Mastering Research During Residency: A Guide to Clinical Informatics

Clinical informatics sits at the intersection of medicine, data, and technology—and residency is one of the best times to build a strong research foundation in this rapidly evolving field. Whether you aim for a clinical informatics fellowship, an academic residency track, or a health IT leadership role, structured research during residency can set you apart and open doors.

This guide walks you through why informatics research matters, how to design and execute resident research projects, how to collaborate effectively with health IT teams, and how to position yourself for a future in clinical informatics.

---

Why Research During Residency Matters in Clinical Informatics

Clinical informatics is inherently research-driven. Every workflow redesign, decision-support rule, or data dashboard is, at its core, a hypothesis about how technology can improve care. Residency is the ideal time to develop the mindset and skills to ask and answer those questions systematically.

1. Building core skills for a clinical informatics fellowship

If you’re considering a clinical informatics fellowship, programs will look closely at your ability to:

• Identify meaningful clinical and operational problems
• Ask researchable questions about health IT and data
• Design feasible projects within real-world constraints
• Work with interdisciplinary teams (IT, data science, nursing, administration)
• Disseminate findings through presentations and publications

Having completed one or more substantial resident research projects—particularly in informatics—signals that you can navigate that entire lifecycle.

2. Strengthening your application for academic and leadership tracks

Even if you’re not certain about a formal fellowship, informatics research is valuable if you’re eyeing:

• An academic residency track with protected research time
• A junior faculty role with a focus on quality improvement or digital health
• Departmental roles in EHR optimization, population health, or data analytics
• Health system leadership in health IT or quality & safety

Research demonstrates your ability to handle complex systems, work across departments, and generate evidence to support change—key traits for physician leaders in informatics.

3. Developing a systems-thinking lens

Informatics research forces you to think beyond the individual patient to:

• Workflow and team dynamics
• Information flow and documentation patterns
• User interface and usability issues
• Data quality, completeness, and bias
• Regulatory and privacy constraints (HIPAA, data sharing)

This systems-thinking lens is a hallmark of effective clinical informaticians and will influence how you practice medicine, even in a purely clinical role.

4. Creating tangible impact on patient care

Many traditional resident research projects are retrospective chart reviews with limited immediate impact. Informatics projects, by contrast, often:

• Redesign order sets or documentation flows
• Implement clinical decision support tools
• Improve test ordering, handoffs, or discharge processes
• Optimize dashboards for sepsis, readmissions, or throughput

Your research can directly improve efficiency, safety, or patient outcomes during your own training—making the process more rewarding and helping you build a track record of impact.

---

Understanding the Landscape: Types of Clinical Informatics Research You Can Do in Residency

There is no single template for informatics research. Think of opportunities along three intersecting axes: research type, data source, and implementation scope.

1. By research type

a. Quality improvement (QI) with an informatics focus

Many residency programs require QI projects; adding an informatics dimension is a natural fit.

Examples:

• Redesigning an EHR order set to reduce unnecessary labs or imaging
• Implementing alerts for high-risk medications or renal dosing
• Streamlining discharge documentation to improve follow-up scheduling

These projects often use PDSA (Plan–Do–Study–Act) cycles, with clear, measurable outcomes—highly publishable if done rigorously.

b. Clinical data analytics and outcomes research

Using EHR or registry data, you can study patterns in:

• Treatment variation
• Diagnostic delays
• Predictors of readmission or ED revisits
• Population health metrics (e.g., vaccination rates, chronic disease control)

Examples:

• Building and validating a simple risk score from routine EHR data
• Examining how time of day or provider type affects ordering behavior
• Studying adherence to clinical guidelines using structured data

These projects sharpen your skills in health IT training: data extraction, basic statistics, and interpretation of real-world clinical data.

c. Human factors and usability research

How do clinicians interact with technology, and how does it affect their performance or well-being?

Examples:

• Studying alert fatigue and override rates for EHR warnings
• Observing and coding workflow disruptions due to new software
• Surveying residents about EHR usability and burnout symptoms

These projects may use qualitative methods (interviews, focus groups) and mixed methods (combining survey data with EHR logs).

d. Implementation science and digital health interventions

These projects focus on how to get a technology intervention adopted, sustained, and scaled.

