Analysis Report Two - Systems Integration and Decision Support

Executive Summary

Healthcare organizations depend on information systems to collect data from daily operations and turn it into useful information for managers and executives. Transaction Processing Systems record routine activities such as patient registration, billing, appointments, supply orders, and payments. Management Information Systems then organize this information into reports that managers can use to monitor performance and make decisions. Enterprise Resource Planning systems go a step further by connecting areas such as finance, human resources, purchasing, inventory, and patient services into one system. When these systems work together, healthcare leaders are able to see a more complete picture of how the organization is performing.

However, having more data available does not always mean that better decisions will be made. Too much information can become overwhelming and make it harder for managers to determine what is actually important. Newer technology, including artificial intelligence, will also not be successful if an organization has inaccurate data, outdated systems, or employees who are not properly trained to use it. Healthcare organizations should first focus on improving the quality of their data, connecting their systems, and making sure employees understand how the information should be used.

This report shows how SQL joins can connect information from different database tables and how visualizations can make the information easier for managers to understand. Healthcare leaders should focus on using reliable and relevant data instead of simply collecting more information. By placing useful information in the hands of the people who need it, healthcare organizations can make better decisions, improve efficiency, reduce risks, and provide better patient care.

Introduction

This week’s readings focused on how information systems connect data from different areas of an organization and turn it into information that employees, managers, and executives can actually use. Transaction Processing Systems, Management Information Systems, and Enterprise Resource Planning systems each serve a different purpose, but they work together. Transaction Processing Systems collect the routine data created through daily activities, while Management Information Systems organize that data into reports that help managers make decisions. Enterprise Resource Planning systems provide an even broader view by connecting information from departments such as finance, human resources, purchasing, inventory, and operations.

One of the main ideas from the readings is that simply having access to more data does not always lead to better decisions. Organizations now collect large amounts of information, but too much data can make it harder to identify what is truly important. In some cases, a smaller amount of relevant information can produce results that are more accurate, easier to understand, and less likely to include unfair or misleading factors. This shows why managers must carefully decide which information should be included when creating reports or making predictions.

The readings also explained why most organizations should not rush into developing a separate artificial intelligence strategy. AI depends on accurate and organized data, updated technology, trained employees, and clearly defined business needs. If a company has poor data quality or systems that do not work well together, an AI strategy will not correct those problems. Instead of starting with the question of how AI can be used, organizations should first identify the problem they are trying to solve and then decide which technology is the best fit.

These ideas are especially important in healthcare because information is collected from many different areas, including patient registration, billing, scheduling, staffing, purchasing, and supply management. Healthcare leaders need systems that can connect this information and present it in a clear and useful way. The overall lesson from the readings is that technology and data only create value when they are properly integrated, carefully selected, and placed in the hands of people who can use them to make better decisions.

Outside Research

Healthcare organizations often use electronic health records and health information exchanges to connect information across departments, providers, and locations. Research shows that this integration can support organizational goals, but the results depend heavily on data quality and how the technology is used.

Research has found that interoperable electronic health records and health information exchanges can create positive results for healthcare organizations, although outcomes are sometimes mixed depending on the setting and how the systems are used (dobrow2019interoperable?).

A separate review found that interoperability can improve medication safety, reduce certain patient-safety events, and lower costs. However, improvements in workflow and time savings were less consistent (li2022impact?).

Integrated systems can also create problems when the data they produce is inaccurate. One study found that errors were introduced into several emergency-department timestamps after an electronic health record system was implemented. This shows why managers should carefully review data quality before using system-generated information to make operational decisions (ward2013operational?).

Information sharing may also support some strategic business goals without improving every measure of organizational performance. One study found that increased health-information-exchange capability was associated with changes in payer mix, but not with improvements in market share or operating margin (vest2022strategic?).

Overall, these studies show that integration can improve access to information, safety, and certain operational outcomes. At the same time, technology alone does not guarantee better performance. Healthcare organizations must also focus on accurate data, system compatibility, employee training, and how the information will be used to support decisions.

Data Visualizations

The following visualizations demonstrate how information from separate database tables can be connected to provide a more complete view of organizational activity. The first visualization uses a two-table join and a boxplot to compare quantitative data. The second uses a three-table join and a stacked bar chart to compare two qualitative variables.

