OUP user menu


DOI: http://dx.doi.org/10.1093/alcalc/agh185 388-393 First published online: 25 July 2005


Aims: To explore the attitudes of Swedish general practitioners (GPs) and nurses to secondary alcohol prevention (early identification of, and intervention for, alcohol-related problems) and compare it to their attitudes to other important lifestyle behaviours such as smoking, stress, exercise, and overweight. Methods: An adjusted version of The WHO Collaborative Study Questionnaire for General Practitioners was posted to all GPs and nurses in the County of Skaraborg, Sweden; 68 GPs and 193 nurses responded. Results: The importance of drinking alcohol moderately, counselling skills on reducing alcohol consumption and perceived current effectiveness in helping patients change lifestyle behaviours ranked lower than working with all the other lifestyle behaviours. The nurses rated their potential effectiveness in helping patients change lifestyle higher than that of GPs for all the lifestyle behaviours. Nurses receiving more alcohol-related education had more positive attitudes than nurses with less education. For alcohol, the GPs assessed their role adequacy, role legitimacy and motivation higher than that of the nurses. The main obstacles for the GPs to carry out alcohol intervention were lack of training in counselling for reducing alcohol consumption, time constraints, and the fact that the doctors did not know how to identify problem drinkers who have no obvious symptoms of excess consumption. Conclusion: GPs and the nurses estimated their alcohol-related competence as lower than working with many other health-related lifestyles. These results can be explained by lack of practical skills, lack of training in suitable intervention techniques, and unsupportive working environments. All these elements must be considered when planning secondary alcohol prevention programs in primary health care.

(Received 3 March 2005; first review notified 21 March 2005; accepted in revised form 2 June 2005)


In primary health care, patients consult general practitioners (GPs) and nurses for a number of health-related problems, many of which can be alcohol-related (Thakker, 1998). Over a 2-year period, people to a large extent visit their primary health care, making screening for alcohol problems in primary health care practicable (Fiellin et al., 2000); treatment does not have to be extensive. Thus, various brief interventions (BIs) have been shown to be effective (Fleming et al., 1997; Poikolainen, 1999). Very brief (5–10 min) advice and counselling by GPs or nurses can reduce alcohol consumption in high-risk drinkers (Ockene et al., 1999), and has been shown to have middle to high ranking on the list of efficient means to reduce drinking hazards (Babor et al., 2003). Benefits have been shown for up to 48 months (Fleming et al., 2002) but one study following ∼500 patients over 10 years, with assessment at the beginning and after 9 months, did not show reduction in drinking behaviour (Wutzke et al., 2002). The method has the potential to be a cost-effective means of intervention with a benefit–cost ratio of 5.6:1 (Fleming et al., 2000).

The attitudes of GPs to patients with alcohol-related problems have been described (Rush et al., 1994); the issue is very complex with positive attitudes for role legitimacy and role adequacy, but lack of training, practical skills, and self-efficacy are negative factors which make implementation difficult (Aalto et al., 2001, 2003). In a study in England, 83% of the GPs felt prepared to counsel excessive drinkers but only 21% felt effective in helping patients to reduce alcohol consumption (Kaner et al., 1999). The GPs who had received more alcohol-related education and had higher total score for role security and therapeutic commitment, were more likely to carry out work related to alcohol problems in their practice (Anderson, 1985; Anderson et al., 2003).

In many countries, nurses in the primary health care system play an important role in promoting health. By providing BI, the nurses can complement the GPs role and give the patients follow-up visits. The effectiveness of using primary care physician–nurse teams has been evaluated when patients with previous trauma were screened for hazardous drinking; a follow-up visit by the nurse was superior to simple advice after 12 months follow-up (Israel et al., 1996). BI offered by a nurse can give the same results as BI given by a GP (McIntosh et al., 1997).

However, in-depth interviews with 24 nurses in general practice in the north-east of England showed that they received little or no preparation for the task of alcohol intervention although they have many opportunities to engage in that work (Lock et al., 2002).