Examples:

• Rolling out a secure messaging tool for inpatient communication
• Implementing a telehealth follow-up clinic and tracking outcomes
• Piloting patient portals or remote monitoring for a specific population

You’ll learn about change management, stakeholder engagement, and evaluation frameworks (e.g., RE-AIM, CFIR).

2. By data source

Common data sources you can access during residency include:

• EHR transactional data (orders, meds, labs, vitals, encounters)
• Clinical documentation (notes, problem lists) for chart review or NLP
• Audit logs (who clicked what, when)
• Registries or disease-specific databases
• Patient-reported outcomes or portal usage
• Operational data (bed management, throughput, staffing)

Understanding what data exists—and how to legally and safely access it—is a core skill in clinical informatics research.

3. By scope and feasibility

Residency time is limited. Scope wisely:

• Single-unit or single-service projects: Easiest to implement and evaluate
• Multi-site or multi-specialty studies: High value but require strong mentorship and institutional support
• Retrospective studies: More feasible for tight schedules; good starting point
• Prospective or interventional studies: Higher impact but require careful planning and often IRB approval well in advance

Choose a project type and scope that fits your training schedule and available resources.

---

Designing and Launching Resident Research Projects in Clinical Informatics

The most common pitfall is picking a project that is too broad, too technical, or too dependent on others to succeed in the short timeline of residency. Here’s a structured approach to get it right.

1. Start with a real, painful problem

The best informatics ideas usually come from daily frustrations:

• Repeatedly ordering the same complex panel of tests
• Confusing or redundant documentation requirements
• Communication breakdowns during handoffs
• Delays due to missing pre-op information or consents

Ask yourself:

• What slows me down or puts patients at risk?
• What do my co-residents complain about constantly?
• What do nurses or pharmacists find most burdensome in the EHR?

Turn a complaint into a researchable question:

• “How can we reduce redundant lab ordering by improving the order set?”
• “Can we decrease discharge delays by redesigning discharge instructions and follow-up workflow?”
• “Does a new sepsis alert reduce time to antibiotics without increasing alert fatigue?”

2. Find the right mentors and collaborators early

In clinical informatics, you rarely work alone. Key players may include:

• An informatics-oriented faculty mentor (clinical informatics board-certified, CMIO, or similar)
• A data analyst or data warehouse staff member
• A statistician or epidemiologist
• IT analysts or EHR builders (Epic, Cerner, etc.)
• Quality improvement staff and nurse leaders

Practical steps:

• Ask your program director who leads health IT training or quality initiatives
• Attend informatics or QI committee meetings and introduce yourself
• Email the hospital’s clinical informatics or analytics team expressing interest in a small pilot project

Aim for at least:

• 1 clinical mentor with informatics experience
• 1 person who can help you obtain and interpret data

3. Clarify your project type and endpoints

Decide early: Is this QI? Research? Hybrid?

• QI: Focused on local improvement; often doesn’t need full IRB review if not generalizable research—but check your institution’s policies.
• Research: Designed to produce generalizable knowledge; typically requires IRB review or exemption.
• Hybrid: Many informatics projects start as QI but are written up as research; planning for publication from the beginning helps.

Define SMART endpoints:

• Specific
• Measurable
• Achievable
• Relevant
• Time-bound

Examples:

• “Reduce unnecessary daily BMP orders by 30% within 6 months”
• “Increase appropriate VTE risk assessment documentation from 60% to 85% in 4 months”
• “Improve median time from ED arrival to first dose of antibiotics for septic shock by 20% over 1 year”

4. Map your data and methods

For each endpoint, ask:

• Where does this data live? (EHR module, data warehouse, local registry)
• How is it structured? (Discrete fields vs. free text)
• Who can pull it for me? (Analyst, informatics fellow, self-service tools)
• What is the appropriate analysis? (Run charts, before–after comparisons, regression)

Design simple but robust methods:

• Pre/post study with statistical tests (e.g., t-test, chi-square)
• Interrupted time series for implementation projects
• Control groups if feasible (e.g., unit with vs. without intervention)
• Qualitative feedback to complement quantitative metrics

5. Secure approvals and align with institutional priorities

Most institutions will be more supportive if:

• Your project aligns with existing initiatives (e.g., sepsis, readmissions, opioid stewardship)
• You’re not asking for major EHR changes without strong justification
• You demonstrate awareness of patient privacy and data security

Action steps:

• Present a 1-page summary to your mentor and relevant committee (e.g., Pharmacy & Therapeutics, QI committee, Informatics steering group)
• Determine if you need:
  • IRB approval or exemption
  • Approval from the EHR change control board
  • Departmental or IT sign-off

Starting this process early avoids delays that can derail projects within residency timelines.

