Recommendations for effective collaboration and capacity building in epidemiological studies on the effect of alcohol and drug use on traffic safety in low- and middle-income countries

Abstract

Objective

Alcohol or drug impairment is a major risk factor for road traffic crashes, and studies on this issue are essential to provide evidence-based data for policymakers. In low- and middle-income countries (LMICs), such studies are often conducted in partnership with one or more organizations in high-income countries (HICs). The aim of this article is to provide recommendations for improving project planning and decision-making processes in epidemiological studies on alcohol, drug and traffic safety in LMICs involving HICs.

Methods

We searched Pubmed, Google Scholar, and Google Search for articles and reports in English about lessons learned when conducting collaborative research in LMIC as well as papers presenting recommendations for effective research collaboration with partners in LMICs.

Results

Based on the search results, we selected 200 papers for full text examination. Few were related to studies on the effect of alcohol or drug use on road traffic safety. However, several conclusions and recommendations from other studies were found to be relevant. We combined the findings with our own experience in a narrative review. We also present a checklist for risk and quality assessment.

Conclusions

Many papers presented similar recommendations, which included the importance of addressing local needs, ensuring adequate resources, local project ownership and leadership, establishing strong partnerships among all involved stakeholders, promoting shared decision-making and planning, and implementing strategies to translate research findings into policy, practice, and publications. It is also important to avoid HIC bias, which prioritizes the interests or perspectives of HICs over those of LMICs.

Keywords:

• Low- and middle-income countries
• collaborative research
• alcohol
• drugs
• driving under the influence
• traffic safety

Introduction

Low- and middle-income countries (LMICs) have approximately 60% of the world’s vehicles, but account for 93% of the world’s fatalities on the roads (World Health Organization 2021). Some LMICs, particularly in Africa, have 30 or more fatalities per 100,000 inhabitants per year, compared to less than three per 100,000 in several European countries (World Health Organization 2018; International Transport Forum 2020). The main reasons for higher relative crash rates in LMICs are the poorer quality of roads and motor vehicles than in high-income countries (HICs); also, the knowledge and attitudes toward road safety and risk factors may be weaker in LMICs (Stewart et al. 2012; Pechansky et al. 2016; World Health Organization 2018). The number of motor vehicles in LMICs is increasing. We expect that without adequate preventive actions, the number of road traffic crash injuries and deaths will also increase.

Driving under the influence (DUI) of alcohol or drugs is an important contributing factor to road traffic crashes worldwide (World Health Organization 2021). Based on research findings at the beginning of the twentieth century, the first countries introduced legal blood alcohol concentration limits for motor vehicle drivers in the 1930s and 1940s, and others followed. Later, many HICs have improved their legislation on alcohol and drug impaired driving and increased the enforcement, strengthened the information about the crash risks associated with DUI, implemented random breath alcohol testing or sobriety checkpoints, and treatment or rehabilitation of drivers with alcohol or drug problems (European Commission 2021; Venkatraman et al. 2021). Combined efforts have resulted in a significant decline in the number or proportion DUI-related road traffic crashes and fatalities in several countries, such as Australia, Canada, Chile, New Zealand, Norway, and the USA (Christophersen et al. 2016; Nazif-Muñoz et al. 2020; Harris et al. 2022; Vanlaar et al. 2022).

The authorities in many LMICs have not had a similar focus on reducing the incidence of DUI. As a result, the prevalence of alcohol-impaired drivers in random road traffic is higher in some LMICs than in HICs (Damsere-Derry et al. 2016; Oyono et al. 2021; Rabelo-da-Ponte et al. 2021). The prevalence of alcohol and drugs among drivers injured in road traffic crashes is also relatively high in several countries (Sundet et al. 2020; Divsalar et al. 2021; Bombana et al. 2022).