We have found only a few studies which compare alcohol-related attitudes between GPs and nurses (Bendtsen and Akerlind, 1999; Andreasson et al., 2000; Aalto et al., 2001, 2003; Kaariainen et al., 2001; Johansson et al., 2002). In none of these studies are attitudes to alcohol compared with attitudes to other lifestyle behaviours such as smoking, exercise, overweight or stress levels. The present study was planned to further explore alcohol-related attitudes among GPs and nurses, and compare them to attitudes to other lifestyle behaviours. The GPs' and nurses' own alcohol consumption was assessed to see if it had an influence on the alcohol-related attitudes.


A postal survey was carried out from December 2001 to February 2002 on a sample of GPs and nurses in the County of Skaraborg in south-west Sweden. Skaraborg County has 15 municipalities with 254 000 inhabitants and has a mixture of rural and urban developments. Primary Health care sector has a long tradition and is comparatively strong in a Swedish comparison. In Skaraborg the proportion of acute visits to the GPs is 60%. An average visit has duration of 20–30 min. In general, preventive work is not contemplated among the centre's duties.

All physicians and nurses in the 25 public health care centres in the region were sent an anonymous questionnaire with a covering letter that explained the background to the survey. Two written reminders were given at monthly intervals.

The WHO Collaborative Study Questionnaire for GPs was translated into Swedish by the authors. Its original form is described elsewhere (Kaner et al., 1999). It was abbreviated and adjusted to local conditions. The response option ‘as indicated’ to the question ‘to what extent do you obtain information on your patients’ about different lifestyle behaviours was changed to ‘often’, as this better maintains the ordinal scale response categories. Further, the Shortened Alcohol and Alcohol Problems Perception Questionnaire (SAAPPQ) (Anderson and Clement, 1987) was assessed for ‘problem drinking’. ‘Problem drinking’ refers to subjects with hazardous or harmful alcohol use, but excludes subjects dependent on alcohol. We asked about the respondents' current self-perceived effectiveness in helping patients change lifestyle behaviours separately for male and female patients. In addition to the WHO questionnaire, we added four items concerning treatment impact and treatment resources to the GPs, taken from a survey conducted in Philadelphia, USA, (Spandorfer et al., 1999), with the exception that we asked about treatment results separately for ‘problem drinkers’ and alcohol-dependent persons.

The first three AUDIT questions were included to estimate the participants' own drinking habits (Bush et al., 1998). Binge drinking was defined as ≥5 drinks for males and ≥4 drinks for females. In Sweden, one standard drink is usually 12 g of alcohol. The respondents were divided into two groups: abstainers or low consumers who scored 2 points or less and moderate to high consumers who scored 3 points or more.

The nurses received a shortened version of the questionnaire adjusted to their professional role in the Swedish primary health care system leaving out questions about diagnostic and management skills, incentives, and disincentives for brief alcohol intervention as well as questions about treatment impact and treatment resources. This was done because nurses do not refer patients for treatment in Sweden. Thus, the questionnaire consisted of 115 items for the GPs and 72 items for the nurses. The questionnaire was pre-tested on 12 GPs outside the study region. Pilot testing of the questionnaire used for nurses was not done.

The Ethical Committee of The Sahlgrenska Academy at Göteborg University approved the survey (reference number Ö 406–00).


Data were entered into a Microsoft Access database. Stat View (version was used for statistical analyses. Descriptive statistics and 95% confidence intervals (CIs) were calculated, and the Mann–Whitney U-test was used to test the significant difference between items.


Sample characteristics

Sixty-eight GPs and 193 nurses answered the questionnaire; the response rates were 52 and 67%, respectively. The GPs' mean age was 47 years (SD = 9.1), 68% were male, 72% were specialists in general practice, and their mean number of years in practice was 12.6 (SD = 7.5). The corresponding data for the nurses were 48 years (SD = 8.2), 5.2% male, and 14.3 (SD = 8.6) years in practice. The average time working in general practice was 35 h/week (SD = 8.6) for the GPs; 19% of them saw up to 29 patients per week, 54% saw between 30 and 59 patients, and 25% saw 60–90 patients. They devoted 7.8% (SD = 8.2) of their time to preventive work.

