Introduction
In all global comparative risk assessments, alcohol use is amongst the 10 leading risk factors for both deaths and disability adjusted life years (Rehm and Imtiaz, Reference Rehm and Imtiaz2016; GBD 2016 Alcohol Collaborators, 2018) and has been estimated to cause about 3 million deaths annually (Shield et al., Reference Shield, Manthey, Rylett, Probst, Wettlaufer, Parry and Rehm2020). It has been linked with increasing the risk of a number of diseases including alcohol use disorders (Grant et al., Reference Grant, Goldstein, Saha, Patricia Chou, Jung, Zhang, Pickering, June Ruan, Smith, Huang and Hasin2015), cancers (Bagnardi et al., Reference Bagnardi, Rota, Botteri, Tramacere, Islami, Fedirko, Scotti, Jenab, Turati, Pasquali, Pelucchi, Galeone, Bellocco, Negri, Corrao, Boffetta and La Vecchia2015), liver disease (Rehm et al., Reference Rehm, Taylor, Mohapatra, Irving, Baliunas, Patra and Roerecke2010), infectious diseases (Imtiaz et al., Reference Imtiaz, Shield, Roerecke, Samokhvalov, Lönnroth and Rehm2017) and ischaemic (for heavy drinking occasions)(Roerecke and Rehm, Reference Roerecke and Rehm2014) as well as non-ischaemic cardiovascular disease (Rehm and Roerecke, Reference Rehm and Roerecke2017). Although the highest levels of per capita alcohol consumption are found in the European region (World Health Organisation, 2018), the pattern of high levels of alcohol consumption is also prevalent in the Latin American region (Manthey et al., Reference Manthey, Shield, Rylett, Hasan, Probst and Rehm2019), along with a high level of negative consequences (World Health Organisation, 2018). In Colombia, Mexico, and Peru, the three Latin American countries included in this study, alcohol use ranked as the fifth (in Mexico) and sixth (in Colombia and Peru) highest risk factor for death and disability in 2017 (Institute for Health Metrics and Evaluation, 2019a, 2019b, 2019c). The estimated percentages of deaths attributable to alcohol in the three countries ranged between 6.4 and 11% for males and 1.2–2.1% for females, and percentages of total attributable disability adjusted life years were above the world average at 7.6–12% for males and 2.1–3% for females (Gakidou et al., Reference Gakidou, Afshin, Abajobir, Abate, Abbafati, Abbas, Abd-Allah, Abdulle, Abera and Aboyans2017; GBD 2016 Alcohol Collaborators, 2018). These estimations show that the three countries could benefit from widespread implementation of measures to decrease heavy drinking in order to reduce the alcohol-related harm.
There is a large and robust evidence base demonstrating positive impacts for alcohol screening and brief advice (SBA) programmes, particularly when delivered in primary health care (PHC) settings. Over 70 randomised controlled trials suggest these simple interventions are both clinically and cost-effective at helping clinicians to identify patients drinking excessively and to provide short, structured advice to those needing to reduce their alcohol consumption (O’Donnell et al., Reference O’Donnell, Anderson, Newbury-Birch, Schulte, Schmidt, Reimer and Kaner2014; Kaner et al., Reference Kaner, Beyer, Muirhead, Campbell, Pienaar, Bertholet, Daeppen, Saunders and Burnand2018). While evidence for the effectiveness of alcohol SBA in PHC comes mainly from studies in high-income countries (HIC) (O’Donnell et al., Reference O’Donnell, Wallace and Kaner2014), emerging evidence points to its effectiveness also in middle-income countries (MIC) (Joseph and Basu, Reference Joseph and Basu2017), including in the Latin American region (Ronzani et al., Reference Ronzani, Mota and de Souza2009; Moretti-Pires and Corradi-Webster, Reference Moretti-Pires and Corradi-Webster2011). Evidence from PHC settings in HIC also shows that despite the established effectiveness of alcohol SBA, uptake in routine care remains low (Colom et al., Reference Colom, Scafato, Segura, Gandin and Struzzo2014; O’Donnell et al., Reference O’Donnell, Wallace and Kaner2014). Likewise, although there are on-going efforts to introduce SBA in Latin American countries (Gelberg et al., Reference Gelberg, Natera Rey, Andersen, Arroyo, Bojorquez-Chapela, Rico, Vahidi, Yacenda-Murphy, Arangua and Serota2017), widespread implementation has still not been achieved.
