Week 1 · Module 1
Where is your English actually weak?
A 24-question diagnostic that finds the specific error patterns costing you interviews and manuscript acceptances — and builds your personal study order for the next four weeks
Your scenario
You are taking the pre-employment editing assessment at a top tier-1 CRO in Mumbai. The test consists of an unedited draft of an international multi-center oncology trial report. The clinical science is robust, but the draft contains subtle grammatical, tense, prenominal modifier, and register errors. The hiring lead tells you: "We don't expect you to be a Shakespeare scholar; we need to know if you can spot the 12 classic error patterns before our global sponsor sees them." This 24-question diagnostic assesses your baseline across all 12 domains and builds your personalized 4-week study curriculum.
Read this before you start
This is not an English test. It produces no grade and it is not a judgement of your intelligence or your science. It is a diagnostic instrument — the same logic as a lab panel. It tells you which of twelve specific error classes you produce, so you spend four weeks on the three or four that actually hold you back instead of re-reading rules you already follow correctly.
Almost every medical writer trained in India makes errors from a small, highly predictable set. There are twelve classes. Most people have serious trouble in three or four and are fine in the rest. Studying all twelve equally is a waste of your time.
How to take it honestly
Do not look anything up. Do not ask anyone. A wrong answer here is worth more to you than a right one, because it is the only way the diagnostic can find the gap. If you cheat this test you will study the wrong units for four weeks.
—
—
Your twelve areas — study the red ones first
Before you move on
Write down your three weakest areas, in order. Come back to this box at the end of Week 4 and note what changed.
People who write this down finish the course at roughly twice the rate of people who do not.
Next module
1.2 — Articles: a, an, the, and nothing
Week 1 · Module 2
Articles: a, an, the, and nothing
The single largest error class in Indian scientific English — and the one that never clears by itself, because your first language gave you nothing to transfer
Your scenario
A global regulatory affairs specialist reviews your Investigator's Brochure draft for an upcoming US FDA Investigational New Drug (IND) filing. She leaves a note on Section 4: "Your safety warnings drop articles in generic plurals and add extraneous definite articles before general disease names. In regulatory documents, imprecise article use creates ambiguity about whether an adverse event applies to all patients or a specific subset." This module builds the 3-question mental model you need to eliminate article errors permanently.
Why you make this error
Telugu, Tamil, Kannada, Malayalam, Hindi, Bengali, Marathi and Gujarati have no article system. Definiteness is carried by word order, case marking, or context. When you write English there is nothing in your first language to transfer — you are not being careless, you are working without a template.
This is why article errors survive long after every other error in your writing is fixed, and why reading more English does not repair them. Reading builds recognition. Articles are a production skill. Only drilling fixes them.
This is why article errors survive long after every other error in your writing is fixed, and why reading more English does not repair them. Reading builds recognition. Articles are a production skill. Only drilling fixes them.
The three-question test
Figure 1.2 — The article decision. Two questions, answered in order, resolve the great majority of cases. The second is the one that goes wrong: the writer knows which specimen, which cohort, which trial — and forgets that the reader has not been told yet.
Forget the grammar-book rules. Every article decision in scientific writing resolves by asking three questions in order. Run them on the noun, not on the sentence.
| Ask | If yes | Worked example |
|---|---|---|
| 1. Can I count this noun? Can I say “two of them”? | If no → no article. Stop here. | Evidence supports this approach. Not an evidence, not evidences. |
| 2. Am I talking about the whole class in general, using a plural? | If yes → no article. Stop here. | Patients with diabetes have a higher risk. Diabetic patients everywhere, not one cohort. |
| 3. Does the reader already know exactly which one I mean? | Yes → the. No → a / an. | We enrolled a patient with EGFR mutation. The patient received osimertinib. |
The reader-already-knows rule, in detail
The reader knows which one you mean when: you have already mentioned it; you define it in the same phrase (the patients enrolled in this study); there is only one in the world (the World Health Organization, the liver); or it is a superlative or ordinal (the primary endpoint, the highest dose). Everything else on first mention takes a or an.
A or an — listen, do not read
The choice follows the sound of the next word, not its spelling. This matters constantly in medical writing because abbreviations are everywhere and their spelling lies about their sound.
Wrong
Right
a MRI scanM is a consonant letter
an MRI scanbut it is said “em” — a vowel sound
an one-year follow-upO is a vowel letter
a one-year follow-upbut it is said “wun” — a consonant sound
a hour
an hourthe h is silent
an SAE report / a IV line
an SAE report / an IV line“ess-ay-ee”, “eye-vee”
an unit of insulin
a unit of insulinsaid “yoo-nit”
The four traps in clinical text
1 — Generic plurals
2 — First vs later mention
3 — Uncountables
4 — Defined by the phrase
The trap
Adding the to a plural noun that means the whole class. Highest-frequency article error in Indian manuscripts.
Wrong
“The immune checkpoint inhibitors have transformed the treatment of the melanoma.”
Right
“Immune checkpoint inhibitors have transformed the treatment of melanoma.”
Checkpoint inhibitors as a drug class takes no article. Melanoma as a disease takes no article. But the treatment keeps its the — because “of melanoma” tells the reader exactly which treatment is meant.
The trap
Using the on first mention, before the reader can know which one you mean.
Wrong
“We report the case of the 62-year-old man with the rare urothelial carcinoma. The patient received velparimab.”
Right
“We report the case of a 62-year-old man with a rare urothelial carcinoma. The patient received velparimab.”
First mention introduces — a man, a carcinoma. Second mention refers back — the patient. Note “the case of” is correct, because “of a 62-year-old man” identifies it.
The trap
Treating an uncountable noun as countable — the instinct that produces informations and equipments.
Wrong
“A recent research provides an evidence that the therapy is effective. We reviewed the literatures.”
Right
“Recent research provides evidence that the therapy is effective. We reviewed the literature.”
If you need a countable version, add a counting word: a piece of evidence, a body of research, a line of therapy, two studies. Full list in Module 2.1.
The trap
Dropping the when a following phrase has already pinned the noun down to one specific thing.
Wrong
“Primary endpoint of study was overall survival. Median duration of follow-up was 24 months.”
Right
“The primary endpoint of the study was overall survival. The median duration of follow-up was 24 months.”
Whenever a noun is followed by of plus something specific, or preceded by primary, first, highest, median or overall, it almost always takes the.
Recognition drills
Drill 1 of 3
Which version is correct?
A“Biomarkers are increasingly used to select patients for immunotherapy in the CASTELLAN cohort.”
B“The biomarkers are increasingly used to select patients for immunotherapy in CASTELLAN cohort.”
C“Biomarkers are increasingly used to select patients for immunotherapy in CASTELLAN cohort.”
D“The biomarkers are increasingly used to select the patients for immunotherapy in the CASTELLAN cohort.”
Drill 2 of 3
Complete: “Each participant underwent ___ ECG at baseline and ___ 24-hour Holter recording at week 4.”
Aa / a
Ban / a
Can / an
Da / an
Drill 3 of 3
A Methods section reads: “Blood samples were collected in ___ EDTA tube and stored at −80 °C until ___ analysis.”
Aan / the
Ba / the
Can / no article
Da / no article
Production drill — this is where the learning happens
Rewrite the paragraph below, fixing every article error. There are nine.
“The lung cancer is leading cause of the cancer death worldwide. In present study, we evaluated efficacy of the pembrolizumab in patients with advanced disease. A total of 312 patients were enrolled. The primary endpoint was overall survival. We collected the informations on adverse events at each visit. A median follow-up was 18 months. The results suggest that pembrolizumab may provide the survival benefit in this population.”
Corrected: “Lung cancer is the leading cause of cancer death worldwide. In the present study, we evaluated the efficacy of pembrolizumab in patients with advanced disease. A total of 312 patients were enrolled. The primary endpoint was overall survival. We collected information on adverse events at each visit. The median follow-up was 18 months. The results suggest that pembrolizumab may provide a survival benefit in this population.”
The nine fixes:
1. The lung cancer → Lung cancer — a disease named in general takes no article.
2. is leading cause → is the leading cause — superlative, only one.
3. the cancer death → cancer death — generic, uncountable.
4. In present study → In the present study — present identifies it.
5. evaluated efficacy → evaluated the efficacy — “of pembrolizumab” pins it down.
6. the pembrolizumab → pembrolizumab — drug names never take an article.
7. the informations → information — uncountable, no article, no s.
8. A median follow-up → The median follow-up — median means there is only one.
9. the survival benefit → a survival benefit — first mention, unspecified, hedged.
Score yourself: 8–9 means you have the rule. 5–7 means recognition without production — repeat this drill in three days. Below 5, do not move on yet.
The nine fixes:
1. The lung cancer → Lung cancer — a disease named in general takes no article.
2. is leading cause → is the leading cause — superlative, only one.
3. the cancer death → cancer death — generic, uncountable.
4. In present study → In the present study — present identifies it.
5. evaluated efficacy → evaluated the efficacy — “of pembrolizumab” pins it down.
6. the pembrolizumab → pembrolizumab — drug names never take an article.
7. the informations → information — uncountable, no article, no s.
8. A median follow-up → The median follow-up — median means there is only one.
9. the survival benefit → a survival benefit — first mention, unspecified, hedged.
Score yourself: 8–9 means you have the rule. 5–7 means recognition without production — repeat this drill in three days. Below 5, do not move on yet.
Reflection
Open something you wrote in the last month — an email, a CV, a manuscript draft. Read only the first paragraph and mark every article decision. How many would you now change?
Auditing your own old text beats any exercise, because these are the errors your hand actually produces.
Next module
1.3 — Tense: the IMRAD tense map
Week 1 · Module 3
Tense: the IMRAD tense map
Scientific English uses tense to signal how certain a statement is. Get it wrong and you accidentally claim more — or less — than your data allow
Your scenario
You draft the Discussion section of a Phase II post-hoc biomarker analysis. A senior medical director leaves feedback: "You wrote 'This biomarker predicts treatment response' in the present tense, implying an established universal biological fact. We only observed this in an exploratory cohort of 42 patients—it belongs in the past tense or with a modal hedge." In scientific writing, tense is not merely grammar; it is a legal and scientific assertion of truth value.
Why you make this error
Indian languages mark tense reliably, so this is not a grammar gap — it is a gap in convention. Scientific English loads extra meaning onto tense that everyday English does not: past means we observed this once, present means this is known to be true, present perfect means the field has established this. Nobody teaches it, because native speakers absorb it by reading hundreds of papers.
The second problem is transfer: Hindi and the Dravidian languages allow the progressive with verbs of state, which is why “I am having a doubt” and “We are knowing the result” feel natural to you and mark you instantly to an international reader.
The second problem is transfer: Hindi and the Dravidian languages allow the progressive with verbs of state, which is why “I am having a doubt” and “We are knowing the result” feel natural to you and mark you instantly to an international reader.
The map — nine decisions, in order
Figure 1.3 — The IMRAD tense map. Four sections, one underlying rule. The Discussion is the only section that switches within a paragraph, and the switch carries meaning — past marks what you observed, present marks what you claim follows from it.
Print this table. Keep it beside you for the next ten documents you write. After that you will not need it.
| Where | What you are saying | Tense | Example |
|---|---|---|---|
| Introduction | Established background knowledge | present | Lung cancer is the leading cause of cancer death. |
| Introduction | What the field has found so far | present perfect | Several trials have shown a survival benefit. |
| Introduction | What one named study did | past | Paik et al. reported a response rate of 46%. |
| Introduction | Your objective | past | We aimed to assess efficacy in this subgroup. |
| Methods | What you did, once, in the past | past | Patients were randomised in a 1:1 ratio. |
| Results | What you found | past | Median progression-free survival was 8.2 months. |
| Results | What a table or figure displays | present | Table 2 shows baseline characteristics. |
| Discussion | Restating your own finding | past | In this study, response was higher in the biomarker-positive group. |
| Discussion | Interpreting it, or drawing an implication | present | These findings suggest that biomarker testing may guide selection. |
The one that gets manuscripts rejected
The last two rows are the whole game. Your finding stays in the past. Your interpretation moves to the present — but must be hedged. Writing “This biomarker predicts response” puts your single result into the tense of established fact, and a reviewer reads that as overclaiming. Write “In this cohort, biomarker positivity was associated with response,” then “These data suggest that the biomarker may predict response.” Module 3.2 covers the hedging half.
The two habits to break
Wrong
Right
I am having a doubt about the endpoint.have is a state, not an action
I have a question about the endpoint.stative verbs never take the progressive
We are knowing the results already.
We already know the results.know, believe, understand, contain, consist, belong, need, prefer, seem
The sample is containing three analytes.
The sample contains three analytes.
We have performed the assay in March 2025.present perfect cannot take a finished time
We performed the assay in March 2025.a named past time forces the simple past
The study has been completed last year.
The study was completed last year.
Since 2019, we published four papers.
Since 2019, we have published four papers.since opens a window still running
Recognition drills
Drill 1 of 2
You are writing the Discussion. Your trial found a numerically higher response rate in one subgroup, but was not powered for that comparison. Which sentence is correct?
A“MET exon 14 alteration predicts a higher response rate to tepotinib.”
B“MET exon 14 alteration has predicted a higher response rate to tepotinib.”
C“MET exon 14 alteration is predicting a higher response rate to tepotinib.”
D“Response rate was numerically higher in this subgroup; these data suggest that alteration status may influence response.”
Drill 2 of 2
Which correctly opens an Introduction paragraph?
A“Urothelial carcinoma is the second most common skin cancer. Several trials have evaluated PD-1 blockade in advanced disease.”
B“Urothelial carcinoma was the second most common skin cancer.”
C“Urothelial carcinoma has been the second most common skin cancer. Several trials evaluated PD-1 blockade.”
D“Urothelial carcinoma is being the second most common skin cancer.”
Production drill
Fix the tense in every sentence. There are seven errors.
“Diabetes was a major global health burden. Several studies demonstrated an association between glycaemic control and cardiovascular outcomes. In the present study, we are aiming to evaluate a novel GLP-1 agonist. Patients are randomised in a 2:1 ratio between March 2024 and August 2024. Median HbA1c reduction has been 1.4% at 24 weeks. Table 3 showed the full safety profile. These findings are suggesting that the agent improves glycaemic control.”
Corrected: “Diabetes is a major global health burden. Several studies have demonstrated an association between glycaemic control and cardiovascular outcomes. In the present study, we aimed to evaluate a novel GLP-1 agonist. Patients were randomised in a 2:1 ratio between March 2024 and August 2024. Median HbA1c reduction was 1.4% at 24 weeks. Table 3 shows the full safety profile. These findings suggest that the agent may improve glycaemic control.”
The seven fixes:
1. was → is — standing background fact.
2. demonstrated → have demonstrated — accumulated literature.
3. are aiming → aimed — your objective is a past decision.
4. are randomised → were randomised — Methods, and a named past window forbids anything else.
5. has been → was — a Result you observed, finished.
6. showed → shows — the table displays it every time it is read.
7. are suggesting → suggest — stative verb, no progressive. Note the added may: interpretation in the present must be hedged, which is Module 3.2.
Score yourself: 6–7 means the map is working. Below 5, print the table and write your next three documents with it open.
The seven fixes:
1. was → is — standing background fact.
2. demonstrated → have demonstrated — accumulated literature.
3. are aiming → aimed — your objective is a past decision.
4. are randomised → were randomised — Methods, and a named past window forbids anything else.
5. has been → was — a Result you observed, finished.
6. showed → shows — the table displays it every time it is read.
7. are suggesting → suggest — stative verb, no progressive. Note the added may: interpretation in the present must be hedged, which is Module 3.2.
Score yourself: 6–7 means the map is working. Below 5, print the table and write your next three documents with it open.
