Impact of text and email messaging on the sexual health of young people: a randomised controlled trial
- Megan S C Lim1,2,
- Jane S Hocking3,
- Campbell K Aitken1,2,
- Christopher K Fairley3,4,
- Lynne Jordan5,
- Jennifer A Lewis1,
- Margaret E Hellard1,2
- 1Centre for Population Health, Macfarlane Burnet Institute for Medical Research and Public Health, Melbourne, Australia
- 2Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia
- 3Melbourne School of Population Health, University of Melbourne, Australia
- 4Melbourne Sexual Health Centre, Carlton, Australia
- 5Family Planning Victoria, Box Hill, Australia
- Correspondence to Dr Margaret E Hellard, Macfarlane Burnet Institute for Medical Research and Public Health, Burnet Institute, Commercial Road, Melbourne 3004, Australia;
Contributors Please note that all authors, external and internal, had full access to all of the data (including statistical reports and tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. Guarantor: A/Prof Margaret Hellard, Burnet Institute.
- Accepted 5 July 2010
- Published Online First 17 March 2011
Objective To carry out a randomised controlled trial on the effect of a new method of health promotion—email and mobile phone text messages (short messaging service (SMS))—on young people's sexual health.
Methods 994 people aged 16–29 were recruited at a music festival to a non-blinded randomised controlled trial. Participants were randomised to either receive sexual health promotion messages (n=507) or the control group (n=487). The 12-month intervention included SMS (catchy sexually transmissible infections prevention slogans) and emails. Participants completed questionnaires at the festival at baseline and online after 3, 6 and 12 months. Outcomes were differences between the control and intervention groups in health-seeking behaviour, condom use with risky partners (new or casual partners or two or more partners within 12 months) and STI knowledge.
Results 337 (34%) completed all three follow-up questionnaires and 387 (39%) completed the final questionnaire. At 12 months, STI knowledge was higher in the intervention group for both male (OR=3.19 95% CI 1.52 to 6.69) and female subjects (OR=2.36 95% CI 1.27 to 4.37). Women (but not men) in the intervention group were more likely to have had an STI test (OR=2.51, 95% CI 1.11 to 5.69), or discuss sexual health with a clinician (OR=2.92, 95% CI 1.66 to 5.15) than their control counterparts. There was no significant impact on condom use. Opinions of the messages were favourable.
Conclusion This simple intervention improved STI knowledge in both sexes and STI testing in women, but had no impact on condom use. SMS and email are low cost, popular and convenient, and have considerable potential for health promotion.
Clinical trial registration number Australian Clinical Trials Registry - ACTRN12605000760673.
- Cellular phone
- electronic mail
- health promotion
- sexual behaviour
- sexually transmitted diseases
- sexual health young
In recent years, rates of sexually transmissible infections (STIs) have increased in almost every country in the world. Annual chlamydia notifications doubled in Australia between 2002 and 2007.1 Young people frequently engage in high-risk sexual behaviour; a recent survey found that up to 50% of Australian 16–29 year olds have had multiple partners in the past year and a significant proportion report using condoms inconsistently.2 Young Australian adults have poor knowledge of STIs other than HIV,2 3 and only about 7% of 16–24-year-old women are tested for chlamydia each year.4 Increasing STI rates in Australia and world wide strongly suggest that new ways to combat STIs are required.
New technologies have the potential to play an important role in sexual health promotion, particularly among young people who are the greatest users of new technologies and at greatest risk of contracting STIs.5 One of the most promising developments is short messaging service (SMS), also known as mobile phone text messaging. SMS is fast, portable, convenient, can be sent to multiple recipients simultaneously, and the cost of sending messages is low. SMS is also hugely popular among young people. Australian mobile phone subscriber data show that almost 90% of Australians aged between 16 and 34 are mobile phone users and that over 10 billion SMS are sent annually in Australia.6 Mobile phones are also an important mode of communication in developing countries, with broader reach than other forms of communication such as fixed-line telephones, television or the internet.7–9
Electronic mail (email) also has the potential to be used for sexual health promotion. Although it does not have the portability and convenience of SMS, email can transmit much more detailed messages and contain links to instantly take readers to websites of interest. Email is also widely embraced by young people—85% of Australians aged 18–24 used the internet in 2006, and email is one of the most common internet activities.10
SMS and email are already being used to improve health in a variety of ways,11–15 but their usefulness for health promotion in the STI field has not been evaluated using robust epidemiological methods.5 We conducted a randomised controlled trial to determine the impact on young people of sending regular email and SMS on condom use, knowledge of STIs and STI testing behaviour.