A key factor in tackling the road traffic injury burden in LMICs is the implementation of preventive measures based on a good knowledge base, which requires increased research capacity across a range of interlinking sectors. This should be backed by strong political commitment and adequate and sustainable resources (Peden et al. 2004; Heydari et al. 2019). Not all preventive measures implemented in HICs are easily transferable to LMICs, and some measures and interventions that are effective in HICs may not be as effective in LMICs. Therefore, studies are needed to collect local data and evaluate which efforts can be successfully implemented (Peden et al. 2004; Pechansky and Chandran 2012; Stewart et al. 2012; Heydari et al. 2019; Shuey et al. 2020). For example, drivers’ compliance with legislation and police enforcement might be lower. Also, drivers may ignore signs and engineering measures, and the impact of education and information might be limited (King 1999). High-technology solutions that may be part of the approach in HICs – such as intelligent roads or high levels of digitalization of traffic networks, are not the reality of LMICs.

A point to bear in mind when planning the development of collaborative studies in LMICs is the issue of budgeting and financing. Local authorities often do not have the financial capacity to fund research projects on road safety. Therefore, several governmental and non-governmental organizations in HICs contribute to improving road traffic safety in LMICs by supporting research initiatives. It is not uncommon that even when these resources are available, professionals from LMICs may not be aware of them or have the knowledge to approach, prepare or submit grant applications aimed at these institutions. Collaboration with experienced researchers in HICs may therefore be a key component in project initiation, development, and funding applications. This enables capacity building by training and sharing knowledge. The research process itself may be more important than the actual research findings, as the research experience can enable project team members to conduct good-quality research activities to improve road safety in the future. However, more or less serious challenges may occur in collaborative projects.

The authors of this article have many years of experience with coordinating and implementing research projects on alcohol, drugs and traffic safety with partners from HICs and LMICs. Some mistakes and challenges that we observed, as well as their consequences and possible mitigation efforts, are presented in Supplemental Material Table A1. The consequences were in most cases delays and additional costs, and in some cases reductions of scope or revision of study aims, and in a few cases project termination. Generally, projects that were initiated, led, and funded from an LMIC encountered fewer challenges than projects initiated, led, and funded from a HIC but performed in an LMIC.

In our perspective, many errors and challenges could be prevented by a thoughtful establishment of the project team, aiming at eliminating the potential power imbalance between partners. Also, improved project planning, execution, communication, and follow-up would avoid some challenges. We also experienced problems getting sufficient funding, difficulties allocating personnel, and lack of local researchers with relevant experience. Challenges were also related to the recruitment of participants (study sample), and data collection and analysis. Serious challenges occurred because obtaining import permits for equipment or export permits for biological test samples was impossible or extremely bureaucratic and time-consuming. Difficulties in translating research findings into policy and practice occurred as some stakeholders did not support study conclusions which would be potentially against government or agency policies at the time.

This article aims to provide recommendations for collaborative epidemiological studies on the effects of alcohol and drug use on road traffic safety in LMICs. We also suggest ways to address or overcome some challenges that may arise. In some of the recommendations, we primarily address partners in countries with advanced technologies and economies. The article does not cover all aspects of this topic, as each project is unique and specific to the country and situation.

Methods

We conducted a literature search using PubMed, Google Scholar, and Google Search for articles and reports in English presenting experiences when conducting research in LMICs, particularly collaborative research, as well as papers presenting recommendations for collaborative studies and quality assurance in research. We first screened the titles of papers, then read selected abstracts, and finally selected papers for the examination of the full text. We also examined reference lists for additional papers. We focused our attention on papers within the fields of medical sciences and traffic safety.

Results

The literature search revealed a large number of potentially relevant articles and reports. We selected about 200 of the papers for full text examination. Few articles were related to studies on alcohol, drugs, and traffic safety. However, published recommendations from some studies within other research fields were also relevant. As many papers gave similar recommendations, we chose to refer to a selection of articles in this narrative review. In this article we give recommendations based on our own experience combined with those found in the literature review.

