Prospective vs. Retrospective Cohort Studies: What's the Difference?

A cohort study follows a defined group of people over time to observe who develops a particular outcome. The distinction between prospective and retrospective cohort studies comes down to when the researcher starts the study relative to when the exposure and outcome data were collected. In a prospective cohort study, you enroll participants now, measure their exposures, and follow them forward in time to see who develops the outcome. In a retrospective cohort study, you identify a group of people who were already exposed or unexposed in the past, then look back through existing records to determine their outcomes. The choice between these two designs affects data quality, cost, time to completion, susceptibility to bias, and the strength of causal claims you can make. This article explains the practical differences so researchers, students, and life-science professionals can select the appropriate design for their question.

Defining Cohort Studies and Their Core Purpose

A cohort is a group of people who share a defining characteristic or experience within a defined period. Cohort studies are observational, meaning the researcher does not assign exposures or interventions. Instead, the researcher observes who is exposed and who is not, then tracks outcomes over time. This design is well suited for studying incidence, natural history of disease, risk factors, and prognosis.

Cohort studies occupy a middle position in the evidence hierarchy. They sit above case reports and case series but below randomized controlled trials for questions of treatment effectiveness. For questions where randomization is unethical or impractical, such as studying the health effects of smoking, air pollution, or surgical techniques that are already established in practice, cohort studies are often the strongest feasible design.

The defining feature of any cohort study is the temporal direction. You must establish that the exposure occurred before the outcome. Both prospective and retrospective cohort studies maintain this temporal sequence, but they do so in different ways. A prospective study establishes the sequence by following people forward. A retrospective study establishes the sequence by reconstructing it from historical records.

Prospective Cohort Studies: Design and Workflow

In a prospective cohort study, you define your study population, assess baseline exposures, and then follow participants forward in time. The follow-up period can last months, years, or decades. During follow-up, you periodically assess participants for the development of outcomes.

Enrollment and Baseline Data Collection

Prospective enrollment requires you to identify and recruit participants before any outcomes have occurred. You must define clear inclusion and exclusion criteria. At baseline, you collect information on exposures, potential confounders, and demographic characteristics. This baseline assessment happens before you know who will develop the outcome, which is a key strength of the design.

Baseline data collection can include questionnaires, physical examinations, laboratory tests, imaging, and biological samples. Because you control the data collection process, you can standardize measurements across all participants. You can also train staff to use consistent techniques and calibrate instruments.

Follow-Up Procedures

After enrollment, you follow participants at regular intervals. Follow-up visits may occur annually, biannually, or at other prespecified times. At each visit, you reassess exposure status, collect outcome information, and update covariate data. You must maintain contact with participants to minimize loss to follow-up.

The Experimental Design Assistant from NC3Rs provides a structured tool for planning animal studies, but its principles of randomization, blinding, and allocation concealment apply to the design of any longitudinal study. For human cohort studies, you should also consult reporting guidelines available through the EQUATOR Network to ensure your study design and eventual manuscript meet field standards.

Real-World Example of a Prospective Cohort

A prospective controlled cohort study evaluated whether hyaluronic acid injections into the chin caused bone resorption. Researchers recruited 78 patients and collected computed tomographic scans at baseline and again at 6 to 12 months after injection. They compared bone measurements before and after exposure in the injection group and against a control group. The study found a significant association between hyaluronic acid injection and bone resorption, with discernible bone resorption visible on imaging in 35.90% of patients. This design allowed the researchers to measure bone thickness before any exposure occurred, providing a clean baseline for comparison.

Another prospective cohort study compared robot-assisted thyroidectomy with open thyroidectomy in 306 patients with papillary thyroid carcinoma. Researchers enrolled patients, assigned them to surgical groups, and followed them postoperatively to measure rates of hypoparathyroidism, quality of life, and scar outcomes. The prospective design allowed standardized postoperative assessment using validated scales at consistent time points.

Retrospective Cohort Studies: Design and Workflow

In a retrospective cohort study, you identify a group of people who were exposed or unexposed in the past, then use existing records to determine their outcomes. The study begins after both the exposure and the outcome have already occurred. You reconstruct the timeline from historical data.