Reflection
Take any abstract from a journal in your therapy area. Label the tense of every sentence and match it against the nine rows. Where does the published abstract deviate, and can you work out why?
Deviations are usually deliberate. Finding out why is how you move from following the rule to controlling it.
Next module
1.4 — Subject–verb agreement under pressure
Week 1 · Module 4
Subject–verb agreement under pressure
You know the rule. You break it anyway — because scientific sentences put fifteen words between the subject and its verb, and because half the nouns in medicine are irregular Latin plurals
Your scenario
Your CSR draft goes to QC. It comes back with a single repeated comment: “Agreement — data is plural.” Forty-one instances. You protest that everyone writes the data shows. The QC lead replies that everyone is not submitting to a regulator, and that the sponsor style guide has said data show since 2011. She is right, and the fix takes you two hours you did not have.
Why you make this error
Two separate causes, and they need separate fixes.
Cause one — distance. Indian languages are typically verb-final, so a long gap between subject and verb is normal and the listener holds the subject in memory. English puts the verb early, and when a scientific sentence inserts three prepositional phrases in between, your ear latches onto the nearest noun instead of the real subject.
Cause two — the Latin plurals. Nothing to do with your first language. Native speakers get these wrong too. It is pure memorisation, and regulators enforce it.
Cause one — distance. Indian languages are typically verb-final, so a long gap between subject and verb is normal and the listener holds the subject in memory. English puts the verb early, and when a scientific sentence inserts three prepositional phrases in between, your ear latches onto the nearest noun instead of the real subject.
Cause two — the Latin plurals. Nothing to do with your first language. Native speakers get these wrong too. It is pure memorisation, and regulators enforce it.
The distance problem
The verb agrees with the head noun of the subject — never with whatever noun happens to sit closest to it. Find the head, bracket everything between, then choose the verb.
Wrong
Right
The incidence of treatment-emergent adverse events in elderly patients were higher.verb grabbed “patients”
The incidence … was higher.head noun is incidence — singular
A list of all reported serious adverse events are provided in Appendix 2.
A list … is provided.one list
The results of the interim analysis was presented to the DSMB.
The results … were presented.
Each of the 312 patients were followed for 24 months.each is always singular
Each of the 312 patients was followed.
Neither the sponsor nor the investigators was aware of allocation.
Neither the sponsor nor the investigators were aware.with neither/nor, agree with the nearer subject
The Latin and Greek plurals
Medicine inherited its vocabulary from Latin and Greek and kept the original plurals. In regulatory and journal contexts the classical plural is still enforced, even where general English has drifted.
| Singular | Plural | How it behaves |
|---|---|---|
| datum | data | The data show… Enforced in CSRs, regulatory submissions and most journals. In casual prose data is is now accepted, but not in your documents. |
| criterion | criteria | The inclusion criteria were… Never “a criteria”. |
| bacterium | bacteria | The bacteria were isolated… |
| medium | media | The culture media were prepared… |
| phenomenon | phenomena | These phenomena are… |
| analysis | analyses | Sensitivity analyses were performed… Note the pronunciation shift. |
| hypothesis | hypotheses | Both hypotheses were tested… |
| index | indices | Indexes is acceptable for book indexes only. |
| appendix | appendices | In documents. Anatomical: appendixes. |
| regimen | regimens | Regular — but note it is regimen, not regime. A regime is a government. |
The pair that reverses
“A number of” takes a plural verb. “The number of” takes a singular verb.
Both appear in almost every Results section. Learn them as a pair or you will keep swapping them.
A number of patients were lost to follow-up. — here a number of just means several, so the real subject is patients.The number of patients was 312. — here the subject is the number itself, and there is one of it.Both appear in almost every Results section. Learn them as a pair or you will keep swapping them.
Recognition drills
Drill 1 of 3
Which is correct?
A“The proportion of patients discontinuing treatment across both arms were higher than expected.”
B“The proportion of patients discontinuing treatment across both arms was higher than expected.”
C“The proportions of patients discontinuing treatment across both arms was higher than expected.”
D“The proportion of patients discontinuing treatment across both arms were being higher than expected.”
Drill 2 of 3
Which sentence is correct in a regulatory document?
A“The data suggests that a criteria for exclusion was not met.”
B“The datas suggest that one criteria for exclusion were not met.”
C“The data suggest that one criterion for exclusion was not met.”
D“The data is showing that a criteria for exclusion was not met.”
Drill 3 of 3
Complete both blanks: “A number of participants ___ withdrawn consent, although the number of withdrawals ___ within the anticipated range.”
Ahas / were
Bhave / were
Chas / was
Dhave / was
Production drill
Fix every agreement error. There are eight.
“The data from the three sensitivity analysis was consistent with the primary result. A number of patient who met the inclusion criterias was subsequently excluded. The incidence of treatment-emergent adverse events in the two arms were comparable. Each of the investigators were blinded to allocation. The number of protocol deviations were 14. Neither the sponsor nor the site staff was aware of the code. These phenomenon has been reported previously.”
Corrected: “The data from the three sensitivity analyses were consistent with the primary result. A number of patients who met the inclusion criteria were subsequently excluded. The incidence of treatment-emergent adverse events in the two arms was comparable. Each of the investigators was blinded to allocation. The number of protocol deviations was 14. Neither the sponsor nor the site staff were aware of the code. These phenomena have been reported previously.”
The eight fixes:
1. analysis → analyses — three of them.
2. data … was → data … were — plural of datum.
3. a number of patient was → a number of patients were — plural on both.
4. criterias → criteria — already plural; there is no s form.
5. The incidence … were → was — head noun is incidence, not arms.
6. Each … were → Each … was — each is always singular.
7. The number of … were → was — the number is the subject.
8. staff was → staff were — with neither/nor, agree with the nearer subject, which is plural. And phenomenon has → phenomena have.
Score yourself: if you missed 5, you have the distance problem — practise bracketing the middle of every subject before you choose a verb. If you missed 1, 4 or 8, it is pure vocabulary. Write the ten Latin plurals out by hand.
The eight fixes:
1. analysis → analyses — three of them.
2. data … was → data … were — plural of datum.
3. a number of patient was → a number of patients were — plural on both.
4. criterias → criteria — already plural; there is no s form.
5. The incidence … were → was — head noun is incidence, not arms.
6. Each … were → Each … was — each is always singular.
7. The number of … were → was — the number is the subject.
8. staff was → staff were — with neither/nor, agree with the nearer subject, which is plural. And phenomenon has → phenomena have.
Score yourself: if you missed 5, you have the distance problem — practise bracketing the middle of every subject before you choose a verb. If you missed 1, 4 or 8, it is pure vocabulary. Write the ten Latin plurals out by hand.
Reflection
Search your last document for “data”. How many times did you write “data is” or “data shows”? Note the count — it is usually higher than people expect.
Find-and-replace will not fix this safely. You have to read each one, because “the data set is” is correct.
Next week
2.1 — Countable and uncountable nouns
Week 2 · Module 1
Countable and uncountable nouns
The s that must never appear — and the counting words that let you say what you actually meant
Your scenario
A recruiter forwards your writing sample to the hiring manager with one line: “Science is good, English needs work.” She has not read past your second paragraph, which contained “several researches,” “the evidences suggest,” and “all the necessary equipments were calibrated.” Three words. Each one is a single letter too long.
Why you make this error
In Hindi, Telugu and most other Indian languages, the nouns behind research, information, equipment and advice can be pluralised freely. There is nothing odd about counting them. English arbitrarily marks a subset of nouns as mass nouns and refuses to count them, and there is no rule that predicts which ones — it is historical accident.
Because it is arbitrary, understanding will not help you. This module is a memorisation task. Learn the list, and learn the counting words that give you a way around it.
Because it is arbitrary, understanding will not help you. This module is a memorisation task. Learn the list, and learn the counting words that give you a way around it.
The never-s list
These nouns never take a plural s and never take a or an. They are the ones that actually appear in medical and scientific writing — there are others, but these are the ones that will cost you.
Never write
Always write
researches / a research
researchneed to count? two studies, three research projects
evidences / an evidence
evidencea piece of evidence, two lines of evidence
informations / an information
informationa piece of information, two items of information
equipments / an equipment
equipmenta piece of equipment, two instruments
literatures
the literaturetwo papers, several reviews
advices / an advice
advicea piece of advice, a recommendation
feedbacks
feedbacktwo comments, a set of comments
trainings
trainingtwo training sessions, a training programme
personnels / staffs
personnel / staffboth already plural: the staff were trained
softwares
softwaretwo software packages, three applications
knowledges / a knowledge
knowledgean understanding of, an awareness of
progresses
progressan advance, two developments
The escape route: counting words
You are never stuck. Every uncountable noun can be counted by putting a countable word in front of it.
When you feel the urge to add an s to one of the never-s nouns, that urge is telling you a counting word is missing.
a piece of evidence · a body of research · a line of therapy · an item of information · a course of treatment · a set of criteria · a unit of insulin · a strand of DNAWhen you feel the urge to add an s to one of the never-s nouns, that urge is telling you a counting word is missing.
The medical words that go both ways
Some nouns are countable in one sense and uncountable in another. Getting these wrong changes your meaning, not just your grammar.
| Word | Uncountable sense | Countable sense |
|---|---|---|
| therapy | the general approach — patients receiving therapy | a specific regimen — three therapies were compared |
| treatment | the process — during treatment | a specific intervention — two treatments were available |
| experience | accumulated skill — nine years of experience | a single event — a distressing experience |
| time | duration — little time remained | occasions — dosed three times daily |
| disease | the state — progressive disease | named entities — rare diseases |
| work | effort or output — previous work has shown | never countable in this sense; use studies, not works |
Much, many, less, fewer, amount, number
Each pair splits on exactly the same line: one member for things you count, one for things you measure.
| Countable | Uncountable | Example |
|---|---|---|
many | much | many adverse events · much variability |
fewer | less | fewer discontinuations · less bleeding |
number of | amount of | the number of patients · the amount of drug |
each / every | all of the | each patient · all of the tissue |
The one exception worth knowing
Measurements of time, money and distance take less even though they look countable, because you are measuring a quantity rather than counting units: less than 6 months, less than 50 mg, less than 2 cm. But fewer than 20 patients — those are counted individuals.
Recognition drills
Drill 1 of 3
Which sentence is correct?
A“Recent research provides evidence and additional information on this pathway.”
B“Recent researches provide evidences and additional informations on this pathway.”
C“A recent research provide an evidence and additional information on this pathway.”
D“Recent researches provide evidence and additional information on this pathway.”
Drill 2 of 3
Complete: “There were ___ discontinuations in the treatment arm and ___ variability in exposure.”
Aless / less
Bfewer / less
Cfewer / fewer
Dless / fewer
Drill 3 of 3
You want to say that your team examined two separate strands of proof. Which is correct?
A“We examined two evidences.”
B“We examined two evidence pieces.”
C“We examined two lines of evidence.”
D“We examined the evidence.”
Production drill
Fix every countability error. There are nine.
“Previous researches have generated substantial evidences regarding this mechanism. We reviewed the available literatures and collected informations from twelve sites. All equipments were calibrated by trained personnels. The site staffs received two trainings before enrolment began. There were less protocol deviations than anticipated, and the amount of patients withdrawing consent was small. We are grateful for the useful feedbacks from the reviewers.”
Corrected: “Previous research has generated substantial evidence regarding this mechanism. We reviewed the available literature and collected information from twelve sites. All equipment was calibrated by trained personnel. The site staff received two training sessions before enrolment began. There were fewer protocol deviations than anticipated, and the number of patients withdrawing consent was small. We are grateful for the useful feedback from the reviewers.”
The nine fixes:
1. researches have → research has — uncountable, so the verb goes singular too.
2. evidences → evidence.
3. literatures → literature.
4. informations → information.
5. equipments were → equipment was.
6. personnels → personnel; staffs → staff.
7. two trainings → two training sessions — the counting word does the work.
8. less protocol deviations → fewer — deviations are counted.
9. the amount of patients → the number of patients; feedbacks → feedback.
Score yourself: this drill is pure recall. If you missed more than two, write the never-s list out by hand and put it beside your keyboard for a fortnight. There is no shortcut and no logic to reason from.
The nine fixes:
1. researches have → research has — uncountable, so the verb goes singular too.
2. evidences → evidence.
3. literatures → literature.
4. informations → information.
5. equipments were → equipment was.
6. personnels → personnel; staffs → staff.
7. two trainings → two training sessions — the counting word does the work.
8. less protocol deviations → fewer — deviations are counted.
9. the amount of patients → the number of patients; feedbacks → feedback.
Score yourself: this drill is pure recall. If you missed more than two, write the never-s list out by hand and put it beside your keyboard for a fortnight. There is no shortcut and no logic to reason from.
Reflection
Which three words from the never-s list do you personally add an s to? Be honest — most people have a consistent set of two or three, not all twelve.
Naming your own three is worth more than memorising all twelve. Those are the ones to set as autocorrect entries.
Next module
2.2 — Prepositions and fixed verb patterns
Week 2 · Module 2
Prepositions and fixed verb patterns
No rules, no logic, no shortcuts — but a short list of pairings that covers almost everything you will write in clinical and regulatory documents
Your scenario
In your second-round interview, the hiring manager asks you to talk through a paper. You say: “The authors have discussed about the limitations and emphasised on the need to cope up with confounding.” Your analysis is sharp. The manager writes one word in her notes: prepositions. Nothing about your reasoning was wrong. She has still learned that your written drafts will need line editing.
Why you make this error
Indian languages use postpositions — the marker follows the noun rather than preceding it — and the mapping to English prepositions is not one-to-one. Telugu -gurinchi and Hindi -ke baare mein both attach to the noun and both translate as about, so discussed about feels like a faithful translation. It is. English simply built about into the verb discuss already.
There is no rule here. Native speakers cannot explain why you comply with but adhere to. Learn the pairings as fixed units, the way you learned drug names.
There is no rule here. Native speakers cannot explain why you comply with but adhere to. Learn the pairings as fixed units, the way you learned drug names.
Verbs that already contain the preposition
These are the highest-frequency errors in Indian scientific English, and the easiest to fix, because the correction is always the same: delete the preposition.
Wrong
Right
discuss about the results
discuss the results
emphasise on the need
emphasise the needor: place emphasis on
comprise of three arms
comprise three armsor: consist of / be composed of
cope up with confounding
cope with confounding
revert back
replyrevert means return to a former state
return back / repeat again
return / repeat
investigate about / study about
investigate / study
request for a sample
request a samplebut: make a request for
reach to a conclusion
reach a conclusion
accompany with
accompanybut: accompanied by
The clinical pairings — learn these as units
| Pairing | How it is used |
|---|---|
associated with | Never associated to. The workhorse of observational reporting. |
compared with / compared to | With for like-for-like comparison of groups. To for metaphor or analogy. In clinical text it is nearly always with. |
superior to / inferior to | Never superior than. Same for preferable to. |
different from | Different than is US-informal; different to is UK-informal. In scientific writing use from. |
incidence of · prevalence of | Always of the event, in the population: the incidence of neutropenia in elderly patients. |
risk of an event · risk in a group · at risk of | The risk of relapse was higher in the control arm; these patients are at risk of infection. |
result in vs result from | Result in = leads to (cause → effect). Result from = is caused by (effect → cause). Reversing these reverses your causal claim. |
consistent with · in line with | Not consistent to. |
comply with but adhere to | Both mean follow the rule. Different prepositions. No reason. |
indicated for · contraindicated in | Indicated for the treatment of X; contraindicated in patients with Y. Regulatory wording — get these exact. |
treated with a drug · treated for a condition | Patients treated with velparimab for advanced UC. |
attributable to · due to | Due to should modify a noun, not a whole clause. Use because of or owing to for clauses. |
The pair that changes your meaning
Result in and result from point in opposite directions.