---

Working with EHRs and Health IT: Practical Tips for Residents

Accessing and using data is where informatics projects succeed or stall. Learning to navigate this landscape is itself a form of health IT training.

1. Understand your institution’s data infrastructure

Key questions to ask early:

• Do we have a clinical data warehouse or analytics environment?
• Are there “standard reports” already built for common metrics?
• Who can create custom data extracts?
• Is there a self-service analytics tool (e.g., Tableau, Power BI embedded in the EHR)?

Schedule a short meeting with a data analyst to:

• Review what data fields are available
• Learn basic report parameters you can request
• Clarify data lag (e.g., real-time vs. 24-hour delay)

2. Manage expectations about EHR changes

Major EHR changes are complex, so:

• Start with small, contained changes: order set tweaks, adding a decision support rule, modifying a flowsheet.
• Expect a testing and approval process: build → test environment → end-user feedback → production deployment.
• Align your project timelines with EHR build cycles.

If your project depends on an EHR change:

• Build in 2–3 months of lead time
• Get a named IT analyst as a collaborator, not just a “contact”
• Clarify who owns maintenance after your residency

3. Be realistic about data quality and limitations

EHR data are messy:

• Missing values, inconsistent coding, and documentation bias are common.
• Free-text notes may be rich but hard to analyze without natural language processing (NLP) support.
• Time stamps can be misleading (e.g., orders entered late, back-charted documentation).

Strategies:

• Start with well-structured data elements (orders, meds, labs, ICD codes).
• Cross-check metrics with a small manual chart review to validate data accuracy.
• Document limitations transparently in your methods and discussion.

4. Protect patient privacy and comply with regulations

Research during residency must comply with:

• HIPAA and local privacy laws
• Institutional policies on data access and transfer
• IRB and information security guidelines

Basic principles:

• Use de-identified or limited datasets whenever possible.
• Never export identifiable data to personal devices or unapproved storage.
• Use institution-approved tools for analysis (e.g., secure servers, encrypted laptops).
• Complete required research and privacy training modules.

Clarify early:

• What level of data (identified vs. de-identified) do you truly need?
• Where can the data live during analysis?
• Who else may access or work with the dataset?

---

Balancing Research with Clinical Duties and Career Planning

Residency is intense. To succeed with informatics research, you need both tactics for day-to-day balance and a strategy aligned with your long-term goals.

1. Protect time intentionally

Even in busy programs, you can:

• Block regular research hours on lighter rotations (e.g., ambulatory, electives).
• Use small time windows:
  • 15–20 minutes to review articles or update your protocol
  • 30–60 minutes to work on analysis or writing
• Schedule recurring check-ins with your mentor (e.g., twice monthly); external accountability keeps projects moving.

If your program offers an academic residency track:

• Apply early, especially if it comes with protected research time.
• Use that time strategically for high-value tasks: data extraction, analysis, manuscript drafting.

2. Stage your projects across residency years

A suggested trajectory:

PGY-1: Exploration and skill-building

• Identify interests and possible informatics mentors.
• Participate in a small QI or data project as a co-investigator.
• Complete basic courses or workshops in statistics, R, or Python if feasible.
• Learn your institution’s EHR and reporting tools.

PGY-2: Lead a defined informatics project

• Design a feasible study with clear endpoints.
• Secure IRB/QI approvals.
• Implement or extract data and begin analysis.
• Present locally (departmental conference, hospital QI day).

PGY-3+: Deepen and disseminate

• Complete analysis and draft manuscripts.
• Submit abstracts to regional/national conferences (e.g., AMIA, specialty meetings).
• If aiming for a clinical informatics fellowship, align your projects with fellowship interests and mentors.