Materials and methods
Participants were recruited on 28 January 2006 at the Big Day Out (a day-long music festival held annually in several cities in Australia and New Zealand http://www.bigdayout.com/) in Melbourne, which was attended by approximately 40 000 young people. Participants were recruited at a market stall within the festival grounds. Participants either approached the stall or were approached by recruitment staff and asked if they would be interested in taking part. Eligible participants were aged between 16 and 29, were current residents of Victoria or Tasmania, had sufficient English skills to complete the questionnaire and had a working email address and mobile phone number.
At the Big Day Out music festival, participants completed a self-administered, paper-based baseline questionnaire, provided contact information and signed an informed consent form. The intervention group received regular email and SMS messages. SMS messages were sent every 3–4 weeks (a total of 14 over 12 months), while emails were sent less than monthly (a total of eight over 12 months). The SMS were short and catchy pieces of advice or information about STI or safe sex—for example, ‘Chlamydia: hard to spell, easy to catch—Use a condom’. SMS messages were tested in a focus group for understanding, relevance and amusement with a convenience sample of people aged 16–29. The emails were longer and contained two to five short paragraphs about a different safe sex or STI topic each month and links to other sexual health websites. Messages were sent at various times and on different days, with the SMS concentrated on Friday and Saturday evenings and the emails usually sent during weekday working hours. Participants completed follow-up questionnaires online at 3, 6 and 12 months after recruitment regardless of whether they had completed the previous questionnaire. Participants who completed all three follow-up questionnaires were given a CD voucher valued at approximately $A25.
Study hypothesis and objective
The study hypothesis was that young people's STI knowledge, health-seeking behaviour and use of condoms could be improved by the receipt of regular email and SMS messages.
The study's objective was to measure if sending email and SMS to a group of young people (intervention group) increased STI knowledge, health-seeking behaviour and use of condoms compared with a group of young people who did not receive the email and SMS (control group).
The main outcome measures were differences between the control and intervention groups in health-seeking behaviour (having an STI test in the past 6 months or talking to a general practitioner about sexual health or contraception), condom use with risky partners (defined as new or casual partners, or two or more partners within 12 months) and STI knowledge (true/false statements directly related to concepts included in the messages). Examples of statements used to test knowledge include: ‘Chlamydia can cause infertility’, ‘Chlamydia can be diagnosed by a urine test’, ‘People infected with chlamydia often don't have any symptoms and won't know they have the infection’. Participants' opinions of the SMS and email messages were also an outcome measure.
Sample size calculations
Three per cent of people in the target age groups are tested for chlamydia each year.16 We estimated that the testing rate in the intervention group would increase to 8% and remain stable among the control group at 3%. A sample size of 130 in each group was required to detect a 5% change in testing rates for the intervention group with 80% power and 5% significance. To account for an estimated 50% loss to follow-up, we initially aimed to recruit at least 520 participants in total—260 intervention, 260 control.
After recruitment, participants were randomly allocated to either the control or intervention group (using Microsoft Excel's random number function) by a study researcher. Owing to the nature of the intervention it was not possible to blind participants as to whether they were in the intervention group or the control group.
All randomised participants were included in the analysis. Our analysis was based on an intention to treat. Clustered weighted estimating equations were used to compare outcomes by intervention group at each time point; these were clustered by participant ID to allow for within-subject correlation with more than one measure on the same person. This method was chosen as it is able to account for participants dropping out at one time point and then returning to complete a later questionnaire.17
A weight for the missing data was calculated using methodology relating to post-stratification in sample surveys according to Carlin et al.17 Among participants with a missing time point, the fixed factors that were predictive of missing response were identified. Baseline factors associated with not completing all questionnaires were not living with one's parents, lower STI knowledge, more frequent binge drinking and illicit drug use in the past month. The number of participants at baseline for each of these factors (set in 4×4 cells) was determined, and at each time point the response rate for each cell was calculated and the reciprocal used to determine the weight for each respondent.
Interaction terms were included for time point and intervention group in the weighted estimating equation analysis. Gender was also included as an interaction term, but is only discussed when significant. All analyses included gender, age, education, drug use and alcohol frequency as potential confounding variables, and a p value of <0.05 was considered significant. Comparison of loss to follow-up between groups used a χ2 test. Analysis was performed with Stata 9.
Recruitment and follow-up
Nine hundred and ninety-four participants completed baseline questionnaires on 28 January 2006 and were randomised—487 to the control group and 507 to the intervention group (figure 1). Before completing a follow-up questionnaire, 97 participants (10%) were found to have insufficient contact details, defined as the owner of a supplied mobile phone number informing us that he/she was not a study participant or an email ‘bouncing’ four times in a row (figure 1).
The follow-up questionnaires were made available to participants on 1 May 2006, 31 July 2006 and 31 January 2007, and were closed after 6 weeks. Five hundred and eighty-seven (59%) participants completed at least one follow-up questionnaire, including 337 (34%) who completed all three follow-up questionnaires and 387 (39%) who completed the final questionnaire. There was no significant difference in the proportion of people in the intervention group (40%) and the control group (46%) who completed the 12 month questionnaire (p=0.07).
