Abstract
Aims:
The aims of this study were to evaluate and compare popular beliefs and attitudes regarding alcohol conversations in healthcare in Sweden and Norway; and to explore which factors were associated with different levels of support for alcohol-prevention work in the two countries.
Methods:
Population-based cross-sectional surveys were conducted in Sweden (n = 3000) and Norway (n = 1208). Logistic regression was used to identify the characteristics of participants who were supportive of routine alcohol screening and brief intervention delivery.
Results:
A higher proportion of Swedish respondents agreed to a large extent that healthcare professionals should routinely ask about alcohol consumption. In addition, a higher proportion of Swedish respondents compared to respondents from Norway agreed that healthcare providers should only ask about patient’s alcohol consumption if this was related to specific symptoms. There were similar correlates of being supportive of routine alcohol screening and brief intervention delivery in both countries. Support was lower in both countries amongst moderate and risky drinkers, and among single adults or those on parental leave, but higher amongst older individuals. Having had an alcohol conversation in healthcare increased the level of support for alcohol prevention in routine healthcare among risky drinkers.
Conclusions:
There is a high level of support for preventative alcohol conversations in routine healthcare in Norway and Sweden, although there was a lower proportion of respondents who were positive to alcohol prevention in routine healthcare in Norway compared to Sweden. Experiencing alcohol conversation may positively affect risky drinkers’ attitudes towards and support for alcohol prevention. Thus, more frequent alcohol conversations in routine healthcare may also result in increased level of support for alcohol prevention among risky drinkers.
Despite strong evidence for the effectiveness of brief alcohol interventions to reduce hazardous and harmful drinking (Kaner et al., 2018), their adoption in routine healthcare has been slow (Johnson et al., 2011; Nilsen, 2010; Vendetti et al., 2017). Factors affecting implementation have been researched extensively from the perspective of healthcare practitioners. Barriers include: scepticism about intervening with patients who do not have alcohol-related symptoms; lack of time, knowledge and training; and concerns about patients’ resistance to alcohol-related discussions (Beich et al., 2007; Johnson et al., 2011; Nilsen, 2010; Rapley et al., 2006).
However, there has been far less research exploring attitudes within the wider patient population towards the routine delivery of alcohol prevention in healthcare. Although previous evidence suggests that patients are generally comfortable to discuss alcohol issues with healthcare providers, most studies have been relatively small in scale, and/or focused on the views of clinical populations, such as problem drug users (Hutchings et al., 2006; Lundin et al., 2017). Two national population surveys have explored population attitudes towards alcohol-related discussions in healthcare settings. Nilsen et al (2012) found that although there was considerable support in the general population in Sweden for the delivery of brief interventions (BI) in healthcare, attitudes were less positive amongst hazardous drinkers, and many respondents felt that healthcare providers should only ask about a patient’s drinking if they presented with alcohol-related symptoms. However, this study was conducted nine years ago, and beliefs and attitudes may have changed since then. In a similar, more recent, English study, O’Donnell et al. (2018) also found that most respondents supported the routine implementation of alcohol prevention in healthcare, although around one in ten saw alcohol as a personal matter, and not something that healthcare practitioners should ask their patients about. Respondents from lower socio-economic groups were also less supportive of alcohol being addressed in routine healthcare compared to other respondents.
Sweden and Norway share many social, political and economic characteristics. Sweden is a part of the European Union (EU), while Norway, although not a member, is included in the EU internal market via the European Economic Area. The two countries have similar alcohol consumption patterns (Moskalewicz et al., 2016), with an average 9.5 litres of pure alcohol consumed by drinkers aged 15 years and older in Sweden in 2010, compared to 9 litres in Norway the same year, dropping to 9.2 litres in Sweden and 7.5 litres in Norway in 2016 (WHO, 2018). However, a higher proportion of adults in Norway report having been drunk in the past 12 months compared to those in Sweden (82.8% vs. 40.1%) (WHO, 2018). Both Sweden and Norway also have restrictive government alcohol policies to limit alcohol consumption and prevent alcohol-related harm, focussing on limited access, high taxes and a marketing ban. In Sweden, the alcohol retail monopoly (Systembolaget) reports to the Ministry of Health and Social Affairs and operates without a profit incentive. Similarly, the Norwegian retail monopoly (Vinmonopolet) reports to the Ministry of Health and Care Services. Since 1996, Vinmonopolet has been a strictly retail monopoly, whilst prior to that date, the monopoly also controlled the import of all wine and spirits, as well as some production of spirits.
