How to Write the Review of Literature for an MD/MS Thesis

The Review of Literature (ROL) is usually the longest chapter of an MD, MS or DNB thesis, and it is often the one written in the biggest hurry. Many residents collect a folder of PDFs, copy the abstracts into Word and call it a review. Examiners notice. A good ROL shows that you understand what is already known about your topic, where the gaps are, and why your study is worth doing.

This guide walks you through the whole process: searching the literature properly, screening and organising papers in an evidence matrix, structuring the chapter, summarising each study in three or four sentences, and writing in your own words. Follow the steps in order and the chapter becomes far less overwhelming.

1What the Review of Literature Must Do

Before you search anything, be clear about the job of this chapter. A thesis ROL is not a textbook chapter and not a list of abstracts. It should:

• Set the background – history of the disease or procedure, and the basic science your study depends on.
• Show what previous studies found – especially studies with a similar design, population or outcome to yours.
• Compare and contrast – where do studies agree, where do they disagree, and why might that be (sample, setting, method, definitions)?
• Identify the gap – for example, few studies from India, small samples, a different age group, or no data on the outcome you are measuring.
• Feed the Discussion – the studies you review here are the studies you will compare your results with later.

Keep your aims and objectives open while you work. Every section of the ROL should connect, directly or indirectly, to a variable or outcome in your study. If you have not finalised these yet, read our guide on writing aims and objectives first.

2Searching the Literature: PubMed, MeSH and Google Scholar

Start with PubMed

PubMed is the main database for biomedical literature and should be your first stop. Break your topic into concepts, usually population, exposure or intervention, and outcome. For a study on serum vitamin D levels in children with recurrent wheeze, the concepts are “children”, “vitamin D” and “wheeze”.

For each concept, combine free-text words with the matching MeSH term. MeSH (Medical Subject Headings) is the controlled vocabulary NLM uses to index articles, so a MeSH search finds papers even when authors used different words. Use the MeSH Database inside PubMed to find the right heading, then build a search like this:

("Vitamin D"[MeSH] OR "vitamin D" OR cholecalciferol OR "25-hydroxyvitamin D") AND ("Respiratory Sounds"[MeSH] OR wheez*) AND (child* OR paediatric OR pediatric)

• OR joins synonyms within a concept; AND joins different concepts.
• The asterisk (wheez*) truncates a word to catch wheeze, wheezing and wheezy.
• Include both British and American spellings (paediatric/pediatric, haemoglobin/hemoglobin, anaemia/anemia).
• Use filters for article type, species (humans) and language only after you see the unfiltered result count.

Save your final search string and the date you ran it; some guides and examiners ask how the literature was searched. For a detailed walkthrough of filters, clipboard and saved searches, see how to search the PubMed database.

Add Google Scholar and other sources

Google Scholar picks up papers from Indian journals, conference proceedings and theses that PubMed may miss. Its “Cited by” link is very useful: open a key older paper and see who has cited it since. Also check the Cochrane Library for systematic reviews on your topic, and the reference lists of the most relevant papers you find (backward searching).

3Screening and Organising Papers in an Evidence Matrix

A search can return hundreds of results. Screen them in two rounds. First, read titles and abstracts and keep only papers relevant to your population, exposure and outcome. Second, get the full text of the shortlisted papers and decide which ones you will actually use.

As you read each full-text paper, enter it into an evidence matrix (also called a summary table or literature grid) in Excel or Google Sheets. One row per study, one column per feature. A practical set of columns:

The matrix does three things for you. It forces you to read each paper properly, it lets you sort studies by theme, design or year when you write, and it becomes your ready reckoner when you write the Discussion months later.

At the same time, import every paper into a reference manager such as Zotero, Mendeley or EndNote. Doing this from day one saves days of reformatting at the end. Our Vancouver referencing guide covers setting this up.