Scaling up SBA programmes will increase the number of patients detected to drink excessively and receiving advice on how to cut down, which could in turn lead to reduced alcohol consumption among the identified risky drinkers and its associated individual and wider societal harms. When aiming to scale up alcohol SBA in a new context however, it is beneficial to engage and consult with local stakeholders in order to adapt the intervention and increase the likelihood of successful and widespread implementation (Theobald et al., Reference Theobald, Brandes, Gyapong, El-Saharty, Proctor, Diaz, Wanji, Elloker, Raven, Elsey, Bharal, Pelletier and Peters2018). This study assessed the perspectives of key local stakeholders in three municipalities in Colombia, Mexico and Peru on two aspects relevant for successful implementation of SBA in practice: perceived appropriateness of the intervention and barriers to adoption.
First, appropriateness has been defined as the perceived fit, relevance or compatibility of the evidence-based programme for a given practice setting, provider or consumer and/or the perceived fit of the intervention to address a particular issue or problem (Proctor et al., Reference Proctor, Silmere, Raghavan, Hovmand, Aarons, Bunger, Griffey and Hensley2011). Assessment of appropriateness can provide an insight to the social validity of the intervention as perceived in the local context (World Health Organisation, 2016) and to help understand the implementation processes once the intervention is implemented (Proctor et al., Reference Proctor, Silmere, Raghavan, Hovmand, Aarons, Bunger, Griffey and Hensley2011). There is currently a lack of information on perceived appropriateness of alcohol SBA in PHC settings in the Latin American context, and no other studies assessing this issue have been identified in the literature.
Second, studying existing or potential barriers to delivery can help identify the reasons behind the evidence-practice gap for a specific intervention or initiative, and thus support the development of more effective strategies to improve successful implementation (World Health Organisation, 2016). A large body of literature on barriers to alcohol SBA in PHC exists, suggesting lack of time, lack of training, providers’ attitudes and lack of organisational support, as core factors affecting delivery (Johnson et al., Reference Johnson, Jackson, Guillaume, Meier and Goyder2011; Rahm et al., Reference Rahm, Boggs, Martin, Price, Beck, Backer and Dearing2015; Abidi et al., Reference Abidi, Oenema, Nilsen, Anderson and van de Mheen2016; Derges et al., Reference Derges, Kidger, Fox, Campbell, Kaner and Hickman2017; Vendetti et al., Reference Vendetti, Gmyrek, Damon, Singh, McRee and Del Boca2017), However, most of this evidence comes from HIC (eg, UK, US, Finland, Sweden, Australia) (Johnson et al., Reference Johnson, Jackson, Guillaume, Meier and Goyder2011; Derges et al., Reference Derges, Kidger, Fox, Campbell, Kaner and Hickman2017), and there is less knowledge of whether the barriers are the same in low- and middle-income countries (LMIC). In Latin America, for example, the few published studies have focussed on barriers to SBA implementation in specialised rather than PHC settings (Hoffman et al., Reference Hoffman, Beltrán, Ponce, García-Fernandez, Calderón, Muench, Benites, Soto, McCarty and Fiestas2016; Isela et al., Reference Isela, Martínez, Yolanda, Trejo, Echeverría, Vicente and Medina-Mora2016), and identified factors such as lack of standardised guidelines, lack of training of the providers, lack of providers’ perceived role responsibility, lack of time, lack of proper infrastructure and diversity of users affecting their delivery. These barriers echo some of those found in HIC (Johnson et al., Reference Johnson, Jackson, Guillaume, Meier and Goyder2011; O’Donnell et al., Reference O’Donnell, Wallace and Kaner2014; Derges et al., Reference Derges, Kidger, Fox, Campbell, Kaner and Hickman2017). However, the evidence suggests there are also some region-specific barriers, such as the lack of proper facilities to deliver the intervention.
In order to facilitate the assessment and comparison of barriers between countries, the Tailored Implementation for Chronic Diseases (TICD) framework was used (Flottorp et al., Reference Flottorp, Oxman, Krause, Musila, Wensing, Godycki-Cwirko, Baker and Eccles2013). This framework groups the determinants of practice into seven domains: guideline factors, individual health professional factors, patient factors, professional interactions, incentives and resources, capacity for organisational change and social, political and legal factors (Flottorp et al., Reference Flottorp, Oxman, Krause, Musila, Wensing, Godycki-Cwirko, Baker and Eccles2013). The latter five domains can be further framed as contextual factors (Nilsen and Bernhardsson, Reference Nilsen and Bernhardsson2019). The added value of using such a framework is the recognition of different levels of influence on practice, including the importance of context, going beyond the individual-level factors which are often overly prominent in alcohol SBA implementation studies (Vendetti et al., Reference Vendetti, Gmyrek, Damon, Singh, McRee and Del Boca2017).
