The role of alcohol in the management of hypertension in patients in European primary health care practices – a survey in the largest European Union countries

Background Even though addressing lifestyle problems is a major recommendation in most guidelines for the treatment of hypertension (HTN), alcohol problems are not routinely addressed in the management of hypertension in primary health care. Methods Internet based survey of 3081 primary care physicians, recruited via the mailing lists of associations for general practitioners (GPs) in France, Germany, Italy, Spain and the UK. Clinical practice, attitudes, knowledge, education and training were assessed. Logistic regression to predict screening, brief intervention and treatment for alcohol dependence in the management of hypertension were assessed. Results Overall, about one third of the interviewed GPs reported sufficient screening in cases with HTN (34.0 %, 95 % confidence interval (CI):32.1–35.8 %). One out of five GPs screened and delivered brief interventions in HTN patients with hazardous consumption (22.2 %, 95 % CI: 20.6–23.8 %) and about one in 13 GPs provided treatment for HTN patients with alcohol dependence other than advice or brief intervention (7.8 %, 95 % CI: 6.8–8.9 %). Post-graduate training and belief in their effectiveness predicted interventions. There were marked differences between countries. Conclusions While current interventions were overall low, marked differences between countries indicate that current practices could be improved. Education and post-graduate training seems to be key in improving clinical practice of including interventions for problematic alcohol consumption and alcohol dependence in primary health care. Electronic supplementary material The online version of this article (doi:10.1186/s12875-016-0529-5) contains supplementary material, which is available to authorized users.


Background
Hypertension (HTN) is the single most important risk factor for mortality and burden of disease, globally and especially for high income countries in Europe [1]. Its main effect is on cardiovascular outcomes, and consequently, reduction of blood pressure is among the risk factor targets of the World Health Organisation 'Global Action Plan for Prevention and Control of Non-communicable Diseases' for the period 2013-2020 [2]. Primary health care has traditionally had a key role in the detection and the management of HTN [3]. Part of this management involves advice and interventions on lifestyle factors underlying HTN, and guidelines recommend lifestyle changes as important means to reduce blood pressure, prevent and/or avoid medication for HTN [3][4][5]. Both epidemiology and randomized trials converge in demonstrating that alcohol consumption, in particular heavy drinking, is one of the most important lifestyle based risk factors for HTN [6][7][8][9].
However, the mortality and disease burden attributable to HTN has increased globally since 1990 [1] and large European surveys still show a large proportion of adults with uncontrolled HTN (http://apps.who.int/gho/data/ ?theme=home), indicating the need for further action. Of all lifestyle factors, alcohol seems to be the least intervened in the management of HTN [10][11][12][13], which is no surprise given the low screening and intervention rates for hazardous drinking and alcohol use disorder in primary health care [14,15]. Interventions for hazardous drinking are scarce [15][16][17]; and alcohol use disorders have the lowest treatment rate of all mental disorders [18][19][20], despite evidence that there are effective interventions available for both hazardous drinking and for alcohol use disorders [21,22], which could be implemented at the primary care level [23,24].
Thus, improving alcohol interventions in primary health care promises to yield substantial health benefits [10][11][12][13]25]. The main question to realize this potential is how to best implement such interventions [23], both for hazardous drinking and for alcohol use disorders, as part of routine management of HTN. Together with primary care associations in the five largest countries in the European Union (France, Germany, Italy, Spain, and the United Kingdom (UK)), we developed a survey of general practitioners (GPs) to explore knowledge, attitudes and clinical practice of lifestyle interventions in the management of HTN and to help a potential implementation of alcohol interventions (Baseline Alcohol Screening and Intervention Survey (BASIS)).

Design of the BASIS survey and pilot
All authors were involved in drafting and finalizing the survey, originally in English. After an empirical pilot study in five countries (N = 41 respondents), the survey was translated into French, German, Italian, and Spanish and the national versions were again tested and finalized with the help of local experts. A brief summary of the survey and its subsections are given in Additional file 1. It contained 28 core items (in addition to a few countryspecific items) and was put online in all languages using SurveyMonkey (http://www.surveymonkey.com). The English version of the survey can be found online (Additional file 2). The theoretical basis was the Information-Motivation-Behavioural Skills model [26,27], which stipulates that information and education is not sufficient to adopt behaviours; in addition there needs to be motivation and behavioural skills. This model had been adopted to care of non-communicable diseases [28].