Establishing successful international research collaboration

Research and capacity building in LMICs should be driven by national needs and priorities identified by local authorities and collaboration partners rather than driven by donors, and firmly anchored in the local community (Costello and Zumla 2000; Peden et al. 2004; World Health Organization 2014; Franzen et al. 2017; Faure et al. 2021). Therefore, local researchers should be included when setting aims and milestones for the research project. Local researchers should have principal roles, as this may increase the impact of the study on national policy-making and contribute to building the local academic infrastructure (Costello and Zumla 2000; Bowsher et al. 2019). The probability of success is higher if the leadership also is local (Datta et al. 2012; Matenga et al. 2019).

All partners in collaborative research projects should obtain institutional commitments, and top-level management should attend important meetings and sign collaboration contracts (Evers and Lokhoff 2012). A large project may need a steering committee, which is an advisory body of key stakeholders who oversee and support the project to ensure its success. The committee may include representatives from management, stakeholders, and local authorities. It provides guidance and assistance throughout the project’s lifecycle, and may contribute to translating research findings into improvements in policy and practice.

Successful collaboration requires commitment, trust, good and open communication, flexibility, shared decision-making, and mutual respect and equity among all investigators (Costello and Zumla 2000; De Grijs 2015; Stöckli et al. 2018; Bowsher et al. 2019; Matenga et al. 2019; Faure et al. 2021). Therefore, it is advisable to organize face-to-face meetings and spend time together, particularly in the early stages, to help research teams learn to know each other, share experience and knowledge, discuss suggestions and alternatives, and build a strong and trusting relationship (Datta et al. 2012; Evers and Lokhoff 2012; De Grijs 2015; Gewin 2018). Annual face-to-face meetings are advisable if the budget permits.

One key element to consider is that there will be culture-related differences in leadership and working styles, response times, and attitudes toward deadlines. These issues are common and can only be solved through flexibility and open discussions to avoid conflicts, and by trying to understand the local leadership and work culture. To avoid conflicts and misunderstandings, the research team should obtain consensus on some key elements.

Costello and Zumla (2000) published a checklist for LMIC research partnerships, and other organizations and research groups published recommendations on how to obtain successful partnerships by listing several challenges and suggestions on how to avoid or mitigate them (World Health Organization 2014; Stöckli et al. 2018; Bowsher et al. 2019). Those reports should be reviewed when initiating collaborative projects.

Establishing a research team

Finding a dedicated project leader (or principal investigator), researchers, and collaborators is an important early step that will define the project’s progress. An aim should be to avoid power imbalance between research partners when discussing priorities, timelines, expenditures, and more, especially if the research in an LMIC is funded by a HIC (Matenga et al. 2019). The project leader should be a local researcher; HIC researchers should have a capacity-building role.

The project team may need competence within several scientific fields, including traffic safety, social science, traffic psychology, epidemiology, statistics, pharmacology, toxicology, medical sciences, and/or drug analysis, depending on the study aims and methodology, to avoid errors and weaknesses. Recruiting local personnel with sufficient background knowledge and providing proper training might be a challenge. Therefore, some projects may need to involve experienced researchers from other countries to provide training and mentoring or contribute to project activities. Some projects may provide a technical exchange of professionals from LMICs to train and develop methods at HIC institutions.

Individuals from the local community who have no research background but are respected and trusted by various linguistic and cultural groups and authorities can provide essential knowledge for project planning and implementation. They can also help with tasks such as project management, record collection and maintenance, data monitoring, tracking, reviewing, and auditing.

For capacity building, local students (master’s or Ph.D. level) may be involved and use parts of the project for their theses. The aim is that these trained students contribute with their new experience in the same country or organization in the future.

Project planning and follow-up

Engagement by local collaborators is needed in the early planning phase, including the discussion of aims, procedures, opportunities, challenges, and budget. It is important to establish a clarity of ownership, roles and responsibilities, priorities, tasks, expectations, and timelines (De Grijs 2015). An agreed written document may be useful to help clarify doubts about these topics.

