My career as a postgraduate guide began in 1988 with Monica Ahuja. She was an MGIMS student from the 1982 batch, and her task was to decipher the prognostic significance of treadmill tests in patients who had recently survived the wreckage of a myocardial infarction. At the time, I was what one might call a “raw” guide. I possessed the title and the white coat, but I was entirely untrained in the formal architecture of a thesis. I knew little of statistics and even less of epidemiology. Like many of my contemporaries, I was a bush-pilot flying by sight, learning the process through a messy series of trials and errors.
The early years were defined by technical frustration. I harbored a long-cherished desire to master the tools of modern research, but they remained just out of reach. That changed when Madhukar Pai entered my life. It was through our association that I finally developed a firm grip on biostatistics and the digital instruments that make sense of clinical chaos. I learned to navigate the sharp corners of Stata, to organize a sprawling EndNote library, and to coerce Microsoft Word and Excel into doing my bidding. This newfound fluency allowed me to direct my students toward clinical questions that actually mattered in the wards of Sevagram.
The topics we tackled were as diverse as the patients in our waiting rooms. We interrogated the human body and the systems we use to measure it. P. Sarkar looked at the utility of palpation in assessing cardiomegaly, while M. Singh tested the simple validity of a hand on a forehead to screen for fever. We explored the grim realities of rural life through the clinical profiles of snake bites and pesticide poisoning—work carried out by residents like Pranay Taori and Abhay Kelkar. Later, we moved into the realm of neurological scoring, where Priya Badam and Vaishali Solao evaluated the Siriraj and Guy’s hospital stroke scores to see if they held water in a rural Indian setting.
Yet, as I look back on these thirty-two residents, regret colors the memory. The workflow followed a predictable, flawed pattern. My postgraduates would type their drafts and send them to me via email. I would then sit at my desk, delete their clunky phrasing, retype their observations, and polish the prose until it shone. I would deliver them a final, pristine copy. In doing so, I committed a pedagogical sin. I robbed them of the freedom to stumble. By smoothing over their mistakes, I prevented them from developing the very skills I had worked so hard to acquire. Most of my postgraduates wrote their theses passively; they emerged as degree-holders, but few emerged as confident researchers.
The Integrity of the Byline
In the competitive world of Indian medical academia, there is an unwritten rule: the names of heads of departments or senior guides must occupy the prestigious first or second authorship slots. I found this practice a quiet form of intellectual theft. If a student spends two years laboring over a project, their name should lead the charge in the journals.
I took a different path. It is a matter of quiet satisfaction that many of my MD residents saw their thesis work transition from a bound volume to high-impact medical journals as first authors. Monica Ahuja led the way with her work on treadmill tests. Parimal Sarkar and Dhanraj Singh Chongtham published rigorous interrogations of the physical exam—Sarkar on the utility of palpation in cardiomegaly and Singh on the clinical signs of splenomegaly and ascites. These were not mere student exercises; they were contributions to the global understanding of bedside medicine.
Others followed. Manoj Singh published on the screening of fever, while Priya Badam and Vaishali Solao saw their evaluations of stroke scores reach a wider audience. Ravindra Indra contributed to the diagnosis of hypothyroidism, and Samir Patil documented risk factors for myocardial infarction in our rural setting. Even complex cross-sectional studies reached publication under the primary authorship of the residents: Geetha Devi on brain lesions, Amandeep Singh on snake bite mortality, and Trunal Lokhande and Swati Waghdhare on the accuracy of physical signs in pleural effusion and meningitis. Madhuri Meena rounded out this tradition with her work on undifferentiated fever.
By stepping back and taking the final slot, I hoped to teach them that the true reward of research is not the hierarchy of the byline, but the integrity of the evidence. My Google Scholar profile serves as a ledger of these shared successes—a record of what happens when a guide acts as a steward rather than a master.
The Evidence of the Wards
My research has never been about the vanity of citation counts; it has been a systematic attempt to turn the daily observations of a rural hospital into data that could help others. The research was born from the dilemmas we faced at the bedside. We looked at the prevalence of tuberculosis infection among our own healthcare workers in a study published in JAMA, and we interrogated simple tools, such as using pallor to detect anaemia, finding that our eyes are rarely as precise as the laboratory. We didn’t just look at the high-tech; we also documented the “diagnostic uncertainty” of non-malarial fevers to reduce the irrational use of antimalarial drugs in rural OPDs.
In the mid-2000s, I began a collaboration with Madhukar Pai that eventually yielded some three dozen papers, including systematic reviews on phage-based tests and line probe assays for rapid drug resistance.
The education of a researcher is often a spectator sport. In August 2004, before I moved into a studio in Berkeley, Madhu graciously hosted me for a fortnight. I would watch him work, his eyes fixed on every line of a manuscript, revising and refining until the prose passed his own particular acid test. He was a man of extreme meticulousness; he labored to ensure the final draft was flawless, aiming for the exacting standards of journals like JAMA.
It was this rigor that saw his PhD thesis on tuberculosis—work actually performed in the wards of Sevagram—find its home in the pages of JAMA with Madhu as the first author. When the news arrived, the joy in his eyes was impossible to describe. It was the look of a debutant test cricketer who had just struck a century in his first innings.
My work has also been a tool for advocacy—from exploring the crisis in access to essential medicines to campaigning against industry sponsorship of medical conferences. During the pandemic, we insisted on the rigor of well-designed trials for Hydroxychloroquine and Inhaled budesonide over the noise of anecdotal “cures.” This body of work is the “paper trail” of my life in Sevagram—evidence that we tried to leave the science a little clearer than we found it.
The Fall of the Bound Volume
This realization eventually grew into a public critique. I have argued that we must stop the mandatory PG thesis. In its current form, it has devolved into a “Double-Blind Trial”: the student is blind to the methodology, and the guide is often blind to the science. And both have no idea what is going on until the thesis is submitted to the university. You can read my full indictment in The Double-Blind MD Thesis.
The tragedy of the modern thesis is that instead of teaching ethics, we are inadvertently instilling the habits of data manipulation and plagiarism. It is a system running toward a cliff, as I noted in The Fall of the MD Thesis. We force students to “cook” data to satisfy a disinterested examiner, teaching intellectual dishonesty before the career has even begun.
We would serve our students better if we abandoned this bound volume of fiction. Instead, we should teach them to critically appraise a paper on diagnostics or clinical trials. If we taught them how to read a study and apply its truth to the patient in front of them, we would produce physicians who practice with wisdom. A doctor who can interpret a likelihood ratio at two in the morning is far more valuable than one who has spent three years assembling a thesis that will never be read again.