Identifying the Cohort From Existing Records

The first step in a retrospective cohort study is to define the source population and identify eligible participants from existing databases, medical records, registries, or administrative data. You must establish clear inclusion and exclusion criteria, just as you would in a prospective study. The key difference is that you are working with data that were collected for other purposes, often clinical care or administrative billing.

You must verify that exposure status can be determined from the available records. If exposure information is missing, incomplete, or inconsistently recorded, the study may not be feasible. You should conduct a preliminary review of the data source to confirm that the necessary variables exist and are complete enough to answer the research question.

Reconstructing Exposure and Outcome Data

After identifying the cohort, you extract data on exposures, covariates, and outcomes from the records. This extraction requires a standardized data collection form or electronic database. You must define each variable precisely and apply the same definitions to all participants.

Outcome ascertainment in retrospective studies depends on the quality of the records. Outcomes that are routinely documented, such as death, hospitalization, or surgical complications, are easier to capture reliably. Outcomes that require specific testing or clinical judgment may be underascertained if testing was not performed consistently.

Real-World Example of a Retrospective Cohort

A retrospective cohort study using the MIMIC-IV database examined the association between stress hyperglycemia ratio and delirium in 1,111 elderly surgical ICU patients. Researchers calculated the stress hyperglycemia ratio from admission glucose and HbA1c values already recorded in the database. They identified delirium using the CAM-ICU tool, which was documented as part of routine clinical care. The study found that higher stress hyperglycemia ratio was independently associated with greater odds of delirium, with an adjusted odds ratio of 2.79 for the highest quartile compared with the lowest.

A national retrospective cohort study in Denmark included 25,720 patients with stage IA to IV melanoma diagnosed between 2008 and 2019. Researchers used nationwide population-based registry data to estimate stage-specific risks of recurrence and melanoma-specific mortality. Patients were followed from primary treatment until December 2021, with a median follow-up of 5.9 years. This study demonstrates how retrospective designs can leverage large, complete national registries to answer questions that would be impractical prospectively.

Key Differences Between Prospective and Retrospective Designs

The choice between prospective and retrospective cohort designs affects nearly every aspect of the study. The table below summarizes the most important differences.

Bias and Validity Considerations

Bias is a systematic error that distorts the true association between exposure and outcome. The two designs differ in their susceptibility to various types of bias.

Selection Bias

Selection bias occurs when the participants included in the study are not representative of the target population. In prospective studies, selection bias can arise from differential loss to follow-up. If participants who drop out differ systematically from those who remain, the observed association may be distorted. You should compare baseline characteristics of completers and dropouts to assess this risk.

In retrospective studies, selection bias can arise from the way you identify the cohort from records. If the records do not capture all eligible individuals, or if certain groups are more likely to have complete records, the cohort may not represent the target population. You should carefully define the source population and document the sampling frame.

Information Bias

Information bias occurs when data on exposure or outcome are measured inaccurately. Prospective studies generally have better information quality because you control the measurement process. You can use standardized instruments, train data collectors, and implement quality control procedures.

Retrospective studies depend on data that were collected for clinical or administrative purposes. These data may be incomplete, inconsistent, or recorded in ways that are difficult to interpret. For example, a retrospective study of chronic endometritis in women with repeated implantation failure and recurrent pregnancy loss analyzed 392 women with repeated implantation failure and 119 women with recurrent pregnancy loss who underwent endometrial biopsy between 2016 and 2024. The study relied on CD138 immunohistochemistry results from clinical biopsies, which were performed at the discretion of treating physicians instead of according to a research protocol.

Confounding

Confounding occurs when a third variable is associated with both the exposure and the outcome, distorting the observed association. Both prospective and retrospective cohort studies are susceptible to confounding. You can address confounding through study design, such as restriction or matching, or through statistical adjustment in the analysis.

A retrospective cohort study comparing transperineal and transrectal prostate biopsy techniques in 127 patients found a significantly higher cancer detection rate with the transperineal approach. However, the authors noted that the non-randomized design and protocol differences between groups meant the findings should be interpreted with caution and warranted confirmation in prospective multicenter studies. This example illustrates how residual confounding can limit retrospective findings.