In a safety narrative these are two completely different events. This is not a style point.
Treatment interruption resulted in disease progression. — the interruption caused the progression.Treatment interruption resulted from disease progression. — the progression caused the interruption.In a safety narrative these are two completely different events. This is not a style point.
Recognition drills
Drill 1 of 3
Which sentence contains no preposition error?
A“The authors discussed about a trial that comprised of two arms and was superior than placebo.”
B“The authors discussed a trial that comprised of two arms and was superior than placebo.”
C“The authors discussed about a trial that comprised two arms and was superior to placebo.”
D“The authors discussed a trial that comprised two arms and was superior to placebo.”
Drill 2 of 3
A safety narrative states that the patient stopped treatment because their disease got worse. Which is correct?
A“Treatment interruption resulted from disease progression.”
B“Treatment interruption resulted in disease progression.”
C“Treatment interruption resulted due to disease progression.”
D“Treatment interruption resulted to disease progression.”
Drill 3 of 3
Complete this label wording: “The product is indicated ___ the treatment of advanced disease and is contraindicated ___ patients with severe hepatic impairment.”
Ain / for
Bfor / in
Cto / for
Dfor / with
Production drill
Fix every preposition error. There are ten.
“We discussed about the findings with the steering committee. The regimen comprised of three cycles and was superior than the comparator. Neutropenia was associated to prior chemotherapy and was different than the pattern reported earlier. The incidence for febrile neutropenia in this cohort was 12%. Dose reduction resulted from improved tolerability. Investigators were asked to comply to the protocol and to adhere with the visit schedule. Patients were treated by pembrolizumab. We request for your comments on the same.”
Corrected: “We discussed the findings with the steering committee. The regimen comprised three cycles and was superior to the comparator. Neutropenia was associated with prior chemotherapy and was different from the pattern reported earlier. The incidence of febrile neutropenia in this cohort was 12%. Dose reduction resulted in improved tolerability. Investigators were asked to comply with the protocol and to adhere to the visit schedule. Patients were treated with pembrolizumab. We would welcome your comments.”
The ten fixes:
1. discussed about → discussed.
2. comprised of → comprised.
3. superior than → superior to.
4. associated to → associated with.
5. different than → different from.
6. incidence for → incidence of.
7. resulted from improved tolerability → resulted in — the reduction caused the improvement, not the reverse. This one changes the meaning.
8. comply to → comply with.
9. adhere with → adhere to — note 8 and 9 take opposite prepositions for the same idea.
10. treated by → treated with; and request for your comments on the same rewritten (see Module 3.3).
Score yourself: number 7 is the one that matters. Missing a preposition looks careless; reversing a causal direction in a safety document is a different order of problem.
The ten fixes:
1. discussed about → discussed.
2. comprised of → comprised.
3. superior than → superior to.
4. associated to → associated with.
5. different than → different from.
6. incidence for → incidence of.
7. resulted from improved tolerability → resulted in — the reduction caused the improvement, not the reverse. This one changes the meaning.
8. comply to → comply with.
9. adhere with → adhere to — note 8 and 9 take opposite prepositions for the same idea.
10. treated by → treated with; and request for your comments on the same rewritten (see Module 3.3).
Score yourself: number 7 is the one that matters. Missing a preposition looks careless; reversing a causal direction in a safety document is a different order of problem.
Reflection
Pick the five pairings from the table above that appear most often in the documents you actually write. Write each in a full sentence from your own therapy area.
Five pairings you use daily will do more for you than thirty you recognise but never write.
Next module
2.3 — Word order and modifier placement
Week 2 · Module 3
Word order and modifier placement
Where you put a word decides what it modifies. In clinical writing, a misplaced only or a dangling participle is not untidy — it states something you did not mean
Your scenario
A medical reviewer flags one line in your promotional piece: “The drug was only approved for second-line use.” You meant the approval covers second line and nothing else. The sentence as written says the drug was merely approved — as though approval were a small thing. The reviewer cannot tell which you meant, so the piece goes back. One word, three positions available, three different meanings.
Why you make this error
Indian languages are largely verb-final and flexible in the order of everything before the verb. Case marking tells the listener who did what, so word order carries less of the load and can move for emphasis without changing meaning.
English has almost no case marking left. Position is meaning. Move a modifier and you have modified something else. This is why English feels rigid to you — it is, and the rigidity is doing grammatical work your first language does with endings.
English has almost no case marking left. Position is meaning. Move a modifier and you have modified something else. This is why English feels rigid to you — it is, and the rigidity is doing grammatical work your first language does with endings.
Dangling participles
When a sentence opens with a participle phrase — having adjusted…, after adjusting…, using…, based on… — English attaches it to the subject of the main clause. Whatever noun comes first after the comma is understood to be doing the action. If that noun cannot perform it, the modifier dangles.
Dangling
Fixed
After adjusting for age, the risk was lower.risk cannot adjust for age
After adjusting for age, we found a lower risk.or: After adjustment for age, the risk was lower
Using SPSS version 27, the data were analysed.data cannot use SPSS
We analysed the data using SPSS version 27.
Having completed screening, informed consent was obtained.consent did not complete screening
Having completed screening, patients provided informed consent.
Based on these findings, the trial was expanded.the trial is not based on the findings
On the basis of these findings, the sponsor expanded the trial.
When treated with velparimab, tumour size decreased.the tumour was not treated; the patient was
When patients were treated with velparimab, tumour size decreased.
The one-second test
Read the opening phrase, then read the first noun after the comma. Ask: can that noun do that thing? If not, either change the noun or turn the participle into a full clause with its own subject. Two seconds per sentence, and it catches every dangling modifier you will ever write.
Only, also, however — the movable words
Only attaches to whatever follows it. Every position is grammatical and every position means something different, so the reader cannot correct you — they can only guess.
| Sentence | What it means |
|---|---|
| Only the treatment group received the active drug. | No other group received it. |
| The treatment group only received the active drug. | They received it and did nothing else with it — they did not, say, return it. |
| The treatment group received only the active drug. | They received nothing besides the active drug — no concomitant medication. |
| The treatment group received the active drug only in cycle 1. | They received it in cycle 1 and in no other cycle. |
The Indian-English only
In Indian English, only also works as an emphasis particle at the end of a clause — “I am telling you only,” “That is the reason only.” This is a direct calque of Hindi hi and Telugu -ē, and it is completely natural in Indian speech.
An international reader does not have this reading available. They will parse the only as a restriction and get a meaning you never intended. Delete it, or move it to where the restriction actually belongs.
An international reader does not have this reading available. They will parse the only as a restriction and get a meaning you never intended. Delete it, or move it to where the restriction actually belongs.
Noun stacks
Figure 2.3 — Unpacking a noun stack. Read right to left: the last noun is the head, everything before it modifies. Three nouns is usually tolerable; four is heavy; five should be rebuilt with prepositions, which is longer but readable in one pass.
English lets nouns modify other nouns, and medical writing abuses this freedom badly. Three nouns in a row is usually readable. Four is hard. Five is a puzzle — the reader has to guess which noun modifies which.
Stacked
Unpacked
treatment emergent adverse event onset time analysissix nouns; at least two readings
analysis of the time to onset of treatment-emergent adverse events
patient reported outcome questionnaire completion rate
the rate of completion of patient-reported outcome questionnaires
study drug dose modification criteria
criteria for modifying the study drug dose
The unpacking method
Read the stack backwards and insert of, for, or to where the relationship needs stating. Then hyphenate any pair that genuinely acts as one modifier — treatment-emergent, patient-reported. You will use more words. The sentence will be read correctly the first time, which is the only measure that counts.
Recognition drills
Drill 1 of 3
Which sentence has no dangling modifier?
A“Using liquid chromatography–mass spectrometry, plasma concentrations were determined.”
B“Using liquid chromatography–mass spectrometry, the plasma concentrations of the analyte were determined.”
C“Using liquid chromatography–mass spectrometry, we determined plasma concentrations.”
D“Having used liquid chromatography–mass spectrometry, plasma concentrations were determined.”
Drill 2 of 3
You want to say: the drug is approved for second-line use and for no other line. Which sentence says that?
A“The drug was only approved for second-line use.”
B“The drug was approved for second-line use only only.”
C“Only the drug was approved for second-line use.”
D“The drug was approved only for second-line use.”
Drill 3 of 3
Best unpacking of: “study site staff training completion documentation”
A“documentation that staff at study sites have completed training”
B“study-site staff-training completion-documentation”
C“site staff training documents”
D“training documentation completion of study site staff”
Production drill
Fix every misplaced or dangling modifier and unpack the noun stack. There are six problems.
“After excluding patients with prior therapy, the analysis population was defined. Using a Cox proportional hazards model, hazard ratios were estimated. The intervention only reduced events in the high-risk subgroup. Based on these results, further study is warranted. We report the treatment emergent adverse event onset time distribution. Patients were followed for 24 months only.”
Corrected: “After patients with prior therapy were excluded, the analysis population was defined. We estimated hazard ratios using a Cox proportional hazards model. The intervention reduced events only in the high-risk subgroup. On the basis of these results, we consider further study warranted. We report the distribution of time to onset of treatment-emergent adverse events. Patients were followed for 24 months.”
The six fixes:
1. After excluding … the analysis population — the population did not do the excluding. Turned into a full clause.
2. Using a Cox model, hazard ratios were estimated — ratios cannot use a model. Subject supplied.
3. only reduced events → reduced events only in — the restriction belongs to the subgroup, not the verb. As written it hinted the intervention merely reduced events.
4. Based on these results, further study is warranted — the study is not based on the results. Rewritten with a real subject.
5. Six-noun stack unpacked by reading backwards and inserting of and to, with treatment-emergent hyphenated.
6. Trailing only deleted — the Indian emphasis particle. If you genuinely meant no longer than 24 months, write for no more than 24 months.
Score yourself: 3 and 6 are the ones that carry real risk. Both change what the sentence claims, and both survive every spellcheck and grammar checker you will ever run.
The six fixes:
1. After excluding … the analysis population — the population did not do the excluding. Turned into a full clause.
2. Using a Cox model, hazard ratios were estimated — ratios cannot use a model. Subject supplied.
3. only reduced events → reduced events only in — the restriction belongs to the subgroup, not the verb. As written it hinted the intervention merely reduced events.
4. Based on these results, further study is warranted — the study is not based on the results. Rewritten with a real subject.
5. Six-noun stack unpacked by reading backwards and inserting of and to, with treatment-emergent hyphenated.
6. Trailing only deleted — the Indian emphasis particle. If you genuinely meant no longer than 24 months, write for no more than 24 months.
Score yourself: 3 and 6 are the ones that carry real risk. Both change what the sentence claims, and both survive every spellcheck and grammar checker you will ever run.
Reflection
Search your last document for every sentence beginning with a word ending in -ing or with the word “Based”. Apply the one-second test to each. How many dangle?
Opening participles are where almost all danglers live. Checking that one pattern catches most of them.
Next module
2.4 — Punctuation for scientific text
Week 2 · Module 4
Punctuation for scientific text
Six marks, used properly, cover everything you will write — and one of them, the hyphen, is the fastest way for a reviewer to judge whether you have been edited before
Your scenario
You are shortlisted for a senior writer role and given a 300-word editing test. You fix the science, tighten the prose, and hand it back. You do not get the job. The feedback, when you finally ask, is one line: “Candidate did not correct any of the seeded hyphenation errors.” There were nine. You had read past every one of them.
Why you make this error
Devanagari and the Dravidian scripts historically used the danda and very little else — comma, semicolon, apostrophe and hyphen are all borrowings, and none of them maps onto a native mark with a fixed function. English school teaching in India also tends to present the comma as “a short pause,” which is a description of speech, not of grammar.
That is exactly what produces the comma splice: if a comma is a pause, then joining two sentences with one feels right. In English, a comma cannot join two complete sentences. Ever.
That is exactly what produces the comma splice: if a comma is a pause, then joining two sentences with one feels right. In English, a comma cannot join two complete sentences. Ever.
The comma splice
A comma splice is two complete sentences joined by a comma alone. It is the most common punctuation error in Indian scientific drafts, and it has four correct fixes — pick whichever suits the relationship between the two ideas.
| Fix | Example | Use when |
|---|---|---|
| Full stop | The drug was well tolerated. No serious events occurred. | The two ideas are independent. |
| Semicolon | The drug was well tolerated; no serious events occurred. | They are closely linked and you want to show it. |
| Comma + conjunction | The drug was well tolerated, and no serious events occurred. | You want an explicit connective. |
| Subordinate one clause | Because the drug was well tolerated, no dose reductions were required. | One idea genuinely depends on the other. |
The trap: however, therefore, moreover
These are adverbs, not conjunctions. They cannot join two sentences with a comma, however natural it feels.
Wrong: The endpoint was met, however the effect was small.
Right: The endpoint was met; however, the effect was small.
Right: The endpoint was met. However, the effect was small.
Right: The endpoint was met, but the effect was small.
But, and, or, so, yet are conjunctions and take a comma. However, therefore, moreover, nevertheless, thus, hence need a semicolon or a full stop.
Wrong: The endpoint was met, however the effect was small.
Right: The endpoint was met; however, the effect was small.
Right: The endpoint was met. However, the effect was small.
Right: The endpoint was met, but the effect was small.
But, and, or, so, yet are conjunctions and take a comma. However, therefore, moreover, nevertheless, thus, hence need a semicolon or a full stop.
Hyphenation — the reviewer's tell
One rule covers almost every case: hyphenate a compound modifier when it comes before the noun; leave it open when it comes after the verb.
Before the noun — hyphenate
After the verb — open
a well-tolerated regimen
the regimen was well tolerated
a dose-dependent effect
the effect was dose dependent
first-line therapy · second-line use
used in the first line
open-label extension · double-blind phase
the study was open label
treatment-emergent adverse events
events that emerged during treatment
patient-reported outcomes · placebo-controlled trial
outcomes reported by patients
Three exceptions to the hyphen rule
1. Adverbs ending in -ly are never hyphenated. A statistically significant difference, not statistically-significant. The -ly already signals that the word modifies the adjective.
2. Established compounds stay closed. Follow-up as a noun takes a hyphen (at follow-up); as a verb it is two words (we followed up). Baseline, healthcare and lifespan are now single words.
3. Suspended hyphens keep their hyphen. First- and second-line therapy, not first and second-line therapy.
2. Established compounds stay closed. Follow-up as a noun takes a hyphen (at follow-up); as a verb it is two words (we followed up). Baseline, healthcare and lifespan are now single words.
3. Suspended hyphens keep their hyphen. First- and second-line therapy, not first and second-line therapy.
Semicolon, colon, dash, parentheses
| Mark | Job | Example |
|---|---|---|
; | Joins two closely related complete sentences; also separates list items that already contain commas. | Sites were in Pune, India; Lyon, France; and Boston, USA. |
: | Introduces a list, an explanation or an expansion. What comes before it must be a complete sentence. | Three endpoints were assessed: ORR, PFS and OS. |
— em dash | Sets off an interruption more sharply than commas. Use sparingly in formal writing. | The effect — unexpected at this dose — persisted for 12 weeks. |
– en dash | Number ranges and paired names. Not a hyphen. | weeks 4–12 · the Kaplan–Meier estimate |
( ) | Your own supplementary material. | Median PFS was 8.2 months (95% CI 6.1–10.4). |
[ ] | Your insertion into someone else's quoted words. | The authors noted that “[the effect] was durable”. |
The serial comma
The comma before and in a list — ORR, PFS, and OS — is optional in general English but strongly preferred in scientific writing, because lists of drugs and endpoints are frequently ambiguous without it. AMA style requires it. Pick one convention and apply it to the whole document; inconsistency is worse than either choice.