Survey implementation
In each of the five countries, regional or nationwide GP associations disseminated the web link to the survey to their members, mainly via electronic mail (for details see Table 1). The median completion time was 8.8 min, with a span from under 2 min to over an hour). Four responses were removed from the data set due to suspicion of being duplicates. The entire survey was answered by 2468 respondents (80.1 % of those who started: 3081) between September 29 and December 1, 2015.
The survey included a number of free text items, including descriptions how alcohol problems were managed. A coding scheme based on free text responses given in Germany and the UK was developed and subsequently all such responses were classified by two independent raters for each language. Kappa agreement coefficients were calculated and ranged from 0.31 to 1 in the variables analyzed. Non-concordant ratings were revisited and a final decision was made by JM.

Statistical analyses
Three different indicators for good practice alcohol management in patients with HTN were derived from responses given in the questionnaire: a) sufficient screening for alcohol use (at least 7 out of 10 HTN patients); b) sufficient screening (as above) in addition to management of alcohol problems in hypertensive patients with hazardous drinking levels by the GP themselves or within the same practice usually with brief interventions (for rationale see care [21,29]; c) sufficient screening and management of alcohol dependence in hypertensive patients by the GPs themselves or within the same practice. Indicator c was only met if GPs did not only offer brief advice or counselling as management for alcohol dependence but also reported other interventions, such as psychotherapy, or pharmacotherapy. This operationalization was chosen, as current guidelines do not recommend brief advice only as a treatment intervention for dependence [30,31].
Logistic regressions on each indicator were computed with Stata 14.0 [32], using the following variables as covariates (specifications in parentheses): age (categories as dummy variables with '70 or older' as reference category), sex, country (dummy coded with UK as reference category), beliefs about success of different lifestyle interventions for hypertension (questionnaire items 3 and 4: dummy variables, each scored 1 if rated (highly) successful, else 0), knowledge (questionnaire item 1: dummy variable, scored 1 if alcohol was selected as important risk factor for HTN, else 0), education (questionnaire items 24 and 27: dummy variables, each scored 1 for at least 4 out of 5 points on Likert scale regarding adequacy of graduate education on alcohol/HTN, else 0; questionnaire items 25 and 28: dummy variables, each scored 1 if post-graduate education on alcohol, HTN was received, else 0), and workload (questionnaire item 7: continuous variable containing number of daily patient contacts, z-standardized for each country to achieve comparability). A measure of the respondents' drinking patterns (questionnaire items 32-34) was also considered for inclusion in the models. However, it was decided against it as it would have overly limited the generalizability of the findings by reducing the sample size by 24 % (from 2468 to 1885) because these items were not assessed among UK respondents and responses were not required to complete the survey in the remaining countries.

Results
Two thousand four hundred sixty eight health professionals participated in the survey (for details see Table 1).
With respect to the indicators of good practice alcohol management (= main dependent variables), Table 2 gives the prevalence by country.
The overview of influencing variables for good practice alcohol management are given in Table 3, where the reference country was always the UK. Clearly screening for alcohol was best implemented in the UK and Spain, management of hazardous drinking levels was best implemented in Spain (87 % of all identified GPs treated An advertisement in a nationwide newspaper was placed during the period of data collection with very little response. Out of 103 GPs, 98 were from Bavaria and the remaining five from Hesse (1), North Rhine-Westphalia (2) and Saxony (2). The response rate was calculated by omitting the 54,000 potential readers from the denominator  For age and country, the p-values refer to an omnibus test for the entire variable, i.e., testing the global hypotheses that the coefficient for any age category or country deviates from the null hypothesis of no difference above chance ("French paradox"; see [33]; see also the official training materials of the French cardiologists [34]). The lack of knowledge and training in Italian GPs with respect to screening and brief interventions has been found in several other European studies (INEBRIA: AMPHORA: [35]; see also http://www.epicentro.iss.it/alcol/apd2013/presentazioni/9.Cuffari.pdf ), and has seemingly not improved over the past years. Before we discuss potential conclusions of the results, we would like to highlight limitations. First, response rates are relatively low. While it is hard to compare response rates across physicians' surveys, as there are different sampling frames and several web-based surveys do not even give response rates [36,37], and even though web-based surveys have comparably lower response rates [38], an overall response rate of 6 % must be considered low. As a consequence, while the national/regional sampling frames can be considered as representative, the low response rates suggest that a convenience sample of GP's being more motivated and interested in the topic has been drawn [39]. Thus the screening and intervention rates reported are likely to be overestimates (for intervention rates in samples of GP's with representative sampling and a considerably higher response rate [14,40]). Second, all answers were self-reports and social desirability bias may have shifted some of our key results upwards [41]. In other words, based on the two major limitations of this study, the rates for screening and interventions among hypertensive primary health care patients in Europe are most likely lower than described in this study. However, given the low response rate, we cannot fully rule out that we have underestimated the GPs' involvement, e.g., if engaged GPs were too busy to participate in our survey.

