Alcohol Text Messages: A Developmental Study

A convenience sample of 51 adult men (> age 18 years) was recruited through social media (Facebook, Twitter and LinkedIn) from June through July 2016. Entry criteria included being a current or expectant Australian father, having ever consumed alcohol and having access to the Internet. Participants were entered into a draw to receive a $100 store gift voucher after completing the survey. Ethical approval for the study was granted by the University of Newcastle, Australia Human Research Ethics Committee, approval number H-2016-0136.

Thirty short messages (< 160 characters) were developed by the first author (MR). The messages addressed alcohol-related harms in relation to infant health, the spousal relationship, father-infant relationship and the fathers’ health. Messages were based on national alcohol guidelines (National Health and Medical Research Council 2009) regarding alcohol consumption. Motivational Interviewing (MI) techniques were considered in the construction of the text message language (Miller and Rollnick 2012), with a primary focus of encouraging self-reflection (Miller and Rollnick 2012). The Stages of Change model (Prochaska and DiClemente 1983; Prochaska et al. 2013) underpinned the text message design. This model focuses on self-directed intentional change, as opposed to imposed change for the modification of substance use (DiClemente 1993), and has proved effective in treating substance use disorders for over 30 years (Prochaska et al. 2013). Individuals are thought to move through a series of stages: precontemplation, contemplation, preparation, action and maintenance when changing unhealthy behaviours. Providing information about alcohol-related harm during the perinatal period may prompt some fathers to contemplate reducing the amount they drink or take action by ceasing alcohol. Information-seeking behaviour (by pressing on a hypothetical hypertext link relating to a particular text topic) indicates a proactive response according to this model.

Messages were refined using a modified Delphi system (Dalkey and Helmer 1963) with feedback from co-authors (ALB, AJD) and an alcohol researcher at the University of Newcastle. The text messages were presented using two types of message voice: a child’s voice, as if the father’s own child were addressing him e.g. ‘Hey dad, do you know how many drinks you can have before it affects your health?’ or a second person voice e.g. ‘Do you know how many drinks you can have before it affects your health?’ The messages had a Flesch-Kincaid grade level of 5.4 to ensure ease of comprehension.

All participants viewed and rated the same 30 messages during an on-line survey using the Qualtrics platform (Qualtrics software 2015). Responses were entered by participants via smart phone, tablet or computer into the on-line survey. The messages were not sent to participants’ mobile phones but rather were developed so they could be sent as SMS text messages in future research. Participants rated messages for importance using two 3-point Likert scales (Important, Can’t Say, Not Important) and whether participants would use a hypothetical hypertext link for more information on the message topic (Likely, Can’t say, Unlikely) (active links were not inserted). The order of the SMS text messages was randomised for each participant using an inbuilt randomisation feature of the program. The text message responses were then coded into one of two categorical variables: For importance, 0 = ‘not important’ or ‘can’t say’ and 1 = ‘important’, and for ‘likelihood’ of seeking further information, 0 = ‘not likely’ or ‘can’t say’ and 1 = ‘likely’. This produced a continuous variable for importance of the second person voice (0–15) and child voice (0–15), summed together for a total importance score (0–30), and a continuous variable for likelihood of seeking further information for the second person voice (0–15) and child voice (0–15) summed together for a total likelihood score (0–30).

National Health and Medical Research Council Guidelines for alcohol consumption (National Health and Medical Research Council 2009) were used in the design of the SMS text messages. Alcohol guidelines one to four regarding recommended alcohol consumption were presented to fathers in the survey before they answered the two final questions in the on-line survey. Fathers self-rated the importance and difficulty of drinking alcohol at levels recommended to reduce short-term and long-term harm based on the Australian alcohol guidelines using two 10-point Likert scales (with 1 equal to not important through to 10 being extremely important, and 1 being equal to not at all difficult through to 10 being very difficult to drink at recommended levels).