Recognition drills
Drill 1 of 3
Which is correctly punctuated?
A“The primary endpoint was met, however the magnitude of benefit was modest.”
B“The primary endpoint was met; however, the magnitude of benefit was modest.”
C“The primary endpoint was met: however the magnitude of benefit was modest.”
D“The primary endpoint was met, the magnitude of benefit was however modest.”
Drill 2 of 3
Which is correctly hyphenated throughout?
A“A statistically-significant benefit was seen in first line patients; the regimen was well-tolerated.”
B“A statistically significant benefit was seen in first line patients; the regimen was well-tolerated.”
C“A statistically significant benefit was seen in first-line patients; the regimen was well tolerated.”
D“A statistically-significant benefit was seen in first-line patients; the regimen was well-tolerated.”
Drill 3 of 3
The study ran sites in three cities across three countries. Which is correctly punctuated?
A“Sites were located in Pune, India, Lyon, France, and Boston, USA.”
B“Sites were located in: Pune, India; Lyon, France; Boston, USA.”
C“Sites were located in Pune, India — Lyon, France — and Boston, USA.”
D“Sites were located in Pune, India; Lyon, France; and Boston, USA.”
Production drill
Fix every punctuation error. There are nine.
“The regimen was well-tolerated, no dose-limiting toxicities were observed. A statistically-significant difference was seen in first line patients, however the effect in second line patients was smaller. Three endpoints were assessed, ORR, PFS and OS. Median PFS was 8.2 months (95% CI 6.1-10.4). Patients were enrolled at sites in Chennai, India, Milan, Italy, and Toronto, Canada. Follow up continued for 24 months.”
Corrected: “The regimen was well tolerated; no dose-limiting toxicities were observed. A statistically significant difference was seen in first-line patients; however, the effect in second-line patients was smaller. Three endpoints were assessed: ORR, PFS, and OS. Median PFS was 8.2 months (95% CI 6.1–10.4). Patients were enrolled at sites in Chennai, India; Milan, Italy; and Toronto, Canada. Follow-up continued for 24 months.”
The nine fixes:
1. well-tolerated → well tolerated — follows the verb, so open.
2. Comma splice after tolerated → semicolon.
3. statistically-significant → no hyphen after an -ly adverb.
4. first line patients → first-line patients — modifier before the noun. Same for second-line.
5. , however → ; however, — however is an adverb, not a conjunction.
6. assessed, → assessed: — a colon introduces the list.
7. Serial comma added before and OS.
8. Hyphen in the range → en dash: 6.1–10.4.
9. City-country list → semicolons; and Follow up → Follow-up as a noun.
Score yourself: 7–9 is strong. If you missed 1, 3 and 4, you have the hyphen gap specifically — and that is the one an editing test is built to detect.
The nine fixes:
1. well-tolerated → well tolerated — follows the verb, so open.
2. Comma splice after tolerated → semicolon.
3. statistically-significant → no hyphen after an -ly adverb.
4. first line patients → first-line patients — modifier before the noun. Same for second-line.
5. , however → ; however, — however is an adverb, not a conjunction.
6. assessed, → assessed: — a colon introduces the list.
7. Serial comma added before and OS.
8. Hyphen in the range → en dash: 6.1–10.4.
9. City-country list → semicolons; and Follow up → Follow-up as a noun.
Score yourself: 7–9 is strong. If you missed 1, 3 and 4, you have the hyphen gap specifically — and that is the one an editing test is built to detect.
Reflection
Open your last document and search for “however”. Look at the mark immediately before it. How many are commas?
This single search catches more comma splices than reading the whole document does.
Next week
3.1 — Concision: cutting a quarter of your words
Week 3 · Module 1
Concision: cutting a quarter of your words
Journals cap abstracts at 250 words and regulators cap nothing but read everything. Learning to cut 25% without losing meaning is the most commercially valuable skill in this course
Your scenario
Your abstract comes in at 340 words. The limit is 250. You spend an afternoon deleting findings — the subgroup result, one safety signal, half the methods — and submit something weaker than what you had. You never needed to cut a single fact. Ninety-four of those words were doing no work at all.
Why you write long
Two habits, and neither is a language deficit.
Formal register in Indian academic English signals respect through length. Longer constructions read as more careful and more deferential. In English-language scientific writing the convention runs the opposite way: brevity signals confidence and command of the material. Same intention, opposite execution.
Nominalisation feels more scientific. We made an evaluation of sounds weightier than we evaluated. It is not — it is the same claim with three extra words and a buried verb. Once you can see nominalisations, you cannot unsee them.
Formal register in Indian academic English signals respect through length. Longer constructions read as more careful and more deferential. In English-language scientific writing the convention runs the opposite way: brevity signals confidence and command of the material. Same intention, opposite execution.
Nominalisation feels more scientific. We made an evaluation of sounds weightier than we evaluated. It is not — it is the same claim with three extra words and a buried verb. Once you can see nominalisations, you cannot unsee them.
Nominalisation: dig out the buried verb
A nominalisation is a verb turned into a noun — evaluate becomes evaluation, analyse becomes analysis. Once the verb becomes a noun, the sentence needs a weak filler verb (perform, conduct, make, carry out) plus a preposition to hold it up. Delete the scaffolding and put the verb back.
Buried
Freed
performed an evaluation of4 words
evaluated1 word
conducted an analysis of
analysed
made a comparison between
compared
carried out an investigation into
investigated
provided confirmation of
confirmed
is in agreement with
agrees with
gave rise to an increase in
increased
reached the determination that
determined that
How to spot them fast
Search your document for these six words: perform, conduct, make, carry, provide, undertake. Almost every hit is a nominalisation with a real verb trapped inside it. This one search typically removes 5–8% of a draft in ten minutes.
Empty frames
There is, there are and it is at the start of a sentence usually hide the real subject and the real verb further along. Find them and promote them.
Framed
Direct
There are several studies which have shown that…8 words before the content
Several studies show that…
It is important to note that the effect was small.
The effect was small.if it were not worth noting you would not write it
It has been reported by Kumar et al. that…
Kumar et al. reported that…
There was an observation of increased toxicity.
Toxicity increased.
The deadweight list
Each phrase on the left can be replaced by the word on the right in almost every sentence you will ever write. None of them carries meaning the shorter form lacks.
| Delete | Write |
|---|---|
| due to the fact that · owing to the fact that | because |
| in order to | to |
| in the event that | if |
| at this point in time · at the present time | now — or delete |
| a majority of | most |
| a number of | several — or give the number |
| in close proximity to | near |
| prior to · subsequent to | before · after |
| with regard to · in respect of | about · for |
| it is to be noted that · it should be mentioned that | delete entirely |
| in the present study, we have made an attempt to | we + verb |
| the vast majority of the patients in the study | most patients |
Concision is not deletion
Cutting words must never cut information. If your shortened sentence has lost a qualifier, a population, a timepoint or a hedge, you have not been concise — you have been inaccurate.
Too far: “The drug worked.”
Right: “Median PFS was longer with the drug than with placebo (8.2 vs 5.1 months).”
The second is longer and far better. Target 25% fewer words at the same information content, not 25% fewer facts.
Too far: “The drug worked.”
Right: “Median PFS was longer with the drug than with placebo (8.2 vs 5.1 months).”
The second is longer and far better. Target 25% fewer words at the same information content, not 25% fewer facts.
Passive vs. Active Voice — The Medical Writer's Decision Framework
Many Indian universities teach students to write exclusively in passive voice (e.g., "An observation was made by us that..."). In professional medical writing, both voices have precise, complementary roles:
- Use Passive Voice in Methods & CSRs (ICH E3): When the procedure or the patient is the primary focus and the investigator's identity is irrelevant.
Example: "Patients were randomised in a 1:1 ratio." (Correct) • "Blood samples were collected at baseline." (Correct) - Use Active Voice in Introductions, Results & Discussions: When describing what the drug, the trial, or the data accomplished. It eliminates nominalisations and cuts words instantly.
Bloated Passive: "A reduction in progression risk was demonstrated by the treatment." (10 words)
Direct Active: "Treatment reduced the risk of progression." (6 words)
Recognition drills
Drill 1 of 3
Best revision of: “It is to be noted that in the present study we have made an attempt to perform an evaluation of the safety profile.”
A“We evaluated the safety profile.”
B“In the present study we attempted to evaluate the safety profile.”
C“An evaluation of the safety profile was performed.”
D“Safety was noted.”
Drill 2 of 3
Which revision is concise without losing information?
AOriginal: “Due to the fact that a majority of the patients in the study who were treated with the combination regimen experienced a reduction in tumour size, it was concluded that the regimen has activity.” → “The regimen works.”
B→ “Because most patients receiving the combination had tumour reduction, we concluded that the regimen has activity.”
C→ “Due to the fact that most of the patients experienced a reduction in tumour size, activity was concluded.”
D→ “The combination regimen reduces tumour size.”
Drill 3 of 3
Which sentence contains a nominalisation that should be freed?
A“We analysed the pharmacokinetic data.”
B“Sensitivity analyses were prespecified in the statistical analysis plan.”
C“We undertook a comparison of the two regimens.”
D“The central laboratory confirmed all positive results.”
Production drill — the 25% target
This paragraph is 118 words. Rewrite it in under 90 without losing a single fact.
“It is to be noted that in the present study, we have made an attempt to carry out an investigation into the efficacy and the safety of the combination regimen. Due to the fact that a majority of the patients in our cohort had received prior therapy, there was a necessity for the performance of a stratified analysis. Prior to the initiation of treatment, an assessment of baseline characteristics was undertaken. It is important to note that there were several patients who experienced a discontinuation of treatment. At this point in time, it is our conclusion that further investigation of the regimen in a randomised setting is warranted.”
Model rewrite (67 words, 43% shorter):
“We investigated the efficacy and safety of the combination regimen. Because most patients had received prior therapy, we performed a stratified analysis. We assessed baseline characteristics before treatment. Several patients discontinued treatment. We conclude that the regimen warrants further investigation in a randomised setting.”
What went, and why:
1. It is to be noted that — carries no information. Deleted twice.
2. we have made an attempt to carry out an investigation into → we investigated. Ten words to two.
3. Due to the fact that → Because. A majority of the patients in our cohort → most patients.
4. there was a necessity for the performance of → we performed.
5. Prior to the initiation of treatment, an assessment … was undertaken → We assessed … before treatment.
6. there were several patients who experienced a discontinuation of treatment → several patients discontinued treatment.
7. At this point in time, it is our conclusion that → We conclude that.
Check your own version: every fact in the original is still present — the objective, the prior-therapy majority, the stratified analysis, the baseline assessment, the discontinuations, and the call for randomised study. If any of those six is missing from your rewrite, you cut facts rather than words. If you came in under 90 with all six, you have the skill.
“We investigated the efficacy and safety of the combination regimen. Because most patients had received prior therapy, we performed a stratified analysis. We assessed baseline characteristics before treatment. Several patients discontinued treatment. We conclude that the regimen warrants further investigation in a randomised setting.”
What went, and why:
1. It is to be noted that — carries no information. Deleted twice.
2. we have made an attempt to carry out an investigation into → we investigated. Ten words to two.
3. Due to the fact that → Because. A majority of the patients in our cohort → most patients.
4. there was a necessity for the performance of → we performed.
5. Prior to the initiation of treatment, an assessment … was undertaken → We assessed … before treatment.
6. there were several patients who experienced a discontinuation of treatment → several patients discontinued treatment.
7. At this point in time, it is our conclusion that → We conclude that.
Check your own version: every fact in the original is still present — the objective, the prior-therapy majority, the stratified analysis, the baseline assessment, the discontinuations, and the call for randomised study. If any of those six is missing from your rewrite, you cut facts rather than words. If you came in under 90 with all six, you have the skill.
Reflection
Take the last 200 words you wrote. Run the six-word search (perform, conduct, make, carry, provide, undertake) and cut what it finds. What is your word count before and after?
Most first drafts lose 8–12% from this search alone, in under ten minutes.
Next module
3.2 — Hedging and claim calibration
Week 3 · Module 2
Hedging and claim calibration
The most commercially important unit in this course. In scientific and regulatory writing the verb you choose is a statement about your study design — and choosing the wrong one is a compliance failure, not a style slip
Your scenario
You draft a slide for a medical science liaison. It reads: “Real-world data show that early initiation reduces mortality by 30%.” The data are from a retrospective chart review. Medical-legal-regulatory review rejects the deck. The reviewer's comment is one line: “Causal language, observational design.” Your number was right. Your verb was not.
English marks certainty through verb choice and modal verbs, and scientific English does it with unusual precision. Every rung of the ladder below corresponds to a level of evidence. This is not politeness or modesty — it is a technical vocabulary for reporting how much your design actually supports.
The claim ladder
Figure 3.2 — The claim ladder. Calibration works in both directions. Writing “may be associated with” about an adequately powered randomised trial that met its primary endpoint understates evidence clinicians need in order to act — and reads as a writer who cannot tell strong evidence from weak.
| Verb | What it claims | What you need to say it |
|---|---|---|
proves · confirms | Settled. No further evidence needed. | Almost never appropriate for a single study. Reserve for mathematical proof or overwhelming consensus. |
demonstrates | This study establishes it. | An adequately powered, prespecified primary endpoint met in a randomised controlled trial. |
shows | The data display it directly. | A directly observed result. Safe for describing figures and tables. |
indicates | The data point this way. | A consistent result, possibly secondary, or from a well-conducted observational study. |
suggests | Compatible with this reading. | Exploratory analyses, subgroups, small samples, hypothesis-generating work. |
may suggest | One possible reading among several. | Underpowered, post hoc, or where confounding is plausible and unaddressed. |
Association is not causation, and the language must say so
Only a randomised design licenses causal verbs. Everything else takes association language.
Causal — needs randomisation: reduced, caused, led to, resulted in, improved, prevented, protected against.
Associational — safe for observational data: was associated with, correlated with, was accompanied by, occurred more frequently in, was observed alongside.
Writing “the drug reduced mortality” from a retrospective cohort is not a stylistic slip. It is a claim your data cannot support, and in promotional copy it is the most common single reason for rejection at review.
Causal — needs randomisation: reduced, caused, led to, resulted in, improved, prevented, protected against.
Associational — safe for observational data: was associated with, correlated with, was accompanied by, occurred more frequently in, was observed alongside.
Writing “the drug reduced mortality” from a retrospective cohort is not a stylistic slip. It is a claim your data cannot support, and in promotional copy it is the most common single reason for rejection at review.
The "Significant" Guardrail — Precision in Medical Writing
In everyday conversational English, "significant" means important, notable, or large. In scientific and regulatory medical writing, "significant" is a technical term strictly reserved for alpha-controlled, pre-specified hypothesis tests (p < 0.05).
The Two Inviolable Rules:
The Two Inviolable Rules:
- Never use "significant" for magnitude alone: If you observed a 40% reduction that did not achieve p < 0.05, write "a marked reduction", "a substantial decrease", or "a clinically meaningful trend" — never "a significant reduction".
- Always pair with exact statistics: When you write "a statistically significant improvement", you must immediately supply the test statistic, 95% CI, and p-value in parentheses: (HR 0.68; 95% CI: 0.52–0.89; p = 0.004).