Visualization One - Two Table Join

This query connects the Products and Categories tables using the shared CategoryID field. It provides the category name from the Categories table and the unit price from the Products table.

SELECT Categories.CategoryName,
       Products.UnitPrice
FROM Products
INNER JOIN Categories
ON Products.CategoryID = Categories.CategoryID

ggplot(data = myquery1,
       aes(x = UnitPrice, y = CategoryName)) +
  geom_boxplot() +
  labs(
    title = "Distribution of Product Prices by Category",
    x = "Unit Price",
    y = "Product Category"
  )

Visualization Two - Three Table Join

SELECT Shippers.CompanyName,
       Customers.City,
       Customers.Country
FROM Orders
INNER JOIN Shippers
ON Orders.ShipVia = Shippers.ShipperID
INNER JOIN Customers
ON Orders.CustomerID = Customers.CustomerID
WHERE Customers.Country LIKE "%USA%"

ggplot(data = myquery2,
       aes(y = CompanyName, fill = City)) +
  geom_bar() +
  labs(
    title = "Orders Handled by Shipping Company and Customer City",
    x = "Number of Orders",
    y = "Shipping Company"
  )

Recommendations for Industry

Organizations that want to improve integration should first make sure their existing data is accurate, organized, and consistent across departments. Connecting systems will not create better results if the information being shared is incomplete or incorrect. In healthcare, this means departments such as billing, purchasing, human resources, scheduling, and patient services should use consistent data standards so that information can be combined and understood correctly.

Organizations should also focus on the business problem before choosing a new technology. AI or other advanced systems should not be used simply because they are popular. Leaders should first decide what they are trying to improve, such as reducing supply costs, improving delivery times, preventing shortages, or coordinating patient services. Once the goal is clear, the organization can decide which technology and information are actually needed.

The findings from the visualizations also show the importance of connecting data from multiple areas. By joining product and category data, managers can better understand differences in pricing. By connecting orders, customers, and shipping companies, managers gain a clearer view of delivery activity and customer locations. Healthcare organizations could use similar connections to compare medical supply costs, monitor delivery companies, and identify which hospitals or clinics are placing the most orders.

Another recommendation is to keep reports and visualizations focused and easy to understand. More data does not always lead to better decisions, especially when managers are given too much information at one time. Organizations should provide decision makers with the most relevant information instead of including every available data point. Simple reports and visualizations can often make patterns and unusual results easier to identify.

Finally, organizations should provide employee training and clear guidelines for using integrated systems. Technology can only support strategic goals when employees understand how to enter data correctly, interpret reports, and use the information in their daily work. By focusing on data quality, clear business goals, system compatibility, and employee readiness, organizations can improve integration while reducing the risks of information overload, poor decisions, and unsuccessful technology investments.

References

(article?){dobrow2019interoperable, title={Interoperable electronic health records and health information exchanges: systematic review}, author={Dobrow, Mark J and Bytautas, Jessica P and Tharmalingam, Sukirtha and Hagens, Simon}, journal={JMIR medical informatics}, volume={7}, number={2}, pages={e12607}, year={2019}, publisher={JMIR Publications Inc., Toronto, Canada} }

(article?){li2022impact, title={The impact of electronic health record interoperability on safety and quality of care in high-income countries: systematic review}, author={Li, Edmond and Clarke, Jonathan and Ashrafian, Hutan and Darzi, Ara and Neves, Ana Luisa}, journal={Journal of medical Internet research}, volume={24}, number={9}, pages={e38144}, year={2022}, publisher={JMIR Publications Toronto, Canada} }

(article?){ward2013operational, title={Operational data integrity during electronic health record implementation in the ED}, author={Ward, Michael J and Froehle, Craig M and Hart, Kimberly W and Lindsell, Christopher J}, journal={The American Journal of Emergency Medicine}, volume={31}, number={7}, pages={1029–1033}, year={2013}, publisher={Elsevier} }

(article?){vest2022strategic, title={Strategic use of health information exchange and market share, payer mix, and operating margins}, author={Vest, Joshua R and Freedman, Seth and Unruh, Mark Aaron and Bako, Abdulaziz T and Simon, Kosali}, journal={Health care management review}, volume={47}, number={1}, pages={28–36}, year={2022}, publisher={LWW} }