When you walk into a medical clinic in Mozambique or Laos, there could be any level of data management behind-the-scenes from rudimentary paper patient files to highly integrated information systems that connect patient data to supply chain data and beyond.
For those who work in global health, it is important to understand what can be gained from integrating datasets to support global health goals. How does data integration help your organization better serve the needs of patients and the global community? The benefits can be impressive and inexpensive.
What Types of Datasets Can Be Integrated with Health Care Supply Chain Management Systems?
Dataset #1—Patient Health Data
Gathering clean, digitized patient data is the start. Every time a patient receives a treatment, the supply chain management system has immediate “point-of-sale” data which shows supplies that need to be replenished, aiding in better inventory management.
Integrating these systems can act as an early warning system. If there is a sudden uptick in the usage of a certain supply, is there an outbreak? An extreme case would be a situation like Ebola. If there is a very rapid trend of patients seeking stomach flu treatments, there could be a flag in the systems that suggests ordering medication for the flu or other flu-like diseases like Ebola.
It is important to analyze long-term usage patterns. For example, if a supply is often not available, it may cause people to use an alternative supply or not receive treatment. It may mean that chronic illnesses such as diabetes are not managed properly, and thus there is an increase of emergencies. This will be reflected in high usage of other supplies or nights in hospital beds.
Dataset #2—Human Resources or Budgeting Data
Supply chain systems can be integrated with human resources data to predict which staff are likely to be needed. What specific skills are needed regularly and is there planning for buffer staffing in more extreme cases? Connecting with budget systems allows for earlier notifications of budget shortfalls or the ability to shift funds based on needs.
Dataset #3—Insurance Claim Systems
Claims data that can give a good idea of what is happening with the patients. They are a hybrid between patient health records and budgeting systems because they also have the financial data. They show where funding flows are going. This could help with better budgeting based on trends.
Dataset #4—Climate Change Data
Climate change will impact the availability of clean water. When there is an unexpected amount of water—too little or too much—there are health impacts. It is possible to predict health trends in a region based on calculations of more rain, a longer rainy season, or a heavier rainy season, such as an increased chance of insect-borne diseases.
Climate change causes major weather events. In areas highly impacted by these events, there will be more hunger- and stress-related health issues. Climate data today is accurate enough to calculate risk in any part of the world and thus aid in supply chain planning.
Best Practices of Integrating Datasets
Data integration occurs at three levels: people, process, and technology.
- Technology—Technology tends to be the easiest part. One way to gain the benefits of data integration without a huge build or constant connectivity is a periodic data pull. This is not necessarily a two-way, real-time data flow. It could be as simple as a monthly data report—whatever that right frequency is.
- People—The most benefit will be gained from a collaborative culture built around the data integration effort. People need to be trained on entering clean data and how to respond when something breaks. Do people collaborate or do they start pointing fingers? When people across all the systems see the value gained, that’s when the greatest leaps will occur in interpreting data for positive results.
- Process—It is important that the system is engineered to reinforce good data management behaviors. This could be clean data entry or creative thinking about how to interpret data. It could be calculating the best frequency for data sharing. How does the full team administratively share and use that data?
It all goes back to strong data stewardship. If patient data is not reliable, the ability to analyze that data will be limited. For example, there might be a diagnostic code of “general” or “none of the above.” Some people might not want to look up the correct code and may always choose that. This creates non-specific data. How could the correct data entry behavior be incentivized? How could the process be redesigned to get more accurate results? It’s important to start from a place where we can say “If you collect better data, I can provide better service in the data analysis.”
Trends in Data Integration
Within most health systems, there are three important trends:
- Increased emphasis on implementing systems to automate processes
- Increased culture of getting and using more data from patients
- Increased commitment to sharing data and analyzing it for better planning.
In particular, in countries with a universal healthcare national insurance program, data standards are becoming more unified. The ability to leverage that data for stronger supply chain analysis is more advanced. This aids in managing the financial flows of healthcare. It also yields rich data on patient behavior, diagnoses, and prescriptions, leading to better health trend analysis and high-level supply chain management.
Bringing more data into your management practices gives a better ability to manage your supply chain. In situations where there is clean patient-level data available, data integration with other datasets can improve supply chain capability within six months.