15.
The statistical methods and trial design were specified a priori in a protocol paper and in a further detailed Statistical Analysis Plan. All analysis was performed in STATA (V.12). The primary outcome, mean weekly alcohol consumption, was compared between the randomised groups using random-effects linear regression, adjusted for age, sex and harmful alcohol use at baseline. The random-effects model takes into account clustering by sexual health clinic and, in the intervention arm, by treating clinician. Despite the skewed distribution of the outcome data, we used ordinary parametric models, which enables inference to be made about the arithmetic mean and are sufficiently robust to skewed outcome in a large sample. Robustness of the result was assessed by various sensitivity analyses, including non-parametric bootstrapping and non-hierarchical linear models. The main secondary outcome, proportion of participants reporting any unprotected sex, was analysed using random-effects logistic regression, adjusted for age, gender, and unprotected sex in the previous 6 months at baseline. Other secondary outcomes were analysed by linear, logistic or negative binomial regression. For rare outcomes, exact tests were used, and CIs calculated using mid-p method. All analyses were carried out according to the allocated randomisation arm, and two-sided p values were considered significant when below 0.05.
Patients with missing data were excluded in primary 'complete case' analyses, with multiple imputation using chained equations performed as a sensitivity analysis. Outcomes, covariates, predictors of outcomes and predictors of missingness were included in the imputation model, with both arms imputed separately in order to allow for interactions. A range of Missing Not at Random mechanisms were then considered in further sensitivity analyses to assess the robustness of the primary results.
The economic evaluation took a NHS/Personal Social Service perspective and had a 6-month time horizon. Standard parametric tests were used for the analysis of cost data, as recommended, with the robustness of the test confirmed using non-parametric bootstrapping. Effectiveness was assessed in terms of health-related Quality Adjusted Life Years (QALY) derived from the EQ-5D. The cost-effectiveness of the brief advice was assessed through the generation of cost-effectiveness acceptability curves (CEACs), which present the probability that the advice is cost-effective for different values a decision maker might be willing to pay for an improvement in outcome.
Methods
The trial was a two-arm, parallel group, assessor-blind, randomised controlled trial. Ethical approval was obtained from West London Research Ethics Committee 3 (10/H0706/29), and the study protocol was registered with Controlled Clinical Trials (ISRCTN 99963322) prior to data collection. The detailed study protocol has been published elsewhere, and key features are described here.
Study participants were recruited in three sexual health clinics in central and west London. To take part in the study, potential participants had to be aged 19 years or above, be drinking excessively according to the Modified-Single Alcohol Screening Question (M-SASQ), and be willing to provide written informed consent. According to the M-SASQ, men who drink more than eight standard drinks on one occasion once a month or more, and women who drink more than six standard drinks on one occasion once a month or more are considered to be drinking excessively. We excluded any person who was unable to communicate in English sufficiently to complete baseline questionnaires, anyone who did not have an address or contact telephone number, and anyone who believed they may not be contactable again 6 months later.
Study Procedures
Clinic staff gave all those attending the service a postcard with information about the study. Those who agreed to meet a researcher were given information about the study and asked to provide written informed consent. Following assessment of eligibility, baseline data were collected using a computer-assisted self-completion questionnaire. Participants were then randomised via a remote telephone service provided by an independent Clinical Trials Unit using permuted blocks, stratified by site. Block size was randomly assigned between four and six, with an equal allocation probability between arms. Researchers who collected follow-up data at 6 months were blinded to allocation status. Participants who completed the follow-up interview were offered a £15 honorarium in recognition of their time and any inconvenience related to their involvement in the study.
Interventions
The SHEAR study had two treatment conditions. Those randomised to control treatment received a general health information leaflet with advice about smoking, alcohol, diet and exercise. In the brief advice group, participants were given feedback from the treating clinician (lasting 2–3 min) which consisted of information about the possible health consequences of excessive alcohol consumption, written information about alcohol and health and an offer of an appointment with an Alcohol Health Worker (AHW). Brief intervention with the AHW lasted up to 30 min and used the 'FRAMES' approach which combines active listening and feedback about risks associated with excessive alcohol consumption and emphasises personal responsibility for change. For any participant who was drinking at a harmful or dependent level, the AHW had the option of arranging a follow-up appointment or referring them on to local alcohol services for individual alcohol counselling, detoxification, or other treatments. If the participant was unable to attend an appointment on the day they were seen, then the AHW offered them an appointment at a later date or the option of telephone-based information and advice.