Post-graduate education in the alcohol field

Over half of the respondents had received no education (GPs 25%, nurses 41%) or <4 h of post-graduate education (GPs 29%, nurses 17%) on alcohol and alcohol-related problems. A further 28% of the GPs and 19% of the nurses had between 4 and 10 h education; 3% of the GPs and 9% of the nurses had between 11 and 40 h, and 3% >40 h for both groups. Ten percent of the participants did not remember whether they had received any post-graduate education.

The respondents' drinking habits

Forty-seven percent of GPs and 62% of the nurses were abstainers or low consumers; 52% of GPs and 37% of nurses were moderate or high consumers. One GP and two nurses did not answer questions about their drinking habits.

Obtaining information on health-related behaviours

Table 1 illustrates how often GPs and nurses obtain information from patients for some health-related behaviours. Obtaining information about alcohol consumption ranked lowest for both groups: the difference for all items compared with alcohol consumption was statistically significant for the nurses, but there was only a statistically significant difference between drinking and smoking and exercise, in the case of GPs. The responses for alcohol consumption did not differ by years in practice or by the respondents' own drinking habits. Nurses who had received 4 h or more of education on alcohol obtained information on alcohol more often than nurses who had received less education (mean 2.45; CI 2.24–2.66 vs 2.01; CI 1.85–2.16; P = 0.002); there was no difference for the GPs.

View this table:
Table 1.

Mean ratinga and 95% CI of respondents' efforts to obtain information about lifestyle

Attitudes to intervening for various health-related behaviours

The respondents rated the importance of lifestyle behaviours in promoting the health of the average person. The results are shown in Table 2. Drinking alcohol moderately ranked lower than the other behaviours, except for ‘not drinking alcohol at all’ for both GPs and nurses. The results were not related to respondents' drinking habits. Only 29% of the GPs believed that moderate drinking was very important for promoting health compared with 50% of the nurses.

View this table:
Table 2.

Proportion of GPs and nurses rating specific lifestyle behaviours as important or very important in promoting the health of the average person (%)

The respondents also rated how effective they currently ‘feel in helping patients achieve change’ in lifestyle behaviours separately for male and female patients. The results were very similar for both GPs and nurses for all items and the results are thus shown together. Fourteen nurses did not answer the question for male patients. The results for self-estimated counselling skills, current effectiveness, and potential effectiveness in helping patients change their lifestyle behaviours after ‘given adequate information and training’ are shown in Table 3.

View this table:
Table 3.

Mean ratingsa and 95% CI of respondents' counselling skills and perceived current and potential effectiveness in helping patients change lifestyle

Both GPs and nurses rated their counselling skills and current effectiveness for reducing alcohol consumption significantly lower than for counselling on smoking, exercising regularly, and avoiding excess calories. The nurses rated their potential effectiveness higher than the GPs in all lifestyle behaviours studied. The results for reducing alcohol consumption were not related to the participants' own drinking habits, but the nurses who had received 4 h or more of post-graduate education on alcohol scored significantly higher for counselling skills (mean 2.68; CI 2.48–2.89 vs 2.35; CI 2.19–2.52; P = 0.04) and for current effectiveness for female patients (mean 2.54; CI 2.36–2.73 vs 2.21; CI 2.08–2.33; P = 0.01).

Attitudes to working with problem drinkers

The results from the five variables of role acceptance according to the SAAPPQ are shown in Table 4. The GPs rated significantly higher on role adequacy, role legitimacy and motivation than the nurses. GPs with moderate to high alcohol consumption scored significantly higher on role adequacy than light consumers (mean 4.84; CI 4.46–5.22 vs 4.22; CI 3.84–4.60; P = 0.03). The nurses who had received 4 h or more of post-graduate education on alcohol, scored higher on role adequacy (mean 4.03; CI 3.69–4.38 vs 3.55; CI 3.32–3.78; P = 0.03), role legitimacy (mean 5.78; CI 5.54–6.02 vs 5.12; CI 4.89–5.34; P = 0.0003), and task-specific self-esteem (mean 4.66; CI 4.28–5.02 vs 4.10; CI 3.86–4.34; P = 0.02) than nurses who had received less education.