Cost, Time, and Resource Considerations

Prospective cohort studies require substantial resources. You must fund recruitment, baseline assessments, follow-up visits, data management, and staff salaries. The longer the follow-up period, the greater the cost. Studies with decades of follow-up require sustained funding commitments and institutional support.

Retrospective cohort studies are generally faster and less expensive. The data already exist, so the main costs are data extraction, cleaning, and analysis. A retrospective study can often be completed in months instead of years. This efficiency makes retrospective designs attractive for preliminary investigations, hypothesis generation, and questions where prospective data are not yet available.

However, the lower cost of retrospective studies comes with tradeoffs in data quality and control. You cannot go back and collect additional variables that were not recorded. You cannot standardize measurements that were performed inconsistently. You must work within the limitations of the existing data.

Combined Retrospective-Prospective Designs

Some studies combine retrospective and prospective components to leverage the strengths of both designs. These hybrid designs are increasingly common in clinical research.

A study of symptomatic hemorrhagic transformation in acute ischemic stroke patients used a combined approach. In the first part, researchers screened 18 patients with symptomatic hemorrhagic transformation and 128 without it from a retrospective cohort. They measured 92 cerebrovascular disease-related proteins in baseline blood samples using Olink proteomics technology. In the second part, they selected 28 patients with symptomatic hemorrhagic transformation and 130 without it from a prospective cohort and used ELISA to confirm the associations found in the retrospective phase. This two-stage design allowed efficient biomarker discovery in the retrospective phase followed by validation in the prospective phase.

A study of thromboembolic risk in patients undergoing elective electrical cardioversion used a combined retrospective-prospective design. The retrospective cohort included 220 patients who underwent cardioversion without routine transesophageal echocardiography. The prospective cohort included 85 patients who underwent transesophageal echocardiography before cardioversion. The study found that thromboembolic events were rare in both groups, with 3 of 220 patients in the retrospective cohort experiencing an event and none in the prospective cohort.

A study of oropharyngeal administration of own mother's colostrum in preterm very low birth weight infants used a prospective-retrospective observational design to examine short-term health outcomes. Similarly, a study of double-layer versus conventional mucoperiosteal flap closure for medication-related osteonecrosis of the jaw used a combined retrospective-prospective cohort design. These hybrid approaches allow researchers to build on existing data while adding prospective elements to address specific limitations.

Choosing Between Prospective and Retrospective Designs

The choice between prospective and retrospective cohort designs depends on your research question, available resources, and the current state of knowledge.

When to Choose a Prospective Design

Choose a prospective cohort study when you need precise, standardized measurement of exposures and outcomes. Prospective designs are preferable when:

• The exposure can be measured accurately at baseline
• You need biological samples collected under controlled conditions
• The outcome requires active surveillance to detect
• You want to minimize recall bias
• You have the resources and time for long-term follow-up
• The research question involves rare exposures that require careful baseline characterization

A prospective cohort study of parathyroid function after thyroidectomy enrolled 306 patients and followed them postoperatively with standardized quality of life scales. The prospective design allowed consistent administration of the Short Form-36, Visual Impairment Scale, Swallowing Impairment Scale, Neck Impairment Scale, and Scar Questionnaire at defined time points.

When to Choose a Retrospective Design

Choose a retrospective cohort study when you need answers quickly, when the data already exist, or when a prospective study would be impractical. Retrospective designs are preferable when:

• The outcome has a long latency period
• You need a large sample size that would be difficult to recruit prospectively
• The exposure and outcome data are already recorded in reliable databases
• You are conducting preliminary or hypothesis-generating research
• Resources for prospective follow-up are not available

A retrospective cohort study of incidental pulmonary nodules in oral squamous cell carcinoma included 372 patients treated with curative intent between 2011 and 2018. Researchers extracted clinical, demographic, and radiographic data from electronic records. The study found pulmonary nodules in 25.6% of patients with baseline thoracic imaging, but nodules were not independently associated with metastatic disease-related death or reduced overall survival.

