Building a Robust Data Strategy for a Hospital

There are many different perspectives on what a data-driven hospital should look like. If you ask a doctor, they typically respond with clinical decision support algorithms; nurses tend to prefer a lower administrative burden; and support staff focus on logistical and financial models. Our team at Amsterdam Data Collective (ADC) helped OLVG, a large hospital in Amsterdam, align its departments to increase organisational efficiency through a robust data strategy.

The Challenge

Our client, OLVG, is a large hospital in Amsterdam that asked for help to become more data-driven. Within the hospital there are countless employees, such as doctors, nurses, support staff, and board members who work with data. Consequently, it was often unclear who did what, who to approach for data requests and support, or whose responsibility it was to manage the quality of data. Various employees were unknowingly working on the same tasks. This made it necessary to improve the way of working with data between departments.

What is our goal with becoming more data–driven? Who is responsible for what and how should we work together? What use cases are important to start working on now? And what is the best data foundation that helps us work better together? The overarching goal was to answer these questions to help the hospital become more data-driven by building a robust data strategy and implementing an organisation system.

The Approach

Firstly, we aggregated a comprehensive understanding of the challenges from the perspective of each department within the hospital. To do so, we interviewed approximately 50 employees from different departments, including those in medical care, finance, procurement, education, and the board of directors. This provided a better sense of how the data in the hospital is used on a day-to-day basis and how the organisation could be improved.

After gathering enough information, we held a series of design sprints. With a small group from OLVG, we tackled the problem in four steps:

define the vision, data strategy, and end goal of working with data in the hospital,
determine how we can best organise the different departments working with the data,
identify use cases for the various data challenges within the hospital, and
deliver a framework for the best data foundation to be used by all the stakeholders

During the third design sprint, numerous valuable use cases were developed. These included methods of facilitating data transfers to other hospitals and using Artificial Intelligence to predict short-term clinical capacities. For example, the probable number of upcoming surgeries, emergency visits, and available hospital beds. Having this information helps OLVG better allocate its employees and resources.

The Solutions

Overall, we helped OLVG understand where they stand in the process of building a strong data strategy, as well as what the hospital can improve on. Improving governance to create a more efficient department structure was an integral part of the solution. For example, by combining all the data engineers and data scientists into one centralised team led by a CDO, the hospital can use its employees as efficiently as possible. This includes preventing redundancies and increasing continuity in their work.

In addition, we assisted the hospital with its decision-making process to best determine which use cases had the highest priority. This was done through the Quadruple Aim framework that measures the impact of a use case through four criteria: patient experience, employee well-being, health outcomes, and cost of care. Combined with the feasibility, it gave OLVG a good overview and prioritised to-do list of data science projects.

Impact for OLVG

Based on our advice, OLVG can efficiently organise its employees who work with their data by increasing standardisation, documentation, and collaboration. This not only helps multiple employees from different departments find each other efficiently, but also prevents redundancies and do-overs. Additionally, the hospital improved the alignment of its many departments and stakeholders, giving them a clear path going forward.

”ADC effectively helped us with our vision and story on how to become more data-driven.
Loren Kruseman (EPD Service Manager, OLVG)

Learnings

OLVG’s challenge to organise its data and related departments is not unique; many hospitals struggle to develop a robust data strategy that amalgamates all internal departments. Hospitals generate vast amounts of data, and this only continues to increase in a digital world. As a result, it is interesting to consider different ways technology can improve the healthcare system and ease the burden on its employees.

How can hospitals best use their employees and data? How can they best prioritise their projects? What use cases are most beneficial to develop? These are all questions to consider when working towards becoming a more data-driven hospital. A clear place to start is improving the flow of information between hospitals. For example, when developing use cases and algorithms that involve patient outcomes, it could be beneficial to share the development costs and workload between several regional hospitals that all have varying resources and data sets.

Overall, there are many different, and oftentimes conflicting, perspectives on what a data-driven hospital should look like. Each department has varying priorities, such as predicting patient outcomes, lowering administrative loads, or improving KPIs. If all stakeholders in the hospital are not aligned before working on a data strategy, it greatly reduces its chance of being successfully adopted by employees. One method of aligning stakeholders is to follow a design sprint process that includes members from each department to determine their shared goals and priorities.