How to Write the Results Chapter of a Medical Thesis

The Results chapter is where your thesis stops planning and starts answering. Examiners turn to it to see whether you actually answered every objective, whether the numbers add up, and whether you understand what your statistics mean. A well-organised Results chapter makes the Discussion easy to write; a disorganised one, full of SPSS output pasted as it came, makes even good data look weak.

This guide covers the order that works for most MD, MS and DNB theses, rules for tables and figures, how to report numbers correctly (n (%), mean ± SD, median (IQR), effect sizes with 95% confidence intervals and exact p-values), how to choose the right graph, and what does not belong in Results. Your university may prescribe its own format; follow it where it differs.

1What the Results Chapter Must Do

The Results chapter presents your findings objectively, in a logical order, without interpretation. Think of it as a structured report: what you found, how many, how much, and how certain. Three principles guide everything else:

• Answer every objective. Each objective in your synopsis should have at least one table or figure that answers it.
• Follow your analysis plan. The tests you report should be the ones described in your Methods chapter. If you changed something, explain it in Methods.
• Every number must be traceable. Totals in each table should match your sample size or explain the difference (missing data, losses to follow-up).

If you have not yet chosen your tests, start with our guide to statistical analysis for a medical thesis.

2The Order That Works

Step 1: Participant flow

Begin with how many people were screened, how many were eligible, how many were enrolled, and how many were analysed, with reasons for exclusion and loss at each step. For a randomised trial, a CONSORT-style flow diagram is the clearest way to show this; for observational studies, a STROBE-style flow description or diagram works well. Even a cross-sectional study benefits from one sentence: "Of 214 patients screened, 180 met the eligibility criteria and were included."

Step 2: Baseline characteristics

Next, describe your participants: age, sex, and the key clinical and demographic variables. In a comparative study, present the groups side by side in one table (often called Table 1). In a randomised trial, many reporting guidelines advise against testing baseline differences with p-values, because any differences are due to chance; describe them instead. Follow your guide's preference, but know the reasoning.

Step 3: Objective-wise results

Then present results in the same order as your objectives: primary objective first, then secondary objectives. Use subheadings that echo each objective, so the examiner can find the answer to Objective 2 in seconds.

3Table Rules

• Number and title every table (Table 1, Table 2...). The title goes above the table and should say what, who and, where relevant, when: "Table 3: Comparison of postoperative pain scores between groups (n = 120)".
• State units and summary measures in column headings, for example "Age (years), mean ± SD".
• Keep decimals consistent: usually one decimal place for percentages and means of clinical measures, more only when the measure demands it.
• Define abbreviations in a footnote below the table, and name the statistical test used for each p-value.
• Do not paste SPSS output. Rebuild the table in Word with only the numbers the reader needs. See our SPSS guide for where these numbers come from.
• Use a clean format: many universities prefer horizontal lines only (top, below the header, bottom), without vertical gridlines.

In the text, do not repeat every number from a table. Point to the table and highlight the key findings: "Mean pain scores at 6 hours were lower in Group A than in Group B (Table 3)."

Writing the text around a table

A useful pattern for each table is three sentences: what the table shows, the main finding with its key numbers, and any secondary finding worth noting. For example: "Table 4 shows the distribution of anaemia by severity. Moderate anaemia was the commonest category (58 of 150 participants, 38.7%). Severe anaemia was more frequent in participants from rural areas than in those from urban areas (Table 4)." Keep the tense consistent; most theses describe results in the past tense.

Missing data

If some participants have missing values for a variable, say so. Give the denominator in the table ("n = 172 with complete data") or in a footnote, and state whether missing values were excluded or handled another way, as described in your Methods. Silent changes in denominators from one table to the next are a common source of examiner queries.

4Reporting Numbers Correctly

Choose the summary measure according to the type of variable and its distribution:

Common slips

• Reporting mean ± SD for a variable that is clearly skewed (for example, when the SD is larger than the mean for a value that cannot be negative).
• Giving percentages without the numerator and denominator.
• Using SEM instead of SD to describe variability among participants. SD describes spread; SEM describes precision of the mean.
• Percentages that do not add up to 100% without a note explaining rounding or missing data.

5p-Values, Effect Sizes and 95% Confidence Intervals

A p-value alone tells the reader very little. Report three things for each main comparison:

1. The effect size: the difference in means or medians, the risk difference, relative risk, odds ratio, hazard ratio or correlation coefficient, as appropriate.
2. Its 95% confidence interval: the range of values compatible with your data.
3. The exact p-value: write p = 0.032, not p < 0.05. For very small values, write p < 0.001.

Example of a complete sentence: "The proportion of patients with postoperative nausea was lower in the ondansetron group (12/60, 20.0%) than in the placebo group (24/60, 40.0%); risk difference 20.0 percentage points (95% CI 4.0 to 36.0), p = 0.017." (Numbers here are illustrative only.)

For a deeper explanation of what p-values do and do not mean, read our post on understanding p-values in medical research.

6Choosing the Right Graph

Use figures when they show a pattern more clearly than a table, not to duplicate a table. Never show the same data in both a table and a graph unless your university requires it.

Number figures separately from tables (Figure 1, Figure 2...), put the title below the figure, label both axes with units, and make sure the figure is readable when printed in black and white.

7What Does Not Belong in Results

• Interpretation and explanation: "This may be due to better pain control..." belongs in the Discussion.
• Comparison with other studies: also Discussion material; see our guide to writing the Discussion chapter.
• Methods details: descriptions of how tests were done belong in Material and Methods.
• Claims beyond the data: "Drug A is superior" when the confidence interval includes no difference.
• Raw data: the master chart goes in the annexures, not in the chapter.
• Outcomes not in your plan, presented as if they were planned. Exploratory analyses can be reported, but label them as exploratory.

8Final Checklist Before Submission

Finally, read the chapter once against your master chart, checking a few numbers by hand. Errors that examiners spot in Results damage confidence in the entire thesis.

9Getting Help with Your Results Chapter

If your analysis is done but the chapter is not taking shape, or you want your tables, graphs and reporting checked before your guide sees them, the PubMedico team can help organise objective-wise results and present them clearly and correctly. Message us on WhatsApp at +91 78787 33004.