Decision Framework

Use the following questions to guide your choice:

1. Do the exposure and outcome data already exist in reliable records? If yes, a retrospective design may be feasible.
2. Can you measure the exposure accurately at baseline? If not, a prospective design may be needed.
3. How long is the expected latency between exposure and outcome? Longer latency favors retrospective designs using existing data.
4. What resources are available for follow-up? Limited resources favor retrospective designs.
5. How important is standardized outcome ascertainment? If critical, a prospective design is preferable.
6. Is the research question about a rare exposure or a rare outcome? Rare exposures favor prospective designs, while rare outcomes may be more efficiently studied retrospectively.

Common Failure Patterns and How to Avoid Them

Both prospective and retrospective cohort studies can fail in predictable ways. Recognizing these failure patterns helps you design a study that produces valid results.

Failure Pattern 1: Incomplete Exposure Data

In retrospective studies, exposure data may be missing or recorded inconsistently. For example, a retrospective study of residual myometrial thickness after cesarean delivery initially included 80 pregnant women, but 15 were excluded because of incomplete records or missing follow-up data, leaving 65 patients in the final analysis. This loss of 18.75% of the sample could introduce selection bias if the excluded women differed systematically from those included.

To avoid this failure, conduct a thorough feasibility assessment of the data source before committing to the study. Document the completeness of each variable you plan to use. If key variables are missing for a substantial proportion of participants, consider whether the study is feasible or whether you need to supplement the data with additional sources.

Failure Pattern 2: Loss to Follow-Up

In prospective studies, loss to follow-up is a major threat to validity. Participants may move, withdraw, or die from causes unrelated to the study outcome. If loss to follow-up is differential between exposure groups, the observed association may be biased.

To minimize loss to follow-up, maintain multiple contact methods, schedule regular follow-up visits, and collect contact information for family members or friends who can help locate participants who move. Monitor retention rates throughout the study and investigate reasons for dropout.

Failure Pattern 3: Outcome Misclassification

In retrospective studies, outcomes may be misclassified if the records do not capture all events or if diagnostic criteria changed over time. For example, a retrospective study of chronic endometritis relied on CD138 immunohistochemistry performed on clinical biopsies. If some women with chronic endometritis were not biopsied, they would be misclassified as not having the condition.

To avoid outcome misclassification, use validated outcome definitions, apply the same criteria to all participants, and consider using multiple data sources to confirm outcomes. If possible, have outcomes adjudicated by reviewers who are blinded to exposure status.

Failure Pattern 4: Confounding by Indication

In observational studies of treatments, the reason a patient received a particular treatment may be associated with the outcome. This is called confounding by indication. For example, a retrospective cohort study comparing sintilimab and pembrolizumab for advanced non-small cell lung cancer used propensity score matching with height and treatment regimen as covariates to balance between-group differences. The study found no significant differences in hepatic-related adverse events between the two groups.

To address confounding by indication, collect detailed data on the reasons for treatment decisions, use statistical methods such as propensity score matching or multivariable adjustment, and interpret results cautiously when residual confounding is possible.

Records and Measurements

Accurate record keeping is essential for both prospective and retrospective cohort studies. The quality of your data determines the validity of your conclusions.

Data Collection Instruments

For prospective studies, develop standardized data collection forms or electronic case report forms before enrollment begins. Pilot test these forms to ensure they are clear, complete, and feasible to administer. Train all data collectors to use the forms consistently.

For retrospective studies, develop a data extraction form that specifies the exact variables to be collected, the definitions of each variable, and the source of each data element. Use a codebook to document variable names, coding schemes, and any transformations applied.

Data Quality Checks

Implement quality control procedures to identify and correct errors. These procedures may include:

• Range checks to identify implausible values
• Consistency checks to identify contradictory data
• Double data entry for a sample of records to assess error rates
• Periodic audits of data collection procedures
• Documentation of all data cleaning decisions

Documentation Standards

Maintain a study protocol that describes the study design, population, exposure and outcome definitions, and analysis plan. Document any deviations from the protocol and the reasons for them. Keep a log of all data queries and resolutions.