Data were analysed using SPSS version 22 software (IBM Software 2016). Analyses were performed using a 2 × 2 × 2 mixed ANOVA design. The first factor was a within-subjects factor of voice, with two levels, child voice or a second person voice. The second factor was a between-subjects factor of ‘at-risk’ alcohol consumption, with two levels, operationalized by summing self-reported AUDIT C scores (Bush et al. 1998) as low risk (0–3) or high risk (4–12). The third factor was a between-subjects factor of psychological distress, with two levels, operationalized by summing self-reported Kessler-6 (Kessler et al. 2002) scores as low distress (6–12) and moderate (13–18) to high distress (19–30). The moderate and high distress groups were collapsed into one group due to the small sample size.

After completing the on-line survey, participants were asked if they were interested in being contacted by telephone for a brief qualitative interview (N = 17) Data were transcribed verbatim and pseudonyms were assigned prior to coding using Template Analysis (King 1998). A priori themes included the acceptability of the messages, the importance of the alcohol messages, receptiveness to the design of messages in second person and child voice format and willingness to seek additional information. Themes were identified, coded and summarised into a template.

Seventeen participants (17/51, 33%) who agreed to be contacted for a telephone interview included both younger fathers (8/17) and older fathers (9/17). The representation of participants who consumed alcohol at high-risk levels (10/17, 59%) was similar to the quantitative sample of high-risk drinkers (65%). The proportion of low-risk drinkers (6/17) were evenly spread between the age groups. The final participant was a younger father who consumed alcohol at high levels when his first child was born but was now an ex-drinker.

Most interview participants responded positively to the text messages and noted they found them to be non-judgmental in style. Jim, an older father and risky drinker said, ‘I didn’t find them preachy. I thought they were straight up statements’. Craig, a younger father and risky drinker said, ‘I like the way it did not imply you have a drinking problem’. Alex, a younger father and low-risk drinker, also indicated acceptability and potential benefit: ‘I will be absolutely honest …for my first child, I probably needed them’. However, one participant, John, a younger father and risky drinker, did not like the tone of the messages stating: ‘I found that the messages were quite condescending’. The potential impact of a text messaging intervention for fathers was highlighted by Ryan: ‘For a number of years I was a very heavy drinker… at that time having an objective external voice prompting me to think about my choices would have been really useful.’

Participants also consistently indicated that they thought the alcohol text messages were important. The messages rated as most important among participants, included the effects of paternal alcohol consumption on paternal bonding, sleep deprivation, developmental problems, behavioural issues and FASD. Participants repeatedly stated they thought messages that directly connected the father’s behaviour with their child’s health and emotional wellbeing had greater impact than the messages about their own health. For example, Craig a younger father and risky drinker, said:

I would try and tailor it more to the relationship - like the ones purely about your health needs to be related back to your relationship with the baby. Especially if you are targeting the niche market for fathers where the emotion is fresh … like the one about bonding, like the ones about how alcohol affects your family.

Cameron, an older father and risky drinker, felt that ‘the ones that discuss the importance of connecting on an emotional level with your child’ were particularly important for him. Kelvin, an older father and risky drinker, said ‘It is really important to know that alcohol can damage the baby and this could bring awareness to men about fatherhood…you know link the two up.’

The effects of drinking on sleep and ability to cope with stress were frequently raised as issues for new fathers. As Marc, an older father and low-risk drinker, said:

I think the things that struck me as particularly useful were … regular and or heavy alcohol use can negatively affect your sleep, coping with stress, energy levels … most blokes I know still think that alcohol helps them sleep, cope with stress… I think the big one is the sleep deprivation you and mum go through for the first year or two with each child. I think that is a big selling point for change in alcohol use.

Interestingly, the ten participants interviewed with at-risk AUDIT-C scores who responded to the qualitative survey provided positive feedback about the messages. Jim, an older father who scored in the high range on both the AUDIT-C (8) and the K6 (20) said, ‘Look I probably scored every single one of those statements as important’. He then went on to say they were ‘sort of confronting …like the statements about the impact of the drinking on the child in a sort of future sense’ and that made the messages ‘…more important in a sense’.