However, although both countries have implemented a range of preventive initiatives to reduce alcohol-related harm, including measures aimed at increasing the delivery of screening and brief interventions in healthcare, approaches have differed in context and focus. In Sweden, national guidelines were introduced in 2011 to encourage universal delivery of lifestyle advice, including moderate alcohol consumption, across the entire population by practitioners working in routine primary, child, maternity and occupational healthcare (Swedish National Board of Health and Welfare, 2011). In contrast, recommendations on alcohol screening and brief intervention issued by the Norwegian Health Directorate target specific patient populations, including antenatal care, combined drug problems and mental disorders, detoxification, and health regulations for drivers’ licenses (Helsedirektoratet, 2018). Further, the Norwegian Ministry of Health and Care Services has issued mandates ordering the regional health trusts (state-owned bodies running the hospitals) to implement strategies in somatic hospital wards, mental health services and drug treatment services to identify and treat alcohol and drug problems affecting the patients’ health (HOD, 2015). Unlike Sweden, however, these do not include specific advice on which strategies to implement, nor do they apply any incentives.
Despite the similarities in context and drinking patterns between Norway and Sweden, a key difference is that Sweden has taken a broader population approach to alcohol prevention with the ambition to address alcohol issues with patients who visit primary, child, maternity and occupational healthcare. In contrast, Norway’s approach to alcohol prevention is more targeted, focused on specific patient populations (Helsedirektoratet, 2018; Swedish National Board of Health and Welfare, 2011). Using population-based surveys in Sweden and Norway, the aims of this study were therefore: (1) to evaluate and compare beliefs and attitudes regarding alcohol conversations in healthcare in the two countries; and (2) to explore demographic and socio-economic factors associated with different levels of support for alcohol-preventive work in both countries.
Methods
Study population and design
Cross-sectional surveys were performed in Sweden in 2017 (Abidi et al., 2020; Karlsson et al., 2019) and Norway in 2018. In Sweden, recruitment was based on a web-panel administered by EnkätFabriken, a company specialising in survey research (www.enkatfabriken.se). The Swedish sample consisted of 5900 nationally representative panel members (i.e., representative of the age, sex and region of residence of the Swedish population aged 18–64 years: Karlsson et al., 2019).
In Norway, a sample of 6000 adults aged 18–88 years were randomly drawn from a nationally representative web-panel of 30,000 participants (i.e., representative of age, sex and region of residence of the Norwegian adult population), administered by Respons Analyse, a company specialising in survey research (www.responsanalyse.no). Cut-off was set at 1000 respondents, but as closing the survey is a manual procedure, 1208 had responded when the survey was closed.
Data collection
The Swedish data were collected by means of an electronic questionnaire, which was distributed via a web-panel in August–September 2017. Of the 5900 survey recipients, 489 individuals answered only the initial background questions, three opened the survey but did not respond to any questions, and 2413 did not answer at all. Therefore, the study population in the Swedish survey consisted of the 3000 individuals who answered the complete survey questionnaire, yielding a response rate of 50.7% (Karlsson et al., 2019).
The Norwegian data were collected in 2018 by means of a web-based questionnaire, accessed by the participants via a unique link provided in an email invitation. No reminders were sent, and individuals not activating the link before the cut-off was reached were non-responders. Comparison with the most recent national data on the time of the survey (December 2018) from Statistics Norway, performed by Respons Analyse, showed that the gender balance of the responders was identical to the national gender balance in this age group (49.7% women). There were fewer younger respondents than expected (18–24 years, 11.5% in the survey vs. 14.7% nationally) and slightly more responders in the oldest age group (65+ years, 20.8% in the survey vs. 19.4% nationally). For the other age groups, the differences between the sample and national data were less than 1%.
Questionnaire
The questionnaire consisted of questions on: socio-demographic characteristics; alcohol consumption; and attitudes towards and experiences with strategies for addressing alcohol in routine healthcare. The Norwegian version of the questionnaire was translated from English, and the translation was tested by comparing with the Swedish version, due to similarities between Swedish and Norwegian. The Norwegian translation was adjusted until it fitted well with both the English and the Swedish versions. The Norwegian and Swedish authors then jointly approved the final version.
