The Medical Research Guide

Introduction

After learning essential statistical concepts, let’s explore publicly available databases for conducting clinical studies. Publicly available databases are essential resources for conducting large-scale studies without the need for new data collection. These databases, ranging from patient registries to national health surveys, offer a cost-effective way to explore research questions, especially for medical students and residents who may not have extensive resources. Learning to navigate these databases allows researchers to gain valuable insights into healthcare trends, treatment outcomes, and disease patterns. This chapter introduces key publicly available databases, with a focus on how to leverage them for meaningful studies, including the simpler approach of descriptive studies using the MAUDE database.

Common Publicly Available Databases in Medical Research

1. National Inpatient Sample (NIS)
   • Overview: Part of the Healthcare Cost and Utilization Project (HCUP), NIS is the largest all-payer inpatient care database in the U.S. It includes data on patient demographics, diagnoses, procedures, costs, and hospital information.
   • Use Cases: Studies on hospitalizations, procedure trends, healthcare utilization, and patient outcomes.
2. Nationwide Readmissions Database (NRD)
   • Overview: Also part of HCUP, NRD provides data on hospital readmissions within a calendar year, including patient demographics, diagnoses, and procedures.
   • Use Cases: Ideal for analyzing readmission rates, understanding risk factors, and assessing the impact of interventions on readmission.
3. Medical Device Reporting Database (MAUDE)
   • Overview: Managed by the FDA, the MAUDE database contains adverse event reports involving medical devices. Data includes device information, event descriptions, and outcomes, making it especially valuable for monitoring device safety.
   • Use Cases: Used for identifying trends in device-related adverse events, assessing device efficacy and safety, and exploring complications associated with various devices.
4. SEER (Surveillance, Epidemiology, and End Results Program)
   • Overview: Maintained by the National Cancer Institute, SEER collects comprehensive data on cancer incidence, survival, and mortality.
   • Use Cases: Oncology research focusing on cancer incidence, survival analysis, and treatment outcomes.
5. NHANES (National Health and Nutrition Examination Survey)
   • Overview: Conducted by the CDC, NHANES provides data on health status, nutrition, lifestyle, and biomarkers in a representative sample of the U.S. population.
   • Use Cases: Studies on disease prevalence, health risk factors, and public health trends.
6. ClinicalTrials.gov
   • Overview: A registry and results database of publicly and privately supported clinical studies conducted globally.
   • Use Cases: Useful for research on clinical trial trends, interventions, and outcomes of registered clinical studies.

How to Access Publicly Available Databases

Accessing each database typically involves a few standard steps, though specific requirements may vary. Here is a general overview of the process:

1. Visit the Database Website
   • Access the official website of the database, such as HCUP for NIS and NRD, or the FDA’s MAUDE database. Each site provides guidance on accessing data and complying with usage regulations.
2. Check Access Requirements
   • Some databases require users to register, complete a data use agreement (DUA), or pay a fee. For instance, NIS and NRD require payment and a signed DUA, while SEER data is freely accessible.
3. Complete Any Necessary Training
   • Certain databases, like HCUP, require users to complete an online training course to ensure data is used ethically and responsibly, protecting patient privacy.
4. Submit Data Use Agreement (DUA)
   • Most databases require a signed DUA that outlines how data will be handled, stored, and reported. Adhering to the DUA is essential for legal and ethical compliance.
5. Download Data
   • Once approved, you can download the data in specified formats (often CSV or SAS). It’s crucial to store and handle the data according to the database’s requirements, especially for privacy and security.

Tip: Maintain records of all steps taken to access and comply with the database’s requirements, as these may be necessary for journal publication or audits.

Conducting a Study Using the MAUDE Database

The MAUDE database, managed by the FDA, provides a straightforward way to conduct descriptive studies on medical devices, making it an excellent option for medical students and residents. MAUDE contains adverse event reports related to medical devices, including essential details such as device information, event description, and patient outcomes. Because of its simplicity and accessibility, MAUDE is a valuable starting point for descriptive studies.

Why MAUDE is Ideal for Descriptive Studies

• Studies on the MAUDE database are typically descriptive, focusing on identifying trends or patterns in device-related adverse events rather than complex statistical analyses. This makes MAUDE an excellent choice for researchers without extensive statistical training, allowing them to create impactful studies on device safety and efficacy.