View this table:
Table 4.

Mean ratingsa and 95% CI of GPs and nurses on the SAAPPQ for problem drinkers

Treatment resources and the success of alcohol treatment

The GPs opinion of treatment resources and the success of alcohol intervention are shown in Table 5. Concerning appreciation of present services, it could be said that it probably reflects the present reality both in the uptake area and in Sweden. Resources are very strained in the Primary Health Care sector in general, and in the uptake area the resources for specialized addiction treatment are very scarce as well. This is also the case in most of Sweden.

View this table:
Table 5.

GPs' responses to statements regarding resources for and success of alcohol treatment approaches

Incentives and disincentives for brief alcohol intervention in primary care

Incentives and disincentives are listed in Table 6. The most commonly endorsed disincentive was lack of training in counselling for reducing alcohol consumption and the most common incentive was readily available support services to refer patients to.

View this table:
Table 6.

GPs agreement with suggested disincentives and incentives for alcohol intervention


Strengths of this postal survey were: all the GPs and nurses in the County of Skaraborg were invited to participate; there was a high response rate from the nurses (67%) and an acceptable response rate from the GPs (52%) (Barclay et al., 2002); answers were received from all the health centres. Although, in the Skaraborg region, there is an officially adopted intention to work preventively, also in the health sector, there is no study that can prove that this is done to a greater extent than in other regions of Sweden. Concerning alcohol prevention, we believe this to be an equally undeveloped area in this region as in others. Thus we believe that our results on these respects can be generalized to Swedish primary health care in general.

The main results from our study are that the GPs and nurses obtain information about alcohol consumption relatively seldom (see Table 1); they rated their counselling skills and perceived current effectiveness in reducing alcohol consumption as lower than for all the other lifestyle behaviours investigated (see Table 3). The nurses scored their potential effectiveness considerably higher than that of GPs for all of the lifestyle behaviours, and their alcohol competence may therefore increase more than that of GPs if they acquire more post-graduate education on alcohol. This implies the high potential of the nurses in promoting preventive work in primary health care; this is in agreement with findings in England and Wales where the practice nurse is considered ‘a major under-utilized resource within primary care for screening and BI with the non-problematic patient drinking above sensible limits’ (Deehan et al., 1998).

The GPs results for perceived current effectiveness in helping patients change lifestyle can be compared with the results of a recent survey among 2082 GPs in 11 European countries (Brotons et al., 2005). In the Brotons study, the effectiveness in helping patients reduce alcohol consumption ranked lowest compared to the effectiveness in helping patients reduce tobacco, achieve or maintain normal weight or practice regular physical exercise. This is in accordance with our results. However, the 296 Swedish GPs in the Brotons study ranked their effectiveness in helping patients reduce alcohol consumption in second place after effectiveness in reducing tobacco use. Nevertheless, for reducing alcohol consumption, the results were comparable. To the statement ‘minimal effective or ineffective in helping patients reduce alcohol consumption’, 54% of Swedish GPs endorsed ‘yes’. In our study, 50% endorsed the option ‘ineffective’ and 6% ‘very ineffective’ to the statement ‘how effective do you feel you are in helping patients reducing alcohol consumption’.