The Research Data Framework from the National Institute of Standards and Technology provides guidance on managing research data throughout its lifecycle. Following such frameworks helps ensure that your data are findable, accessible, interoperable, and reusable.

Quality and Welfare Controls

For studies involving human participants, ethical oversight is mandatory. You must obtain approval from an institutional review board or research ethics committee before starting the study. For prospective studies, you must obtain informed consent from participants. For retrospective studies using existing data, the ethics committee may waive the requirement for informed consent if the research involves no more than minimal risk and cannot practically be conducted without the waiver.

For animal studies, the Experimental Design Assistant from NC3Rs helps researchers plan experiments that minimize animal use and suffering while maximizing scientific validity. The tool guides researchers through key design decisions including randomization, blinding, sample size calculation, and statistical analysis.

Patient welfare in clinical cohort studies requires attention to the risks and burdens of study procedures. Baseline and follow-up assessments should be designed to minimize discomfort and inconvenience. If the study involves additional tests or procedures beyond routine care, you must justify these in the protocol and obtain appropriate consent.

Safety and Regulatory Context

Cohort studies are observational and do not involve the administration of investigational treatments. However, they are still subject to regulatory oversight. In many jurisdictions, research involving human participants must comply with data protection laws that govern the collection, storage, and sharing of personal health information.

For retrospective studies using clinical data, you must ensure that the data are de-identified or that you have appropriate authorization to access identifiable data. The National Center for Biotechnology Information provides access to literature and databases that can help researchers understand the regulatory landscape for health data research.

For prospective studies, you must comply with requirements for reporting serious adverse events that occur during follow-up, even if these events are not caused by the study procedures. You should have a plan for managing incidental findings, such as abnormalities detected on baseline imaging or laboratory tests.

Professional Escalation Criteria

Researchers should know when to seek additional expertise or escalate concerns. Consider consulting a biostatistician or epidemiologist when:

• You are uncertain about the appropriate sample size or statistical power
• The analysis requires complex methods such as propensity score matching, competing risk analysis, or time-varying exposures
• You encounter unexpected patterns in the data that may indicate bias or confounding
• You are considering stopping the study early due to concerns about participant safety or data quality

Consider consulting an ethics committee or research integrity office when:

• You discover that the study protocol was not followed
• You identify errors in data collection or analysis that could affect the conclusions
• You have concerns about the ethical conduct of the research
• You are uncertain about whether a planned analysis is appropriate

For clinical questions, consider consulting a clinician with relevant expertise when:

• You need help interpreting clinical outcomes or diagnostic criteria
• You are uncertain about the clinical significance of your findings
• You need guidance on how to present results to clinical audiences

Limitations of Cohort Studies

Both prospective and retrospective cohort studies have inherent limitations that you should acknowledge when interpreting results.

Residual Confounding

Observational studies cannot fully eliminate confounding. Even with careful measurement and statistical adjustment, unmeasured or imperfectly measured confounders may distort the association between exposure and outcome. This limitation is particularly important when studying treatments, because patients who receive different treatments may differ in ways that are not fully captured in the data.

Measurement Error

All measurements contain some error. In prospective studies, measurement error can arise from imperfect instruments, inconsistent data collection, or participant recall. In retrospective studies, measurement error can arise from incomplete or inaccurate records. Nondifferential measurement error tends to bias associations toward the null, while differential measurement error can bias associations in either direction.

Generalizability

Cohort studies are conducted in specific populations, and the results may not generalize to other populations. For example, a cohort study conducted in a single academic medical center may not represent patients treated in community settings. A cohort study conducted in one country may not represent populations with different genetic backgrounds, environmental exposures, or healthcare systems.

Changing Definitions and Practices

Over long follow-up periods, diagnostic criteria, treatment practices, and outcome definitions may change. These changes can complicate the interpretation of results. For example, a study of melanoma recurrence that spans 2008 to 2021 must account for changes in staging systems and treatment approaches over that period.

Practical Steps for Conducting a Cohort Study

Whether you choose a prospective or retrospective design, the following steps provide a framework for conducting a rigorous cohort study.