The main aim of the study was thus twofold. First, the study aimed to assess and compare the perceived overall appropriateness of the alcohol screening and brief advice from the perspective of local stakeholders in three municipalities in Colombia, Mexico and Peru. Second, the study aimed to assess and compare the key stakeholders’ perspective on the barriers to implementation of SBA in the three countries and explore any differences based on their occupations.
Methods
Design and setting
The study was carried out as part of a larger research project testing implementation strategies for SBA implementation in Colombia, Mexico and Peru (SCALA – Scale up of Prevention and Management of Alcohol Use Disorders and Comorbid Depression in Latin America) (Jane-LLopis et al., Reference Jane-LLopis, Anderson, Piazza, O’Donnell, Gual, Schulte, Gomez, Vries, Rey, Kokole, Bustamente, Braddick, Trujillo, Solovei, Leon, Kaner, Matrai, Manthey, Mercken, Pelayo, Rowlands, Schmidt and Rehm2020). A cross-sectional survey was disseminated in municipalities in the cities of Bogota, Lima and Mexico City. In order to maximise feasibility, the local researchers selected the municipalities based on their location in the country and existing networks. To further characterise the setting, main demographic and health care system characteristics of the three countries are presented in Table 1.
Table 1. Demographic and health system characteristics in Colombia, México and Perú
1 DANE (2018). Censo nacional de población y vivienda. Proyecciones de población. Available from: https://www.dane.gov.co/index.php/estadisticas-por-tema/demografia-y-poblacion/proyecciones-de-poblacion [accessed 23.9.2020]
2 INEGI (n.d.). Banco de indicadores, 2015. Available from https://www.inegi.org.mx/app/indicadores/?t=0070&ag=09014##D00700060 [accessed 23.9.2020]
3 INEI (2017). Censos nacionales 2017: XII Censo de Población, VII de Vivienda y III de Comunidades Indígenas. Sistema de Consulta de Base de Datos. Available from: http://censos2017.inei.gob.pe/redatam/ [accessed 23.9.2020]
4 OECD (2015). OECD Reviews of Health Systems: Colombia 2016. Paris: OECD Publishing.
5 Báscolo, E., Houghton, N., & Del Riego, A. (2018). Lógicas de transformación de los sistemas de salud en América Latina y resultados en acceso y cobertura de salud. Revista Panamericana de Salud Pública, 42, e126.
6 WHO (n.d.) Global Health Expenditure database: https://apps.who.int/nha/database/ [accessed 7.10.2020]
7 WHO (2017). Primary health care systems (PRIMASYS): case study from Mexico, abridged version. Geneva: World Health Organization.
8 WHO (2017). Primary health care systems (PRIMASYS): case study from Peru, abridged version. Geneva: World Health Organization.
9 WHO (2017). Primary health care systems (PRIMASYS): case study from Colombia, abridged version. Geneva: World Health Organization.
10 WHO (n.d.) Global Health Workforce Statistics, the 2018 update, Available from: https://apps.who.int/gho/data/node.main.HWFGRP?lang=en [accessed 7.10.2020]
Participants
In order to ensure the information was gathered from participants who were familiar with the intervention and/or setting, only stakeholders from the three countries who fulfilled at least one of the following inclusion criteria were invited to participate in the study: experience in the field of alcohol (prevention), experience in implementing any kind of intervention in PHC, currently working in a PHC centre. In each country, a local research group with knowledge of the local context identified the stakeholders in their network fitting these criteria and invited them to take part in the survey via e-mail. Both health professionals and professionals from other occupations (eg, regional health administrators) were invited to participate in the survey. Eighty-three stakeholders were invited to participate and in total; 55 stakeholders responded to the survey (66% response rate): 16 from Colombia (53% response rate), 18 from Mexico (75% response rate) and 21 (72% response rate) from Peru.
Instrument
The survey was disseminated online and questions covered demographic characteristics (gender, country, occupation) and 24 items regarding appropriateness and barriers of alcohol SBA. All the survey questions were developed by the authors, as no instruments based on the TICD framework to study implementation outcomes and barriers were found in the literature.