The four hedging tools
Hedging is not one device. You have four, and precise writing uses them in combination rather than stacking one of them repeatedly.
| Tool | Examples | What it hedges |
|---|---|---|
| Modal verbs | may, might, could, can | The possibility that the claim holds. May improve ≠ improves. |
| Reporting verbs | suggests, indicates, appears to | The strength of the inference from data to conclusion. |
| Scope limiters | in this cohort, among patients aged over 65, at 24 weeks | The population, setting or timepoint the claim covers. |
| Quantifiers | a proportion of, some, numerically higher, a trend towards | How much and how many. Numerically higher flags a difference that did not reach significance. |
You can also under-claim, and that is also a fault
Calibration works both ways. If your randomised, adequately powered trial met its prespecified primary endpoint, writing “these data may suggest a possible benefit” is inaccurate in the other direction — it understates evidence that clinicians need in order to act.
Excessive hedging also reads as a lack of confidence, and in a job assessment it looks like you cannot tell strong evidence from weak. Match the language to the design. That is the entire skill.
Excessive hedging also reads as a lack of confidence, and in a job assessment it looks like you cannot tell strong evidence from weak. Match the language to the design. That is the entire skill.
Recognition drills
Drill 1 of 3
A single-arm Phase II study of 42 patients met its primary endpoint of overall response rate. You are drafting the abstract conclusion. Which verb is defensible?
A“These results demonstrate that the agent is superior to standard care.”
B“These results prove the efficacy of the agent.”
C“These results may suggest possible activity of the agent.”
D“These results show promising activity; randomised evaluation is warranted.”
Drill 2 of 3
A large, adequately powered, double-blind randomised trial met its prespecified primary endpoint of overall survival. Which is correctly calibrated?
A“The regimen improved overall survival; these data demonstrate a benefit in this population.”
B“The regimen may be associated with a possible survival benefit.”
C“This trial proves that the regimen prolongs life.”
D“The regimen was associated with longer survival.”
Drill 3 of 3
A prespecified secondary endpoint showed a difference that did not reach statistical significance (p = 0.09). How should it be reported?
A“The difference approached statistical significance.”
B“The rate was numerically higher in the treatment arm, but the difference was not statistically significant (p = 0.09).”
C“A benefit was seen on the secondary endpoint.”
DOmit it — non-significant results do not need reporting.
Production drill
Recalibrate every claim. The study is a retrospective chart review of 118 patients, unadjusted for treatment-selection bias.
“Our analysis proves that early initiation of therapy reduces mortality and definitely improves quality of life. The results confirm that this approach should be adopted as standard of care. Early treatment prevents disease progression and protects against hospitalisation. We have demonstrated that the benefit is causal. These findings establish early initiation as the preferred strategy for all patients.”
Model rewrite:
“In this retrospective analysis of 118 patients, earlier initiation of therapy was associated with lower mortality and with higher patient-reported quality-of-life scores. These findings suggest that the approach may warrant prospective evaluation. Earlier treatment was also associated with a lower rate of progression and fewer hospitalisations. Because the analysis was retrospective and unadjusted for treatment-selection bias, a causal relationship cannot be inferred. Confirmation in a randomised setting is required before any change in practice.”
Every change and why:
1. proves → was associated with. A retrospective design licenses association language only. This is the fix that matters most.
2. reduces mortality → was associated with lower mortality. Causal verb removed.
3. definitely improves → deleted, and quality of life became the measurable thing actually recorded (patient-reported scores).
4. confirm … should be adopted as standard of care → suggest … may warrant prospective evaluation. A chart review cannot change standard of care.
5. prevents and protects against → both are causal. Replaced with associational phrasing.
6. We have demonstrated that the benefit is causal → replaced with the opposite and correct statement: causality cannot be inferred. Naming the specific bias is what a reviewer looks for.
7. establish … for all patients → scope limiter added and the claim reduced to a call for randomised confirmation.
Score yourself: if you softened the tone but left proves, prevents or protects standing, you fixed the cosmetic problem and left the compliance problem. Those three verbs are the test.
“In this retrospective analysis of 118 patients, earlier initiation of therapy was associated with lower mortality and with higher patient-reported quality-of-life scores. These findings suggest that the approach may warrant prospective evaluation. Earlier treatment was also associated with a lower rate of progression and fewer hospitalisations. Because the analysis was retrospective and unadjusted for treatment-selection bias, a causal relationship cannot be inferred. Confirmation in a randomised setting is required before any change in practice.”
Every change and why:
1. proves → was associated with. A retrospective design licenses association language only. This is the fix that matters most.
2. reduces mortality → was associated with lower mortality. Causal verb removed.
3. definitely improves → deleted, and quality of life became the measurable thing actually recorded (patient-reported scores).
4. confirm … should be adopted as standard of care → suggest … may warrant prospective evaluation. A chart review cannot change standard of care.
5. prevents and protects against → both are causal. Replaced with associational phrasing.
6. We have demonstrated that the benefit is causal → replaced with the opposite and correct statement: causality cannot be inferred. Naming the specific bias is what a reviewer looks for.
7. establish … for all patients → scope limiter added and the claim reduced to a call for randomised confirmation.
Score yourself: if you softened the tone but left proves, prevents or protects standing, you fixed the cosmetic problem and left the compliance problem. Those three verbs are the test.
Reflection
Find one published abstract in your therapy area whose conclusion you think overclaims relative to its design. Write the sentence, name the design, and rewrite the claim.
Learning to see this in published work is what makes you useful in a review meeting. It is also the most common interview question for senior writing roles.
Biomarker, Pharmacodynamic & Acronym Conventions for Medical Writers
Precision in scientific English requires standard formatting of biological metrics, gene symbols, and medical abbreviations:
- Gene Symbols vs Proteins: Human gene symbols are italicised and uppercase (e.g., EGFR, KRAS, BRCA1, HLA-Cw6), whereas the corresponding protein products are written in non-italicised plain text (EGFR, KRAS, BRCA1, HLA-Cw6).
- Dosing & Posology Units: Always use a non-breaking space between numerical values and standard SI units (e.g.,
100 mg,5 mL,250 μg). Never write “100mg” without a space. - Laboratory Ratios & Expressions: Use × ULN (e.g., “ALT > 3 × ULN” with spaces around the multiplication sign), not “3xULN” or “3X ULN”.
- First Mention Acronym Rule: Spell out every clinical score or compound on first mention in each major deliverable section (e.g., “Major Adverse Cardiovascular Events (MACE)”, “Expanded Disability Status Scale (EDSS)”), then use the acronym consistently.
Next module
3.3 — Register and Indianisms
Week 3 · Module 3
Register and Indianisms
Correct English that costs you the job — the administrative vocabulary that marks a draft on sight, and how to match the register international scientific communication actually runs on
Your scenario
An international peer reviewer on a Lancet oncology submission writes in the confidential editor remarks: "The author rebuttal and revision letter are polite but heavily congested with archaic administrative jargon ('undersigned', 'do the needful', 'intimate'). It distracts from the authors' genuine scientific rebuttal points." Learning to write in standard modern international scientific English ensures your manuscripts and correspondence are judged purely on scientific merit.
Be clear about what this unit is
Indian English is a legitimate variety with hundreds of millions of fluent speakers and its own established conventions. There is nothing deficient about it, and this module is not about correcting who you are.
The point is narrow and practical: international scientific publishing and pharmaceutical communications run on a particular house register, and matching it is a professional skill exactly like learning ICH formatting or AMA style. Use the register the document requires. Use whatever you like everywhere else.
The point is narrow and practical: international scientific publishing and pharmaceutical communications run on a particular house register, and matching it is a professional skill exactly like learning ICH formatting or AMA style. Use the register the document requires. Use whatever you like everywhere else.
The kill list
Every phrase below is correct English and is understood across India. Almost none is understood the same way elsewhere in professional scientific communication, and international editors, recruiters and clients read them as a signal about the writer — usually before reading anything else.
Delete
Write instead
Kindly do the needful
Please review and approvesay what the action actually is
Please revert back
Please replyrevert means return to a former state
Please find attached the same
I have attached the protocolthe same as a pronoun is legal English from 1890
at the earliest
by Friday, 14 Marchgive a date, not a feeling
Respected Sir / Dear Sir
Dear Dr Anand / Dear Editoruse the name; Sir alone reads as junior and dated
I have a doubt
I have a questiona doubt means you suspect it is false
prepone the meeting
move the meeting earlierprepone does not exist outside South Asia
discuss about / emphasise on
discuss / emphasiseModule 2.2
It is a matter of great pleasure to present this humble work
This study evaluated…ceremonial openings signal an unedited draft
the above said / the aforementioned study only
this studyonly as an emphasis particle — Module 2.3
out of station · intimate you · needful action
away · inform you · the action required
myself R. Menon / I am having 9 years experience
My name is R. Menon / I have nine years of experience
The three registers you need
Register is not a single dial from casual to formal. Scientific work needs three distinct settings, and using the wrong one is as visible as a grammar error.
| Register | Where | What it sounds like |
|---|---|---|
| Formal impersonal | Manuscripts, CSRs, protocols, regulatory documents | Third person or first-person plural. No contractions. No direct address to the reader. Patients were randomised… |
| Professional direct | Emails to editors, clients, recruiters, colleagues | First person, named addressee, one clear request, a specific date. Contractions acceptable but not required. I have attached the revised draft; could you confirm by Friday? |
| Plain | Patient information leaflets, lay summaries, consent forms | Short sentences, everyday words, direct address to the patient. Tell your doctor if you notice yellowing of your skin. |
The most common register failure
Writing an email in the formal impersonal register. This is what produces “It is requested that the needful may kindly be done at your end.” The formal impersonal register belongs in documents, not correspondence. Emails take professional direct: you, I, a name, a request, a date.
Recognition drills
Drill 1 of 3
You are emailing a journal editor about a manuscript submitted six weeks ago. Which version is appropriate?
A“Respected Sir, kindly update on the status of the same and revert at the earliest.”
B“Sorry for disturbing you. Kindly do the needful regarding my humble submission.”
C“Dear Dr Mehta, I am writing about manuscript EJC-2026-0412, submitted on 3 June. Could you let me know its current status?”
D“Status of my manuscript. Please respond immediately.”
Drill 2 of 3
Which sentence belongs in a Patient Information Leaflet?
A“This product is contraindicated in patients with hepatic impairment or concomitant administration of CYP3A4 inhibitors.”
B“Kindly avoid this medicine in case of hepatic insufficiency.”
C“This medicine is safe for most people.”
D“Do not take this medicine if you have liver problems. Tell your doctor about all other medicines you take.”
Drill 3 of 3
You are opening the Discussion section of a manuscript. Which is correct?
A“In this cohort, earlier initiation was associated with lower mortality.”
B“It gives us immense pleasure to present the findings of this humble study.”
C“As you can see from the above said results, our finding is quite interesting.”
D“We have not seen anything like this before — it is a remarkable result.”
Production drill — the one that gets you hired
Rewrite this email to a hiring manager after a first-round interview.
“Respected Sir, It is a matter of great pleasure that I got the opportunity to interact with your good self yesterday. Myself having 9 years experience in medical writing domain. I have a doubt regarding the next steps of the process. Kindly do the needful and revert back at the earliest regarding the same. Also kindly intimate me if any further informations are required from my end. Thanking you in anticipation.”
Model rewrite:
“Dear Mr Anand,
Thank you for taking the time to speak with me yesterday. I enjoyed the discussion about the oncology publications workstream, particularly the question of how the team handles author queries at the review stage.
Could you let me know what the next steps in the process are, and roughly when I might expect to hear back? I am happy to provide additional writing samples or references if that would be useful.
Best regards,
R. Menon”
What changed and why:
1. Respected Sir → the manager's name. Sir alone reads as junior and dated.
2. Ceremonial opening and your good self deleted — replaced with a specific detail from the conversation, which shows you were engaged and makes the email memorable.
3. Myself having 9 years experience deleted entirely. They have your CV. Restating your experience in a follow-up reads as anxious.
4. I have a doubt → a direct question. A doubt means you suspect something is false.
5. Kindly do the needful and revert back at the earliest regarding the same → one clear question with a timeframe. Four kill-list items removed in one sentence.
6. kindly intimate me if any further informations are required from my end → a concrete offer. Note informations was also a Module 2.1 error.
7. Thanking you in anticipation → Best regards. The phrase thanks someone for something they have not done yet.
The underlying shift: the original asks the reader to do something vague. The rewrite asks one specific question and offers something concrete. That is what professional direct register actually is — not colder, just clearer about what you want.
“Dear Mr Anand,
Thank you for taking the time to speak with me yesterday. I enjoyed the discussion about the oncology publications workstream, particularly the question of how the team handles author queries at the review stage.
Could you let me know what the next steps in the process are, and roughly when I might expect to hear back? I am happy to provide additional writing samples or references if that would be useful.
Best regards,
R. Menon”
What changed and why:
1. Respected Sir → the manager's name. Sir alone reads as junior and dated.
2. Ceremonial opening and your good self deleted — replaced with a specific detail from the conversation, which shows you were engaged and makes the email memorable.
3. Myself having 9 years experience deleted entirely. They have your CV. Restating your experience in a follow-up reads as anxious.
4. I have a doubt → a direct question. A doubt means you suspect something is false.
5. Kindly do the needful and revert back at the earliest regarding the same → one clear question with a timeframe. Four kill-list items removed in one sentence.
6. kindly intimate me if any further informations are required from my end → a concrete offer. Note informations was also a Module 2.1 error.
7. Thanking you in anticipation → Best regards. The phrase thanks someone for something they have not done yet.
The underlying shift: the original asks the reader to do something vague. The rewrite asks one specific question and offers something concrete. That is what professional direct register actually is — not colder, just clearer about what you want.
Reflection
Open your sent-mail folder. Find the last five professional emails you wrote. Count the kill-list phrases. Then rewrite the one that mattered most — to a client, editor or recruiter.
This is the single highest-return exercise in the course. Your email register is judged before anyone sees your writing.
Next module
3.4 — Parallelism and cohesion
Week 3 · Module 4
Parallelism and cohesion
Individually correct sentences that do not hold together as a paragraph — why your Discussion reads as a list of true statements rather than an argument
Your scenario
Your protocol goes to QC. The objectives read: “to assess safety, evaluation of efficacy, and measuring pharmacokinetics.” QC rejects it — not on style, but because a reviewer genuinely cannot tell whether evaluation of efficacy is a study objective or a description of a method. Three items, three grammatical shapes, and the ambiguity is real.
Parallelism
Every item in a series must take the same grammatical form. This matters more in regulatory writing than anywhere else, because objectives, endpoints and inclusion criteria are read as legally-operative lists.
Not parallel
Parallel
to assess safety, evaluation of efficacy, and measuring PKinfinitive, noun, gerund
to assess safety, to evaluate efficacy, and to measure PKthree infinitives
Patients were excluded if they had renal impairment, prior therapy, or were pregnant.
Patients were excluded if they had renal impairment, had received prior therapy, or were pregnant.
The regimen is effective, well tolerated, and has convenience.
The regimen is effective, well tolerated, and convenient.three adjectives
not only reduced mortality but also was reducing hospitalisation
not only reduced mortality but also reduced hospitalisationpaired constructions must match on both sides
The vertical test
Write the list items one under the other, aligned at the start. If the first word of each is a different part of speech, the list is not parallel. This takes ten seconds and catches every instance — far more reliable than reading the sentence aloud.
Given–new: the rule that makes paragraphs flow
Figure 3.4 — Given–new chaining. The new information at the end of one sentence becomes the given information opening the next. This is what readers experience as flow, and its absence is usually what they mean when they say a paragraph is hard to follow but cannot say why.
English paragraphs cohere by a single principle: each sentence should start with information the reader already has, and end with the new information. The end of one sentence becomes the beginning of the next. When your writing feels choppy despite every sentence being correct, this is almost always what is missing.
Choppy — every sentence correct
Cohesive — same facts, reordered
What changed
Original
“MET exon 14 skipping alterations occur in 3–4% of non-small-cell lung cancers. Poor outcomes with conventional chemotherapy are seen in these patients. Tepotinib is a selective MET inhibitor. A response rate of 46% was reported in the VISION trial. Older patients are frequently affected by this alteration. Targeted therapy is therefore of particular interest.”