Further, participants reported that messages using the child’s voice made them aware of their paternal responsibilities and how their drinking behaviour may impact upon their child. For example, Jerome, a younger low-risk drinker, said the child voice ‘makes you think there is another person involved… makes it concrete... I think it reminds you are dealing with a person’. Further support for salience of the child’s voice came from Ryan who said, “I thought for me it felt like… this even makes me feel more intensely’. Similarly, Kelvin indicated ‘The ones from the child helped make you think that there was someone else involved. It made you think about someone other than yourself and that you are affecting. I think it worked better than the other messages’. Darren, an older father and low-risk drinker suggested ‘I think that the ‘hey dad’ would work for new fathers especially because they are not used to being called dad yet. I think that could add something extra’. The fathers who previously drank heavily were generally more enthusiastic about the potential of the child’s voice to raise awareness about the impact of their alcohol consumption.

Importantly and in contrast to the quantitative data, the qualitative data also indicated potential for the child’s voice to be seen as manipulative and ineffectual. A minority of participants expressed concerns about feeling manipulated. For example, Marc, an older father and risky drinker, who generally approved of the messages, had an issue with the child voice:

It is probably just me, still with some vestigial oppositional tendencies, but I found the messages with the so called ‘child voice’ got my back up a bit, it almost seemed like a form of emotional blackmail… trying to get at me through my baby… and I felt like I would not buy into it by responding to any of them. Whereas the messages directed to me as one adult to another seemed very reasonable and I was happy to take them on board.

Another area explored during the interview was whether the messages made participants more likely to seek more information. A common theme expressed by participants was that they liked the format of brief messages with the option of receiving further information. For example, Craig said ‘I have always been a fan of if I receive something in communication you’ve got a really short brief message, a snap shot, and if you want more information you go and get more’. Most of the fathers indicated that they would access information if it were directly linked to fatherhood and developmental concerns. For example, as noted by Darren:

I like the ones that link the fathers to the kids in developmental studies. If you go on a website there is heaps of information for mothers but not a huge amount for men, so if you could sort of say one of the benefits of you drinking less, it’s been shown in this study to have this affect that would be pretty interesting.

When fathers were asked for their own ideas regarding alcohol text messages they thought were most important to send to future fathers, most expressed support for messages about reducing excessive paternal alcohol consumption. Interestingly, ex-drinkers and risky drinkers indicated a preference for the strongly worded alcohol messages. Ryan suggested, ‘Is this the person you want to be? Is this the parent you would want to be?’ Jim suggested, ‘Have you thought about how many drinks you’re going to have tonight? It’s not impossible to stop at two!’ Another participant highlighted the importance of paternal health in the post-partum period. For example, Marc suggested: ‘New dads aren’t getting enough sleep, feel tired all the time, and certainly don’t want to do anything that makes it worse’. Aiden, an older father and risky drinker had numerous suggestions including self-health messages ‘Did you realise more than 3-4 drinks a day is considered excessive drinking?’ Suggested text messages also focused on fathers’ psychological health. For example, Jim suggested: ‘Did you know that excessive drinking can be isolating and lead to depression?’

Text messages concerning the effects of drinking alcohol on family life were also offered by a small group of fathers, like Aiden, ‘An hour spent drinking after work is an hour you could have played backyard cricket with your children’. Marc went further by highlighting the importance of prenatal health for fathers: ‘Getting healthy before conception is important… Women have been on to this for a long time’. These text message suggestions highlight the alcohol-related harms that this group of fathers thought were important to raise awareness of in new fathers. These data also demonstrate that the fathers who participated in the study were reflecting on alcohol-related harm and how it impacted their parenting. Many participants openly discussed how the text messages made them reflect on their own drinking.