What a PhD adds to a clinician: a practical look

Key points#

• A research doctorate is not a detour from clinical work but a way of thinking that carries into the exam room: framing a question, weighing what the data can and cannot say, and acting under uncertainty.
• Jasaman Tojjar holds an MD and a PhD from Lund University (PhD 2017 to 2024), with population research using nationwide Swedish cohorts; this is a documented research background, not evidence of individual clinical performance.
• Working with cohort data on thousands of children teaches the difference between association and causation, a distinction that protects everyday judgment when a single study or a confident claim is in front of you.
• Evidence-based practice, as originally defined, integrates the best available research with clinical expertise and patient values. Research training concentrates practice in appraising the first ingredient without replacing the other two.
• Research methods normalize the idea that confidence intervals, not certainties, describe much of medicine and can support clearer communication about uncertainty.
• Asking answerable questions, appraising study design, and communicating risk in plain terms are relevant across family medicine, internal medicine, research, and public health.

What a PhD adds to a clinician: two doctorates, one way of thinking#

What a PhD adds to a clinician is a method, not a separate career. An MD trains you to care for the patient in front of you. A PhD trains you to ask whether what you believe is actually true. The two impulses are not in tension. They sit side by side in the same clinic day, and the second is the quieter of the pair: it works in the background, where a result gets weighed rather than simply accepted.

The claim this piece examines is narrower: research training develops methods for question framing, evidence appraisal, and honest judgment under uncertainty. Jasaman Tojjar holds both degrees. She earned her MD at Lund University and her PhD there as well, completed between 2017 and 2024, with population research built on nationwide Swedish cohorts.

Research training is not a signal of any particular clinical specialty or of individual bedside performance. It develops habits such as asking a sharp question, respecting what a dataset can and cannot show, and staying honest about uncertainty. Those methods are relevant to broad fields including family medicine and internal medicine. What follows is an explanation of that methodological connection.

Research training develops methods that can inform many fields; the degree alone does not establish clinical performance.

What nationwide cohort data teaches that a textbook cannot#

A textbook hands you conclusions. Cohort data hands you the raw material those conclusions were built from. Consider what it means to work with population data on thousands of children. The BDD study (Better Diabetes Diagnostics) gathers data from all 42 pediatric clinics in Sweden, capturing children across an entire country rather than a single center. The ETICS cohort, used in Jasaman Tojjar's first-author work, included 11,050 twelve-year-olds. Numbers like those change what you notice. You stop seeing a clean result and start seeing the machinery underneath it: how each variable was defined, who was included and who fell away unnoticed, how missing data behaves, and where a real signal separates from chance. In a dataset that large, plenty of patterns appear by luck, and part of the work is learning which survive scrutiny.

Two of Jasaman Tojjar's first-author papers grew out of these cohorts: one on parental diabetes and childhood obesity using the ETICS data (PMID 32271617), and one on sex differences at the diagnosis of type 1 diabetes drawn from a nationwide sample (PMID 37699205). Both are population-based, and population-based and cross-sectional designs describe associations; they do not, on their own, establish causation. Naming that limitation is not a knock on the work. It is the correct reading of what the method can support.

The clinic-floor habit this leaves behind is small but durable. Before trusting a number, you pause and ask how it was measured. A lab value, a screening result, a reported symptom: each was produced by some process, and the process shapes what the number can mean. That pause is cheap, and it prevents real mistakes.

Association is not causation, and why that protects patients#

If research training leaves a clinician with one portable lesson, this is it. Two things can travel together without one causing the other.

A neutral example makes it concrete. Suppose a risk factor reliably shows up alongside a worse outcome in observational data. It is tempting to conclude the risk factor drives the outcome. But the people who carry it may differ from those who do not in ways nobody measured (age, access to care, other conditions), and one of those unmeasured differences may be doing the real work. The link is real; the causal story is an assumption layered over it. An observational design, by its nature, cannot tell the two apart.

That restraint shows up clinically in a few concrete ways. You do not rebuild your practice around a single observational study, however striking its headline. You ask what could be confounding an association before acting on it. And you notice when a guideline rests on strong versus weak evidence. The general structure of this appraisal, including how the strength of a recommendation is graded, is laid out by bodies like the U.S. Preventive Services Task Force, and Cochrane is a useful model for how careful reviewers separate a credible effect from a fragile one.

The protective part is for the patient. A clinician who reflexively treats association as causation will chase findings that later reverse and project a confidence the evidence never earned. The restraint is not timidity. It is a way of not spending a patient's time and trust on a story the data cannot back.