The two groups were comparable at baseline with respect to both demographic and behavioural variables (table 1). At baseline the median age of participants was 19 years, 58.0% were female and the majority were born in Australia. Overall, 81.9% had ever had sexual intercourse and 9.6% had had an STI test in the past 6 months. At baseline the median score on a test of STI knowledge was five out of eight questions correct.
Both groups showed improvements in knowledge across the four time points. After 12 months of intervention, knowledge was significantly higher in the intervention group than the control group for both sexes (tables 2 and 3).
At 12 months, female participants in the intervention group were significantly more likely to have had an STI test than female subjects in the control group (table 2). There was no difference in the proportions of male participants who reported testing between the intervention group and the control group (table 3).
Acceptability of the intervention
At 12 months, 96% of participants in the intervention group recalled receiving SMS about sexual health and 84% recalled receiving email compared with 92% recalling SMS and 73% recalling email at 6 months, and 93% recalling SMS and 58% recalling email at 3 months. Over three-quarters of participants who received SMS had shown them to someone, generally a friend or partner; this behaviour was much less common for email (table 4). Sixty-six per cent of those who recalled email and 62% who recalled SMS at 12 months agreed that they learnt something from them; at 12 months, 13% and 24% found the emails and SMS annoying. The SMS were described as interesting or entertaining by 69%, compared with 45% for emails.
As far as we know, this is the first randomised controlled trial evaluating the use of SMS and email messages to improve the sexual health of young people. An important outcome of the study was that young women—but not young men—in the intervention arm were significantly more likely to have had an STI test or talked to their doctor about sexual health compared with the control arm. A second notable finding was that both young men and women in the intervention arm demonstrated increased knowledge after receiving these messages compared with a control group. Third, the intervention had no effect on condom use in either gender.
The increase in reported sexual health-seeking behaviour among young women is a promising result. Chlamydia testing is important because it decreases the duration of infection, reducing time available for transmission and the risk of complications associated with infection. Australia is currently implementing a chlamydia screening programme with the aim of increasing testing among young women. Our trial demonstrated that SMS and email may be an effective way to promote and encourage uptake of this chlamydia screening programme. The lack of impact on young men is of concern, as young men are known to access healthcare, in general, significantly less often than women (70% of young men compared with nearly 90% of young women visit a GP for their own health each year).18 Because much of the STI testing in young men is opportunistic, this reduces their likelihood of being tested. Additionally, at the time of this study, the Royal Australian College of General Practitioners did not recommend annual testing for men—only women.19
When the sample size calculation was made in 2004, the most recent data on chlamydia testing rates showed that 3% of young adults were tested in 2002. By 2004, testing rates in the same age group had increased to 7% and, more recently, annual testing rates have increased to 9%.16 20 In the baseline survey, participants were asked whether they had been tested for any STI in the previous 6 months (the survey was administered in January 2006). At this time, 8% of participants reported a chlamydia test in the past 6 months. Given that at that time annual chlamydia testing rates had already increased to 7–9%, it appears that baseline testing rates in our study population were similar to, or higher than, those in the community generally. At 12 months the proportion of young women in the control group reporting a test remained close to the community rate at 10% while testing amongst the intervention group increased to 18%, suggesting that the intervention, rather than background change in testing rates, was the cause of the increase.
The increase in the proportion of female intervention recipients who reported discussing sexual health issues with a clinician is also important. Such discussions give clinicians opportunities to educate and inform their patients and to offer testing if they consider their patients to be at risk. It also suggests that many young women receiving the intervention took an active interest in their sexual health.
The intervention produced significant increases in STI knowledge, which shows that those who received the messages read, understood and learnt from them. Increases in STI knowledge does not necessarily translate into changes in behaviour as many other complex factors are involved in health and preventive behaviours (eg, self-efficacy, perceived social norms, behaviour motivation and external factors).21–24 Nonetheless, a basic level of knowledge is necessary to enable young people to make more accurate decisions about sexual health, and among women in our study at least, a corresponding change in health-seeking behaviour was seen.
Condom use was not increased by the intervention. A meta-analysis by Albarracin et al showed that although communications recommending condom use can often lead to changes in condom-related knowledge and attitudes, they are unlikely to have a significant effect on actual condom use.25 Other studies show that condom promotion is more effective when condom skills training is provided.26 27 Perhaps a more multifaceted intervention is required to influence this complex behaviour—for example, in future trials we will use video and picture phone messaging.