Steps to Conduct a Study Using the MAUDE Database

1. Define Your Research Question
   • Focus on a specific device or type of adverse event. For example, “What are the most common complications associated with pacemakers reported in the MAUDE database?”
2. Access the Database
   • The MAUDE database is freely accessible on the FDA’s website. Simply navigate to the MAUDE page, where you can query the database by device type, event date, manufacturer, or other filters.
3. Extract and Organize Relevant Data
   • Use the MAUDE query tool to narrow down reports based on your research question. For example, if studying pacemakers, filter for pacemaker-related reports. Once filtered, export the data to organize and analyze event descriptions, reported complications, device information, and patient outcomes.
4. Analyze the Data Descriptively
   • As a descriptive study, focus on summarizing the findings. Count and categorize adverse events (e.g., mechanical failure, infection, malfunction) and report trends. Use tables or charts to present the frequency of specific complications, outcomes, and types of devices involved.
5. Interpret Findings in Context
   • Discuss the observed trends and patterns in your data. Compare with findings from other studies or published reports on device safety to provide context.

Tip: Because MAUDE studies are descriptive, they typically do not require complex statistical analyses, allowing for straightforward interpretation. This simplicity makes MAUDE an ideal database for medical students and residents to gain research experience in device safety.

Need for Collaborating with a Statistician

Working with public databases like NIS, NRD, or SEER often involves handling large and complex datasets. Collaborating with a statistician who is familiar with these databases and their specific requirements can enhance the quality and accuracy of your research.

1. Database-Specific Expertise
   • Each database has unique characteristics. For example, NIS and NRD datasets require specific adjustments, such as weighting, to create national estimates. Statisticians experienced with these databases understand the nuances and can help you manage the data correctly.
2. Advanced Statistical Analysis
   • Many research questions require advanced analyses, such as regression modeling, survival analysis, or handling confounding factors. A statistician can ensure that appropriate statistical tests are applied to your study.
3. Cost-Effectiveness
   • Accessing some of these databases can be costly, and collaborating with a statistician familiar with the database may eliminate the need to purchase separate access or software, saving resources.
4. Ensuring Compliance
   • Statisticians often have experience in complying with data use agreements, privacy standards, and the database’s specific guidelines, which helps to prevent ethical or legal issues.

Tip: When contacting a statistician, clearly outline your research objectives, data sources, and preliminary analysis plan. This information will help the statistician assist you effectively.

Compliance with Database Requirements

Using publicly available databases involves adhering to strict guidelines for data usage, handling, and reporting. To ensure compliance, follow these steps:

1. Adhere to Data Use Agreements (DUA)
   • Follow all terms outlined in the DUA, which specify data handling, storage, and reporting requirements.
2. Acknowledge the Database Source
   • Always cite the database appropriately in your study. Most databases provide citation guidelines; for example, HCUP databases require specific acknowledgment in both the methods section and references.
3. Ensure Privacy Standards
   • Even with de-identified data, privacy standards (such as HIPAA compliance) are essential. Store data securely and restrict access to authorized individuals only.
4. Document Your Analysis Steps
   • Keep thorough records of all data handling and analysis steps, which may be required by journals or during peer review to verify the study’s reproducibility.
5. Submit for Review if Required
   • Some databases, like HCUP, may require researchers to submit study results or publications for review before release. Check specific guidelines to ensure compliance.

Conclusion

Publicly available databases offer a wealth of information that medical students, residents, and clinicians can use to conduct meaningful research. The MAUDE database is particularly accessible, enabling descriptive studies on medical device safety and allowing beginners to gain experience with a relatively straightforward study design. For more complex databases like NIS and NRD, collaborating with a statistician can enhance study quality, ensuring that data handling and analyses meet scientific standards. By following these steps and maintaining compliance with database requirements, researchers can generate valuable insights that contribute to healthcare knowledge and patient safety.

References

1. National Inpatient Sample (NIS): Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality. Available at: https://www.hcup-us.ahrq.gov/nisoverview.jsp
2. Nationwide Readmissions Database (NRD): Healthcare Cost and Utilization Project (HCUP). Agency for Healthcare Research and Quality. Available at: https://www.hcup-us.ahrq.gov/nrdoverview.jsp
3. Medical Device Reporting (MAUDE): U.S. Food and Drug Administration (FDA). Available at: https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfMAUDE/search.CFM
4. SEER (Surveillance, Epidemiology, and End Results Program): National Cancer Institute. Available at: https://seer.cancer.gov/
5. NHANES (National Health and Nutrition Examination Survey): Centers for Disease Control and Prevention (CDC). Available at: https://www.cdc.gov/nchs/nhanes/index.htm
6. ClinicalTrials.gov: U.S. National Library of Medicine. Available at: https://clinicaltrials.gov/