I've mentioned in the 2026 Goals post that I was pursuing a research accelerator in order to improve my ability to publish a journal article next year. I signed up in late October for The Good Research Project. The founder is Dr. Shreya Deoghare, a dermatologist with more than a dozen papers to her name: https://pubmed.ncbi.nlm.nih.gov/?term=Shreya+Deoghare. This project helps students do secondary data analysis with mentors with the goal of writing a publishable journal article. I'd seen some dissatisfied students online, but the complaints really lacked detail and I wanted to see it for myself. It's inexpensive when compared with other accelerators like Lumiere ($4,800) or Polygence ($3,000) so I figured it was worth it to learn something and report back. First step, pay the fee, currently $275 USD. I was added to a WhatsApp group where I was asked which specialty I wanted to write an article about. I chose psychiatry, but they didn't have enough students sign up for psych. Psychiatry is not a research-heavy specialty so that makes sense: most of the students are International Medical Graduates (IMG) trying to build a publication record to secure a US residency. They offered me the chance to wait until they had enough students, or transfer to neurology. I said I would change to neuro so I was reassigned to neurology. Introduction We are given one week to collect our data. Failure to collect data or write your section by the deadline (or a second revised deadline) results in ineligibility to be on the article or a drop in author order depending on the totality of your contributions. After they got the students together, they had us sign some introductory documents to make sure we understood what was expected of us, including the COPE Authorship Guidelines and some training videos. The WhatsApp group is like an alumni group, for all members. In-project, we communicate through what looks like a white-labeled WhatsApp mobile app. My group is 7 people: one is a master's student, there's me (doctoral level), and several physicians. In addition to students there is a statistician, a research mentor, and the founder overseeing the whole thing. They sent us a link to a project tracker to keep track of the amount of data collection and writing, in order to properly assign credit. I'd read online that authorship by TGRP is assigned based on the order you join the project, however that does not appear to be the case. Instead, they use the combination of the different tasks completed by each student plus the sections that students write to determine authorship. For example, authoring the Discussion and Limitations section usually goes to first author, Results, Tables and Figures merits second authorship, authoring the Methodology merits a third author position while the Introduction 4th author. The first video we had to watch was an overview of the full research process. What I thought was interesting is that although we are doing the manuscript writing and data collection, the actual statistical analysis is done by a professional statistician. The second video was a data collection video. They shared several templates, such as a title page, cover letter, and the IMRAD format typical of medical journal articles. The research topic for my group is to understand the trends in mortality related to Parkinson's and depression. The dataset used for this project is the CDC WONDER (Wide-ranging Online Data for Epidemiologic Research). Data Collection The first milestone is data collection. We have one week to collect the data we need, and the data collection video explains exactly how to collect the data, with screenshots and video based on an older study they've conducted with a different population, so that we can follow the steps. One criticism from people online is that the research projects are formulaic, which is true because it's impossible to do research at scale without it being formulaic. As a way to build our skills I like it, but I think they can be rightfully criticized for increasing the flood of articles without actually contributing novelty. As an example of this recycling, several past papers assessed the quality and reliability of YouTube videos about various medical conditions. Several 2022 papers were bibliometric analyses based on gender representation in academic publications. The data collection video was 30 minutes long and quite detailed. I learned how to use the CDC Wonder database, collected my data. Writing Then we watched a third video, on paper writing. This accompanied a template where we walked through the different sections of a journal article. I felt like this was useful, especially for those who have never written before. It set out specific things you should aim to have in each section, approximate lengths, examples of how to state hypotheses, or structure the introduction, etc. Nothing you can't learn elsewhere but it's nice to have an outline. In the future I'd like to apply the Creating a Research Space (CARS) model to an intro as well as it seems like a good framework to use. I wrote my section a few days ago, and the deadline for everyone to have their writing in is today. I believe the next step is to prepare the paper for publication, once everyone has written their parts. Unfortunately, in the process of writing my section, I discovered an identical study to the one we are working on published in NIHR Open Review. Reviewers noted that this work substantially overlapped with an older paper using a different dataset as well, but voted to publish anyway. My hope is that once all the written parts are in, there will be an editing period. During that editing period I will see if I can cite that previous work and myself and the coauthors can identify an angle that make this paper unique. If that is not possible, this might be a $275 lesson in writing a journal article where I take my name off at the end.