Every sentence is grammatically correct and factually true. The paragraph still feels like six separate statements, because each one starts with new information the reader has not been given yet.
Rewritten
“MET exon 14 skipping alterations occur in 3–4% of non-small-cell lung cancers. These alterations are more frequent in older patients, who respond poorly to conventional chemotherapy. This poor response has made targeted therapy particularly attractive in this group. One such agent, tepotinib, is a selective MET inhibitor that achieved a response rate of 46% in the VISION trial.”
Same six facts, one fewer sentence, and it now reads as a single argument leading to tepotinib. Nothing was added.
The chain
Sentence 1 ends on alterations → sentence 2 begins These alterations
Sentence 2 ends on poor response to chemotherapy → sentence 3 begins This poor response
Sentence 3 ends on targeted therapy → sentence 4 begins One such agent
Sentence 4 ends on the new information: the 46% result.
Sentence 2 ends on poor response to chemotherapy → sentence 3 begins This poor response
Sentence 3 ends on targeted therapy → sentence 4 begins One such agent
Sentence 4 ends on the new information: the 46% result.
Each sentence hands the reader a handle at the start and gives them something new at the end. The most important new fact — the response rate — lands in the final position, which is where English puts emphasis.
Transition signals — use the right direction
Connectives are not interchangeable. Each one promises the reader a particular relationship, and using an additive connector across a contrast makes the reader lose the thread.
| Direction | Words | Promise to the reader |
|---|---|---|
| Adding | moreover, furthermore, in addition, also | More of the same kind is coming. |
| Contrasting | however, nevertheless, by contrast, although | What follows cuts against what preceded. |
| Concluding | therefore, thus, consequently, hence | What follows is caused by or inferred from what preceded. |
| Exemplifying | for example, in particular, notably | A specific instance of the general point. |
| Sequencing | first, subsequently, finally | Steps in a defined order. |
Sentence length
Aim for an average of 20–25 words, with a hard ceiling around 40. Vary the length deliberately — several long sentences in a row exhaust the reader, and several short ones in a row read as abrupt. A useful rhythm is two medium sentences followed by one short one to land the point.
Recognition drills
Drill 1 of 3
Which set of inclusion criteria is parallel?
A“aged 18 years or older; histological confirmation of disease; must have adequate organ function”
B“age 18 years or older; histologically confirmed disease; adequate organ function”
C“having age 18 or above; must have histological confirmation; adequate organ function”
D“patients must be 18 or older; they should have confirmed disease; adequate organ function”
Drill 2 of 3
“The primary endpoint was met. ___ , the secondary endpoint was not.”
AMoreover
BTherefore
CHowever
DFurthermore
Drill 3 of 3
Which second sentence follows the given–new principle after: “Median progression-free survival was 8.2 months in the treatment arm.”
A“Grade 3 neutropenia occurred in 14% of patients.”
B“The control arm achieved a median of 5.1 months.”
C“Baseline characteristics were balanced between the arms.”
D“This figure compares with 5.1 months in the control arm.”
Production drill
Fix the parallelism error in the objectives, then rewrite the paragraph so the sentences chain.
“The objectives of this study were to assess the safety of the regimen, evaluation of its efficacy, and measuring pharmacokinetic parameters. A total of 240 patients were enrolled. Two arms were used in the randomisation. Adverse events were the reason for discontinuation in 18 patients. The most common adverse event was neutropenia. Moreover, the primary endpoint was not met.”
Model rewrite:
“The objectives of this study were to assess the safety of the regimen, to evaluate its efficacy, and to measure pharmacokinetic parameters. We enrolled 240 patients and randomised them to two arms. Of these, 18 discontinued because of adverse events, most commonly neutropenia. Despite this manageable safety profile, the primary endpoint was not met.”
What changed:
1. Parallelism. Three infinitives — to assess, to evaluate, to measure — replacing an infinitive, a noun and a gerund.
2. Chaining. 240 patients → Of these, 18 → most commonly neutropenia. Each sentence now opens on a handle from the previous one.
3. Combining. Six sentences become four. Two arms were used in the randomisation was a whole sentence carrying one fact; it now rides along in the enrolment sentence.
4. Connector direction. Moreover promised more good news and delivered a failed endpoint. Despite this signals the contrast correctly.
5. End position. The most important fact — the primary endpoint was not met — now lands in the final position of the final sentence, where English puts emphasis.
Score yourself: the parallelism fix is mechanical and most people get it. The connector fix in point 4 is the one that separates competent writers from good ones, because it requires you to notice what the paragraph is actually arguing.
“The objectives of this study were to assess the safety of the regimen, to evaluate its efficacy, and to measure pharmacokinetic parameters. We enrolled 240 patients and randomised them to two arms. Of these, 18 discontinued because of adverse events, most commonly neutropenia. Despite this manageable safety profile, the primary endpoint was not met.”
What changed:
1. Parallelism. Three infinitives — to assess, to evaluate, to measure — replacing an infinitive, a noun and a gerund.
2. Chaining. 240 patients → Of these, 18 → most commonly neutropenia. Each sentence now opens on a handle from the previous one.
3. Combining. Six sentences become four. Two arms were used in the randomisation was a whole sentence carrying one fact; it now rides along in the enrolment sentence.
4. Connector direction. Moreover promised more good news and delivered a failed endpoint. Despite this signals the contrast correctly.
5. End position. The most important fact — the primary endpoint was not met — now lands in the final position of the final sentence, where English puts emphasis.
Score yourself: the parallelism fix is mechanical and most people get it. The connector fix in point 4 is the one that separates competent writers from good ones, because it requires you to notice what the paragraph is actually arguing.
Reflection
Take one paragraph from your own writing. Underline the first four words and the last four words of each sentence. Does the end of each sentence connect to the start of the next?
If the underlined pairs have nothing in common, that is exactly why the paragraph feels choppy despite every sentence being correct.
Next week
4.1 — Numbers, units and abbreviations
Week 4 · Module 1
Numbers, units and abbreviations
The conventions that make a draft look professionally edited — small, mechanical, entirely learnable, and the first thing a QC reviewer checks
Your scenario
Your manuscript is returned before peer review. The editorial office letter says the submission does not meet formatting requirements. You scan the paper expecting a structural problem and find none. The issues are: 5mg with no space, p values that do not follow the journal’s own style, hyphens where en dashes belong, and four abbreviations never expanded. Two hours of work, and a four-week delay.
None of this is grammar and none of it is hard. It is convention, and conventions are enforced by people whose job is to enforce them. Learn the twenty rules below and you will never lose a submission to formatting again.
Numbers
| Rule | Example |
|---|---|
| Use numerals for all measurements and data, even below ten. | 3 patients, 5 mg, 2 cycles |
| Spell out a number that begins a sentence — or better, rewrite so it does not. | Twelve patients withdrew. Better: In total, 12 patients withdrew. |
| Never begin a sentence with a numeral or an abbreviation. | Not 12 patients withdrew. Not PFS was 8.2 months. at paragraph start after a heading |
| Use a leading zero for values below 1. | 0.05, not .05 |
| Match decimal places to the precision of your measurement, and be consistent. | 8.2 and 5.1 months, not 8.2 and 5 months |
| Percentages: numeral plus the symbol, no space. | 46%, not 46 % or 46 percent |
| Give both the count and the percentage for proportions. | 18/240 (7.5%) |
| Ranges take an en dash, not a hyphen. | 6.1–10.4, weeks 4–12 |
| Do not use an en dash after from or between. | from 4 to 12 weeks, between 4 and 12 weeks |
Units
| Rule | Example |
|---|---|
| A space separates the number from the unit — including before °C. | 5 mg, 37 °C, 10 mL |
| Exception: no space before % or °. | 46%, a 90° angle |
| Units never take a plural s. | 5 mg, not 5 mgs |
| Units take no full stop unless ending a sentence. | 10 mL, not 10 mL. |
| Capitalisation matters. mg ≠ Mg. mL is preferred over ml. L is capital. | mg milligram · Mg magnesium |
| Use a solidus for rates, and repeat the unit in ranges. | 5 mg/kg, from 5 mg to 10 mg |
| Repeat the unit on both sides of a comparison. | 8.2 vs 5.1 months — the unit belongs once at the end here, but 5 mg and 10 mg takes it twice |
Statistics conventions
How a p value is written depends on the target journal. House styles differ on capitalisation, spacing around the operator and the leading zero, so check the journal’s instructions for authors before you format. For example, AMA style (JAMA and the JAMA Network journals) uses an italic capital
Report exact values down to three decimals; below that use
Confidence intervals:
P with spaces and no leading zero: P = .03, P < .001. The Lancet uses a lower-case p with no spaces: p=0·03. Many other journals use p = 0.03. None of these is wrong. What is wrong is mixing styles within one document or ignoring the journal’s own rule. The examples in this course use p = 0.03 as a neutral default.Report exact values down to three decimals; below that use
p < 0.001. Never write p = 0.000 — a probability is not zero.Confidence intervals:
95% CI 6.1–10.4. State the interval, not just the point estimate. Give the hazard ratio with its CI: HR 0.68 (95% CI 0.51–0.91).Abbreviations
Wrong
Right
PFS (progression-free survival) was the endpoint.introduced backwards
Progression-free survival (PFS) was the endpoint.
…PFS… Progression-free survival was 8.2 months.reverting after defining
…PFS… PFS was 8.2 months.once defined, always use it
P.F.S. · C.S.R. · e.t.c.
PFS · CSR · etc.no internal full stops
Defining an abbreviation used only twice
Write it out both timesan abbreviation used fewer than three times costs the reader more than it saves
Abbreviations in the title or in the abstract without redefining
Expand in the title, and expand again on first use in the abstract and in the bodythe abstract is read independently
e.g. and i.e. used interchangeably
e.g. = for example · i.e. = that isboth take a comma after in US style
Which style guide?
Journals in medicine mostly follow AMA or their own house style derived from it. Regulatory documents follow the sponsor style guide, which usually derives from ICH conventions plus in-house rules. Life sciences journals often use CSE.
You do not need to memorise all three. You need to know that they differ, ask which one applies before you start, and apply it consistently. Consistency inside a document matters more than which convention you picked.
You do not need to memorise all three. You need to know that they differ, ask which one applies before you start, and apply it consistently. Consistency inside a document matters more than which convention you picked.
Recognition drills
Drill 1 of 3
Which follows standard scientific style?
A“Patients received 5 mg twice daily, stored at 37 °C, with p = 0.03.”
B“Patients received 5mgs twice daily at 37°C storage, with p=0.03.”
C“Patients received 5 mg twice daily, stored at 37 C, with P = 0.03.”
D“Patients received 5-mg twice daily, stored at 37-°C, with p-0.03.”
Drill 2 of 3
Which handles the abbreviation correctly across a paragraph?
A“OS (overall survival) was the primary endpoint. Median OS was 18.4 months.”
B“Overall survival (OS) was the primary endpoint. Median OS was 18.4 months. OS favoured the treatment arm.”
C“Overall survival (OS) was the primary endpoint. Median overall survival was 18.4 months.”
D“O.S. was the primary endpoint. Overall survival (O.S.) was 18.4 months.”
Drill 3 of 3
Which correctly reports a proportion and an interval?
A“Discontinuation occurred in 7.5% of patients (HR 0.68, CI 0.51-0.91).”
B“Discontinuation occurred in 18 of 240 patients; HR .68 (CI .51–.91).”
C“Discontinuation occurred in 18/240 patients (7.5%); HR 0.68 (95% CI 0.51–0.91).”
D“Discontinuation occurred in 18/240 patients (7.5%), p = 0.000.”
Production drill
Fix every style error. There are eleven.
“240 patients received 5mgs of study drug twice daily. Samples were stored at -80°C. PFS (progression-free survival) was the primary endpoint. Median progression-free survival was 8.2 months versus 5 months in the control arm (HR .68, CI 0.51-0.91, p=0.000). Grade 3 neutropenia occurred in 14 percent. Discontinuation occurred in 18 patients. C.S.R. preparation is ongoing.”
Corrected: “In total, 240 patients received 5 mg of study drug twice daily. Samples were stored at −80 °C. Progression-free survival (PFS) was the primary endpoint. Median PFS was 8.2 months versus 5.1 months in the control arm (HR 0.68, 95% CI 0.51–0.91, p < 0.001). Grade 3 neutropenia occurred in 14%. Discontinuation occurred in 18/240 patients (7.5%). Clinical study report (CSR) preparation is ongoing.”
The eleven fixes:
1. Sentence began with a numeral → In total, 240 patients.
2. 5mgs → 5 mg — space added, plural s removed.
3. -80°C → −80 °C — minus sign, and a space before the degree symbol.
4. PFS (progression-free survival) → expanded first, abbreviation in brackets.
5. Median progression-free survival → Median PFS — once defined, use it.
6. 5 months → 5.1 months — match the decimal precision of 8.2.
7. .68 → 0.68 — leading zero.
8. CI → 95% CI — state the confidence level.
9. Hyphen → en dash in the interval.
10. p=0.000 → p < 0.001, italic, with spaces.
11. 14 percent → 14%; 18 patients → 18/240 (7.5%); C.S.R. → expanded on first use as CSR.
Score yourself: 9–11 means you are ready for QC work. Under 6 means you have never been trained on a style guide — pick AMA, download the summary, and keep it open for a month.
The eleven fixes:
1. Sentence began with a numeral → In total, 240 patients.
2. 5mgs → 5 mg — space added, plural s removed.
3. -80°C → −80 °C — minus sign, and a space before the degree symbol.
4. PFS (progression-free survival) → expanded first, abbreviation in brackets.
5. Median progression-free survival → Median PFS — once defined, use it.
6. 5 months → 5.1 months — match the decimal precision of 8.2.
7. .68 → 0.68 — leading zero.
8. CI → 95% CI — state the confidence level.
9. Hyphen → en dash in the interval.
10. p=0.000 → p < 0.001, italic, with spaces.
11. 14 percent → 14%; 18 patients → 18/240 (7.5%); C.S.R. → expanded on first use as CSR.
Score yourself: 9–11 means you are ready for QC work. Under 6 means you have never been trained on a style guide — pick AMA, download the summary, and keep it open for a month.
Reflection
Which style guide governs the documents you write — AMA, ICH plus a sponsor guide, CSE, or a journal house style? If you do not know, that is your first action item.
Asking this question at the start of a project marks you as experienced. Not asking it is what produces two rounds of avoidable QC comments.
Next module
4.2 — The sentence clinic
Week 4 · Module 2
The sentence clinic
Eleven units of rules, applied one sentence at a time. This is where recognition becomes production
Your scenario
You have finished eleven units and you can name every rule. Then you sit down to write, and your hand produces the same sentences it always did. This is normal and it is the hardest part of the whole course. Recognition transfers to production only through repeated diagnosis under mixed conditions — which is what this module is.
Each sentence below contains between one and three errors drawn from anywhere in Weeks 1 to 4. Unlike the unit drills, you are not told which category. That is the point — real editing gives you no category label.
Work in order. Write out each corrected sentence in full before revealing the answer. Reading the answer without writing first teaches you nothing, because recognition is not what you are short of.