All clinicians who treated study patients received training on delivering brief advice prior to the start of the study. In addition to this, the lead researcher RS spoke to front-line clinicians on the days when recruitment was taking place. She provided support and advice to clinicians, gave feedback on their performance, and checked that brief advice was being delivered in accordance with the trial protocol. All AHWs who took part in the study were experienced practitioners who had undertaken specific training in counselling people who misuse alcohol. They each received regular clinical supervision during the trial and were encouraged to discuss work with trial participants along with other patients they saw during these sessions.
Clinicians delivering brief advice and AHWs were asked to complete a treatment proforma for each person they saw. These proforma were based on ones we used in a previous trial, and required clinicians to indicate whether they had delivered each of the four components of brief advice and AHWs to record the number and length of session(s), interventions delivered during the session(s) and further information of referrals that were subsequently made. Proforma were examined at the end of the study to examine treatment fidelity.
Outcome Measures
All outcomes were measured 6 months after randomisation and assessed behaviour in the 3 months prior to the date of the assessment. The primary outcome was mean weekly alcohol consumption (measured using the Form 90), and the main secondary outcome was the proportion of participants who reported any unprotected sex during the previous 3 months. Other secondary outcomes included mean units of alcohol consumed per drinking day, percentage days abstinent (both measured using the Form 90), and whether the participant was drinking excessively (defined as more than eight UK units/64 g of alcohol on one occasion for men, and more than six UK units/48 g for women). Sexual behaviour outcomes included total number of sexual partners, number of unprotected sexual partners, any incidence of regretted sex, any incidence of unprotected sex after drinking alcohol or while drunk, how long they knew their last sexual partner before they had sex with them, unplanned pregnancy and any new diagnosis of a sexually transmitted infection. These were assessed using questions derived from a previous study. Finally, we collected data on health-related quality of life (measured using the EuroQol-5 Dimensions scale; EQ-5D), and health and social care resource use during the past 6 months measured using a modified version of the Adult Service Use Schedule (AD-SUS). The cost of the brief advice was directly calculated from salaries using a microcosting approach, and national UK unit costs for the year 2010–2011 were applied to medication, hospital contacts and community health and social services.
Statistical Methods
The initial sample size calculation was based on identifying differences in mean weekly alcohol consumption as found in our previous trial of brief advice in an emergency department and suggested a minimum of 160 per arm. However, in the first few months of the trial the rate of recruitment was higher than expected and the sample size was therefore increased to provide additional power to test the primary and main secondary outcomes. The final sample size was based on a practical size of 380 per arm (760 in total). If the intervention reduced the proportion of participants who had unprotected sex from 65% to 50%, the power to detect a significant difference would be above 90%, assuming 25% drop out, and a clustering design effect of 1.15.
The statistical methods and trial design were specified a priori in a protocol paper and in a further detailed Statistical Analysis Plan. All analysis was performed in STATA (V.12). The primary outcome, mean weekly alcohol consumption, was compared between the randomised groups using random-effects linear regression, adjusted for age, sex and harmful alcohol use at baseline. The random-effects model takes into account clustering by sexual health clinic and, in the intervention arm, by treating clinician. Despite the skewed distribution of the outcome data, we used ordinary parametric models, which enables inference to be made about the arithmetic mean and are sufficiently robust to skewed outcome in a large sample. Robustness of the result was assessed by various sensitivity analyses, including non-parametric bootstrapping and non-hierarchical linear models. The main secondary outcome, proportion of participants reporting any unprotected sex, was analysed using random-effects logistic regression, adjusted for age, gender, and unprotected sex in the previous 6 months at baseline. Other secondary outcomes were analysed by linear, logistic or negative binomial regression. For rare outcomes, exact tests were used, and CIs calculated using mid-p method. All analyses were carried out according to the allocated randomisation arm, and two-sided p values were considered significant when below 0.05.
Patients with missing data were excluded in primary 'complete case' analyses, with multiple imputation using chained equations performed as a sensitivity analysis. Outcomes, covariates, predictors of outcomes and predictors of missingness were included in the imputation model, with both arms imputed separately in order to allow for interactions. A range of Missing Not at Random mechanisms were then considered in further sensitivity analyses to assess the robustness of the primary results.
The economic evaluation took a NHS/Personal Social Service perspective and had a 6-month time horizon. Standard parametric tests were used for the analysis of cost data, as recommended, with the robustness of the test confirmed using non-parametric bootstrapping. Effectiveness was assessed in terms of health-related Quality Adjusted Life Years (QALY) derived from the EQ-5D. The cost-effectiveness of the brief advice was assessed through the generation of cost-effectiveness acceptability curves (CEACs), which present the probability that the advice is cost-effective for different values a decision maker might be willing to pay for an improvement in outcome.