Appropriateness was assessed with three questions covering: fit of intervention to the problem, fit to the local setting and fit of the provider. Respondents were asked to rate their agreement with alcohol SBA being an appropriate approach to reduce heavy alcohol use, and the PHC centre being a suitable setting to conduct alcohol SBA on 5-point Likert scales (1 = completely disagree to 5 = completely agree). Additionally, they had to indicate which health professionals they considered suitable to carry out alcohol SBA in primary care.
The development of a list of barriers to the implementation of SBA was guided by the TICD framework (Flottorp et al., Reference Flottorp, Oxman, Krause, Musila, Wensing, Godycki-Cwirko, Baker and Eccles2013), based on prior research identified through an examination of reviews in this area (Johnson et al., Reference Johnson, Jackson, Guillaume, Meier and Goyder2011; O’Donnell, et al., Reference O’Donnell, Wallace and Kaner2014; Derges et al., Reference Derges, Kidger, Fox, Campbell, Kaner and Hickman2017), and on recommendations of an expert panel with experience in the topic. The barriers identified in the literature have been extracted and categorised in the TICD framework under relevant domains and determinant headings. The list was shared with the expert panel, which selected additional determinants considered important based on their knowledge and experience. The full list of barrier items based on literature review and expert panel recommendations consisted of 46 items. This initial list was then shared with the local research teams in the three countries. Based on their feedback, the full list was shortened in order to increase the likelihood of response. Next, the most relevant determinants were selected by the central research team based on consultation with the local research teams in the three countries. The final, shortened list contained 21 items, with each categorised into the corresponding TICD framework determinants in one of the domains: guideline factors, individual health professional factors, patients factors, professional interactions, incentives and resources, capacity for organisational change, social legal and political factors. Questions were rated on a 5-point Likert scale (1 = completely disagree, it is not a barrier to 5 = completely agree, it is a large barrier). Both the long and shortened lists of barriers are available as supplementary material.
The survey was developed in English, translated to Spanish, and further refined based on feedback from the local research teams. Before dissemination, two to three experts per country piloted the survey.
Data collection
The data were collected in April and May 2019 using Formdesk, an online survey software. Respondents were invited to participate through e-mail by the local researcher and were sent a reminder after a week in case of no response. No identifiable data were collected, and the survey was anonymous. Participants had to sign the informed consent electronically before they were able to participate in the survey. Ethical review was not required for anonymous online surveys in all three countries.
Data analysis
IBM SPSS Statistics 24 was used for data analysis. Data was first analysed separately for each of the countries (Colombia, Mexico, Peru), and for barriers, also by occupation. To obtain the percentages of respondents agreeing with the statements, the number of participants agreeing or completely agreeing with an item were divided by the number of all participants. Medians and interquartile ranges were computed. Due to the small sample size and non-normal distribution, as tested with one-way Kolmogorov–Smirnov test, non-parametric tests (Kruskal–Wallis H for medians and Chi square for percentages) were used for comparisons. Where additional post-hoc tests (Mann–Whitney U) were used, Bonferroni correction was applied.
Results
In total, 55 respondents participated in the survey. Their demographic characteristics are presented in Table 2. Approximately half of the participants were health care providers, out of which the majority were general practitioners (GPs) and psychologists.
Table 2. Characteristics of key local stakeholders included in the study
* midwife, social worker.
** PHC centre manager, national public policy advisor, national consultant and private treatment centre employee.
Appropriateness
As seen in Table 3, there were high proportions of respondents (75% or above, with one exception) considering alcohol SBA to be an appropriate approach to reduce heavy alcohol use (fit to the problem), and the PHC centre being a suitable place to perform alcohol SBA (fit to the setting). Considering the fit of provider, respondents in all three countries indicated four types of professionals to be appropriate to carry out alcohol SBA (all percentages above 80%): GPs, nurses, psychologists and social workers.
Table 3. Response rates and comparison of perceived appropriateness of alcohol SBA in Colombia, México and Perú
† Me–Median, IQR-Interquartile range.
* % summed responses Agree and Completely agree for the first two rows, % Yes for the latter six rows.
** Kruskal–Wallis H test for the first two rows, Chi square test for the latter six rows.
a Post-hoc test showed significant difference between Peru and Colombia (Mann–Whitney U = 15.440, P = 0.007).
Kruskal–Wallis H test showed a significant difference between countries’ perception of alcohol SBA as an appropriate approach to reduce heavy alcohol use, with post-hoc tests revealing a significant difference between Colombian (most endorsements) and Peruvian respondents (least endorsements). No other county differences were found.