Fifty percent of the nurses believed that ‘drinking alcohol moderately’ was very important in promoting health compared to 29% of the GPs; however, this fact did not prompt them to ask about alcohol consumption more often than the GPs. A possible explanation is the low score for role adequacy compared to GPs, who also scored significant higher for role legitimacy and motivation. Both occupational groups, according to this study, have low counselling skills and low capability in helping patients to reduce alcohol consumption, compared to the other lifestyle changes, calling for more training in this field. These findings are in accordance with previous results from implementations studies in Sweden, in which it was suggested that nurses need more training and motivational efforts than doctors (Bendtsen and Akerlind, 1999; Andreasson et al., 2000). In the present study, the nurses who received 4 h or more of alcohol-related education, more often obtained information on alcohol, scored higher for counselling skills, and perceived current effectiveness in helping patients to reduce alcohol consumption. They also rated higher for role adequacy, role legitimacy, and task-specific self-esteem. According to these findings, it seems that the nurses' alcohol-related competence will increase more than for GPs if they acquire more post-graduate education on alcohol.

When the results from this study were compared with results from England (Kaner et al., 1999), 57% of the English GPs are prepared and 26% very prepared for counselling (Kaner et al., 1999) compared with 41 and 3% respectively in Sweden. In the tables, we have reported the data as means with CI. Here we report the proportions who responded in the various categories, in order to compare our results with those reported by Kaner and other authors. The proportion of English GPs agreeing with the statements relating to role adequacy for ‘problem drinkers’ was 71%, compared to only 49% of the Swedish GPs. On the other hand, the Swedish GPs scored higher on motivation (38 vs 23%) and task-specific self-esteem (38 vs 19%), whereas there was no difference for role legitimacy and role satisfaction. Only 20% of the English GPs felt effective at helping patients to reduce alcohol consumption compared to 40% of the Swedish GPs. This difference in attitudes implies that Swedish GPs estimate their own competence in the alcohol field as lower than the English GPs, but they estimate better their chances of promoting a change in their patients drinking habits. We have no explanations for these results, but speculate that the GPs in Sweden might have better opportunity to use paramedical staff to help problem drinkers and/or to refer them to community alcohol service.

It is noteworthy that there was a significant association between drinking more and scoring higher in role adequacy for the GPs. This was somewhat surprising and we have no definite answer as to why. Perhaps, a higher familiarity with alcohol consumption makes GPs more open towards discussing alcohol issues with a patient. We have found only two studies that relate respondents' drinking habits to alcohol-related attitudes (Anderson, 1985; Kaariainen et al., 2001), and they found no correlations between the respondents' drinking habits and their attitudes.

The majority of the GPs (74%) think that intervention has a positive impact on alcohol consumption. This should stimulate them to undertake alcohol-related work in their practice. At the same time, only 28% of them endorse that treatment is successful in at least 50% of problem drinkers. This ambivalence can be explained by the fact that they are working in an environment where there is lack of treatment resources in the alcohol field; only 6% endorse strongly agree/agree that there are adequate resources for treatment of early problem drinkers and 4% for alcohol-dependent patients. Furthermore, 67% endorse that there are time constraints and 65% that there are lack of knowledge in identifying problem drinkers who have no obvious symptoms of excess alcohol consumption. This, in combination with insufficient counselling skills for reducing alcohol consumption, can explain why they do not identify more patients with alcohol problems and do not even ask about it (Aalto et al., 2002; Andreasson and Graffman, 2002).

In summary, this study shows that GPs have more positive attitudes than nurses towards working with problem drinkers from the viewpoint of role legitimacy and role adequacy; both occupational groups are lacking motivation, role satisfaction and task-specific self-esteem. The GPs and nurses estimate their counselling skills and effectiveness in helping patients achieve change in lifestyles behaviours as lower for reducing alcohol consumption than for many other health-related lifestyles. The nurse's alcohol competence is likely to increase more than that of GPs if they acquire more post-graduate education on alcohol and can, therefore, be a major resource in promoting secondary alcohol prevention in primary care. Lack of practical skills, lack of training in suitable intervention techniques and an unsupportive working environment are the major obstacles for primary health care to better take care of patients with problem drinking. All these elements must be considered when planning secondary alcohol prevention programs in primary health care.


This study was funded by the Skaraborg institute for research and development, R&D Centre Skaraborg Primary Care and the Primary Health Care, Skövde.


View Abstract