After we completed our sections, someone edited it (one of the staff, not a student.) They sent us the completed article for review. Unfortunately, it includes the line "To the best of our knowledge, no previous population-level studies have specifically analyzed long-term mortality trends in patients with PD complicated by depression." That is just not true. The already-published study I linked above says "To the best of our knowledge, this is the first study to investigate the trends in PD-related mortality in the US" which is effectively the same thing. I can't in good conscience put my name on that. Disappointing. Here are multiple papers that appear to cover the same area but published in different journals, likely using the templates produced by TGRP. All of these are published in 2024 and 2025. Since all they're doing is swapping out the ICD-10 codes for the pair of diseases and running these analyses over and over, you'd think they would design just one study for each group. There are thousands of ICD-10 codes in each branch of medicine so there is no need to recycle. Colorectal Cancer: Temporal Trends in Racial and Gender Disparities of Early Onset Colorectal Cancer in the United States: An Analysis of the CDC WONDER Database | Journal of Gastrointestinal Cancer Temporal trends in colorectal cancer mortality rates (1999–2022) in the United States - Ilyas - 2024 - Cancer Reports - Wiley Online Library Stroke and Atrial Fibrillation: Trends in stroke-related mortality in atrial fibrillation patients in the United States: Insights from the CDC WONDER database - ScienceDirect Temporal Trends in Mortality Related to Stroke and Atrial Fibrillation in the United States: A 21‐Year Retrospective Analysis of CDC‐WONDER Database - Ahmad - 2024 - Clinical Cardiology - Wiley Online Library Cervical Cancer: Temporal trends of cervical cancer demographics: a CDC WONDER database study - PMC Temporal Trends and Demographic Patterns in Cervical Cancer Mortality in the United States (1999–2020) - International Journal of Radiation Oncology, Biology, Physics Urinary Cancer: Unveiling trends in urinary tract cancer mortality among older adults in the United States (1999–2022): a CDC WONDER perspective | International Urology and Nephrology Frontiers | Trends in genitourinary cancer mortality in the United States: analysis of the CDC-WONDER database 1999–2020 Crohn's disease: Unveiling the hidden toll: Disparities in Crohn’s diseasemortality – Insights from a CDC WONDER study Temporal Trends in Mortality Related to Crohn’s Disease in the United States: A 21-year Retrospective Analysis of CDC-WONDER Database I also saw this LinkedIn thread from 3 weeks ago where a person notes that their abstract, "Rising Pancreatic Cancer Trends in Asian Americans: CDC Wonder Database Analysis (1999–2020)”, has been accepted for Oral Presentation at the 21st Annual Academic Surgical Congress (ASC 2026). But a search for that name returns this identical paper written by a different group, Trends in pancreatic cancer mortality in the United States 1999–2020: a CDC database population-based study - PMC already published in August 2024. Interestingly, TGRP also helps people write up case reports. I wonder if those patients are also recycled?
This might not all be TGRP's doing, actually. I discovered this person, who is advertising a course in analyzing the CDC WONDER database with the goal of helping Pakistani researchers publish papers: Sumia Medical Research | Sumia Medical Researches. Either way, if these students are not completin
These sound like articles recycled to get someone's publication rate up. Or is an article mill letting them out (selling) to different people with the titles barely changed? Whatever it is, it seems like a case for http://www.Retractionwatch.com to investigate.
It's the first one. The students are doing the writing themselves, but the studies are being designed so that they can be churned out in bulk.