Three drinking status categories were constructed based on answers to the three questions of AUDIT-C, an instrument adapted from the original AUDIT (Alcohol Use Disorders Identification Test) questionnaire developed by the World Health Organization (WHO) for use in primary healthcare settings (Saunders et al., 1993): abstainers, moderate drinkers and risky drinkers. In response to the frequency question, abstainers answered that they have not had a drink in the past 12 months; moderate drinkers had had a drink in the past 12 months but did not reach the risky level. Risky drinking was defined as having a weekly consumption of > nine standard drinks for women and > 14 standard drinks for men and/or engaging in heavy episodic drinking (HED, four standard drinks per occasion for women, five for men) monthly. These are the recommended levels in Swedish guidelines (Swedish National Board of Health and Welfare, 2011). One standard drink in Sweden and Norway equals 12 grams of pure alcohol.
Respondents were asked whether they had visited healthcare services in the past 12 months, with possible answers of “no”, “yes, once” or “yes, more than once”. Beliefs about and attitudes towards being asked about alcohol in routine healthcare were investigated using five questions (see Table 1). Response was on a four-point Likert-type scale, with possible answers of “do not agree”, “agree to some extent”, “agree to a large extent” or “agree completely”.
Beliefs and attitudes about alcohol prevention by country.
Statistical methods
The distribution of sample characteristics (Table 2) and beliefs and attitudes (Table 1) was estimated for each country. Differences in proportions were compared between countries using a chi-squared test. Logistic regression was used to identify the characteristics of those who were “pro-routine” (i.e., those who agreed completely with the statement that healthcare providers should routinely ask about patients’ alcohol consumption) of routine alcohol screening and brief interventions (Table 3). The analysis was unadjusted in Model I, and multivariate adjusted for gender, age, education, occupation, marital status, conversation about alcohol in healthcare, alcohol consumption and country in Model II. Odds ratios (OR) of being “pro-routine” were estimated with 95% confidence intervals. The interaction between country and the determinants of being “pro-routine” was tested using the likelihood ratio test. A sensitivity analysis was performed using multilevel logistic regression analysis with a random intercept to account for clustering effect within country. In order to look further for interaction effects, an analysis was performed among risky drinkers to explore whether an alcohol conversation in healthcare increased the level of support for alcohol prevention in routine healthcare (“pro-routine”) among risky drinkers. A comparison of the percentage of “pro-routine” between risky drinkers who had received a conversation about alcohol in healthcare in the last 12 months vs. those who had not received such conversation was performed with chi-squared test (Table 4). Results were considered statistically significant at p < 0.05 using two-tailed tests. Statistical analyses were performed with SPSS 25.
Sample characteristics by country.
Logistic regression of being “pro-routine” (believing that healthcare providers should routinely ask about patients’ alcohol consumption – agree completely).
Notes. OR = odds ratios; CI = confidence interval.
a Model I is crude. bIn model II, ORs are adjusted for gender, age, education, occupation, marital status, conversation about alcohol in healthcare, alcohol consumption and country.
Comparison of the percentage of “pro-routine” between risky drinkers who received a conversation about alcohol vs. those who did not receive a conversation at any consultation in healthcare in the past 12 months.
Ethical approval
The study was approved by the Swedish National Data Inspection Board and Regional Ethical Review board in Linköping (Dnr. 2017/84-31). The study was assessed by the Norwegian Centre for Research Data, which concluded that a full evaluation was not required (reference code 158794).
Results
Respondent characteristics
Table 2 shows the background characteristics of respondents in Norway and Sweden. There were differences in several socio-demographic characteristics between the two countries, with a higher proportion of women (p = 0.008), older respondents (p < 0.001), and respondents with higher levels of education (p < 0.001) in Norway compared to Sweden.
There was no significant difference in the prevalence of different drinking categories between countries (p = 0.25), or in marital status (p = 0.53). A higher proportion of respondents reported having visited healthcare in the last 12 months in Norway compared to Sweden (79.4% vs. 68.1%; p < 0.001). Amongst those who had visited healthcare in the last 12 months, 16.8% (Norway) and 26.2% (Sweden) reported having had at least one alcohol conversation (p < 0.001).
Beliefs and attitudes about alcohol conversations in healthcare
The distribution of beliefs and attitudes about alcohol prevention in routine healthcare is shown in Table 1.
The outcome “Healthcare providers should routinely ask about patients’ alcohol consumption (“pro-routine”)” was highly statistically different between countries (p < 0.001). Approximately 35% of all respondents agreed completely (“pro-routine” or highly supportive) with minor differences between the countries. However, the proportion agreeing to a large extent was lower in Norway (24.3%) than in Sweden (30.3%).
The outcome “Alcohol consumption is a personal matter and not something healthcare providers should ask about” was statistically different between countries (p = 0.003). Few respondents (3.0%) agreed completely, and the proportion agreeing to a large extent was lower in Norway (3.6%) than in Sweden (6.1%).