A study protocol should be prepared with sufficient details so that the study can be reproduced. The local ethics committee must approve the study, and sometimes also the ethics committee of the collaborating institutions. In some specific cases, institutions may not have a structured ethics committee, and different approaches are employed to circumvent such barriers – such as ad hoc committees formed specifically for projects, or committees from other known institutions in the country that might vouch for project evaluation and oversight. Also, the study must comply with any legal requirements, and necessary approvals must be obtained.

It is usually beneficial to perform a small rehearsal or pilot study to see if there are challenges that were not considered during the planning process, such as technical and logistic problems, difficulties in understanding items from a questionnaire, misunderstandings among project participants or recruited subjects, and institutional difficulties. Although time-consuming, this approach will certainly benefit project development, because methods are rehearsed and reviewed, and staff collaboration is strengthened.

There should be regular follow-up meetings during the project, e.g., once or twice each month or trimester depending on project needs, to ensure that the progress is in accordance with the study protocol, timelines, milestones, and budgets, and that data storage and control are secured. Annual or semiannual progress reports should be distributed to managers, stakeholders, and steering-committee (if appointed).

Funding, budgets, and compensations

Access to sufficient, sustainable, and long-term funding is a key factor. Often, funding from external sources is needed as local institutions (including universities and hospitals) and government agencies may not have sufficient resources allocated to research. Minor to medium-scale projects may receive support from NGOs, whereas large projects may need funding from larger agencies, which can be bureaucratic and restrictive.

Budgets should be realistic but also open to unexpected expenses. LMIC institutions may lack basic equipment and materials and may therefore need support for buying computers, printers, copy machines, color cartridges, and other office supplies in addition to analytical equipment, reagents, and calibration standards. Other indirect costs that are associated with the project should also be included in the budget. (e.g., costs for travel, internet access, video conferences, instrument maintenance, freight deliveries and postage, building operations, administrative offices, purchasing and library services).

Compensation or incentives for participation in a research project is complex, as is compensation to local coordinators and collaborators (Murphy et al. 2006). Their basic salary in some LMICs may be barely sufficient to support their family, and collaborating on international projects may provide hope for financial or material benefits. Costello and Zumla (2000) note that the salaries of academics or doctors employed by an international organization or research station are 5 to 20 times greater than the salaries in government or university positions, making it impossible for local institutions to compete for the best and brightest research talents. Salaries should thus be reasonable in the local context, and not contribute to inflation in salaries provided by international employers. Clear a priori discussion of project and budget limitations, as to what is or is not allowed to purchase as equipment, or considered a proper expenditure, will facilitate the process of funding oversight and the subsequent reporting of expenses to the funders. Also, when planning project funding, it will be important to include the costs of publication in the budget.

Fund management should be transparent. Fear of corruption or fund misuse may cause additional time-consuming discussions; a rigid control procedure may be needed. Eventually, a third-party committee (such as another institution or university) may help defray problems related to budget review and approval.

Recruitment of the study sample and data collection

The plan for recruitment of participants should be designed to ensure that the study samples is a representative selection regarding age, sex, time, geography, and education level, etc. to avoid selection bias. If this is not achieved, adjustment procedures weighting the study results may be used. Successful recruitment depends on an understanding of the community by the researchers; this may include cultural norms and attitudes related to hierarchy, power, gender, race, and religion. Also, trust from potential participants is needed (Murphy et al. 2006). The general population may not always understand the causal factors for RTCs and the effect of risk-taking behavior on road safety, such as speeding and alcohol and drug use (Assailly and Cestac 2018; Heydari et al. 2019). Thoughtful discussions with local project team members can increase all researchers’ understanding of these important issues and help them improve their research plans. Also, efforts may be needed to inform road traffic administration leaders, government officials, and researchers about the importance of the study, even at the planning level, to “pave the way” for different moments of project implementation that will need public support (e.g., collecting data on public areas or highways).