Step 1: Define the Research Question

State the research question clearly, including the population, exposure, comparison, and outcome. Use the PICOT framework if appropriate. Write the question in a way that specifies the temporal relationship between exposure and outcome.

Step 2: Review Existing Evidence

Search the literature to understand what is already known and to identify gaps that your study will address. Use PubMed from the National Library of Medicine to search for relevant studies. Review systematic reviews and meta-analyses to understand the current state of evidence.

Step 3: Select the Study Design

Based on your research question, resources, and the availability of existing data, decide whether a prospective or retrospective cohort design is appropriate. Consider the tradeoffs in data quality, cost, time, and bias described in this article.

Step 4: Define the Study Population

Specify the inclusion and exclusion criteria. Define the source population and the sampling frame. For retrospective studies, document how participants will be identified from existing records.

Step 5: Define Exposure and Outcome Variables

Provide precise definitions for all exposure and outcome variables. Specify how each variable will be measured, when it will be measured, and by whom. For retrospective studies, document the source of each variable and the procedures for extracting data.

Step 6: Develop the Analysis Plan

Specify the statistical methods you will use to estimate the association between exposure and outcome. Identify potential confounders and effect modifiers. Describe how you will handle missing data and loss to follow-up.

Step 7: Obtain Ethical Approval

Submit the study protocol to the appropriate ethics committee. For prospective studies, prepare informed consent materials. For retrospective studies, document the justification for any waiver of consent.

Step 8: Collect and Manage Data

Implement the data collection procedures described in the protocol. Maintain quality control throughout the study. Document all data cleaning decisions.

Step 9: Analyze the Data

Execute the analysis plan. Conduct sensitivity analyses to assess the robustness of your findings. Consider alternative explanations for the observed associations.

Step 10: Report the Results

Report the study according to the appropriate reporting guideline. The EQUATOR Network provides a comprehensive collection of reporting guidelines for health research. Transparent reporting allows readers to assess the validity of your findings and to compare them with other studies.

Frequently Asked Questions

What is the main difference between a prospective and a retrospective cohort study?

The main difference is the timing of the study relative to the exposure and outcome. In a prospective cohort study, you enroll participants before outcomes occur and follow them forward in time. In a retrospective cohort study, you identify participants after both exposure and outcome have occurred and reconstruct the timeline from existing records.

Which type of cohort study is more susceptible to bias?

Retrospective cohort studies are generally more susceptible to information bias because they rely on data collected for other purposes. Prospective cohort studies are more susceptible to selection bias from loss to follow-up. Both designs are susceptible to confounding.

Can a retrospective cohort study establish causality?

A retrospective cohort study can provide evidence of an association and can establish that exposure occurred before outcome, which is necessary for causality. However, retrospective studies are more limited than prospective studies in their ability to control confounding and measurement error, so causal claims should be made cautiously.

How long does a prospective cohort study take?

The duration depends on the latency between exposure and outcome. Some prospective studies follow participants for a few months, while others follow them for decades. The study of hyaluronic acid and bone resorption followed patients for 6 to 12 months, while the Danish melanoma study followed patients for a median of 5.9 years.

What are the advantages of a retrospective cohort study?

Retrospective cohort studies are faster, less expensive, and can include very large sample sizes using existing databases or registries. They are useful for studying outcomes with long latency periods and for conducting preliminary investigations before committing to a prospective study.

What are the advantages of a prospective cohort study?

Prospective cohort studies allow standardized measurement of exposures and outcomes, minimize recall bias, and provide stronger evidence for causal relationships. They are preferable when you need biological samples collected under controlled conditions or when outcomes require active surveillance to detect.

Can a study combine prospective and retrospective designs?

Yes, combined designs are increasingly common. A study might use a retrospective cohort to identify candidate biomarkers and a prospective cohort to validate them, as was done in the study of symptomatic hemorrhagic transformation in stroke patients. Combined designs can also be used when a retrospective cohort is expanded with prospective enrollment.

How do I choose between a prospective and retrospective cohort study?

Consider your research question, the availability of existing data, the latency between exposure and outcome, your resources, and the importance of standardized measurement. If reliable data already exist and you need answers quickly, a retrospective design may be appropriate. If you need precise measurement and can support long-term follow-up, a prospective design is preferable.