Barriers to implementation of alcohol SBA
In Table 4, the percentages concerning perceived barriers for implementation are presented for all the three countries, as well as medians and their comparisons. Four barriers stood out with having high rating (defined as two thirds or more of respondents) in all three countries: heavy drinking patients’ beliefs that their drinking is normal (patient factors TICD domain), lack of on-going support for providers (assistance for clinicians TICD domain), difficulty of accessing referral services (professional interactions TICD domain) and lenient laws and regulations influencing price and availability that encourage cultural tolerance to alcohol (social, political and legal factors TICD domain).
Table 4. Response rates and comparison of perceived barriers to alcohol SBA by country
× Domains 3–7 can also be considered as contextual factors, based on (Nilsen and Bernhardsson, Reference Nilsen and Bernhardsson2019).
† Me–Median, IQR-Interquartile range.
* % responses Agree and Completely agree.
** Kruskal–Wallis H test.
a Post-hoc test showed significant difference between Mexico and Peru (Mann–Whitney U = −18.10, P = 0.001).
b Post-hoc test showed significant difference between Mexico and Peru (Mann–Whitney U = −13.56, P = 0.018) and Colombia and Peru (Mann–Whitney U = −12.82, P = 0.035).
Three barriers had high ratings in two countries: lack of financial (Colombia and Mexico) and non-financial incentives (Colombia and Peru) (both Incentives and Resources TICD domain), and lack of necessary organisational changes (Mexico and Peru) (Capacity of organisational change TICD domain). Certain barriers with high agreement percentages were also country specific: lack of sufficient staff for implementation in the centre as well as patients’ preference not to discuss their alcohol consumption in Peru (the first, social, political and legal factors and the latter, patient factors TICD domain) and lack of providers’ time in Colombia (individual health professional factors TICD domain). The barriers of SBA not being culturally appropriate, not feasible in practice and requiring too much effort (all in Guideline factors TICD domain) were lowest rated in all three countries, with most percentages under 20%.
Country comparison showed two barriers with a statistically significant difference in their ratings: the guidelines for screening and brief advice not being clear enough and instruments for screening not being available. Post-hoc tests indicated that Peruvian respondents were more likely to endorse lack of guideline clarity as compared to Mexican respondents, and more likely to cite lacking availability of SBA instruments as a barrier compared to both Colombian and Mexican respondents. Despite the differences, those were not the most frequently endorsed barriers.
As health professional level barriers are commonly mentioned in previous qualitative research in this area for example (Johnson et al., Reference Johnson, Jackson, Guillaume, Meier and Goyder2011; Derges et al., Reference Derges, Kidger, Fox, Campbell, Kaner and Hickman2017), but were not among the highest rated barriers in our survey (with agreement percentages between 42 and 62%), we decided to further explore barriers by occupation. The available sample allowed us to compare GPs’ responses with responses from psychologists and other occupations (non-health care providers). Comparison showed statistically significant differences in three determinants from the individual health professional factors TICD domain: lack of skills to implement the intervention, providers thinking that alcohol SBA will not help their patients and not considering providing alcohol SBA as their responsibility (Table 5). In all three cases, the GPs rated these barriers significantly lower than psychologists and other professionals.
Table 5. Response rates and comparison of perceived barriers to alcohol SBA by occupation
× Domains 3–7 can also be considered as contextual factors, based on (Nilsen and Bernhardsson, Reference Nilsen and Bernhardsson2019).
† Me–Median, IQR-Interquartile range.
* % responses Agree and Completely agree.
** Kruskal–Wallis H test.
a Post-hoc test showed significant difference between GPs and psychologists (Mann–Whitney U = −14.69, P = 0.023).
b Post-hoc test showed significant difference between GPs and psychologists (Mann–Whitney U = −16.62, P = 0.009) and GPs and other occupations (Mann–Whitney U = −19.72, P = 0.001).
c Post-hoc test showed significant difference between GPs and psychologists (Mann–Whitney U = −19.05, P = 0.002) and GPs and other occupations (Mann–Whitney U = −22.91, P = 0.001)
Discussion
The aim of this study was to assess and compare the perceived general appropriateness of alcohol screening and brief advice and the perceived barriers to its implementation from the perspective of local stakeholders in three municipalities in Colombia, Mexico and Peru.