When approaching the study participants, written information sheets in a simple language can be quite useful to help clarify the goals. Any potential risk should be disclosed as clearly as possible, and adequate time should be allocated for researchers to properly inform potential participants, including the fact that there are no negative consequences for refusals, which in some cultures may be difficult to immediately comprehend.

Another issue to consider when planning how to approach participants is information bias (i.e., flaws in measuring exposure, covariates, outcome variables, or assessments of those), which might occur in different situations: (a) when collecting self-reported information from study participants; (b) when analyzing alcohol and drugs in biological samples (blood, serum, plasma, serum, urine, and saliva), and (c) when assessing the findings, causing incorrect recording or classification (of e.g., impairment, crash involvement, and injuries). If biological samples are not collected and stored correctly, degradation may occur, giving falsely low concentrations in collected samples. If only immunological screening methods are used, the specificity of test results may be poor.

When collecting self-reported data, such as alcohol and drug use, risk-taking behaviors, crash involvement, and attitudes and knowledge about road traffic safety, questionnaires should be available in the local language and be back-translated for quality check, with a specific focus on cultural aspects and local jargon (The World Bank 2021). Socially desirable responding is often a problem, as participants may be unwilling to report accurately, and instead reply in a manner that will be viewed favorably by interviewer, bystanders, or others (Paulhus and Reid 1991). This problem is more common for illegal actions, such as the use of illicit drugs, DUI, and actions causing the feel of guilt or embarrassment, such as causing crash involvement. Social desirability bias may be reduced in surveys based on questionnaires if no information is recorded that can be used to identify a participant, if participants are not observed when filling in the questionnaire, and the questionnaire is delivered in a box without anyone nearby. However, a research assistant may need to explain some of the questions to avoid misunderstandings before the participant fills in the questionnaire.

A lack of knowledge about alcohol-containing beverages and psychoactive substances may be a challenge when collecting self-reported data on alcohol and drug use. Some communities may define only distilled beverages as alcohol while beer is not, and others might not have knowledge about the alcohol content of self-made brews. In order to help solve this issue, a good proposal is to prepare a document with photos of alcohol-containing beverages and present the alcohol content, as done in some projects in LMICs (Reisdorfer et al. 2012; Sundet et al. 2020). It is then easier to calculate the number of standard drinks or grams of alcohol consumed. When asking about the use of drugs, commonly used local slang or street names may be used instead of or in addition to the formal substance names. When asking about the use of botanical drugs that are important and commonly used in some social or cultural settings, such as khat, betel nuts, or kava, the local population should be consulted to describe the questions correctly. In an increasing number of countries, THC-containing beverages, and edibles are available, but study participants may not always regard these products as cannabis. Therefore, questionnaires on alcohol and drug use should be prepared together with local researchers taking the local culture and customs, opinions, and beliefs into consideration. A list of local names and slang for different drugs may be useful in such instances.

For analytical purposes, it may be difficult to find accredited laboratories in LMICs that can analyze alcohol and drug concentrations in biological samples. Instruments for analyzing alcohol in exhaled breath (alcohol breathalyzers) can be easily obtained as there are many manufacturers of accurate instruments, and these can be purchased by collaborating partners in HICs and brought to the study site. Obtaining equipment for drug testing might be more difficult. Disposable devices for drug screening of urine or saliva (oral fluid) samples can be purchased if import licenses are granted. Equipment and reagents for confirmatory analysis of individual drugs may not be available and may be difficult to import if the manufacturer does not have an agreement with a local vendor. Also, some countries do not allow the import of calibration standards for illicit drugs making the determination of drug concentrations impossible. The purchase of expensive instrumentation should also be avoided if there is no capacity in the country to service the equipment (World Health Organization 2014).