Conclusions
While our findings are susceptible to sample distortion, they are sufficiently robust to demonstrate that the GPs' involvement in alcohol screening and management among patients with HTN is generally poor in the largest European countries. Thus, the situation for HTN patients is likely not better than for other primary care patients with respect to detection of and interventions for heavy drinking and alcohol use disorders [14,15,17,40]. What can be done about this? First, medical education at universities have to put more emphasis on alcohol as one of the main risk factors for many disease conditions GPs see in their daily practice [40]. The lack of education seems a common problem in all five countries, and was also highlighted in some of the qualitative answers. Moreover, post-graduate training was shown to increase screening and intervention rates [16,42], and this is, where GP associations can contribute. Secondly, given the high overall workload of GPs, and the overall health burden attributable to alcohol in countries in the European Union [20], alcohol interventions need to be prioritized and this could be done by financial incentives. A recent cluster randomized trial with 746 providers in 120 primary health care centers from five European countries has shown that modest financial incentives increase screening and intervention rates. Interestingly, there is a synergistic effect when financial incentives, training and support are offered together [17]. We hope that the involvement of several GP associations in the current study will help overcome these barriers in the future.

Funding
The study was financially supported by an investigator initiated grant to the first author and the GWT-TUD (Gesellschaft für Wissens-und Technologietransfer der TU Dresden mbHcompany with limited liabilities for transferring knowledge and technology of the Dresden University of Technology) by Lundbeck. The study sponsor has no role in study design, collection, analysis, and interpretation of data. The study sponsor also had no role in writing of the report; and the decision to submit the paper for publication. The corresponding author confirms that the authors had full access to the data in the study at all times, and had final responsibility for the decision to submit for publication. The corresponding author hereby states that no author has been reimbursed for writing this manuscript.   H-JA: reports sponsorship to attend scientific meetings, speaker honoraria and consultancy fees from Bioprojet, D&A Pharma, Ethypharm, Lundbeck, Merck-Serono, Novartis, and Pfizer. JAAP: reports grants and personal fees from Lundbeck during the conduct of the study. JM: reports personal fees from Lundbeck, outside of the submitted work. JR: reports grants from Lundbeck during the conduct of this study, personal fees and serving as board member (Nalmefene) for Lundbeck. JZ: reports personal fees from Lundbeck, and consultancy fees and salaries from Menarini, Lilly and Gilead outside of the submitted work. LK: reports funding from Lundbeck for a research project on alcohol epidemiology unrelated to this study. MBa: reports consulting fees received from Lundbeck (through consulting business Copentown) for organisational and process support in connection with the project. Copentown, which is owned by MB, also has other pharmaceutical/healthcare clients. MBe: reports no direct support from Lundbeck but for foundation of BHÄV (with the aim to promote Junior GPs).

Consent for publication
Not applicable.

Ethics approval and consent to participate
This study was an anonymous survey and exempt from respective research ethics committee approval in each of the countries. All respondents were presented a brief description of the aims of the study before the actual survey started. Consent to participate was therefore a precondition of taking the survey.