The study showed that delivering alcohol SBA in PHC setting was generally seen as an appropriate intervention to reduce heavy alcohol use in these three Latin American countries, although there were small differences, with SBA being considered more appropriate among Colombian compared to Peruvian respondents. In all three countries, GPs, nurses, psychologists and social workers were considered suitable for delivery of SBA in primary care. This suggests that scaling up SBA programmes in PHC in the Latin American context might be achieved by expanding the range of providers. Whilst many studies from HIC have shown the effectiveness of SBA with GPs as the intervention provider (O’Donnell et al., Reference O’Donnell, Anderson, Newbury-Birch, Schulte, Schmidt, Reimer and Kaner2014), there is also emerging evidence of effectiveness of non-physician led alcohol interventions (Sullivan et al., Reference Sullivan, Tetrault, Braithwaite, Turner and Fiellin2011), such as nurses (Platt et al., Reference Platt, Melendez-Torres, O’Donnell, Bradley, Newbury-Birch, Kaner and Ashton2016) or social workers in social service settings (Schmidt et al., Reference Schmidt, McGovern, Schulte, O’Donnell, Lehmann, Kuhn, Schäfer, Newbury-Birch, Anderson, Kaner and Reimer2015). Another consideration not explored in the study, but relevant for practice and further investigation, is the possibility of interprofessional approaches, where team members of different occupations work together to improve health outcomes for the patient (Zwarenstein et al., Reference Zwarenstein, Reeves and Perrier2005). In case of alcohol screening in brief advice this could mean screening done by one member of the team (eg, nurse) and advising by another (eg, GP or psychologist). This could enable scaling up via better integration of SBA into the existing workflow. Further research is needed however on the effectiveness and patient acceptability of SBA delivered by non-physicians in the LMIC context.
The assessment of barriers also showed that the pattern in perception of barriers was similar in all three countries. This implies that a similar approach can be used to implement alcohol SBA across these particular countries, with tailoring efforts focussed on the specific parts needed to improve fit in the local context. In general, intervention-related factors (guideline factors TICD domain) such as lack of feasibility or cultural fit were not seen as major barriers, which echo previous evidence from the HIC context. Yet countries differed concerning SBA guideline clarity: at least a third of Colombian and Peruvian respondents mentioned lack of clarity as a barrier; whereas the percentage among Mexican respondents was much lower. This reflects the differing national contexts with regard to the existing guidelines: in Mexico, official standards establish the obligatory procedures and criteria for mandatory prevention, treatment and control of addictions, which include asking questions on alcohol use (Norma Oficial Mexicana NOM-028-SSA2-2009 para la prevención, tratamiento y control de las adicciones, 2009), and including this information in the patient’s history (Norma Oficial Mexicana NOM-004-SSA3-2012 del expediente clínico, 2012), specifically in primary health care context. In Colombia, the alcohol SBA recommendations are included as part of clinical practice guidelines that focus on detection and treatment of alcohol abuse and dependence on primary, secondary and tertiary care level (Ministerio de Salud y Protección Social, 2013), but there are no official standards as in Mexico. Finally, in Peru, recommendation for providers to deliver alcohol screening can be considered implicitly included in general recommendations to perform mental health-related screening (alcohol use disorder being considered as one of subcategories) (Ministerio de Salud Peru, 2018), therefore making the alcohol SBA guidelines potentially less clear. However, when considered in light of other higher rated barriers, improving clarity of guidelines (at least in Colombia and Peru) is not the main priority.
Looking at the results from the perspective of the TICD framework, the barriers with the highest agreement in all countries can be categorised as contextual (as defined in Nilsen and Bernhardsson, Reference Nilsen and Bernhardsson2019). Specifically, respondents in all three countries highlighted heavy drinking patients’ thinking that their drinking is normal, lack of on-going support for providers, difficulty of accessing referral services and lenient laws and regulations influencing price and availability encouraging cultural tolerance to alcohol, as key factors affecting implementation. Again, these barriers reflect those identified in HIC literature, where patients’ normalisation of heavy drinking, referral issues and organisational factors, including lack of a supportive policy environment, are commonly cited as obstacles to delivery (Anderson et al., Reference Anderson, Kaner, Wutzke, Wensing, Grol, Heather and Saunders2003; Johnson et al., Reference Johnson, Jackson, Guillaume, Meier and Goyder2011; Derges et al., Reference Derges, Kidger, Fox, Campbell, Kaner and Hickman2017; Vendetti et al., Reference Vendetti, Gmyrek, Damon, Singh, McRee and Del Boca2017). To tackle the barrier of patients’ normalised perception of their own heavy drinking, there is a need for communication strategies surrounding SBA programmes to involve a reframing component, which highlights that much alcohol-related harm is experienced by those drinking at non-dependent levels (eg, see (Heather, Reference Heather2006). Lack of restrictions for on/off premise sales of alcoholic beverages or limited restrictions on alcohol advertising in the participating countries might have contributed to the perception of lenient alcohol control policies expressed by the stakeholders in this survey (World Health Organisation, 2018). Indeed, recent research has highlighted the need to address these types of policy factors in LMICs in order to reduce alcohol-related harm (Shield et al., Reference Shield, Manthey, Rylett, Probst, Wettlaufer, Parry and Rehm2020).
