Everything you need to go from "I've never done research" to a submitted thesis. Written the way a good senior would explain it โ plain language, India-specific, no jargon.
Medical research sounds intimidating. It isn't. At its core, research is just answering a question you don't know the answer to โ in a structured, honest, repeatable way.
Here's an example. You operate on a lot of acute cholecystitis. You notice that the patients you operate on early seem to do better than the ones who get operated on after 3 days. Is that actually true? Or is it just your impression? Research is how you find out.
Your thesis exists for two reasons. The official reason: it's a requirement for your MS or DNB degree. The real reason: Indian surgery generates thousands of patients every year whose outcomes are never studied, never published, never used to improve care. Your thesis can change that.
Don't worry about making it "groundbreaking." A well-conducted study that answers one clinical question clearly, honestly, is more valuable than a complicated study done badly.
Every thesis belongs to one of four study design types. Choosing the right one for your question is the single most important decision you'll make. Here's what they mean in plain English:
You look backwards at records that already exist. "Of all patients operated for incisional hernia in the last 5 years, what was the recurrence rate?" Quickest to complete. No ethics approval for patient contact needed. Best for first-time researchers.
You watch patients going forward in time without changing anything. "I will follow all acute pancreatitis admissions for 6 months and record BISAP scores and outcomes." More work, but higher quality evidence. Needs ethics approval.
You randomly allocate patients to two different treatments and compare outcomes. Highest quality evidence. Also the hardest to do. Requires careful ethics approval and randomisation protocol. Good if your unit has high volume.
You test how good a test or score is at diagnosing something. "How accurate is FAST exam compared to CT scan in blunt abdominal trauma?" Needs sensitivity, specificity, AUC analysis. Very publishable.
Ask yourself: "Do I already have this data?" โ Retrospective. "Will I collect this data going forward?" โ Prospective. "Am I allocating treatment randomly?" โ RCT. "Am I testing how good a test is?" โ Diagnostic accuracy.
Every good research question has four parts. PICO helps you structure yours:
A bad research question: "Is laparoscopic surgery good?" โ too vague, can't be studied.
A good research question using PICO: "In adults with acute cholecystitis (P), does early laparoscopic cholecystectomy within 72 hours of admission (I) compared to delayed surgery after 72 hours (C) result in lower complication rates and shorter hospital stay (O)?"
Your thesis title should essentially be a PICO question with the "?" removed. Write your PICO before anything else. Show it to your supervisor before writing a single word of your protocol.
In India, the Institutional Ethics Committee (IEC) is mandatory for any research involving human subjects. Many residents fear this step. You shouldn't โ the committee exists to help ensure your study is ethical, not to stop you.
Any study involving patients โ including retrospective chart reviews that access identifiable patient data. Some purely anonymised retrospective audits may be exempt, but always check with your institution. If in doubt, apply.
Step 1: Submit your protocol (your synopsis) to the IEC. Most institutions have a submission form โ get it from your PG office.
Step 2: IEC meets monthly (usually). You may be called to present your study in 3โ5 minutes.
Step 3: They approve, ask for modifications, or reject. Most first submissions come back with minor modification requests โ this is normal.
Step 4: You get a written approval letter. Keep this โ you need it for thesis submission and publication.
For retrospective studies, also include a waiver of consent request โ explain why collecting consent retrospectively is not feasible and how you will protect patient privacy.
The synopsis is a short document (usually 8โ15 pages) that describes your planned study. It is submitted to the IEC and to your university for thesis registration. Think of it as a map of your research โ you write it before you start collecting data.
Most residents get stuck here because nobody explains what each section actually means. Here it is:
Your PICO question, written as a statement. Include: study population, intervention/comparison, outcome, study design. Example: "Comparison of Early vs Delayed Laparoscopic Cholecystectomy in Acute Cholecystitis: A Prospective Randomised Study."
Tell the story of the problem. Start broad (what is the condition, how common is it globally/India), then narrow (what is the controversy or knowledge gap), then arrive at your question (why this specific comparison needs to be studied). End with: "Hence this study is planned to..."
Aim: One sentence. The overall purpose. "To compare outcomes of early vs delayed laparoscopic cholecystectomy in acute cholecystitis."
Objectives: 3โ5 specific, measurable things you will measure. Each should be answerable with a number.
Summarise 8โ15 key papers that are relevant to your topic. Show what is known, what is contested, and where the gap is. Do NOT copy-paste abstracts. Write 2โ3 sentences per paper in your own words.
The most important section. Must include: study design, setting, duration, inclusion criteria, exclusion criteria, sample size with justification, data collection method, outcome measures, statistical plan.
What software will you use (SPSS/MedCalc/R)? What tests will you use? Continuous data: mean ยฑ SD, t-test. Categorical data: frequency/%, chi-square. Survival: Kaplan-Meier. P < 0.05 = significant.
Most residents see statistics and panic. You don't need to understand everything โ you need to understand enough to apply the right test and interpret the output. Here is that minimum.
The p-value answers one question: "If there were truly no difference between my two groups, how likely is it that I would see a difference this large just by chance?" A p-value of 0.03 means: only 3% chance this result happened by coincidence. We conventionally accept p < 0.05 as "statistically significant."
Comparing two groups on a number (e.g., hospital stay in Group A vs Group B): if data is normally distributed โ independent t-test. If not normal โ Mann-Whitney U test.
Comparing two groups on a category (e.g., complication rate: yes/no): Chi-square test. If any cell <5: Fisher's exact test.
Correlation (does X increase as Y increases?): Pearson's r (normal data) or Spearman's rho (non-normal).
Diagnostic test: Sensitivity, Specificity, AUC (ROC curve).
MedCalc (free trial, most surgical journals use it) ยท JASP (free, simple interface) ยท R (free, powerful, steep learning curve) ยท SPSS (your institution probably has a licence)
Your proforma is the form you fill for every patient you enrol. It is the most practical tool in your entire thesis. A badly designed proforma will ruin your study. A good one makes analysis effortless.
1. Include only what you will actually analyse. Every variable you collect takes time. If you collect 60 variables and only use 12, you've wasted time and introduced data quality issues.
2. Pre-code categorical variables. Don't write "male/female" โ write 1/2. Don't write "yes/no" โ write 1/0. This makes SPSS entry instant.
3. Define every variable in advance. What exactly is "early surgery"? Within 72 hours of admission? From symptom onset? Define it before your first patient, not after 30.
4. Include a column for date of follow-up. You will forget otherwise.
Missing data is the most common reason theses fail at the analysis stage. If more than 20% of your data for any variable is missing, that variable becomes unreliable. Use your proforma as a checklist โ fill it during the patient's admission, not from memory later.
Your thesis does not have to end at submission. Most good theses contain at least one publishable paper. Publishing is not just for career benefit โ it's how your research actually reaches other surgeons and improves patient care.
A thesis chapter becomes a paper by cutting it down: Introduction (500 words) โ Methods (500 words) โ Results (500 words) โ Discussion (1000 words). The thesis chapter may be 15 pages; the paper is 4. You're not throwing anything away โ you're distilling it.
Start realistic: Indian Journal of Surgery, Surgery Today, World Journal of Surgery, Indian Journal of Gastroenterology. These are indexed, respected, and appropriate for first publications.
Check before submitting: Is the journal indexed in PubMed/Scopus? Avoid predatory journals โ if they charge a large fee upfront and promise quick publication, be cautious. Check the BEALL list.
Clear question. Appropriate methodology. Honest limitations. Clinical relevance. Correct statistics. A paper that clearly answers one question well is more publishable than one that answers five questions vaguely.