Your diagnostic checklist
Run these eleven questions on every sentence:
1. Articles — countable? generic? already known?
2. Tense — does it match the IMRAD map?
3. Agreement — what is the head noun of the subject?
4. Countability — any never-s nouns?
5. Prepositions — any of the fixed pairings?
6. Modifiers — does the opening phrase attach to a noun that can do it? Where is only?
7. Punctuation — comma splice? hyphen before the noun?
8. Concision — buried verbs? empty frames? deadweight?
9. Hedging — does the verb match the design?
10. Register — anything from the kill list?
11. Style — numbers, units, abbreviations.
1. Articles — countable? generic? already known?
2. Tense — does it match the IMRAD map?
3. Agreement — what is the head noun of the subject?
4. Countability — any never-s nouns?
5. Prepositions — any of the fixed pairings?
6. Modifiers — does the opening phrase attach to a noun that can do it? Where is only?
7. Punctuation — comma splice? hyphen before the noun?
8. Concision — buried verbs? empty frames? deadweight?
9. Hedging — does the verb match the design?
10. Register — anything from the kill list?
11. Style — numbers, units, abbreviations.
The twelve sentences
Sentences 1–4
Write out a corrected version of each before revealing.
1. “The data shows that a majority of the patients who were treated with the combination regimen experienced a reduction in tumour size.”
2. “After adjusting for age and prior therapy, the risk of progression was only reduced in the biomarker-positive subgroup.”
3. “We have performed an evaluation of the safety profile in March 2025, and the regimen was well-tolerated, no serious adverse events occurred.”
4. “Recent researches provide evidences that the therapy reduces mortality in elderly patients.”
2. “After adjusting for age and prior therapy, the risk of progression was only reduced in the biomarker-positive subgroup.”
3. “We have performed an evaluation of the safety profile in March 2025, and the regimen was well-tolerated, no serious adverse events occurred.”
4. “Recent researches provide evidences that the therapy reduces mortality in elderly patients.”
1. “The data show that most patients receiving the combination had tumour reduction.”
Three errors: data takes a plural verb (1.4); a majority of → most and experienced a reduction in → had … reduction or simply tumours shrank (3.1); who were treated with → receiving (3.1). Twenty-five words to eleven.
2. “After adjusting for age and prior therapy, we found that the risk of progression was reduced only in the biomarker-positive subgroup.”
Two errors: the risk cannot adjust for anything — dangling participle, needs a subject that can (2.3); only reduced → reduced only in, because the restriction belongs to the subgroup (2.3). Note that if the design was observational you would also need was lower in rather than was reduced (3.2).
3. “We evaluated the safety profile in March 2025. The regimen was well tolerated, and no serious adverse events occurred.”
Four errors: present perfect cannot take a finished time (1.3); performed an evaluation of → evaluated (3.1); well-tolerated takes no hyphen after the verb (2.4); comma splice (2.4).
4. “Recent research provides evidence that the therapy is associated with lower mortality in elderly patients.”
Three errors: researches and evidences are uncountable (2.1), and the verb follows to singular; reduces is a causal verb — unless this is randomised evidence it must be association language (3.2).
Three errors: data takes a plural verb (1.4); a majority of → most and experienced a reduction in → had … reduction or simply tumours shrank (3.1); who were treated with → receiving (3.1). Twenty-five words to eleven.
2. “After adjusting for age and prior therapy, we found that the risk of progression was reduced only in the biomarker-positive subgroup.”
Two errors: the risk cannot adjust for anything — dangling participle, needs a subject that can (2.3); only reduced → reduced only in, because the restriction belongs to the subgroup (2.3). Note that if the design was observational you would also need was lower in rather than was reduced (3.2).
3. “We evaluated the safety profile in March 2025. The regimen was well tolerated, and no serious adverse events occurred.”
Four errors: present perfect cannot take a finished time (1.3); performed an evaluation of → evaluated (3.1); well-tolerated takes no hyphen after the verb (2.4); comma splice (2.4).
4. “Recent research provides evidence that the therapy is associated with lower mortality in elderly patients.”
Three errors: researches and evidences are uncountable (2.1), and the verb follows to singular; reduces is a causal verb — unless this is randomised evidence it must be association language (3.2).
Sentences 5–8
Same method — write first, reveal second.
5. “Respected Sir, kindly find attached the revised manuscript and do the needful at the earliest regarding the same.”
6. “The incidence of treatment emergent adverse events in elderly patients were higher, however the difference was not statistically-significant.”
7. “In the present study we have made an attempt to investigate about the efficacy of the drug, which comprised of three cycles.”
8. “240 patients received 5mgs twice daily and each of the investigators were blinded to the allocation.”
6. “The incidence of treatment emergent adverse events in elderly patients were higher, however the difference was not statistically-significant.”
7. “In the present study we have made an attempt to investigate about the efficacy of the drug, which comprised of three cycles.”
8. “240 patients received 5mgs twice daily and each of the investigators were blinded to the allocation.”
5. “Dear Dr Mehta, I have attached the revised manuscript. Could you confirm receipt by Friday?”
Five register errors: Respected Sir, kindly, do the needful, at the earliest, the same (3.3). Note the rewrite also replaces a vague request with a specific one and a date.
6. “The incidence of treatment-emergent adverse events in elderly patients was higher; however, the difference was not statistically significant.”
Four errors: hyphen in the compound modifier before the noun (2.4); the head noun is incidence, singular (1.4); comma splice before however (2.4); no hyphen after an -ly adverb (2.4).
7. “We investigated the efficacy of the drug, which comprised three cycles.”
Four errors: In the present study we have made an attempt to is deadweight (3.1); investigate about takes no preposition (2.2); comprised of takes none either (2.2). Twenty-four words to twelve.
8. “In total, 240 patients received 5 mg twice daily, and each of the investigators was blinded to the allocation.”
Three errors: sentence began with a numeral (4.1); 5mgs — space missing, plural s wrong (4.1); each … were — each is always singular (1.4).
Five register errors: Respected Sir, kindly, do the needful, at the earliest, the same (3.3). Note the rewrite also replaces a vague request with a specific one and a date.
6. “The incidence of treatment-emergent adverse events in elderly patients was higher; however, the difference was not statistically significant.”
Four errors: hyphen in the compound modifier before the noun (2.4); the head noun is incidence, singular (1.4); comma splice before however (2.4); no hyphen after an -ly adverb (2.4).
7. “We investigated the efficacy of the drug, which comprised three cycles.”
Four errors: In the present study we have made an attempt to is deadweight (3.1); investigate about takes no preposition (2.2); comprised of takes none either (2.2). Twenty-four words to twelve.
8. “In total, 240 patients received 5 mg twice daily, and each of the investigators was blinded to the allocation.”
Three errors: sentence began with a numeral (4.1); 5mgs — space missing, plural s wrong (4.1); each … were — each is always singular (1.4).
Sentences 9–12 — the hard four
These carry claim-calibration errors alongside the mechanical ones. Diagnose both.
9. “Our retrospective analysis proves that early initiation of the therapy definitely prevents hospitalisation.”
10. “Using a Cox proportional hazards model, the treatment effect was estimated, and there are several studies which have shown similar findings.”
11. “The objectives were to assess safety, evaluation of efficacy, and measuring PK; the study comprised of two arms and was superior than placebo.”
12. “It is to be noted that a number of patient was lost to follow up, however the amount of withdrawals were less than anticipated.”
10. “Using a Cox proportional hazards model, the treatment effect was estimated, and there are several studies which have shown similar findings.”
11. “The objectives were to assess safety, evaluation of efficacy, and measuring PK; the study comprised of two arms and was superior than placebo.”
12. “It is to be noted that a number of patient was lost to follow up, however the amount of withdrawals were less than anticipated.”
9. “In this retrospective analysis, earlier initiation of therapy was associated with fewer hospitalisations.”
Three errors, one of them serious: proves and prevents are both causal and unavailable to a retrospective design (3.2); definitely has no place in any scientific claim (3.2); the therapy → therapy, generic (1.2). The claim-calibration error here is the one that would fail medical-legal review.
10. “We estimated the treatment effect using a Cox proportional hazards model. Several studies report similar findings.”
Three errors: the effect cannot use a model — dangling participle (2.3); there are several studies which have shown is an empty frame (3.1); splitting into two sentences also fixes the overlong join.
11. “The objectives were to assess safety, to evaluate efficacy, and to measure pharmacokinetics. The study comprised two arms and was superior to placebo.”
Four errors: three different grammatical shapes in the objectives (3.4); comprised of (2.2); superior than → superior to (2.2); PK expanded on first use (4.1). Note also that was superior to placebo is only defensible if this was a randomised superiority trial (3.2).
12. “Several patients were lost to follow-up; however, the number of withdrawals was fewer than anticipated.”
Six errors: It is to be noted that is deadweight (3.1); a number of patient was → plural on both (1.4); follow up as a noun takes a hyphen (2.4); comma splice before however (2.4); the amount of withdrawals → the number of, countable (2.1); the number … were → was, and less → fewer (1.4, 2.1).
Three errors, one of them serious: proves and prevents are both causal and unavailable to a retrospective design (3.2); definitely has no place in any scientific claim (3.2); the therapy → therapy, generic (1.2). The claim-calibration error here is the one that would fail medical-legal review.
10. “We estimated the treatment effect using a Cox proportional hazards model. Several studies report similar findings.”
Three errors: the effect cannot use a model — dangling participle (2.3); there are several studies which have shown is an empty frame (3.1); splitting into two sentences also fixes the overlong join.
11. “The objectives were to assess safety, to evaluate efficacy, and to measure pharmacokinetics. The study comprised two arms and was superior to placebo.”
Four errors: three different grammatical shapes in the objectives (3.4); comprised of (2.2); superior than → superior to (2.2); PK expanded on first use (4.1). Note also that was superior to placebo is only defensible if this was a randomised superiority trial (3.2).
12. “Several patients were lost to follow-up; however, the number of withdrawals was fewer than anticipated.”
Six errors: It is to be noted that is deadweight (3.1); a number of patient was → plural on both (1.4); follow up as a noun takes a hyphen (2.4); comma splice before however (2.4); the amount of withdrawals → the number of, countable (2.1); the number … were → was, and less → fewer (1.4, 2.1).
Score yourself honestly
Count the errors you caught out of the total, which is 44 across the twelve sentences.
35 or more — your production is close to your recognition. Go to Module 4.3.
25–34 — solid, but a specific category is leaking. Look at which unit numbers appear most in your misses and re-do that unit's production drill.
Under 25 — normal at this stage. Repeat this module in three days without looking at the answers first. The gap between recognition and production closes with spacing, not with more reading.
35 or more — your production is close to your recognition. Go to Module 4.3.
25–34 — solid, but a specific category is leaking. Look at which unit numbers appear most in your misses and re-do that unit's production drill.
Under 25 — normal at this stage. Repeat this module in three days without looking at the answers first. The gap between recognition and production closes with spacing, not with more reading.
Reflection
List the unit numbers of the errors you missed. Which two units account for most of your misses?
Two units will account for most of it. Those are the two to re-drill — not all eleven.
Next module
4.3 — The abstract clinic
Week 4 · Module 3
The abstract clinic
Moving from the sentence to the paragraph — where tense, hedging, cohesion and word limits all have to be satisfied at once
Your scenario
The abstract is the only part of your paper most people will ever read. It is also the part a hiring manager will ask you to write in an assessment, because it tests everything at once — tense discipline, claim calibration, cohesion, and the ability to hit a hard word limit without cutting facts.
Cross-Course Portfolio Bridge — Polishing Portfolio Piece 1
In Course 1 (Module 2.3: Abstract Writing Clinic), you drafted your first structured abstract (Portfolio Piece 1). In this clinic, you apply all 12 linguistic rules — eliminating nominalisations, enforcing strict IMRAD tense shifts, calibrating claims against trial design, and cutting filler words — to elevate that draft into an agency-grade, publication-ready portfolio piece.
The structured abstract, section by section
| Section | Words | Tense | What belongs here |
|---|---|---|---|
| Background | 30–50 | present + present perfect | The standing fact, the gap, one sentence of context. No citations. |
| Methods | 50–70 | past | Design, population, intervention, comparator, primary endpoint. Enough to judge the evidence level. |
| Results | 70–100 | past | Numbers with denominators and intervals. Primary endpoint first. Key safety. |
| Conclusions | 25–40 | present, hedged to design | What the data support and nothing beyond. One implication. |
The three failures assessors look for
1. Conclusions that outrun the design. A single-arm study concluding superiority. A retrospective study concluding causation.
2. Results with no denominators. Percentages with no counts, point estimates with no intervals.
3. Methods that omit the comparator. Without it, the reader cannot interpret any of the numbers that follow.
These three are checked before anything else, including your prose.
2. Results with no denominators. Percentages with no counts, point estimates with no intervals.
3. Methods that omit the comparator. Without it, the reader cannot interpret any of the numbers that follow.
These three are checked before anything else, including your prose.
The clinic
Below is a real-shaped abstract at 287 words, containing errors from every unit in this course. Rewrite it to under 250 words, correcting everything, without losing a single fact.
The study is a randomised, double-blind, placebo-controlled Phase III trial. Note that carefully — it determines what your Conclusions section is allowed to say.
The full rebuild
Rewrite in under 250 words. Correct every error. Lose no facts.
Background: It is to be noted that the advanced gastric cancer was a leading cause of the cancer death worldwide. Several studies demonstrated poor outcomes with the standard chemotherapy. In the present study we have made an attempt to carry out an investigation into a novel agent.
Methods: This was a randomised, double-blind, placebo-controlled Phase III trial. 412 patients are randomised in a 1:1 ratio. The regimen comprised of six cycles. Primary endpoint was overall survival. Each of the investigators were blinded to allocation. We collected the informations on adverse events at each visit.
Results: Median overall survival was 14.2 months versus 11 months (HR .74, CI 0.59-0.93, p=0.009). Grade 3 or higher adverse events occurred in 34 percent. Discontinuation due to adverse events occurred in 27 patients. Using a Cox model, subgroup effects were estimated. The benefit was only seen in patients under 65.
Conclusions: Our study proves that the agent definitely reduces mortality and should be adopted as standard of care. It is a matter of great pleasure to present these findings, and we kindly request the readers to revert with their views on the same.
Methods: This was a randomised, double-blind, placebo-controlled Phase III trial. 412 patients are randomised in a 1:1 ratio. The regimen comprised of six cycles. Primary endpoint was overall survival. Each of the investigators were blinded to allocation. We collected the informations on adverse events at each visit.
Results: Median overall survival was 14.2 months versus 11 months (HR .74, CI 0.59-0.93, p=0.009). Grade 3 or higher adverse events occurred in 34 percent. Discontinuation due to adverse events occurred in 27 patients. Using a Cox model, subgroup effects were estimated. The benefit was only seen in patients under 65.
Conclusions: Our study proves that the agent definitely reduces mortality and should be adopted as standard of care. It is a matter of great pleasure to present these findings, and we kindly request the readers to revert with their views on the same.
Model rewrite — 213 words:
Background: Advanced gastric cancer is a leading cause of cancer death worldwide, and outcomes with standard chemotherapy remain poor. We investigated a novel agent in this setting.
Methods: In this randomised, double-blind, placebo-controlled Phase III trial, 412 patients were randomised 1:1 to six cycles of the study regimen or placebo. The primary endpoint was overall survival. Investigators were blinded to allocation, and adverse-event information was collected at each visit.
Results: Median overall survival was 14.2 months with the study regimen versus 11.4 months with placebo (HR 0.74, 95% CI 0.59–0.93; p = 0.009). Grade 3 or higher adverse events occurred in 34% of patients, and 27/412 (6.6%) discontinued because of adverse events. In an exploratory subgroup analysis using a Cox proportional hazards model, the benefit was confined to patients aged under 65 years.
Conclusions: The study regimen improved overall survival compared with placebo in advanced gastric cancer. The apparent restriction of benefit to younger patients was exploratory and requires prospective confirmation.
Check your version against these fourteen:
1. the advanced gastric cancer was → Advanced gastric cancer is — article and tense (1.2, 1.3).