Some projects might decide to ship biological samples to laboratories abroad for testing. However, some countries do not allow the export of biological samples. If export is allowed, it may be difficult to find a courier that can bring the samples to the laboratory. Some samples need shipping while frozen on dry ice, which is an additional challenge. Due to customs controls and approvals, samples may be stranded at airports for a long time at ambient temperature, causing sample degradation. A simpler method for sample shipment is to prepare dried sample spots of blood, plasma, serum, urine, or saliva, which can be transported and stored at ambient temperature, and are easy to send by mail or courier services (Sadones et al. 2014). Dried sample spots can be tested for several drugs, for biomarkers for heavy drinking, but not for alcohol itself. Several kinds of devices are available for the preparation of dried sample spots. This includes filter papers (Sadones et al. 2014) and devices designed to collect accurate amounts of blood to enable better quantitation of drugs (Delahaye et al. 2021).

When studying clinical impairment by alcohol or drugs, systematic test batteries may be used, such as field sobriety tests, or clinical tests of impairment (Bramness et al. 2003; Porath-Waller and Beirness 2014). The use of such tests requires supervised training of the test personnel. In some cases, if police are to collect data for research purposes (as has been done systematically in Brazil and some other countries), researchers should consider whether this type of training should be provided directly by the research team, or whether the “train-the-trainers” approach should be used (when researchers will educate senior members of the force to deliver the training to their peers).

In HICs, registries and databases are often used in research. However, it is well-known that the number of road traffic crashes is not always recorded correctly in police or crash databases in many countries. If existing in LMICs, the databases may be incomplete. As a result, official statistics are not always correct (Stewart et al. 2012; Heydari et al. 2019), and should not always be relied upon. Instead, alternative methods for data collection should be explored.

It is often impossible to obtain unbiased data when studying alcohol and drug use among fatally injured road users. In most LMICs, blood samples are not routinely taken from road users killed in road traffic crashes for alcohol and drug testing. If the police reports alcohol use as a contributing factor for the crash, it might be based on information from passengers or bystanders, which may be incomplete or inaccurate. In cases of road traffic injuries, alcohol breathalyzers may not be available, so data on alcohol use is instead based on smelling alcohol from the road user’s breath, which has its limitations. Therefore, other approaches should be used to study the involvement of alcohol and drugs as contributing factors in serious road traffic crashes. Less biased data may be obtained if systematically studying seriously injured road users admitted to an emergency hospital for treatment and analyzing blood samples taken shortly after arrival for alcohol and drugs.

Original research data should be collected and securely stored in compliance with the ethical approval and local data protection laws. Only authorized persons should have access to the data, therefore a locked drawer or filing cabinet may be used for primary storage. Back-up of electronic data should be stored at a separate location or in a secure “cloud storage” system on the internet.

Original data should also be shared with involved partners. Then, the handling and storage of the data must also comply with the regulations in the partner’s country, such as additional ethical approvals and data protection laws.

Data protection laws do not apply for anonymous, aggregated personal data. There is no universally accepted definition of anonymous data. The guideline from the University of Edinburgh (2020) says that if at least 3–5 individuals share the same combination of personal data, the data can be regarded as anonymous. A guidance document from the Norwegian Ministry of Health (2010) states that data for at least 4–5 individuals are required, whereas Alexander and Jabine (1978) propose that combinations of characteristics for at least five individuals are required.

Ownership, evaluation, and dissemination of research findings

This is a critical topic that should be addressed with care, preferably when planning the research project. Any research agreement should include ownership of collected research data, including biological samples and data generated from sample analysis. Shared ownership is recommended, and all involved research partners should have access to the data and be allowed to analyze the data for the preparation of reports and articles (Matenga et al. 2019; Faure et al. 2021). In some cases, project staff will benefit from technical seminars to prepare for data analysis and dissemination while data are collected and organized.

A plan for data sharing and dissemination of research findings should be agreed upon during project planning, including outlines of the content of articles and reports and principles for coauthorship (De Grijs 2015; Gewin 2018). The International Committee of Medical Journal Editors, commonly referred to as the Vancouver Convention, has published recommendations required for authorship (International Committee of Medical Journal Editors 2023). Each author should contribute to study design or data acquisition, analysis or interpretation, participate in writing and approving the manuscript, and be accountable for the accuracy and integrity of the work. It is sometimes difficult to comply with all these requirements. Therefore, coauthorship should be decided at an early stage, and coauthors should be involved in the processes that are required for authorship.