Barriers from the individual health professional factors TICD domain were neither among the highest nor among the lowest rated barriers. This might have been influenced by differing opinions based on occupation, as shown by the comparison between GPs, psychologists and others. The provider related factors such as lack of skills, lack of responsibility and belief about the intervention not helping the patients were considered much less of a barrier by the GP respondents compared to psychologists and other occupations. Studies from HIC countries however suggest that attitudinal factors do hinder GPs’ implementation of SBA, such as lower role security and therapeutic commitment (Anderson et al., Reference Anderson, Kaner, Wutzke, Wensing, Grol, Heather and Saunders2003), as well as aligning with the disease rather than preventive model of work and valuing individual personal responsibility for protection from alcohol-related harm (Anderson et al., Reference Anderson, Wojnar, Jakubczyk, Gual, Segura, Sovinova, Csemy, Kaner, Newbury-Birch, Fornasin, Struzzo, Ronda, Van steenkiste, Keurhorst, Laurant, Ribeiro, Do rosário, Alves, Scafato, Gandin and Kolsek2014). Whilst the sample is too small to draw definite conclusions, some of the possible reasons for our results may be selection bias (ie, GPs participating in the survey were potentially already more educated and aware about alcohol), GP’s higher self-efficacy when it comes to delivering interventions in PHC, or psychologists seeing the brevity of the intervention as less appropriate to their practice. Nevertheless, these preliminary results point us in direction of the health professional-related barriers potentially being profession-specific and suggest that more research is needed to explore the perspectives of and barriers experienced by other occupations.
Results of this study suggest that multi-level strategies are needed to address barriers to widespread SBA implementation in Colombia, Mexico and Peru. First, although individual health professional level factors were not ranked highest, barriers relating to a perceived lack of skills, self-efficacy, role-legitimacy or and belief in intervention effectiveness can be addressed through means of provider training programmes. The preliminary differences found here between GPs and psychologists suggest that tailoring training might be necessary, using different approaches for providers of different occupations, based on the specific needs, as well as specific strengths, of different health care providers (Wamsley et al., Reference Wamsley, Satterfield, Curtis, Lundgren and Satre2018).
Yet, whilst training can help increase providers’ intervention-related knowledge, skills and self-efficacy, previous research has shown that is unlikely to be sufficient to improve implementation on its own, particularly over the longer term (Anderson, Reference Anderson2004). Looking at the TICD domains of the highest rated barriers in this study, it can be seen that they all relate to the wider social, political and cultural SBA delivery context. Thus, interventions that provide continuous support for the providers (Anderson et al., Reference Anderson, Bendtsen, Spak, Reynolds, Drummond, Segura, Keurhorst, Palacio-Vieira, Wojnar, Parkinson, Colom, Kłoda, Deluca, Baena, Newbury-Birch, Wallace, Heinen, Wolstenholme, van Steenkiste, Mierzecki, Okulicz-Kozaryn, Ronda, Kaner, Laurant, Coulton and Gual2016) and efforts to change the community social norms (Anderson et al., Reference Anderson, Jané-Llopis, Hasan and Rehm2018) related to alcohol (through education or legislation) are also needed to address the perceived relevant barriers in these three countries. This has been shown also through previous work in HIC, where series of multi-country studies concluded that education and support in the working environment are necessary to increase involvement of health care providers (in that case GPs) in managing alcohol problems (Anderson et al., Reference Anderson, Kaner, Wutzke, Wensing, Grol, Heather and Saunders2003, Reference Anderson, Wojnar, Jakubczyk, Gual, Segura, Sovinova, Csemy, Kaner, Newbury-Birch, Fornasin, Struzzo, Ronda, Van steenkiste, Keurhorst, Laurant, Ribeiro, Do rosário, Alves, Scafato, Gandin and Kolsek2014).