3. Use your research to support career next steps

If you are targeting a clinical informatics fellowship:

• Emphasize informatics research in your personal statement:
  • Problem addressed
  • Your role
  • Methods used
  • Impact and lessons learned
• Ask informatics mentors for letters highlighting:
  • Your analytic and problem-solving skills
  • Your collaboration with IT and interdisciplinary teams
  • Your potential as a future informatics leader

If you’re aiming for an academic or leadership role without fellowship:

• Highlight your trajectory of informatics involvement:
  • Resident research projects
  • Committee participation (informatics/QI boards)
  • EHR build or optimization contributions
• Articulate a career plan that integrates clinical practice with informatics responsibilities.

---

Common Pitfalls and How to Avoid Them

Being aware of frequent obstacles can save you months of frustration.

1. Overly ambitious scope

Symptoms:

• Multiple endpoints, multiple units or hospitals, several EHR changes.
• Reliance on multiple external teams with limited capacity.

Prevention:

• Start with a focused question and minimal viable scope.
• Plan follow-up phases you can hand off or expand if time allows.

2. Late engagement with mentors and IT

Symptoms:

• Designing a project in isolation, then discovering it’s not feasible.
• IRB or EHR change approvals delayed for months.

Prevention:

• Involve mentors, data analysts, and EHR builders from the outset.
• Present your idea as a problem statement, not a pre-decided technical solution.

3. Underestimating data and analysis complexity

Symptoms:

• Requesting overly broad datasets that are hard to clean.
• Getting stuck at the analysis stage due to lack of stats support.

Prevention:

• Begin with a well-defined cohort and a limited set of variables.
• Secure statistical support early; use institutional biostatistics resources.
• Accept simpler analyses if they answer your core question.

4. Neglecting dissemination

Symptoms:

• “We did a cool project, but never wrote it up.”
• Missing the chance to present at even local conferences.

Prevention:

• Set explicit dissemination goals at project launch:
  • At least one abstract and one poster or platform presentation.
  • Aim for at least one manuscript submission, even if to a smaller journal.
• Block dedicated writing time as your project nears completion.

---

FAQs: Research During Residency in Clinical Informatics

1. Do I need advanced programming or data science skills to do informatics research in residency?

Not necessarily. Many impactful projects use:

• Standard EHR reporting tools
• Support from data analysts and statisticians
• Simple statistical methods (e.g., before–after comparisons)

Programming skills (R, Python, SQL) can expand what you can do, but they are not mandatory. If you’re interested, start with small, practical tasks—such as cleaning datasets or generating basic plots—rather than trying to build complex models immediately.

2. How is a clinical informatics project different from a typical QI project?

There is overlap. Key distinctions:

• Informatics projects explicitly focus on the design, use, or impact of health IT (e.g., EHR workflows, decision support, digital tools).
• They often involve data extraction and analysis from electronic systems.
• They may require collaboration with IT and informatics teams and sometimes EHR build changes.

Many projects are both QI and informatics-focused, and that dual identity can increase their relevance and publishability.

3. How can I find mentors in clinical informatics if my program doesn’t have a formal fellowship?

Options include:

• Identifying faculty who:
  • Sit on informatics or EHR committees
  • Lead QI or population health initiatives
  • Have published on digital health or health IT
• Contacting your institution’s:
  • CMIO or associate CMIO
  • Director of clinical informatics
  • Data analytics or quality departments
• Reaching out to external mentors:
  • Through professional societies (e.g., AMIA)
  • At conferences where you present your work
  • Via email introductions from your program director

You can also have a local clinical mentor plus an external informatics mentor for specialized guidance.

4. Will doing research during residency really help me match into a clinical informatics fellowship?

Yes. Fellowship programs typically look for:

• Demonstrated interest in informatics (projects, committees, coursework)
• Evidence you can carry a project from design to dissemination
• Comfort working with data and interdisciplinary teams

Having one or more well-executed resident research projects in informatics—especially with presentations or publications—strongly strengthens your application and shows you’re committed to the field.

---

Research during residency in clinical informatics is both feasible and highly rewarding. By choosing a focused, relevant problem, partnering effectively with mentors and IT teams, and planning for dissemination from the start, you can create work that improves care locally while building a strong foundation for a future in informatics, academia, or health IT leadership.