Analysis Report Two - Data, Data Everywhere

Executive Summary

Hospitals today produce and store a very large amount of data. This includes electronic health records, EHRs, laboratory systems, bedside monitoring systems, wearable devices, and administration databases. All of this information has a large potential to improve patient outcomes, organizational performance, and patient satisfaction. However, Hospitals struggle with integration and effectively interpreting the large amounts of data. [@sutton2020overview] and [@raghupathi2014big] both emphasize the true value of how Big Data often lies in the organization’s ability to use the raw information that is available and shape it into meaningful clinical insights, instead of just the quantity of the information available.This analysis report will look at the role Big Data in healthcare serves, and this will be done by looking at different readings and applying SQL to the MIMIC-III database. This report will show that using data that has been combined from multiple hospital systems can help support the decision-making in operational efficiency, as the clinical staff will be better informed. This report will also touch on how continued physician oversight can help minimize risk, like algorithmic bias and information overload from the data [@deloitte2024ai].

Introduction

The healthcare industry has seen a fast digital transformation, which is due to hospitals switching to relying on electronic systems to record and track patient information. Big Data analytics have helped provide healthcare organizations with solutions to improve patient care, outcomes, reduce costs, and help with decision-making based on evidence provided. This is possible because of the ability to analyze very large and complex databases [@raghupathi2014big]. In the past, healthcare organizations would have relied heavily on isolated records, but the new modern analytics can integrate information from laboratory systems, clinical documentation, patient monitoring devices, and administration databases to produce reports and analytics with very comprehensive insights that the healthcare organization can use.There have also been advances made in artificial intelligence, AI, health applications, and patient wearable devices that are helping change the relationship between the physician and patient. The continued availability of data is helping clinical staff and patients have a more collaborative and engaging relationship, as AI tools and systems are helping reduce the admin workload, as well as assist with diagnostics and treatment planning. Although the new advances are helpful to healthcare organizations, the role of the physiciabs remain importnat as they are the ones that provide the ethical oversight and professional judgments to patients that technology cannot [@deloitte2024ai].CDSS are clinical decision support systems that are able to combine patient information with medical knowledge to improve healthcare delivery. When these systems are implemented correctly and effectively, they reduce medication errors, support safer patient care, and improve adherence to established clinical guidelines [@sutton2020overview]. After doing some additional research, there is support for the idea that AI should complement clinicians rather than replace them, as this technology is intended to strengthen, not diminish, the human expertise of physicians. 

The Healthcare Context

Technology integration has completely changed how healthcare organizations are able to operate because of the ability to connect the previously separate systems into a cohesive digital atmosphere. EHRs, patient monitoring equipment, laboratory databases, wearable devices, imaging platforms, and pharmacy systems all generate a large amount of data that is able to be used to improve patient care and organizational performance. The Deloitte article talks about how technologies can create opportunities for physicians and patients to be able to engage in more of a collaborative decision-making process, which reduces administrative workloads as AI is heavily utilized [@deloitte2024ai].Big data analytics have helped healthcare organizations to be able to identify disease and infection trends, which has helped improve operational efficiency and support clinical decisions by analyzing information from a variety of different sources [@raghupathi2014big]. At the same time, clinical decision support systems integrate patient information with available medical knowledge, which helps improve diagnoses, treatment plans, and medication safety [@sutton2020overview].Doing further resrarch tehre is sources that reinforce these findings. AL-Dmour et al. were able to identify that AI technology has helped significantly improve the diagnostic accuracy, treatment planning, operational efficiency, and patient care, especially when it has been supported by the healthcare organizations’ initiatives, like staff training. The study also found that perceived usefulness and the ease of use influence how successful the technology adoption within the healthcare organization is [@aldmour2025]. In addition to this research, Agarwal et al. explain that Big Data analytics are helping healthcare providers in personalized medicine, disease outbreak monitoring, predictive analytics, electronic health record management, and real-time clinical decision making; this is done by processing and deciphering large and complicated databases from multiple healthcare systems [@agarwal2025].Although there are many advantages of integrated healthcare systems and Big Data, there are still challenges that healthcare organizations are facing. The Deloitte article talks about how wearable devices and the continues monoring tecjnologies produce a massive amount of information and data, which can end up overwhelming clinical staff if there are not proper filters set up to work through data that is not relevant [@deloitte2024ai]. Al-Dmour et al. talk about the infrastructure limitations that exist, the lack of organizational readiness, and the lack of training for staff. All of these are barriers to maximizing the benefits of AI and Big Data analytics in healthcare organizations [@aldmour2025]. Going further, Agarwal et al. also highlight concerns that exist regarding the data quality, privacy, and management of the complex databases that are produced; there is an emphasis on the importance of establishing governance frameworks and analytic tools [@agarwal2025].The evidence shows that technology by itself cannot improve healthcare outcomes; hospitals have to be able to combine the existing systems, data governance, clinical education, and AI implementation to be able to use the amount of data and turn it into meaningful insights that help improve patient care and the healthcare organization’s performance.

