To the best of our knowledge, this is the first study to examine how RTC impacts substance use outcomes while controlling for the effect of three different brief interventions delivered in EDs in an African country where infrequent but heavy episodic drinking is typical [2]. Similar to studies conducted in EDs in high-income countries [27–29, 53, 54], this study found that RTC was an important predictor of response to substance use interventions. More specifically, participants who reported greater recognition that their substance use was problematic and of the need for change reported larger reductions in their post-intervention substance use involvement. There are several potential explanations for this association; participants who had more recognition of the need for change at the beginning of the intervention may have paid more attention and been more receptive to the intervention content and intervention materials, may have practiced implementing the skills taught during the intervention more diligently, and may have sought out additional information about how to reduce their substance use involvement than participants with lower levels of recognition. All of these factors may have contributed to participants with higher levels of recognition being more successful at reducing their substance use involvement.
As these findings suggest that people are more likely to be amenable to changing their substance use if they are aware that a problem exists and recognise the need for change, interventions to modify substance use should consider including components that enhance problem recognition and awareness of the need for change. Future adaptations of our interventions may wish to include more use of motivational interviewing techniques, particularly given evidence that the use of motivational interviewing promotes RTC [44, 54]. This seems particularly important for this ED population for whom we found very low levels of problem recognition, suggesting that many participants had not yet considered changing their substance use involvement despite already experiencing moderately severe adverse consequences. These low levels of RTC are not surprising given that participants were seeking health care rather than assistance for substance-related problems. Similar levels of RTC have been observed in other studies of non-treatment-seeking populations [22, 28, 34, 36].
Interestingly, participants with alcohol use problems responded better to the brief substance use interventions than participants with other drug problems. This suggests that although brief interventions may be adequate for helping individuals reduce their alcohol use, they may not be sufficient for helping people change their drug use. This finding is not altogether surprising given evidence from other settings that brief interventions in non-treatment seeking populations are efficacious for reducing alcohol use [13–16], but not for reducing illicit drug use [19–23]. However, it remains unclear why these interventions are effective for reducing alcohol use only. Although not explored here, it is possible that type of substance used may have moderated the effect of problem recognition on substance use outcomes. Future ED studies that are adequately powered should consider testing this hypothesis.
In contrast, participants who had higher baseline scores on the SOCRATES Taking Steps scale (reflecting self-reported action towards changing substance use) reported greater substance use involvement at 3 month follow-up, suggesting that their substance use problems had worsened over time. This finding was surprising as studies with treatment-seeking populations have consistently reported that higher scores on this scale predict greater improvement in substance use outcomes [22, 45, 46]. There are two possible explanations for this finding. First, these earlier studies all measured change in longitudinal substance use outcomes at 18 months to 2 years post-intervention. Changing substance use requires individuals to be motivated to take action and to learn how to overcome barriers to change, which takes time and practice [55]. Consequently, we may not have had a long enough follow-up period to observe a positive association between “Taking Steps†scale scores and substance use involvement. Second, some participants may have already been attempting to reduce their substance use involvement prior to receiving the intervention. It is possible that these participants may have relapsed back to old patterns of substance use during the course of the study, and thus reported greater substance use involvement at follow-up. This second explanation is partially supported by findings that participants with more severe types of substance use problems (specifically poly-substance use or methamphetamine use) scored higher on this component of RTC at baseline, suggesting that they had already made some efforts to change. Therefore the negative association found between action-oriented RTC and substance use outcomes could possibly be accounted for by the fact that people with methamphetamine and other complex types of substance use disorders tend to have poorer responses to substance use interventions than individuals with other types of substance use [25]. Future intervention studies should consider collecting more extensive data on participants’ substance use histories and change attempts to further explore this explanation.
In addition, we explored factors associated with each component of RTC that could potentially serve as targets for future interventions to improve RTC substance use in this population. First, we found that participants presenting with injuries that were related to their substance use had greater recognition of the need for change than those who presented without these injuries. This highlights the value of screening all patients who present with injuries for possible substance-related problems as these injuries may offer a window of opportunity for intervening with patients when they are more likely to recognise the negative consequences of continued substance use. In addition and similar to other studies [22, 36, 55], we found that greater problem severity (as reflected in more symptoms of depression and more complex substance use problems, such as methamphetamine use) was positively associated with more recognition that change was needed. Even though our findings suggest that individuals with greater substance use problem severity may not respond as well to BIs interventions as those with less severe problems, we are not suggesting that these individuals are not provided with a BI. In such instances, it is still useful to intervene given evidence that BIs lead to improved substance use treatment utilisation and lower rates of ED utilisation [24]. Together these findings suggest that a certain level of distress or dissatisfaction with current circumstances is needed before a person will recognise the need to make a change. This is consistent with evidence that depressive symptomatology can sometimes have an adaptive function [56]. Given the key role that mood plays in the recognition of the need for change, interventions to enhance RTC could benefit from not only exploring the potential benefits of changing substance use for health and relationships, but also the potential benefits of change for emotional well-being.
Although some distress may help people acknowledge the need for change, our findings suggest that too much distress may impede action towards change. In this study, more symptoms of depression and complex drug problems (in particular methamphetamine use) were associated with greater ambivalence about whether or not to change substance use involvement. One explanation may lie in findings from previous studies that people with more severe substance use disorders and psychiatric disorders have low self-efficacy for resisting substance use [57]. Lack of confidence in one’s ability to resist substance use has been shown to contribute to ambivalence about change and to negatively influence decisions about whether to reduce substance use [58]. Another explanation may be that people with depression and other mental disorders are more likely to have problem solving deficits and to use substances as a way of avoiding their problems [59]. People who use substances as a coping strategy are more likely to have low self-efficacy for resisting substance use and are less likely to take action towards change [60]. However, as we did not directly assess self-efficacy for resisting substance use or problem solving, these explanations require further evaluation in studies that measure self-efficacy and problem solving as well as RTC. Regardless of the reasons for this finding, there may be some benefits to providing substance-using individuals in psychological distress with additional counselling to help them resolve their ambivalence towards change.
Finally, apart from severity of substance use problems, the only other variable associated with the “Taking Steps†scale was whether participants presented with an injury that was preceded by substance use. Participants who presented with a substance use-related injury had lower scores on this scale than participants who did not present with an injury. The lower levels of action-oriented RTC in this patient population group are not surprising given that they are still experiencing many problems related to their substance use. However, as this is a population at high risk of recurring substance use-related injuries [7], it is important to continue to reach them through interventions that build RTC.
These findings should be considered in the light of some limitations. First, the sample size is relatively small and limited to participants recruited from three EDs in one province; as such findings may not be generalizable to patient populations from other locations. Second, the follow-up period was relatively short and it is not known whether the effects of RTC on substance use involvement persist over time. This is a question for future research. Third, we did not collect detailed information on the type or severity of substance-related injury sustained which may have impacted on our ability to find a significant relationship between injury and change in substance use outcomes. Future studies should consider exploring the relationship between severity of substance-related injury and change in substance use involvement. Fourth, 46Â % of enrolled participants failed to provide follow-up data. Attrition analyses suggests that attrition was not associated with RTC or degree of substance use involvement, increasing confidence in our findings. Nonetheless, future studies in ED settings should consider how participant retention can be improved.