Reading evidence critically as a daily clinical skill#

Move from the dataset to the exam room, and the same skill becomes appraisal of whatever evidence the day puts in front of you.

It helps to remember what evidence-based practice meant when the term was defined. In a 1996 BMJ editorial, Sackett and colleagues described evidence-based medicine as the conscientious integration of the best available research evidence with individual clinical expertise and patient values. Three ingredients, deliberately kept together. The research is one of them, not all of them. A clinician who quotes a trial but ignores the person in the room has missed the definition as badly as one who never reads a trial at all.

What a research background sharpens is the appraisal of that first ingredient. You look past the abstract: what was the sample size, what was the design, how big was the effect rather than merely whether it cleared a significance threshold, and (the question that decides everything) does this finding apply to the patient actually sitting here. A result from a narrow study population may not transfer to an older, sicker, or simply different patient, and recognizing that is part of the appraisal.

This serves a generalist directly. Family medicine and internal medicine ask a clinician to weigh evidence across cardiometabolic risk, infections, mental health, prevention, and more, often within the same afternoon. Nobody can hold deep specialist literature in every one of those areas. What carries over is the method: a reliable way to read an unfamiliar paper and judge how much weight it can bear.

Comfort with uncertainty and honest shared decisions#

Research training does something to a person's relationship with not knowing: it reframes uncertainty as the normal weather of medicine rather than a personal failure to be hidden.

Spend enough time with confidence intervals, base rates, and the gap between absolute and relative risk, and those stop being statistical formalities. They become tools for honest conversation. A 30 percent relative reduction in some risk may translate to a single percentage point of absolute benefit, and a patient deserves to hear it in the terms that actually describe their odds. The uncertainty is not noise to smooth over before talking to the patient; it is part of what the patient needs in order to choose well.

This is the link to shared decision-making and to the prevention counseling that fills so much of primary care. Many real choices (whether to start a medication, whether to screen, whether to wait and watch) come down to a balance the patient has to weigh for themselves. The clinician's job is to lay out what is known, what is uncertain, and what the decision turns on, then decide together. Patient-facing resources from bodies like the NIDDK and the prevention recommendations from the USPSTF exist precisely to support those conversations on common chronic conditions. Said openly, uncertainty tends to build more trust than false precision does. People can tell the difference between a clinician being straight with them and one performing a certainty the evidence does not support.

Asking better questions at the point of care#

There is a research habit so ordinary it barely registers, and it may be the most transferable of all: turning a vague worry into a question you can answer.

In research, a fuzzy hunch (something seems off with this group of patients) only becomes useful once it is framed as a precise question with a defined population, comparison, and outcome. The same move works at the point of care. Instead of a diffuse unease about whether a treatment is right for this person, you sharpen it: in a patient like this one, does this option, compared with that one, change a specific outcome that matters to them. A question framed that way can often be checked against the literature during or shortly after a visit.

None of this requires a doctorate. Asking a structured clinical question, running a quick literature check, and appraising what you find are learnable habits, available to any clinician who decides to build them. Research training is best understood as an accelerator, not a gate: it concentrates years of repetition into the reflex, but plenty of clinicians develop the same reflex at the bedside without ever writing a thesis.

Research as a primary-care asset, not a separate track#

The position, by now, is straightforward. The value of the doctorate is analytical rigor brought back into generalist medicine: a disciplined relationship with evidence, a working fluency in chronic disease and prevention, and the restraint to know what a given study can and cannot prove.

Jasaman Tojjar's diabetes and population research documents experience with cohort methods and evidence appraisal. The broader lesson is that these methods can be relevant to primary care and many other fields; the publication record itself should not be treated as proof of individual clinical skill.

The best use of a PhD in clinic is more understated than its reputation suggests. It is not a stack of papers or a habit of citing studies. It is judgment: knowing what you know, knowing what you do not, and knowing how to find out. That describes good generalist practice as much as it describes research, which is rather the point.

Research as a clinical habit#

For health communication, the practical value is methodological: report the absolute size of a benefit, state when evidence is thin, and frame preference-sensitive choices around the person's priorities rather than as verdicts handed down.

For a trainee or a colleague, the encouraging part is how little of this requires the degree itself. Start with one habit: before you accept a finding, ask how it was measured and whether its study population looks anything like your patient. Applied consistently, that single move carries much of what a doctorate teaches into ordinary clinical reasoning, and it is open to anyone willing to slow down for the length of one good question.