2. demonstrated → the standing fact moves to present (1.3).
3. In the present study we have made an attempt to carry out an investigation into → We investigated (3.1).
4. 412 patients are randomised → sentence recast so it does not open on a numeral, and past tense for Methods (4.1, 1.3).
5. comprised of → of deleted (2.2).
6. Each of the investigators were → Investigators were (1.4).
7. the informations → information, uncountable (2.1).
8. 11 months → 11.4 months — match decimal precision (4.1).
9. .74 → 0.74; CI → 95% CI; hyphen → en dash; p=0.009 → p = 0.009 italic with spaces (4.1).
10. 34 percent → 34%; 27 patients → 27/412 (6.6%) — denominator supplied (4.1).
11. Using a Cox model, subgroup effects were estimated — dangling participle, recast (2.3).
12. only seen in → confined to, and the analysis labelled exploratory — the original implied a prespecified finding (2.3, 3.2).
13. proves … definitely reduces mortality … should be adopted as standard of care → the randomised design licenses improved, but not proves, not definitely, and not a standard-of-care recommendation from one trial (3.2).
14. Ceremonial closing and revert … on the same deleted entirely (3.3).
The one that matters most: number 13. This is a randomised trial, so you were entitled to a causal verb — improved — and you should have used it. Hedging it down to was associated with would have been a calibration error in the other direction (3.2). Getting this right in both directions is the skill the whole course builds towards.
Background: Advanced gastric cancer is a leading cause of cancer death worldwide, and outcomes with standard chemotherapy remain poor. We investigated a novel agent in this setting.
Methods: In this randomised, double-blind, placebo-controlled Phase III trial, 412 patients were randomised 1:1 to six cycles of the study regimen or placebo. The primary endpoint was overall survival. Investigators were blinded to allocation, and adverse-event information was collected at each visit.
Results: Median overall survival was 14.2 months with the study regimen versus 11.4 months with placebo (HR 0.74, 95% CI 0.59–0.93; p = 0.009). Grade 3 or higher adverse events occurred in 34% of patients, and 27/412 (6.6%) discontinued because of adverse events. In an exploratory subgroup analysis using a Cox proportional hazards model, the benefit was confined to patients aged under 65 years.
Conclusions: The study regimen improved overall survival compared with placebo in advanced gastric cancer. The apparent restriction of benefit to younger patients was exploratory and requires prospective confirmation.
Check your version against these fourteen:
1. the advanced gastric cancer was → Advanced gastric cancer is — article and tense (1.2, 1.3).
2. demonstrated → the standing fact moves to present (1.3).
3. In the present study we have made an attempt to carry out an investigation into → We investigated (3.1).
4. 412 patients are randomised → sentence recast so it does not open on a numeral, and past tense for Methods (4.1, 1.3).
5. comprised of → of deleted (2.2).
6. Each of the investigators were → Investigators were (1.4).
7. the informations → information, uncountable (2.1).
8. 11 months → 11.4 months — match decimal precision (4.1).
9. .74 → 0.74; CI → 95% CI; hyphen → en dash; p=0.009 → p = 0.009 italic with spaces (4.1).
10. 34 percent → 34%; 27 patients → 27/412 (6.6%) — denominator supplied (4.1).
11. Using a Cox model, subgroup effects were estimated — dangling participle, recast (2.3).
12. only seen in → confined to, and the analysis labelled exploratory — the original implied a prespecified finding (2.3, 3.2).
13. proves … definitely reduces mortality … should be adopted as standard of care → the randomised design licenses improved, but not proves, not definitely, and not a standard-of-care recommendation from one trial (3.2).
14. Ceremonial closing and revert … on the same deleted entirely (3.3).
The one that matters most: number 13. This is a randomised trial, so you were entitled to a causal verb — improved — and you should have used it. Hedging it down to was associated with would have been a calibration error in the other direction (3.2). Getting this right in both directions is the skill the whole course builds towards.
Reflection
Count the words in your rewrite. Did you come in under 250 with all the facts intact? Which section was hardest to compress?
Results is almost always the hardest, because every number has to keep its denominator and its interval. That is the section where cutting words most easily turns into cutting facts.
Next module
4.4 — Capstone and exit assessment
Week 4 · Module 4
Capstone and exit assessment
A full manuscript section, thirty seeded errors, no category labels — and an honest measure of whether four weeks changed how you write
Your scenario
This is the task an agency will actually set you. A degraded Methods and Results section, no guidance, a clean rewrite expected. The pass mark used in real hiring is roughly 80% of seeded errors caught, with no new errors introduced. That second condition is what most candidates fail on.
How to do this properly
Set a timer for 45 minutes. Write the full corrected text — do not annotate, do not list the errors, write the finished passage as you would submit it. Only then reveal the rubric and score yourself.
If you read the rubric first, this exercise measures nothing and you will have learned nothing from it.
If you read the rubric first, this exercise measures nothing and you will have learned nothing from it.
Capstone task
Rewrite this Methods and Results section in full. The study is a randomised, open-label Phase III trial with a prespecified primary endpoint of progression-free survival.
Methods. It is to be noted that in the present study we have made an attempt to carry out an evaluation of the novel MET inhibitor. 318 patients are randomised in a 1:1 ratio between January 2024 and September 2024. The regimen comprised of the study drug given at 5mgs twice daily. Each of the investigators were blinded to the outcome assessment. Inclusion criterias were age 18 or above, having histological confirmation of disease, and must have adequate organ function. The objectives were to assess safety, evaluation of efficacy, and measuring PK. Using a stratified Cox proportional hazards model, hazard ratios were estimated. We collected the informations on adverse events at each visit and discussed about the findings with the steering committee.
Results. The data shows that a majority of the patients responded. Median progression-free survival was 9.4 months versus 6 months (HR .71, CI 0.55-0.92, p=0.009). PFS (progression-free survival) favoured the treatment arm. The incidence of treatment emergent adverse events in the two arms were comparable, however grade 3 events were more frequent with the study drug. Grade 3 neutropenia occurred in 22 percent. Discontinuation due to adverse events occurred in 31 patients. Treatment interruption resulted from dose reduction in 12 patients. There were less protocol deviations than anticipated. The benefit was only seen in patients under 65, which proves that age definitely predicts response.
Results. The data shows that a majority of the patients responded. Median progression-free survival was 9.4 months versus 6 months (HR .71, CI 0.55-0.92, p=0.009). PFS (progression-free survival) favoured the treatment arm. The incidence of treatment emergent adverse events in the two arms were comparable, however grade 3 events were more frequent with the study drug. Grade 3 neutropenia occurred in 22 percent. Discontinuation due to adverse events occurred in 31 patients. Treatment interruption resulted from dose reduction in 12 patients. There were less protocol deviations than anticipated. The benefit was only seen in patients under 65, which proves that age definitely predicts response.
Model answer
Methods. We evaluated a novel MET inhibitor in this randomised, open-label Phase III trial. Between January and September 2024, 318 patients were randomised 1:1 to the study drug at 5 mg twice daily or to control. Investigators were blinded to outcome assessment. Inclusion criteria were age 18 years or older, histologically confirmed disease, and adequate organ function. The objectives were to assess safety, to evaluate efficacy, and to measure pharmacokinetics. We estimated hazard ratios using a stratified Cox proportional hazards model, collected information on adverse events at each visit, and discussed the findings with the steering committee.
Results. The data show that most patients responded. Median progression-free survival (PFS) was 9.4 months with the study drug versus 6.2 months with control (HR 0.71, 95% CI 0.55–0.92; p = 0.009), favouring the treatment arm. The incidence of treatment-emergent adverse events was comparable between arms; however, grade 3 events were more frequent with the study drug. Grade 3 neutropenia occurred in 22% of patients, and 31/318 (9.7%) discontinued because of adverse events. Dose reduction resulted from treatment interruption in 12 patients. There were fewer protocol deviations than anticipated. In an exploratory analysis, the benefit was confined to patients aged under 65 years; this finding requires prospective confirmation.
The thirty seeded errors — tick each one you caught
Week 1 — articles, tense, agreement (8)
1. It is to be noted that — deadweight opening.
2. we have made an attempt to carry out an evaluation of → we evaluated.
3. are randomised → were randomised — Methods, past tense.
4. Each of the investigators were → Investigators were.
5. The data shows → The data show.
6. the informations → information.
7. criterias → criteria.
8. The incidence … were → was — head noun is incidence.
Week 2 — countables, prepositions, modifiers, punctuation (9)
9. comprised of → comprised or recast.
10. discussed about → discussed.
11. Using a stratified Cox model, hazard ratios were estimated — dangling participle.
12. treatment emergent → treatment-emergent.
13. Comma splice before however → semicolon.
14. less protocol deviations → fewer.
15. Treatment interruption resulted from dose reduction — reversed causality. The interruption caused the reduction, so dose reduction resulted from treatment interruption.
16. only seen in patients under 65 → confined to — misplaced only.
17. a majority of the patients → most patients.
Week 3 — concision, hedging, register, parallelism (7)
18. in the present study — redundant in a Methods section.
19. Inclusion criteria not parallel: adjective phrase, gerund phrase, modal clause → three noun phrases.
20. Objectives not parallel: infinitive, noun, gerund → three infinitives.
21. proves → unavailable; a subgroup finding does not prove anything.
22. definitely → delete.
23. age predicts response → the subgroup analysis was exploratory; label it and call for confirmation.
24. Six short Methods sentences combined into three — cohesion.
Week 4 — style conventions (6)
25. 318 patients are randomised — sentence opens on a numeral; recast.
26. 5mgs → 5 mg.
27. 6 months → 6.2 months — decimal precision must match 9.4.
28. .71 → 0.71; CI → 95% CI; hyphen → en dash; p=0.009 → p = 0.009.
29. PFS (progression-free survival) introduced backwards and after first use of the full term.
30. 22 percent → 22%; 31 patients → 31/318 (9.7%).
Scoring
24 or more caught (80%), no new errors introduced — pass. This is the standard applied in real agency editing tests. Your written English is no longer the thing holding you back.
18–23 — close. Identify which week your misses cluster in and redo that week's production drills. Retake this capstone in one week.
Under 18 — the material has not consolidated yet. Do not repeat the capstone immediately — go back to Module 4.2, the sentence clinic, and work through it twice with three days between. Sentence-level diagnosis has to be automatic before passage-level editing works.
If you introduced new errors: this is the failure mode assessors care about most, because an editor who adds errors is worse than no editor. The commonest cause is over-editing — rewriting sentences that were already correct. Change only what is wrong.
Methods. We evaluated a novel MET inhibitor in this randomised, open-label Phase III trial. Between January and September 2024, 318 patients were randomised 1:1 to the study drug at 5 mg twice daily or to control. Investigators were blinded to outcome assessment. Inclusion criteria were age 18 years or older, histologically confirmed disease, and adequate organ function. The objectives were to assess safety, to evaluate efficacy, and to measure pharmacokinetics. We estimated hazard ratios using a stratified Cox proportional hazards model, collected information on adverse events at each visit, and discussed the findings with the steering committee.
Results. The data show that most patients responded. Median progression-free survival (PFS) was 9.4 months with the study drug versus 6.2 months with control (HR 0.71, 95% CI 0.55–0.92; p = 0.009), favouring the treatment arm. The incidence of treatment-emergent adverse events was comparable between arms; however, grade 3 events were more frequent with the study drug. Grade 3 neutropenia occurred in 22% of patients, and 31/318 (9.7%) discontinued because of adverse events. Dose reduction resulted from treatment interruption in 12 patients. There were fewer protocol deviations than anticipated. In an exploratory analysis, the benefit was confined to patients aged under 65 years; this finding requires prospective confirmation.
The thirty seeded errors — tick each one you caught
Week 1 — articles, tense, agreement (8)
1. It is to be noted that — deadweight opening.
2. we have made an attempt to carry out an evaluation of → we evaluated.
3. are randomised → were randomised — Methods, past tense.
4. Each of the investigators were → Investigators were.
5. The data shows → The data show.
6. the informations → information.
7. criterias → criteria.
8. The incidence … were → was — head noun is incidence.
Week 2 — countables, prepositions, modifiers, punctuation (9)
9. comprised of → comprised or recast.
10. discussed about → discussed.
11. Using a stratified Cox model, hazard ratios were estimated — dangling participle.
12. treatment emergent → treatment-emergent.
13. Comma splice before however → semicolon.
14. less protocol deviations → fewer.
15. Treatment interruption resulted from dose reduction — reversed causality. The interruption caused the reduction, so dose reduction resulted from treatment interruption.
16. only seen in patients under 65 → confined to — misplaced only.
17. a majority of the patients → most patients.
Week 3 — concision, hedging, register, parallelism (7)
18. in the present study — redundant in a Methods section.
19. Inclusion criteria not parallel: adjective phrase, gerund phrase, modal clause → three noun phrases.
20. Objectives not parallel: infinitive, noun, gerund → three infinitives.
21. proves → unavailable; a subgroup finding does not prove anything.
22. definitely → delete.
23. age predicts response → the subgroup analysis was exploratory; label it and call for confirmation.
24. Six short Methods sentences combined into three — cohesion.
Week 4 — style conventions (6)
25. 318 patients are randomised — sentence opens on a numeral; recast.
26. 5mgs → 5 mg.
27. 6 months → 6.2 months — decimal precision must match 9.4.
28. .71 → 0.71; CI → 95% CI; hyphen → en dash; p=0.009 → p = 0.009.
29. PFS (progression-free survival) introduced backwards and after first use of the full term.
30. 22 percent → 22%; 31 patients → 31/318 (9.7%).
Scoring
24 or more caught (80%), no new errors introduced — pass. This is the standard applied in real agency editing tests. Your written English is no longer the thing holding you back.
18–23 — close. Identify which week your misses cluster in and redo that week's production drills. Retake this capstone in one week.
Under 18 — the material has not consolidated yet. Do not repeat the capstone immediately — go back to Module 4.2, the sentence clinic, and work through it twice with three days between. Sentence-level diagnosis has to be automatic before passage-level editing works.
If you introduced new errors: this is the failure mode assessors care about most, because an editor who adds errors is worse than no editor. The commonest cause is over-editing — rewriting sentences that were already correct. Change only what is wrong.
Retake the diagnostic
Go back to Module 1.1 and take the 24-question diagnostic again. Your first score is the baseline. The difference is the only honest measure of whether these four weeks changed anything.
Expect movement of one band. Band D to Band C, or Band C to Band B, is a normal and substantial four-week result. Band A on a retake usually means you were closer to it than the first attempt suggested.
What to do next
Keep drilling in your real work. Pick two units — the two your misses clustered in — and run their checklist over everything you write for the next month. Two units of sustained attention will do more than re-reading all twelve.
Build a portfolio piece. Write one 250-word structured abstract from a published trial in your therapy area, from the paper's data rather than its abstract. Then compare yours with the published version. That comparison is worth a month of exercises, and the piece itself is something you can send to an agency.
Then move to the medical writing course. This course fixed how you write sentences. The IMRAD Academy course teaches what to write, for which document, under which guideline.
Build a portfolio piece. Write one 250-word structured abstract from a published trial in your therapy area, from the paper's data rather than its abstract. Then compare yours with the published version. That comparison is worth a month of exercises, and the piece itself is something you can send to an agency.
Then move to the medical writing course. This course fixed how you write sentences. The IMRAD Academy course teaches what to write, for which document, under which guideline.
Final reflection
Your diagnostic score at the start, and now. Which two units changed your writing most? What is still hardest?
Write this down properly. In six months it is the only record you will have of what you actually fixed.
You have completed the course
Scientific English for Medical Writers
IMRAD Academy — Read like a scientist. Write like a professional.
Scientific English for Medical Writers · v1.0