Reports should be prepared in line with international recommendations on the presentation of research findings, such as the STROBE and SAMPL guidelines (von Elm et al. 2007; Lang and Altman 2013). Writing a comprehensive report that is archived in a library or drawer is of little value, and only presenting results in scientific journals is often not of great importance for improving the local situation (World Health Organization 2005, 2014).

Local researchers may not have the same access to scientific papers as institutions in HICs. Therefore, it is important to support them with relevant literature. Writing academic English is another challenge, so assistance with editing and proofreading their work may be needed. Also, getting a manuscript accepted for publication in a well-recognized scientific journal may be difficult. Publishing in questionable or predatory journals must be avoided, even though this is quite common in LMICs, as publication costs are low. Many of these journals accept manuscripts for publication without proper peer review, and publishing in such journals might have a negative effect on the researchers’ careers.

A delicate topic for the research team is cognitive bias, which may affect the discussion of study findings and suggestions to improve road safety. This may occur when findings are interpreted in a way that matches preconceived or subjective opinions, by paying attention only to other studies that confirm the author’s view or the current research findings and ignoring information that negates them, or when discussing the implementation of actions and procedures commonly used in HICs but that may be less suitable in LMICs. Therefore, open discussions are important, in addition to a thorough review of results, conclusions, and recommendations in similar studies in neighboring countries.

In parallel to the presentation of data in the scientific environment, it is important to disseminate research findings to the local public and authorities. Consider presentations in local seminars and local mass media and social media and publish important findings and recommendations in the local language in a way that is widely distributed. Again, attention should be given to the delicate points of cultural standards. Information using appropriate language should be the goal and therefore needs to be discussed with the local investigators. Involving stakeholders in the dissemination is useful; thereby, a larger audience may be reached. Also, even when the project leader is from a HIC, it might be wise to consider that frontline presentations should be given by local co-investigators.

A lack of confidence in science and expertise may be a challenge when informing the public and politicians about research findings (World Health Organization 2005). In some cases, research findings may indicate that the situation is not in line with the expectations of the authorities, particularly when the local government may be the funder or co-funder of the project. In the worst case, they may not believe that the project findings are correct and therefore do not implement actions to rectify or improve the situation, or simply embargo the dissemination of data, preventing researchers from presenting or publishing results that may be regarded as undesirable. Both censorship and self-censorship may occur and must be avoided. We suggest that this issue is clarified with project team members and the research organizations included in the project during dissemination planning.

Any suggested improvements and interventions should build on the existing capacity to be effective (World Health Organization 2014). This includes traffic safety legislation and enforcement, alcohol, and drug testing facilities, crash examinations, crash databases, and acute injury treatment practices. Previous experience showed that decision-makers in LMICs in many cases do not make effective use of research findings to implement policy changes (World Health Organization 2005; Franzen et al. 2017). Therefore, local leadership and central stakeholders should be actively involved in translating research findings into improvements in policy and practice (Stöckli et al. 2018; Shuey et al. 2020). If a project steering committee has been appointed, it may contribute to the translation of findings to action.

In order to ensure sustainability of the research, local research capacity should be further developed. Maintaining international collaboration and partnerships may therefore be important.

Risk and quality assessment

We have listed some questions for risk assessment and evaluation of study quality below. Those are based on our own experience combined with questions adapted from previously published papers (Costello and Zumla 2000; Stöckli et al. 2018; UK Research Integrity Office 2021).

Authorities

• Is there instability, conflict or authoritarian tendencies that may affect the likelihood of success?

• Can the research be done without political interference?

• Will authorities trust the researchers and research process?

• Is the local institution willing to be assertive with police and authorities to obtain permissions and collaborations?

• Are the research partners (institutions or researchers) willing to confront authorities with information and facts that they do not want to hear or believe?