4Structuring the Chapter

Most Indian universities expect the ROL to move from general to specific. Check your university’s thesis guidelines and a few recently accepted theses from your department, but a structure like this works for most clinical topics:

1. Historical background – a short account of how the disease, test or procedure was described and developed. Half a page to a page is usually enough.
2. Basic science, as relevant – anatomy for surgical and radiology topics, physiology or pharmacology for anaesthesia and medicine topics, pathology and pathogenesis for disease-based topics. Include only what your study needs; do not reproduce a full textbook chapter.
3. Clinical aspects – epidemiology, clinical features, diagnosis, classification or management, again limited to what relates to your variables.
4. Previous studies, grouped by theme – the core of the chapter. Group studies under sub-headings that match your objectives, for example “Studies on prevalence”, “Studies on association with severity”, “Studies comparing technique A and technique B”.
5. Summary and gap – a short closing paragraph stating what is known, what is not, and how your study addresses it.

Within each theme, arrange studies in a logical order: chronologically to show how understanding developed, or international first and then Indian studies. Whatever you choose, be consistent.

5How to Summarise a Study in 3–4 Sentences

Each study in the “previous studies” section needs a short, consistent summary. Three or four sentences is enough for most papers. Use this template and fill it from your evidence matrix:

1. Who, where and what design: “[Author] et al. ([Year]) conducted a [design] study at [setting, city/country] among [number] [population] to [aim].”
2. Method in brief: “[Exposure/intervention] was assessed using [method/tool], and [outcome] was defined as [definition].”
3. Main result with numbers: “They found that [result, with the figure, p-value or confidence interval as reported in the paper].”
4. Conclusion or relevance (optional): “The authors concluded that [conclusion]; however, the study was limited by [limitation].”

Two practical points. First, always copy the numbers from the full text, never from memory or from another thesis, and cite the original paper. Second, vary your opening verbs so the chapter does not read like a form: reported, observed, evaluated, compared, assessed, found, demonstrated. Use “demonstrated” or “proved” only for strong designs; for a small cross-sectional study, “reported” or “observed” is more accurate.

Not every study deserves equal space. A large, well-designed study that closely matches yours can get a full paragraph. Several smaller studies with similar findings can be combined into one sentence with multiple citations.

6Paraphrasing vs Copying

The ROL is where most similarity-index problems come from, because residents paste sentences from abstracts. Most universities now run theses through software such as Turnitin, iThenticate or DrillBit, and the UGC (Promotion of Academic Integrity and Prevention of Plagiarism in Higher Educational Institutions) Regulations 2018 set out levels of plagiarism and the consequences. Your university may have its own acceptable limit, so check it early. Our post on plagiarism checks and similarity index explains how reports are read.

7How Many References and How Long?

There is no single national rule. The expected number of references and the length of the ROL vary by university, department and topic. Some universities specify page limits for each chapter in their thesis guidelines; others leave it to the guide. Ask your guide, read your university’s dissertation regulations, and look at two or three recently accepted theses from your department.

Quality matters more than count. A useful approach:

• Prefer recent studies, but keep the landmark older papers that defined the field.
• Include Indian studies wherever they exist; examiners often ask how your findings compare with Indian data.
• Cite primary sources rather than textbooks for specific findings. Textbooks are fine for the historical and basic-science sections.
• Cite only what you have actually read. Every reference should appear in the text, and every citation in the text should be in the reference list.

Before submission, run through this checklist: every sub-heading links to an objective; each theme ends with a synthesis sentence; numbers match the source papers; the chapter closes with a clear gap statement; references follow your university’s style consistently.

8Getting Help with Your Review of Literature

A well-built ROL makes the rest of the thesis easier, especially the Discussion, which reuses the same studies. If you want a second pair of eyes on your search strategy, evidence matrix or draft chapter, the PubMedico team works with MD, MS and DNB residents on exactly these steps. Message us on WhatsApp at +91 78787 33004. For the bigger picture, see our step-by-step MD/DNB thesis guide.