Details have been modified to protect client confidentiality. This case study represents a composite of common research scenarios encountered by Research Gold.
A systematic review case study illustrating the complete journey from clinical question to journal acceptance. Dr. A., an internal medicine resident preparing for fellowship applications, needed a first-author publication to strengthen their academic portfolio. They had a compelling clinical question about blood pressure management in elderly patients with diabetes but no experience conducting systematic reviews or meta-analyses.
Dr. A. observed conflicting guidance in their hospital's hypertension protocols for diabetic patients over 65. Some attending physicians advocated aggressive targets (below 130/80 mmHg), while others preferred conservative management (below 140/90 mmHg), citing hypotension risks. The clinical uncertainty was genuine and well-suited to evidence synthesis.
However, Dr. A. had never:
- Registered a protocol on PROSPERO
- Developed a multi-database search strategy
- Used a screening tool like Covidence or Rayyan
- Conducted risk of bias assessment with RoB 2 online assessment calculator
- Run a meta-analysis in R or Stata
- Prepared a PRISMA-compliant manuscript
Their fellowship application deadline was 16 weeks away. The typical systematic review timeline of 67 weeks was not an option.
Week 1-2: Protocol Development and Registration
The first step was transforming the clinical observation into a structured research question using the PICO framework:
Population: Adults aged 65+ with type 2 diabetes and hypertension
Intervention: Intensive blood pressure targets (below 130/80 mmHg)
Comparison: Standard blood pressure targets (below 140/90 mmHg)
Outcomes: All-cause mortality, cardiovascular events, serious adverse events, renal outcomes
The protocol was drafted using the PROSPERO registration formatter, specifying planned databases, search strategy outline, eligibility criteria, risk of bias tools (RoB 2 for RCTs), planned subgroup analyses (by diabetes duration, baseline blood pressure, and follow-up length), and the meta-analytic model (random-effects, DerSimonian and Laird).
PROSPERO registration was confirmed within 5 business days. This time-stamped public record established the a priori methodology, a requirement for the target Q2 cardiology journal.
A systematic search strategy was developed across five databases: PubMed/MEDLINE, Embase, Cochrane CENTRAL, CINAHL, and ClinicalTrials.gov. The strategy combined MeSH terms and free-text terms for hypertension, diabetes mellitus, blood pressure targets, elderly/aged, and randomized controlled trials.
Using our search strategy builder to structure the Boolean logic, the search retrieved:
- PubMed: 1,847 records
- Embase: 2,103 records
- CENTRAL: 412 records
- CINAHL: 289 records
- ClinicalTrials.gov: 196 records
After deduplication, 2,847 unique records entered the screening phase.
Week 5-7: Screening and Study Selection
Two independent reviewers screened all 2,847 titles and abstracts against the pre-defined eligibility criteria. Inter-rater agreement was strong (Cohen's kappa = 0.82).
- Title/abstract screening: 2,847 records screened, 127 passed to full-text review
- Full-text assessment: 127 articles reviewed, 23 randomized controlled trials meeting all eligibility criteria included
- Exclusion reasons documented for all 104 excluded full-texts
The PRISMA flow chart generator tool documented each stage with precise numbers, ready for manuscript inclusion.
Week 8-10: Data Extraction and Risk of Bias
Two reviewers independently extracted data from all 23 included studies using a standardized extraction form. Extracted variables included: sample sizes per arm, mean blood pressure achieved, follow-up duration, event counts for mortality and cardiovascular endpoints, and adverse event rates.
Risk of bias was assessed using the Cochrane RoB 2 tool across five domains. Of the 23 studies, 9 were rated low risk of bias overall, 11 had some concerns (primarily in the blinding domain, as blood pressure targets are difficult to blind), and 3 were rated high risk.
Dr. A. participated in the data extraction process, verifying clinical details and providing domain expertise on medication classes and blood pressure measurement protocols. This collaboration ensured both methodological rigor and clinical accuracy.
Meta-analysis was conducted in R using the metafor package. The primary analysis pooled all-cause mortality across 23 studies using a random-effects model (REML estimator), reflecting expected between-study heterogeneity.
Key results:
- All-cause mortality: OR 0.87 (95% CI: 0.76 to 0.99), favoring intensive targets, I-squared = 41%
- Cardiovascular events: OR 0.79 (95% CI: 0.71 to 0.88), I-squared = 28%
- Serious adverse events: OR 1.42 (95% CI: 1.18 to 1.71), higher with intensive targets, I-squared = 52%
- Renal outcomes: No significant difference, OR 0.94 (95% CI: 0.82 to 1.08)
Forest plots were generated for each outcome. Subgroup analyses examined effects by diabetes duration (less than 10 years vs. 10+ years) and baseline blood pressure level, revealing that the mortality benefit was concentrated in patients with diabetes duration under 10 years.
Sensitivity analyses tested robustness: leave-one-out analysis, exclusion of the 3 high risk-of-bias studies, and comparison of random-effects vs. fixed-effect models. Results were consistent across all sensitivity analyses.
Publication bias assessment using funnel plots and Egger's regression showed no significant asymmetry (p = 0.34), supported by trim-and-fill analysis showing no missing studies.
GRADE summary of findings tables rated all-cause mortality as moderate certainty (downgraded one level for serious risk of bias in some studies) and cardiovascular events as high certainty.
This is the kind of project Research Gold handles regularly. Our systematic review and meta-analysis services take researchers from clinical question to publication-ready manuscript. get a tailored research quote with your research question.
Week 13-14: Manuscript Preparation and Submission
The manuscript was prepared following the target journal's author guidelines and complete PRISMA 2020 checklist. Supplementary materials included:
- Full search strategies for all five databases
- List of excluded studies with reasons
- Forest plots for all outcomes including subgroup analyses
- Funnel plots and bias assessment results
- GRADE evidence profile tables
- Annotated R analysis code
Dr. A. wrote the clinical interpretation sections of the discussion, placing the findings in the context of current hypertension guidelines (AHA/ACC, ESC/ESH) and their clinical experience.
The manuscript was submitted to a Q2 cardiology journal at week 14. Peer review returned at week 20 with a decision of minor revisions. Reviewer comments were addressed within one week (clarifying the subgroup analysis rationale and adding a requested sensitivity analysis). The manuscript was accepted for publication at week 22.
Dr. A. included this first-author publication on their fellowship application, which was submitted on time. The review has since been cited 12 times in the first 8 months post-publication.
- A focused PICO question made the review manageable. Dr. A.'s initial question was too broad; narrowing to a specific population and comparison made it publishable.
- Professional support compressed the timeline from the typical 67 weeks to 14 weeks without compromising methodology.
- The researcher retained full authorship and intellectual ownership. Dr. A. conceived the question, directed clinical decisions, wrote interpretation sections, and approved the final manuscript.
- PROSPERO registration was essential for journal submission, establishing methodological credibility.
- Fellowship applications benefit from a published systematic review demonstrating research capability and clinical relevance.
Explore our systematic review services or request a free consultation to discuss your research project. View transparent pricing and see how the process works.
Inspired by this case study? Our step-by-step guide explains how to complete a systematic review during residency alongside clinical duties.
Not sure what to study? Browse 30 research project ideas organized by specialty for inspiration.
If fellowship is your goal, find out how many publications fellowship programs expect and how to reach that target.