There was a larger proportion of respondents in Sweden (34%) compared to Norway (23.0%) that agreed completely that healthcare providers should ask about patients’ alcohol consumption, but only if patients visit healthcare to discuss symptoms that could be related to high alcohol consumption (p < 0.001).
Approximately 10% of respondents agreed completely and 42% agreed to a large or some extent in both Sweden and Norway that healthcare providers should ask about patients’ alcohol consumption, but only if the issue is brought up by the patient (p = 0.17).
The outcome “I believe people answer honestly when they are asked about their alcohol consumption at healthcare visits” was highly statistically different between countries (p < 0.001). Although approximately 5% agreed completely in both countries, the proportion agreeing to a large extent was lower in Norway (13.5%) than in Sweden (19.4%). A similar proportion of respondents (26.0%) disagreed that people answer honestly when they are asked about their alcohol consumption at healthcare visits in both countries.
“Pro-routine” beliefs and attitudes
An overall test of interaction between determinants of being “pro-routine” and country was performed, and the result was not statistically significant (p = 0.93). Accordingly, all subsequent analyses were performed for the pooled data set (Table 3). The odds ratio of believing that healthcare providers should ask routinely about patients’ alcohol consumption was higher for respondents aged 60+ compared to those aged under 30 (OR = 1.42, CI 1.04–1.95). The odds ratio of being “pro-routine” was significantly lower for respondents on parental leave than those who were employed (OR = 0.46, CI 0.27–0.77), and for single compared to married respondents (OR = 0.76, CI 0.64–0.91). The odds ratio of being “pro-routine” was significantly higher for respondents with one conversation about alcohol in healthcare and was more than twice as high for respondents who had two or more such conversations (OR = 2.20, CI 1.57–3.08) compared with those who did not have a conversation about alcohol.
The odds ratio of being “pro-routine” was significantly lower for moderate drinkers (OR = 0.45, CI 0.35–0.58) and for risky drinkers (OR = 0.28, CI 0.21–0.37) compared to abstainers. There were no associations between being “pro-routine” and either gender, level of education or country. The odds ratios of the different predictors estimated in the multilevel logistic regression to take into account the clustering effect within country were very similar to the estimates from the analysis presented in Table 3 and are therefore not shown.
Association between “pro-routine” and alcohol conversation among risky drinkers
In total, approximately 27% of the hazardous drinkers were “pro-routine”. Among the hazardous drinkers who had received a conversation about alcohol in healthcare, 36% were “pro-routine”, a statistically significantly higher (p = 0.002) share than the 25% who were “pro-routine” among hazardous drinkers who had not received a conversation about alcohol (Table 4).
Discussion
This study sought to evaluate and compare beliefs and attitudes regarding alcohol conversations in healthcare in Sweden and Norway, and to explore factors associated with different levels of support for alcohol-preventive work in both countries. Our findings show that, overall, there is widespread support for healthcare providers asking about patients’ alcohol consumption (“pro-routine”) in both countries. However, a lower proportion of respondents in Norway, compared to Sweden, agreed to a large extent that healthcare professionals should routinely ask about alcohol consumption. Furthermore, a significantly higher proportion of Swedish than Norwegian respondents agreed completely that healthcare providers should ask about alcohol only when a patient presents with symptoms related to high consumption (34% compared to 23%) (p < 0.001). Similar correlates of being supportive of routine alcohol screening and brief intervention delivery were found in both countries. However, in both countries support was lower among moderate and risky drinkers, among single participants or those on parental leave, but higher among older individuals.
The high levels of support we found for the delivery of alcohol prevention in routine healthcare are consistent with results from previous research, including the 2010 survey in Sweden (Nilsen et al., 2012). The fact that a lower proportion of respondents were “pro-routine” in Norway compared to Sweden might be linked to the broader healthcare approach to alcohol prevention in Sweden compared to Norway’s more targeted approach, focused on specific patient populations (Helsedirektoratet, 2018; Swedish National Board of Health and Welfare, 2011). In addition, the higher proportion of respondents being “pro-routine” in Sweden may partly be explained by a national project (“the Swedish Risk Drinking Project”) to increase identification of and interventions for risky drinking in routine healthcare (Nilsen, Wåhlin, & Heather, 2011; Reinholdz et al., 2011). While public support for alcohol conversations may be higher in Sweden partly because of a national project to improve practice, it is less clear whether it has also affected alcohol consumption and health (Lundin et al., 2017). However, our findings provide further evidence that patients themselves are broadly in favour of healthcare providers asking about their alcohol consumption in a routine way. This is contrasted by findings indicating that physicians themselves find it challenging to address alcohol with patients who are not consulting for symptoms directly related to risky drinking (Aasland & Johannesen, 2008; Aira et al., 2003; Johansson et al., 2002; Lock et al., 2002; Nygaard & Aasland, 2011; Nygaard et al., 2010; Rush et al., 1995).
