Understanding Barriers in Rural Communities; A Healthcare Case
If we consider the foundation to how communities are built and connected, we must explore the limitations that exist. These limitations can be anything from geographic characteristics such as mountains that split two groups to language barriers that may prevent communication between two individuals. The magnitude of each barrier can be measured by the impact that it creates, such as opportunities that are lost, or more importantly, how the majority within a population has greater access to some resources, while others in the minority do not.
This essentially creates a disconnect within groups of the larger population. The effects are then passed to the level of the individual where they may feel unable to contribute to their community and equally unable to receive support from their community. One of the sectors where this is most felt is in healthcare, although it spans to all aspects of rural and remote life. And these barriers will not only affect individual lives, which may lead to undiagnosed conditions to delayed treatments, but may also add onto a snowballing effect on hospitals having to deal with a larger influx of patients seeking immediate treatment.
To mitigate negative outcomes from happening it is important to identify barriers before they start contributing to the effects, such as those identified above. If for instance patients had access to timely treatment and their conditions were not to progress, then they may not seek emergency hospital treatment. Hospitals then would not be overwhelmed by these cases that could have been treatable in non-emergency clinics. In the same regard, if there was good communication available to patients, then they may be encouraged to attend GP clinics in the first instance. However, the management of these barriers and how we identify them still pose risks to how well we can address them.
Identifying barriers before they exist
One possible solution is to identify barriers before they even become barriers. In the example below, we have detailed a hypothetical scenario in how we can work with our community to address an issue once it has existed. However, the important aspect is that identifying an issue and identifying why it became an issue are two separate questions, and we address them both.
Let’s suppose that a community in a rural part of England has had high hospitalisations with patients with severe allergic reactions, poisonings, and pulmonary diseases. The hospital in that region does not typically handle that many patients in that short amount of time and does not have specialised resources to accommodate the illnesses that are happening. Then, there may be delays to accessing medical equipment, personnel, and medicines. A few months after the disease outbreak, the rural town is back to a perceived normal.
Step 1: Talking with the Community
Seanasol Research deploys an Investigation Team to meet with the community leaders of that town and openly discuss with them the events in the past months. And we ask them very specific questions. Some of these might be:
Was there anything odd that started happening before your community suddenly became ill?
What were members of your community saying in those months; were there any concerns reported?
Was there any changes in habits, such as were people buying more of one item or less of another?
Then, we open up the conversation with those who were directly affected by diseases and ask them similar questions. We ideally would like to get an overview of the event from the community. The reason is that there may have been events that were leading to the onset of the diseases.
Step 2: Understanding the Problem
The Investigation Team returns to the Central Operations Team with the answers of the community. They reported:
The weather was warmer in those months compared to the previous year
The was little rainfall or stormy weather
In the days leading to the outbreak of the diseases, there was a heavy presence of cars and lorries in the main road, which typically doesn’t get a lot of traffic.
People were generally spending more time outside because of the nicer weather, but they would be spending their times in parks away from the congested areas.
The Central Team may then follow with community leaders to get more details or work with the data provided. Then we ask the essential question: which one of these is the hidden indicator?
A “hidden indicator” is what we consider to be an a series of strange happenings (seemingly random coincidences) that play to the direct effects of a barrier that is created, which in turn leads to a serious event or incident happening.
In this case, the barrier (the unknown factors) caused several people to experience potentially life-changing conditions. The hidden indicator here, which could perhaps only be viewed in retrospect, is the increase in congestion in the main road of the town. Let’s not disregard other indicators though, like the increase in temperate, or the increase in footfall and movement in the town in those months. This information is vital because we can look deeper into the computational aspect of event prediction.
Step 3: The Research-Focused Approach
Once we have a hypothesis (or several), the Research Team then approaches the problem using two methods. The complementarity of these approaches is essential, as it allows us to continue to gather evidence and start working on a framework that will help deploy a solution.
The Two Approaches
1) Computational Approach
The computational approach, as the name suggests, is using computer science to build predictive models using data that is collected through time. This is what is referred to as historical data. Historical data can be a couple of weeks to several years, with the caveat that the fewer days that are recorded, the less robust that models can be. This equally applies to data that is collected by other means, with emphasis that access to that information is vital to make predictions.