Our previous analysis suggested that Big Day Out attendees have a slightly greater risk for drug taking than the general population.28 Consistent with this finding, our analysis found that a high proportion of study participants were also at high risk of an STI. A high proportion of participants were sexually active and a large proportion did not always use condoms with casual partners or new partners. This suggests that our intervention was effective in increasing knowledge and health-seeking behaviour in a population that was at high risk of an STI.
The major limitations of this trial are the self-reported nature of our outcome measures and loss to follow-up. The validity of using self-reports of sexual behaviours to assess STI risk has been queried.29 While acknowledging that there are no simple ways to validate self-reported sexual behaviour, our group, in an effort to assess the impact of recall bias, compared the correlation between weekly self-reported sexual health diaries and data collected in a retrospective questionnaire in a group of young people attending the same music festival but in the following year. Data collected in retrospective sexual behaviour questionnaires were found to agree substantially with data collected through weekly self-reported diaries.30 The accuracy of self-reported testing has also been shown to be suboptimal,31 32 but respecting our participants' anonymity meant we were unable to validate their reported testing with their clinician.
Loss to follow-up can introduce bias to a trial if those who withdraw differ from those who remain in the study. Higher levels of loss to follow-up increase the likelihood of respondent bias in later questionnaires, therefore the results of this trial should be interpreted with care as only 34% of participants completed all four questionnaires. However, just 56 participants (5.6%) actively withdrew from the study (figure 1). The attrition observed across the three follow-up questionnaires can be explained in part by loss of contact with the participants and not solely deliberate non-response. In order to maintain participants' anonymity only limited contact information was obtained, so the extensive follow-up techniques usually employed to maintain participation in clinical trials were unavailable. Participants who did not complete the questionnaires after the reminder SMS and emails were not contacted in any other way. There was also no way to determine if email addresses and mobile phone numbers were active or belonged to the enrolled participant. Participants may have changed their mobile number or email address, or provided incorrect or illegible information at recruitment. It was not possible to determine whether the SMS we sent were received, but our software enabled us to determine that only 35–50% of each round of emails were opened. Possibly, many participants never received any contact from us but this could not be confirmed. We used well-recognised statistical methods (weighted analysis) to control for any potential biases due to loss to follow-up.
It is worth noting that our study intervention reflects the real world of community-based public health interventions (such as billboards, posters and television adverts) where the true impact of such intervention is rarely measured. Importantly, an increasing number of public health interventions in young people involve the use of SMS or other mobile technologies. Evaluations of these projects suggest they are feasible, culturally acceptable and well recognised by the groups being targeted.5 15 Therefore it is important to measure the impact of such interventions as rigorously as possible, even though field-based research is methodologically challenging and typically suffers greater loss to follow-up. Recruiting participants from a community venue is a strength of this trial as participants are more likely to be at risk of STI than those recruited from an educational setting, but more representative of the average population than those recruited from a clinical setting (in 2007, 40% of Australians aged 18–24 years had attended a popular music concert in the past year).33 34
In conclusion, this study demonstrated that a simple intervention using SMS and email altered health-seeking behaviour in young people and increased their knowledge of STIs, but did not affect condom use. Future directions for research could include trials on other population groups, whether more broadly or more narrowly defined; implementation in developing countries; separating out the impacts of SMS and email; trialling different styles of message; and testing the method's effectiveness in other areas of health promotion, such as smoking cessation or physical activity. The continued development and uptake of new communications technology provides valuable new opportunities to reach young people and provide them with accurate health information.
What is already known on this subject
Young people's relatively high-risk sexual behaviour increases their chances of acquiring STI, notifications of which have increased in Australia and elsewhere in recent years.
New communications technologies such as short messaging service (text messages) and email are already being used in health settings and are enthusiastically embraced by young people.
These methods of communication hold considerable promise for health promotion related to STI, but their use has not been rigorously evaluated.
What this study adds
This study was the first randomised controlled trial of the use of short messaging service (SMS: text messages) and email messages in sexual health promotion.
An SMS and email intervention can significantly increase reported health-seeking behaviour among young women and young people's knowledge of STI but has no impact on condom use.
The authors are grateful to the participants, recruitment staff, focus group participants, Bianca Fiebiger and the Big Day Out, Emma and Tom's Life Juice, qubeGroup, Qmani, and Andrew, Ben and Cait Symon.
Funding Australian Health Ministers Advisory Council Priority Driven Research Program, 2005. The funders had no role in study design; in collection, analysis, or interpretation of data; in the writing of the report; nor in the decision to submit the article for publication. The researchers are independent of the funders.
Competing interests None.
Ethics approval The study was registered with the Australian Clinical Trials Registry (ACTRN12605000760673 - http://www.anzctr.org.au/) and ethics approval was obtained from the Victorian Department of Human Services Human Research Ethics Committee (project #77/05).
Provenance and peer review Not commissioned; externally peer reviewed.