At the same time, given that around a quarter of respondents (26%) in Norway and Sweden believed that people do not answer honestly when they are asked about their alcohol consumption, our results suggest that some still view alcohol as a sensitive subject for discussion in healthcare (Miller et al., 2006; Nilsen et al., 2012). This proportion is lower than reported in both the 2010 Swedish population survey (34%) (Nilsen et al., 2012), and, interestingly, the aforementioned English survey (55% in 2017) (O’Donnell et al., 2018), a country where alcohol is arguably more culturally embedded (Stewart & McCambridge, 2019). Further, this difference may also reflect the higher levels of trust in public institutions found in Scandinavian countries compared to the UK (Ortiz-Ospina & Roser, 2019).
In both countries, risky drinkers were less supportive of routine delivery of alcohol prevention in healthcare, compared to abstainers. Differences due to drinking status are consistent with results from previous research, including the 2010 survey in Sweden, where abstinence was predictive of being highly supportive of alcohol conversations, whereas hazardous and excessive drinkers were less supportive (Nilsen, McCambridge, et al., 2011). The fact that risky drinkers, the group who stand to benefit most from brief alcohol interventions, are consistently less supportive of their routine delivery, suggests that alcohol prevention remains challenging for healthcare providers in Norway and Sweden. However, an interesting positive finding in our study is that risky drinkers who had had an alcohol conversation in healthcare were more supportive of delivery of alcohol prevention than those who had not had such a conversation. This indicates that with experience (i.e., having a conversation about alcohol), there is less worry about a conversation about one’s alcohol habits, indicating that a non-judgmental conversation reduces the potential shame of such a conversation (Coste et al., 2020; Tam et al., 2015).
We also found that older participants were more likely to be supportive of alcohol prevention in healthcare in both countries, again consistent with findings of the 2010 Swedish survey (Nilsen et al., 2012), and possibly reflecting an increased interest in health issues due to a higher risk of comorbidity in this age group (Gell et al., 2015). However, this finding contrasts with the results from the English survey, where there was no difference in levels of support by age (O’Donnell et al., 2018). As in 2010, we also found that respondents currently on parental leave were less likely to support being asked about their drinking (Nilsen et al., 2012) despite the fact that parents of small children are prioritised for advice provision in the Swedish national guidelines (Swedish National Board of Health and Welfare, 2011).
Findings from an American study showed that patients who initiated a BI themselves after being prompted to do so, decreased their alcohol consumption more than patients who received provider-initiated BI (Rose et al., 2015). Given that more than half of the participants in our study agreed to a large or some extent that “Healthcare providers should ask about patients’ alcohol consumption, but only if the issue is brought up by the patient” it would be interesting to investigate such an approach in future research.
Strengths and limitations
There are some limitations that must be considered when interpreting the findings. The study was based on self-reports and the cross-sectional design does not allow causal inferences. The surveys used two different sampling strategies in each country, and even though the panels are nationally representative on demographic variables, we do not know to what extent they are nationally representative on drinking habits and experiences with alcohol prevention in healthcare (Rehm et al., 2020). However, our samples were comparable in respect of our focus topic in that there were no differences in drinking status between respondents in both countries.
This study also has important strengths. It used a validated screening instrument (AUDIT) to assess drinking status. It is also the first population-based study comparing beliefs and attitudes on addressing alcohol consumption in healthcare between the Swedish and Norwegian populations. By evaluating beliefs and attitudes about alcohol conversation in healthcare and comparing the two countries, our findings could inform more effective implementation strategies for alcohol screening and brief interventions in the future.
Conclusions
There is a high level of support among the Swedish and Norwegian populations for delivery of alcohol prevention in healthcare, although there was a lower proportion of respondents who were positive to alcohol prevention in routine healthcare in Norway compared to Sweden. The support for this work was lower in both countries among risky and moderate drinkers and among those on parental leave. Experiencing an alcohol conversation may positively affect risky drinkers’ attitudes towards and support for alcohol prevention. Thus, more frequent alcohol conversations in routine healthcare may result in an increased level of support for alcohol prevention among risky drinkers.
Footnotes
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was funded by Linköping University, University of Stavanger and Av-og-til, a Norwegian alcohol awareness NGO.
