Is Your Research Topic Doable? The FINER Test Every Trainee Should Run

The Topic Hassan Borrowed

Dr. Hassan Raza arrived early, which should have warned me.

“Sir, I have finished,” he announced, and put a folder on my desk. Inside was a complete UHS synopsis title page, an introduction, even a proforma. The topic: functional outcome after open reduction and internal fixation (ORIF) of distal radius fractures.

“You wrote all this in a week?”

He smiled. “A senior did this topic two years ago. Same department, same format. I only have to change the dates.”

Dr. Sumaira Talib looked up from her notebook. Dr. Hammad Ali leaned over to see the folder, clearly impressed. Dr. Zunaira Malik said nothing. Dr. Bushra Fatima raised an eyebrow. Dr. Junaid Rashid, who had spent fifteen years in medicine OPDs watching people take shortcuts, simply waited.

“Hassan,” I said, “it is a reasonable topic. Distal radius fractures are common, and outcome after fixation matters to patients. But before you change the dates, answer me five questions. If your topic passes all five, you may keep it. If it fails even one, you will find out now, and not eleven months from now when your data collection has collapsed.”

I wrote five letters on the whiteboard: F I N E R.

A Good Question Is Not Always a Good Project

Last week, every doctor in this room left with a PICO question. PICO tells you whether your question is clear. It does not tell you whether you can actually answer it, with your patients, your hospital, your time and your skills.

That is what FINER is for. The criteria come from the textbook Designing Clinical Research by Hulley, Cummings, Browner and colleagues, and they have been used for decades to test research questions before any work begins (1). A good research question should be:

• Feasible: you can recruit enough patients, in the time you have, with the skills and resources available.
• Interesting: to you, to your supervisor, and to the doctors who will read the paper.
• Novel: it confirms, extends or challenges what is known, or answers it in a setting where it has not been answered.
• Ethical: it can be done without exposing patients to unacceptable risk, and an ethics committee can approve it.
• Relevant: the answer could change practice, guide policy or shape future research.

In my experience, most trainees in Pakistan fail on the first letter. Not because their topics are bad, but because nobody asked them to count.

F for Feasible: Count Before You Commit

“Hassan, how many distal radius fractures are fixed with ORIF in your unit each month?”

“Many, sir. Lots.”

“Lots is not a number.”

He called the OT nurse from the corridor and asked her to photograph the register. The answer came back ten minutes later. Over the last six months, his unit had fixed an average of six to seven distal radius fractures a month with ORIF. Many of the others were managed with closed reduction and casting or K-wires, which would not qualify for his study.

Hassan went quiet. We have not calculated his sample size yet (that comes later in this series), but he already knew his senior’s synopsis had a number much larger than six a month could deliver in his study period.

“And the outcome is at three months after surgery?” I asked. “So the last patient must be operated at least three months before your study ends. How many of your patients come from outside Lahore?”

“Quite a few. Kasur, Okara, Sheikhupura.”

“Then some of them will not come back for follow-up. Every patient lost to follow-up is a patient missing from your results. You will need a working phone number for every patient, a follow-up date written on the discharge slip, and a plan for reminder calls. And you should expect some loss, and allow for it.”

Feasibility is a set of practical questions, and you should answer every one of them in writing before you draft a synopsis:

• Patients: How many eligible patients does your unit really see? Count them in the OPD register, OT register or admission record for the last six months. Do not guess.
• Time: How long is your study period, and does it leave room for follow-up, analysis and writing before your submission deadline?
• Follow-up: If your outcome is measured later, how many patients will realistically return?
• Tests and tools: Is every investigation available in your hospital, and free or affordable for patients? Does a validated questionnaire exist for your outcome?
• Skills: Do you, or someone on your team, have the skills the study needs, from the procedure itself to the statistics?
• Cooperation: Will the consultants, anaesthetists, nurses and laboratory staff you depend on actually support the study?

The cheapest time to discover a feasibility problem is today. The most expensive time is after your synopsis has been approved.

Hammad’s Trial Meets Reality

Hammad had been nodding throughout. “Sir, my trial is feasible. Laparoscopic cholecystectomy is done every day in our unit.”

“Numbers are only one part of it,” I said. “Your question compares low-pressure with standard-pressure pneumoperitoneum. Who sets the insufflation pressure?”

“The operating surgeon, sir.”

“And you are a first-year trainee. So every consultant who operates on your patients must agree, in advance, to use whichever pressure the randomisation assigns. The anaesthetists must agree too. Your ethics committee will want to see that. And a randomised trial should be registered in a trial registry before the first patient is enrolled.”

Hammad looked less certain. “So it is not feasible?”

“It may be very feasible. But feasibility for a trial means agreements, not just patients. Speak to your head of department this week. If the consultants agree, write that agreement into your synopsis.”

I for Interesting: You Will Live With This Topic

Interesting sounds like the soft letter. It is not. A dissertation or a first study stays with you for a year or more: synopsis, ethics approval, data collection, analysis, writing, submission and revisions. If you chose the topic only because a template existed, you will struggle to stay with it when data collection gets hard.

