Analysis Report One - What’s Your Data Strategy

Author

Ashlyn Gasperson

Executive Summary

Huge amounts of data are created every day by healthcare organizations; however, many are having difficulty using this data to produce meaningful action. DalleMule and Davenport (2017) describe how organizations must find a balance between competing interests – protecting and using data to create value. Defensive strategies for data are designed around governance, compliance (ensuring patient privacy), and reducing risk; whereas offensive strategies are focused on improving performance, improving the patient experience, and finding ways to improve the health organization. Both types of strategies are necessary for healthcare organizations. The hospitals that comprise our healthcare system operate in a heavily regulated environment requiring the safeguarding of patient safety and confidentiality. Because of the significant regulations and requirements for the healthcare business to operate efficiently while accommodating the volume of patient demand and improving their operational efficiency, there must be a balance between these two competing priorities. This report describes how both offensive and defensive data strategies can be used by healthcare professionals by providing examples from the MIMIC-III database. Based on our findings, we conclude that the most benefit to healthcare organizations occurs when data governance practices are established and when an organization uses its data proactively in order to improve decision-making and operational efficiency.

Introduction

There is more data available to agencies today than at any point in history; however, collecting data alone may not deliver value as far as improved organizational performance or decision-making. The problem is creating a plan to ensure that the data can be relied upon and is accurate, trusted, accessible to decision-makers, and ultimately of value to those who need it. In What’s Your Data Strategy?, DalleMule and Davenport describe two primary approaches to managing organizational data: offensive and defensive strategies (DalleMule and Davenport 2017). The primary focus of a defensive strategy is on data quality, data privacy, data protection, data governance, & data security, with the intent of providing protection against risk and ensuring data integrity & reliability. While an offensive strategy intends to leverage data to maximize organizational productivity, uncover business opportunities, build enhanced customer experiences, and create value through data analysis. The authors argue that organizations should not choose one approach over the other (DalleMule and Davenport 2017). Determining the right strategy combination will depend on industry type, strategic importance of the organization, and willingness to assume risk. Healthcare is especially challenged because patient outcome & reimbursement rates are determined by how accurately the organization collects & maintains patient data. Hence, administrators in the healthcare sphere will need to know how to leverage both offensive & defensive strategies when managing organizational data.

The Healthcare Context

The data strategy discussed in What’s Your Data Strategy? is especially important in the health care environment. Modern health care systems create data via their electronic health record system, laboratory systems, pharmacy systems, billing systems, and operational databases. While health care organizations have a lot of data, most do not have systems that are integrated, nor do they have consistent definitions of their data or ways to convert data into action. Due to the highly sensitive nature of the patient information managed in health care organizations, organizations must have defensive data strategies. Compliance with privacy regulations, maintaining data accuracy, and protecting against cyber threats are all critical responsibilities of health care organizations. According to Kruse et al. (2017), there are continuing challenges related to information security in health care organizations, and organizations must have good safeguards in place to protect electronic health records. (Kruse et al. 2017) The importance of using offensive data strategies for health care leaders cannot be underestimated. Healthcare leaders face competing pressures: improve how patients do things; lower costs; and improve operational efficiency. Through proactive data analysis, leaders can determine what resources they need, identify where inefficiency occurs, improve patient flow, and support decision-making across their organization. As described by Blease et al. (2020), an increasing number of healthcare professionals recognize that data-driven technologies can potentially improve care delivery and help with clinical decision-making. When healthcare organizations utilize both strategies on a regular basis, they will be able to improve health outcomes, remain compliant with regulations, and attain maximum value from the data they collect.(Blease et al. 2020)

Data Visualizations

--I've completed this for you as a placeholder. Replace this with your visualization.
--SQL code in this block

SELECT "admission_type", "subject_id"
FROM "admissions"
WHERE "admission_type" IN (
    "EMERGENCY",
    "ELECTIVE",
    "URGENT",
    "NEWBORN"
);
    
#ggplot visualization in this block. I prettied this up with labels to show what is possible.
ggplot(data = myquery1,
       aes(x = admission_type
)) +
  geom_bar()+
  theme_minimal() +  # Cleans up the background grid lines
  labs(
    title ="Hospital Admissions by Admission Type",
    subtitle ="MIMIC-III Admissions Data",
    x ="Admission Type",
    y ="Number of Admissions",
   
  )

This visualization examines the distribution of admission types within the MIMIC-III database. Understanding admission patterns provides healthcare administrators with valuable information regarding patient demand and organizational resource needs. Admission types such as emergency, elective, urgent, and newborn admissions require different levels of staffing, preparation, and resource allocation. From an offensive data strategy perspective, these findings demonstrate how operational data can be used proactively to improve decision-making. Rather than reacting to fluctuations in patient demand after they occur, administrators can use admission trends to improve capacity planning, optimize staffing models, and enhance operational efficiency.