These predictions are built using indicators that have been established in situations when they have caused an event. Similarly, some indicators that may seem important may not contribute to any outcomes. The combination of these two, plus additional parameters that are added by understanding the dynamics of an event (details from other data sources), are used to build a model, which in simplicity it is a rule.
Now this rule applies only with regards to the data that was collected, and is not a rule to rule them all. It is an algorithm, which is a mathematical description to our understanding of the information that was available and passed through a series of conditions (parameters) that we use from previously published datasets, for instance. In this example, models in the predictive tools would indicate the likelihood of an indicator resulting in a barrier. Then, the barrier needs to be considered, and assessed whether it was significant enough to drive the effects seen in the event and used as a probabilistic measure of whether an event may or may not occur. Testing this model is then essential, and can be done using historical data of a similar occurrence or using an on-going example. And it takes time to build robust models before they are deployed and used to mitigate events.
2) Community Approach
In parallel to the methods above, the community-led approach is working at the ground-level, that is studying the field as a member of the community. It is performing interviews and getting clues as to what led to the barrier being created.
This part is as essential as the computing aspect, because this helps to deliver the strange happenings (hidden indicators) that may not be part of public datasets and are rather information that is given directly by the population. So for instance, it could be information about individual consumer habits, which then can be approached using computer science and ask questions of what the entire community habits were in that time. This can also be cross-checked with local businesses that are supplying those regions.
Step 4: Connecting with the Community
If we consider the information collected from earlier interviews, the increased footfall and automobile congestion has brought two likely hidden indicators. We deploy a dedicated Community Team to follow up with the community and do regular visits and follow up interviews.
Visiting the community over a couple of months, the Community Team reports that some days cars still cause congestion. The Community Team works with officers, local business services, transportation operators, and the general community and over weeks they identify that a few miles down the road there is a traffic diversion sign, which leads cars towards the town centre, but there are no road repairs in the area. Upon contacting the Council, they confirm a traffic repair was carried out in the months leading to the disease outbreak. However, due to an oversight, the sign was never removed, which caused more traffic to be diverted than the Council had expected.
The Community Team works with the Council and community leaders to further understand why this has happened. They declare that discussion around the use of the sign was done and agreed by all members, however, for unknown reasons the sign was not taken down when it should have been. Information collected from transportation operators and local businesses suggest that those who knew the area were more likely to recognise when a traffic diversion is old, but people and services that do not typically serve that area, were not able to recognise this. This matched with the views of the community.
Step 5: Reaching a Verdict
The Community Team reports their findings to the Central Operations and Research Team. They work together to either perform further studies on the effect of increased pollution on general health or cross-check with previous related studies. Then the Research Team collects consumer data. They report that there was an increase in the purchase of antihistamines and cough medicines, which depleted stocks in the months leading to the disease outbreak. The Community Team then works with individuals and identifies that those who did not have access to over-the-counter treatment experienced allergies and respiratory problems.
Those with asthma or other underlying chronic conditions were also more likely to receive urgent treatment and in some cases became seriously ill. However, there were also a few individuals who were desperate to self-treat that they tried alternative medication, which were not be suitable for the symptoms that they are having. In this case, it is the purchase of cough medicines. Ingestion of these in the large amounts that they were taking led to having adverse effects, requiring urgent treatment. So, while these individuals may be trying to cure themselves because of a lack of resources available, they were placing themselves at risk.
Step 6: Informing the Community
The Community Team would collate all these findings into information that is suitable for that community. They would return to the community and deliver the findings. These might help to lead campaigns or create discussion groups where everyone in the community can talk freely or send their opinions to. It may not change what happened, but it may prevent it from happening again. It may also create protocols for members of the transportation network or local businesses to report odd events or road traffic maintenance signs that may not be indicating any traffic issues.
The value that is returned is an example on how to grow and learn together as the community. It also opens up the conversation to future tools that may be needed to address on-going hidden indicators. Instead of having us there to help them, it is about educating the community on the importance of odd events and how these may play out to more serious events.
The Community Team continues to monitor that community for a set period to check the protocols are working and with the larger team may start piloting innovative strategies. Models created will be continually built, trialled, and improved. Impact is then measured by the Research Team in the number of satisfied members of the community, for instance, or dive deeper in reporting a direct effect on hospitalisations when indicators were identified early.