• Are there any potential political or cultural issues limiting academic freedom?

Collaborating institutions

• Is there a written agreement regarding the collaboration?

• Is the project properly anchored in the involved organizations?

• Are there agreements regarding the roles, responsibilities, and authority of management and researchers?

• Is the institutional research interest, knowledge or capacity sufficiently strong?

• Are key personnel available and dedicated to the project?

• Are local supervisors or mentors available at all times?

• Are local stakeholders involved in project planning and execution?

• Is financial support sufficient to handle delays and unexpected costs?

• Is the partner willing to, and able to, spend the money as planned?

• Is the handling of money secure or are actions required to avoid fraud and corruption?

• Is the partner willing to allocate the needed personnel requirements?

• Are there any unrealistic expectations of financial support?

• Are there any conflicts of interest or conflicts between collaborating institutions?

• Will the project team be able to communicate effectively (telephone, video, internet)?

• Are the partners familiar with each other and do they trust each other?

Research planning

• Was the study protocol written with enough detail to ensure that the study can be replicated?

• Are the aims, hypotheses, and beneficiaries clearly defined?

• Is the research design appropriate for the questions being asked?

• Do researchers and participants face any risks to their health, safety, and well-being?

• Will the research comply with all legal and ethical requirements and guidelines?

• Was the application for ethical approval submitted or obtained?

• Were users or researchers with a close understanding of users consulted?

• Do all partners have equal access to scientific, financial and research data?

Research methods

• Is there access to all necessary skills?

• Is there a plan for training researchers and assistants?

• Are office equipment and supplies available, such as laptops, tablets, photocopiers, printers, copy/printer paper, and color cartridges?

• Are technical equipment, reagents, chemicals, and standards available?

• Will test methods and questionnaires be properly validated?

• Is it necessary to import equipment or ship materials abroad for chemical testing? If so, it should be confirmed that shipping is possible, and agreements must be in place.

• Are basic quality assurance principles practiced regarding data collection, handling, secure storage, secondhand control, and audits?

Research sample/participants

• Will participants trust the researchers and research process?

• Are there any risks for information bias or selection bias?

• Are there rumors and misconceptions that may affect the participation rate?

• Could people believe that the researchers are collaborating with the government or foreign powers against the local population?

• May participants be reluctant to reveal sensitive information accurately, such as previous crash involvement and the use of alcohol or drugs?

• Are there any challenges related to informed consent or protection of privacy?

• Is there a risk of information from the data gathering being handed to traffic police, for subsequent criminal prosecution?

• Are language barriers an issue?

Dissemination of research findings

• Are there agreements regarding intellectual property, publication, and authorship?

• Is there a dissemination plan?

• Will results be communicated to users and local stakeholders?

• Is there a plan for balanced authorship?

Discussion

Collaborative projects in LMICs may face challenges ranging from minor issues that are easy to solve to serious situations that cause premature project closure. Common challenges are often related to financial issues, limited resources, lack of experience in similar studies, and insufficient involvement when deciding aims, milestones, and timelines in projects where HIC organizations are involved. Key requirements include addressing local needs, ensuring sufficient resources, local project ownership, capable leadership, strong anchoring in all involved partners, shared decision-making, and plans and willingness to translate research findings into policy, practice, and published recommendations. It is also important to avoid HIC bias, which prioritizes the interests or perspectives of HICs over those of LMICs. Risk and quality assessments should be performed based on factual, direct and continuous cooperative conversation between researchers from all partners.

Limitations

This article does not cover all aspects and challenges that may occur in collaborative projects. It is mainly based on observations and mistakes made by researchers from middle- and high income countries when participating in epidemiological studies in LMIC. Challenges in experimental studies are not discussed.

Disclosure statement

The authors report there are no competing interests to declare. The views expressed are those of the authors and do not necessarily represent the views of their organizations.

Additional information

Funding

The authors did not receive any grants or outside funding in support of the preparation of this manuscript.