--I've completed this for you as a placeholder. Replace this with your visualization.
--SQL code in this block

SELECT "admission_type", "discharge_location"
FROM "admissions"

    
#ggplot visualization in this block. I prettied this up with labels to show what is possible.
ggplot(data = myquery2,
       aes(x = discharge_location
)) +
  geom_bar()+
  theme_minimal() +  # Cleans up the background grid lines
  labs(
    title ="Hospital Discharges by Discharge Location",
    subtitle ="MIMIC-III Admissions Data",
    x ="Discharge Locations",
    y ="Number of Discharges",
   
  )

This visualization examines discharge destinations for patients admitted to the hospital. Accurate documentation of discharge locations is critical because it supports continuity of care, quality monitoring, reimbursement processes, and regulatory reporting. Differences in discharge destinations may also reflect variations in patient acuity and post-acute care needs. From a defensive data strategy perspective, discharge information supports governance and risk reduction. Reliable documentation helps organizations identify trends, maintain compliance standards, and ensure patients receive appropriate follow-up care after hospitalization.

Visualization One - Offensive

--I've completed this for you as a placeholder. Replace this with your visualization.
--SQL code in this block

SELECT "admission_type", "subject_id"
FROM "admissions"
WHERE "admission_type" IN (
    "EMERGENCY",
    "ELECTIVE",
    "URGENT",
    "NEWBORN"
);
    
#ggplot visualization in this block. I prettied this up with labels to show what is possible.
ggplot(data = myquery1,
       aes(x = admission_type
)) +
  geom_bar()+
  theme_minimal() +  # Cleans up the background grid lines
  labs(
    title ="Hospital Admissions by Admission Type",
    subtitle ="MIMIC-III Admissions Data",
    x ="Admission Type",
    y ="Number of Admissions",
   
  )

Visualization Two - Defensive

--I've completed this for you as a placeholder. Replace this with your visualization.
--SQL code in this block

SELECT "admission_type", "discharge_location"
FROM "admissions"

    
#ggplot visualization in this block. I prettied this up with labels to show what is possible.
ggplot(data = myquery2,
       aes(x = discharge_location
)) +
  geom_bar()+
  theme_minimal() +  # Cleans up the background grid lines
  labs(
    title ="Hospital Discharges by Discharge Location",
    subtitle ="MIMIC-III Admissions Data",
    x ="Discharge Locations",
    y ="Number of Discharges",
   
  )

Recommendations for Industry

First, the utilization of admission trends by healthcare organizations is needed to advance staffing and capacity planning. By being aware of the trends of the types of admissions they have, healthcare administrators can better allocate resources to meet changing patient needs and preemptively avoid any operational bottlenecks. Second, a healthcare organization must work to strengthen its discharge planning processes to ensure destinations for discharged patients are being documented consistently and accurately. The effective transition of patients from one level of care to another through proper discharge planning and accurate documentation assists with providing a continuum of care, supports regulatory reporting requirements, and mitigates risks to the organization. Third, healthcare organizations must continue to invest in solid data governance programs. Procedures for standardized documentation, ongoing audits for quality, and clear accountability structures result in a more reliable basis for decision-making throughout the organization. Finally, healthcare leaders must understand that offensive and defensive strategies are complementary and not mutually exclusive. Strong data governance creates confidence in a healthcare organization’s data; therefore, the use of analytics and operational insight to create actionable solutions from the data increases the likelihood of achieving improved patient outcomes, compliance with regulatory requirements, and overall long-term performance.

References

Getting your citations out of Google Scholar and into your references.bib file takes just a few clicks.

  1. Go to Google Scholar and search for the paper you want to cite (for example: “What’s your data strategy”).

  2. Look directly underneath the search result for the Cite button (it looks like a double quotation mark ”). Click it.

  3. A pop-up window will appear showing standard citation styles (APA, MLA, etc.). At the very bottom of that pop-up, click the link that says BibTeX.

  4. A new page or plain text block will open showing some bibtex code. Paste that block of text into your [references.bib] file.

If you have multiple sources, just keep pasting them one after another down the file. You don’t need commas between the different articles; just separate them with a blank line to keep it clean. You can then cite them like this (DalleMule and Davenport 2017) and they will automatically appear in the references section.

Blease, Charlotte, Ted J Kaptchuk, Michael H Bernstein, Kenneth D Mandl, John D Halamka, and Catherine M DesRoches. 2020. “Artificial Intelligence and the Future of Primary Care: Exploratory Qualitative Study of UK General Practitioners’ Views.” Journal of Medical Internet Research 22 (3): e19082.
DalleMule, Leandro, and Thomas H Davenport. 2017. “What’s Your Data Strategy?” Harvard Business Review 95 (3): 112–21.
Kruse, Clemens Scott, Brian Smith, Heather Vanderlinden, and Ann Nealand. 2017. “Security Techniques for the Electronic Health Records.” Journal of Medical Systems 41 (8): 127.