Data Visualizations

Warning

Remember, the practice covers certain specific concepts. Your grade is based on how well you show mastery of these concepts.

Visualization One - Two Table Join

This visualization combines the chartevents and d_items tables to show ICU heart rate measurements. A boxplot is used to show the distribution of the heart rate values. These highlight the median, spread, and the outliers. This is relevant and helpful to healthcare organizations because it is able to show how data from different systems can be integrated to monitor patient health and support clinical decision-making.

SELECT
  CAST(valuenum AS INT) AS val,
  label
FROM chartevents
INNER JOIN d_items
ON chartevents.itemid = d_items.itemid
WHERE d_items.itemid = 220045
AND valuenum IS NOT NULL
LIMIT 5000

ggplot(data = myquery1,
       aes(y = val)) +
  geom_boxplot(fill = "steelblue") +
  theme_minimal() +
  labs(
    title = "Distribution of Heart Rate Measurements",
    subtitle = "MIMIC-III ICU Data",
    x = "",
    y = "Heart Rate"
  )

Visualization Two - Three Table Join

This visualization uses chartevents, d_items, and patients tables to look at the heart rate measurements by the gender of the patient. A violin plot shows the distribution and the density of heart rates for each group. This visual is relevant and helpful to healthcare organizations because integrating demographic and clinical data together helps identify different trends across patients and supports more data driven patient care.

SELECT
  CAST(chartevents.valuenum AS INT) AS val,
  patients.gender
FROM chartevents
INNER JOIN d_items
ON chartevents.itemid = d_items.itemid
INNER JOIN patients
ON patients.subject_id = chartevents.subject_id
WHERE d_items.itemid = 220045
AND chartevents.valuenum IS NOT NULL
LIMIT 5000

ggplot(data = myquery2,
       aes(x = gender, y = val)) +
  geom_violin(fill = "lightgreen") +
  theme_minimal() +
  labs(
    title = "Heart Rate Distribution by Gender",
    subtitle = "Three-table SQL Join",
    x = "Gender",
    y = "Heart Rate"
  )

Recommendations for Industry

Healthcare organizations need to invest in information systems that allow data sharing across all departments and help improve clinical decision making. Hospitals should also utilize AI and clinical support systems that assist instead of replacing healthcare professionals. Healthcare organizations also need to prioritize the importance of data quality, privacy, and staff training so that Big Data can be used effectively and responsibly by the organization’s staff. This will allow for better patient satisfaction, patient outcomes, and operational efficiency.

References

@article{raghupathi2014big,
title={Big data analytics in healthcare: Promise and potential},
author={Raghupathi, Wullianallur and Raghupathi, Viju},
journal={Health Information Science and Systems},
volume={2},
number={3},
year={2014},
doi={10.1186/2047-2501-2-3}
}

@article{sutton2020overview,
title={An overview of clinical decision support systems: Benefits, risks, and strategies for success},
author={Sutton, Reed T. and Pincock, David and Baumgart, Daniel C. and Sadowski, Daniel C. and Fedorak, Richard N. and Kroeker, Karen I.},
journal={npj Digital Medicine},
volume={3},
number={17},
year={2020},
doi={10.1038/s41746-020-0221-y}
}

@article{topol2019high,
title={High-performance medicine: The convergence of human and artificial intelligence},
author={Topol, Eric J.},
journal={Nature Medicine},
volume={25},
number={1},
pages={44–56},
year={2019},
doi={10.1038/s41591-018-0300-7}
}

@article{aldmour2025,
title={Impact of AI and big data analytics on healthcare outcomes: An empirical study in Jordanian healthcare institutions},
author={Al-Dmour, Rand and Al-Dmour, Hani and Amin, Eatedal Basheer and Al-Dmour, Ahmed},
journal={Digital Health},
volume={11},
year={2025},
doi={10.1177/20552076241311051}
}

@article{agarwal2025,
title={Beyond boundaries: Charting the frontier of healthcare with big data and AI advancements in pharmacovigilance},
author={Agarwal, Arohi and Singh, Gagan and Jain, Samyak and Mittal, Piyush},
journal={Health Sciences Review},
volume={14},
pages={100214},
year={2025},
doi={10.1016/j.hsr.2025.100214}
}

@misc{deloitte2024ai,
title={Empowered by new technologies, patients and their physicians can become better at shared decision-making by embracing a health care transformation that is at their doorstep},
author={Abrams, Ken and Fera, Bill},
year={2024},
howpublished={Deloitte Insights}
}