The topic must also interest your supervisor, who has to guide and sign every stage, and it must interest readers. A journal editor asks one question before anything else: will our readers care about this answer?

N for Novel: What It Really Means for a Trainee

Zunaira raised her hand, as quietly as always. “Sir, areca nut chewing and oral submucous fibrosis have been studied many times. Does that mean my topic is not novel?”

It was the most important question of the session.

“Novel does not mean that nobody in the world has ever asked the question,” I said. “Very few trainees will discover a new disease. Novel means your study adds something: data from a population where the answer is not known, a different setting, an updated picture because habits have changed, a better design than earlier studies, or an outcome others did not measure. If the products people chew in your area, or the way they chew them, differ from the settings of earlier studies, that is worth knowing.”

“But you only know what is missing after you have read what exists. That is why the literature search comes next in this series. A topic is not novel because it feels new to you. It is novel because you searched and found a real gap.”

The opposite is also true. Repeating a study that has been done many times in similar hospitals, with the same design and the same outcome, adds very little. Ethics committees and journal editors notice.

E for Ethical: Could an Ethics Committee Say Yes?

Ethics is not a form you fill in at the end. It is a test your question must pass at the start.

• Does any patient receive less than accepted standard care because of your study?
• Are there extra tests, visits, blood samples or radiation that exist only for your research? Are they justified?
• Can every participant give informed consent in a language they understand? If not, who consents for them, and is that acceptable?
• For a trial: is there genuine uncertainty about which option is better? Hammad’s question works because both pressure settings are used in routine practice, so neither group is denied accepted care.
• Will patient data stay confidential, with names and phone numbers kept apart from the data sheet?

An infeasible study is also an ethical problem. If a study cannot reach its sample size, the patients who did take part gave their time and trust to research that can never answer its question. Ethics committees know this, which is why many synopses are returned for feasibility reasons.

R for Relevant: Will the Answer Change Anything?

Bushra, who had been quiet since her difficult session on predatory journals, spoke up. “Sir, my study is about antenatal corticosteroids and breathing problems in preterm babies. Every obstetrician already gives steroids. Is that still relevant?”

“Is every eligible mother in your hospital actually receiving a complete course before delivery?” I asked.

She thought about the labour room at night, the women who arrive already in advanced labour, the referrals from district hospitals. “No, sir. Not all of them.”

“Then your answer is relevant. If your data show how neonatal outcomes differ between babies whose mothers did and did not complete the course in your own hospital, that speaks directly to your labour room, your referral system and your neonatal unit. Relevance does not mean a global discovery. It means the answer matters to someone who can act on it.”

The Topic That Is Too Big

Junaid had been writing steadily. Now he looked up. “Sir, FINER suggests I should narrow things down. My first idea was to study everything about hypertension control in our OPD: adherence, diet, salt intake, follow-up, complications.”

“And?”

“And it fails F. I cannot measure all of that properly, alone, in my OPD time.” He tapped his PICO sheet. “So I will measure one thing well: how common medication non-adherence is among hypertensive patients in the medicine OPD. The rest can be later studies.”

That is the most common lesson FINER teaches. Most first projects are too big, not too small. One clear question, answered properly, is worth more than five questions answered badly.

What Hassan Decided

By the end of the session, Hassan had not abandoned his topic. He had tested it.

His plan was simple: confirm the monthly ORIF numbers with his head of unit, check whether the trauma list from the neighbouring unit could also be included, collect phone numbers and give follow-up dates to every patient, and wait for a proper sample size calculation before promising any number in his synopsis. He also agreed to read his senior’s study as a reference and not as a template.

“It is still the same topic, sir,” he said, a little reluctantly.

“Yes. But now it is your study, and it has a chance of finishing.”

Your FINER Checklist

Before you write a single line of your synopsis, write one or two honest sentences under each heading:

• Feasible: eligible patients per month (from the register, not memory), study period, expected loss to follow-up, tests and tools available, skills, and who must agree.
• Interesting: why you, your supervisor and your readers care about this answer.
• Novel: what your study adds that is not already known for your population and setting (fill this in properly after your literature search).
• Ethical: risks, consent, standard care, confidentiality, and whether an ethics committee could approve it.
• Relevant: who will use your answer, and what they could change because of it.

If you cannot write an honest line for one of these, fix the question now. Changing a question costs you an evening. Changing a study after approval can cost you months.

Key Takeaways

• PICO makes a question clear. FINER tells you whether you can actually answer it.
• Most trainees fail on feasibility. Count eligible patients from registers, and plan for follow-up losses before you commit.
• Novel does not mean never studied. It means your study adds something real for your population or setting, which you can only show after a proper literature search.
• An infeasible study is an ethical problem, not just a practical one.
• A senior’s synopsis is a reference, not a template. Your patients, numbers and timeline are different.

Reference:

1. Browner WS, Newman TB, Cummings SR, Grady DG, Huang AJ, Kanaya AM, Pletcher MJ. Designing Clinical Research. 5th ed. Philadelphia: Wolters Kluwer; 2022.