Related Articles

• Difference Between Dna And Rna
• Difference Between Dna And Rna
• Difference Between Dna And Rna
• Bioinformatics vs. Computational Biology: What's the Difference (and Which Should You Choose)?
• biology study guide

References and Further Reading

• Research Data Framework. National Institute of Standards and Technology.
• EQUATOR Network. EQUATOR Network.
• Experimental Design Assistant. NC3Rs.
• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.
• Association of triglyceride-glucose index trajectory and frailty in urban older residents: evidence from the 10-year follow-up in a cohort study.. Cardiovascular diabetology, 2023.
• Stage-Specific Risk of Recurrence and Death From Melanoma in Denmark, 2008-2021: A National Observational Cohort Study of 25 720 Patients With Stage IA to IV Melanoma.. JAMA dermatology, 2023.
• Active Cerebrospinal Fluid Exchange vs External Ventricular Drainage in the Neurocritical Care Unit: An International, Retrospective Cohort Study.. Neurosurgery, 2025.
• Would hyaluronic acid-induced mental bone resorption be a concern? A prospective controlled cohort study and an updated retrospective cohort study.. International journal of surgery (London, England), 2024.
• Between evidence and expectation: a retrospective cohort study on chronic endometritis in repeated implantation failure and recurrent pregnancy loss.. Reproductive biology and endocrinology : RB&E, 2025.
• The Association of Serum Biomarkers With Symptomatic Hemorrhagic Transformation in Acute Ischemic Stroke Patients: A Combined Retrospective and Prospective Study.. CNS neuroscience & therapeutics, 2025.
• Prospective cohort study of parathyroid function and quality of life after total thyroidectomy for thyroid cancer: robotic surgery vs. open surgery.. International journal of surgery (London, England), 2023.
• Association between stress hyperglycemia ratio and delirium risk in elderly surgical patients: a retrospective cohort study.. BMC geriatrics, 2025.
• Clinical Assessment of Thromboembolic Risk in Patients Undergoing Elective Electrical Cardioversion with or Without Transesophageal Echocardiography: A Real-World Observational Study.. 2026.
• Hepatic safety of sintilimab versus pembrolizumab in advanced non-small cell lung cancer: a retrospective observational cohort study.. 2026.
• Incidental pulmonary nodules in oral squamous cell carcinoma - A retrospective cohort study.. 2026.
• Superior cancer detection with transperineal biopsy in patients with an elevated prostate-specific antigen level and a PI-RADS score of 3-4: A retrospective cohort study.. 2026.
• Morphologic patterns of clinically negative cervical lymph nodes and their association with tumor-induced lymphangiogenesis in Oral Squamous Cell Carcinoma: A retrospective cohort study.. 2026.
• D-dimer and lower limb ultrasound as prognostic factors for recurrent deep venous thrombosis and pulmonary embolism: A systematic review and meta-analysis.. 2026.
• Determinants of First-Trimester Residual Myometrial Thickness After Previous Cesarean Delivery: A Retrospective Cohort Study.. 2026.
• A comparison of the results of prospective and retrospective cohort studies in the field of digestive surgery. Surgery Today, 2017.
• Automatic continuous control of cuff pressure and subglottic secretion suction used together to prevent pneumonia in ventilated patients-a retrospective and prospective cohort study. Journal of Clinical Medicine, 2021.
• Instantaneous Right Ventricular to Pulmonary Artery Systolic Pressure Difference in Cardiac Surgery: A Retrospective and Prospective Cohort Study. Canadian Journal of Cardiology, 2025.
• A Prospective-Retrospective Observational Cohort Study of Short-Term Health Outcomes of Preterm Very Low Birth Weight Infants Receiving Oropharyngeal Administration of Own Mother's Colostrum. Current Therapeutic Research Clinical and Experimental, 2026.
• Double-layer versus conventional mucoperiosteal flap closure in the surgical treatment of medication-related osteonecrosis of the jaw: A combined Retrospective-Prospective cohort study. Journal of Cranio Maxillofacial Surgery, 2026.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.