Strengths and weaknesses
This study contributes to the literature on SBA implementation with evidence from an underexplored region (Latin America) using a quantitative approach that allows for direct comparisons between three countries. The list of barriers to implementation of alcohol SBA was developed within a theoretical framework, combining evidence from previous empirical studies and recommendations from an expert panel. Furthermore, inclusion of a range of key local stakeholders with different occupations and experience in the topic allowed for a broader perspective on barriers to implementation, assessing determinants on various professional and health system levels. We encourage the use of the proposed list of barriers in future SBA barrier assessments in PHC or other occupations across Latin America and elsewhere, if locally adapted.
Beside the abovementioned strengths, the current study also has limitations. One, due to its focus on a municipal context in three Latin American countries and a limited range of eligible stakeholders with enough experience to be consulted, the low sample size limits broader generalisation of the results. Additionally, as the study focussed only on the three countries participating in SCALA project, the results cannot necessarily be generalised to other Latin American countries. While comparison between the three countries points to predominant similarities rather than differences in barriers perception, further local assessment would be necessary before scaling up alcohol SBA beyond Colombia, Mexico and Peru. Two, there are also some general shortcomings of the survey approach to identifying barriers that should be acknowledged: whilst this approach enables us to compare statistically the relative importance of specific barriers to implementation, as these barriers were pre-determined by the team constructing the questionnaire, some other relevant barriers might have been overlooked (Nilsen, Reference Nilsen2015). In our case, the list of barriers had to be considerably shortened in its final form in order to ensure respondents’ completion of the survey, resulting in potentially relevant barrier(s) being excluded. However, it is important to note that this shortcoming was addressed by consulting with the experts and local research partners when determining the final list. Three, the perceived barriers may not necessarily correspond to the actual barriers encountered when implementing the intervention (Nilsen, Reference Nilsen2015). This was beyond the scope of our study, but our findings provide a useful baseline data, whereby future intervention evaluations can compare the encountered barriers to the perceived ones identified in our study. Four, this study did not look at the patient perspective on the implementation of alcohol SBA, which should also be explored in further studies, in line with previous research, such as Lock, Reference Lock2004, or Hutchings et al., Reference Hutchings, Cassidy, Dallolio, Pearson, Heather and Kaner2006. Furthermore, among health professionals our sample predominantly contained perspectives of GPs and psychologists and further perspective from other professionals also considered appropriate to deliver alcohol SBA (nurses and social workers) should be included in any follow-up research.
Future perspectives
Findings of the study point to the necessity of considering barriers on a broader scale than just at the individual provider level. For SCALA project, this means designing process evaluation related data collection in a way to capture the broad spectrum of possible experienced barriers and facilitators. Results will also be used along other data collected in the SCALA project to help explain the outcome on provider level – why did or did not providers implement alcohol SBA in their daily practice. Results may also contribute to wider implementation of alcohol SBA in Latin American countries. We encourage other researchers and practitioners to use the developed instrument (available as the supplementary material) for rapid assessment of appropriateness and barriers in any novel LMIC context and as an aid when tailoring the intervention to the specific local context.
Conclusion
This study investigated local stakeholders’ views of the appropriateness of alcohol SBA, as well as their perceived barriers to its implementation in three municipalities in Colombia, Mexico and Peru. Implementation of SBA in PHC is generally considered as an appropriate means to reduce alcohol-related harm in all three countries. In contrast to evidence from HIC countries, context-related factors were cited as major barriers to SBA implementation, namely lack of support for providers, difficulties with accessing referral services, patients underestimating the danger of their consumption levels and lax alcohol control legislation. Despite the similarities, it is still necessary to be sensitive to existing differences and tailor of the specific SBA programmes for each country.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/S1463423620000675
Financial support
The research leading to these results or outcomes has received funding from the European Horizon 2020 Programme for research, technological development and demonstration under Grant Agreement no. 778048 – Scale-up of Prevention and Management of Alcohol Use Disorders and Comorbid Depression in Latin America (SCALA). Participant organizations in SCALA can be seen at: www.scalaproject.eu. The views expressed here reflect those of the authors only and the European Union is not liable for any use that may be made of the information contained therein.
HLP received funding from the Spanish Ministry of Science, Innovation and Universities, Instituto de Salud Carlos III through a ‘Juan Rodes’ contract (JR19/00025), with the support of the European Social Fund, and IDIBPAS is a CERCA Programme/Generalitat de Catalunya.
Conflict of interest
H.L.P has received travel grants from the laboratories honoraria and travel grants from Janssen and Lundbeck. None of them has relationship with this research.
A.G. has received funding from Novartis for a clinical trial on cocaine, not related to the current work.
Ethics
Ethical review was not required for anonymous online surveys in all three countries.
Informed consent
Informed consent was obtained from all individual participants included.
Open access
