Please see the #updates section at the end for pointers to the latest updates to this page.

An article published by Mark Alipio in the Philippines on 2020-04-08 has gained prominence in the debate about vitamin D and COVID-19.  It has been widely quoted and cited, including by me, as if it was a genuine report of vitamin D levels and COVID-19 symptom severity.  In late June 2020, I formed the hypothesis that this article does not represent real research.

This was prompted by questions which arose about the veracity of a second article, attributed to Prabowo Raharusuna, Sadiah Priambada, Cahni Budiarti, Erdie Agung and Cipta Budi (although the first author is also known as Prabowo Raharusun) of Indonesia. 

This lead me to similar questions about a third article by some authors ("Glicio" et al) in India.

All three articles concerned the relationship between vitamin D levels in the blood and the severity of COVID-19 symptoms.   This is a matter of great interest and importance.  Although, in July 2020, more research is now appearing concerning this question, there was great lack of such research before this, which lead to the Filipino and Indonesian articles being widely discussed and cited.

These three articles were published on the widely used and respected SSRN preprint system, which is owned by Elsevier.   The Indonesian article was withdrawn in mid June and the Indian one ("Glicio") in early July 2020.

In researching these articles I discovered 20 more by Mark Alipio, most of which were about subjects unrelated to COVID-19 and vitamin D.  I have mentioned these below, because I believe that the article of his which gained such prominence, number 16 of the series published in March and April 2020, should be considered in the context of the others.

Despite extensive efforts, I was unable to find any evidence which contradicted my hypothesis that all these articles (21 attributed to Mark Alipio, one to "Raharusun" et al. and one to "Glicio" et al.) were fabricated - that the research they purport to report never took place.  I found numerous reasons to believe they were indeed all fake academic research articles. 

If anyone has arguments for the veracity of these articles, on a public web page, forum or academic article I will be happy to link to these.  I have spent a lot of time on these articles and am not interested in engaging with further debate about their veracity or regarding their authors.

On 2020-07-01 I created this website to host the page you are reading now.  This page was originally focused on both the textual/diagrammatic content of the articles and on questions of who wrote them, since I was not convinced that they were all written by the person to whom their authorship is attributed.

This second set of concerns, about authorship, grew in complexity. 

Since I believe the veracity of the articles can be shown to be less than the standard expected, just based on the textual/diagrammatic content, on 2020-07-12 I completely revised this page to focus jut on this.   However, I do mention the other 20 Mark Alipio articles, one of the organisations he is affiliated with, and the Indonesian hospital mentioned as the contact address for the first author of the Indonesian article.  I believe these are relevant when assessing the veracity of the articles, since they tell me something about the contexts in which the articles were written.  

For instance, if I had discovered that all the other 20 articles by Mark Alipio were of excellent quality, then it would be harder to argue that number 16's deficiencies were the result of the article being fake.  Also, if I had discovered that the hospital was a large, well-resourced, teaching hospital with multiple MDs and other researchers with a solid academic publishing history, then it would be more difficult to argue that the deficiencies I found in the Indonesian article were the result of it being fictional.

The question of authorship is important when evaluating the veracity of academic research articles.  Please see my thoughts on this at the home page: ../#authorship . 

Please see the home page ../#about for notes on how I am not an authority, have no formal qualifications etc.

Below (except for updates prompted by an auricle by three Indonesian MDs and a subsequent document published on an Indonesian government website) I proceed as if there are no questions about the identity of the ostensible authors or whether they wrote the articles.  If these questions interest you - and they may if you find the text and diagrams of the articles meet your standards of academic rigour - then you will need to research this yourself.

The veracity of academic articles in general, and these ones in particular, is of interest to many people and so some of these articles have been mentioned on other sites.  I have not put much effort into finding such sites, but here are two:

 

---

Please link to this page and quote selectively from it rather than copy the whole of its contents.  I may update it and don't want old copies floating around on the Web.

To the ResearchVeracity.com home page: ../

This page established 2020-07-01.  Simplified and revised 2020-07-12. 
Last update                   2020-08-20.
Update history: #updates

Robin Whittle Email: rw@firstpr.com.au .   Daylesford, Victoria, Australia 

Here are the articles in date order:

1

Vitamin D Supplementation Could Possibly Improve Clinical Outcomes of Patients Infected with Coronavirus-2019 (COVID-2019) Mark Alipio  2020-04-08  (Not peer-reviewed.) Davao Doctors College; University of Southeastern Philippines https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3571484  (Withdrawn around 2020-08-19.)

This is the 16th article in a set of 21 articles #MA-list which were published on preprint servers over a period of 5 weeks in late March to late April 2020 with authorship attributed solely to Mark Alipio. 

See: #MA-critiques and #MA-16-critique .

2

Patterns of COVID-19 Mortality and Vitamin D: An Indonesian Study Prabowo Raharusuna, Sadiah Priambada, Cahni Budiarti, Erdie Agung, Cipta Budi 2020-04-30  (Not peer-reviewed.) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3585561 (Withdrawn mid-to-late June 2020.) It is available at: https://archive.org/details/ssrn-id-3585561 and, with an additional page 1 signed by Lorenz Borsche and Dr. Bernd Glauner, June 2020, from https://borsche.de/res/...COVID_19_Mortality_and_Vitamin_D_An_Indonesian_Study.pdf

Important update 2020-07-27:  See the following article in which three Indonesian MDs find no trace of the purported authors of the above article.  This includes searching the Indonesian Medical Council database and contacting the hospital mentioned in the article:

See #R-article and #R-critique .

I am not aware of any other articles by any of Prabowo Raharusun, Sadiah Priambada, Cahni Budiarti, Erdie Agung, Cipta Budi, or by any of the authors of the next mentioned article from India

Important update 2020-08-03:  See the end of #R-critique for three additional critiques of the "Raharusun" article from Dr Raymond Pranata.

Important update 2020-08-13:  See #indon-govt for a link to a document at an Indonesian Government website which attests to the veracity of the COVID-19 and Misinformation article and establishes beyond doubt that there never was a doctor by the name of "Prabowo Raharusun" associated with the hospital mentioned in the article, that the purported research is unknown to the hospital and that the hospital does not and never did have a website.   This means that the website I and others assumed was the hospital website was (since it no longer exists) fake.

3

Vitamin D Level of Mild and Severe Elderly Cases of COVID-19: A Preliminary Report El James Glicio, MD; Siddharth Neelam, MD; Rajeev Rashi, MD; Deepak Ramya, MD No date on the PDF, but 2020-05-05 on the SSRN page. (Not peer-reviewed.) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3593258 (Withdrawn in early July.)

See #G-critique.

I had never encountered academic articles which were largely or wholly fictional.  Now I believe I have, I will be much more cautious in the future. 

The framework in which I am suspicious of some academic articles is partly based on my understanding of the problems of predatory journals.  Here are some pages to get started on this important field, which is a real threat to science and the pursuit of knowledge:


#MA13-similar
Update 2020-08-10:


However the articles I discuss below, which I believe to be fake, are not directly related to predatory publishing.  They exist largely or entirely outside journals - even predatory journals.  They are published primarily or solely on preprint servers, academia.edu (a commercial company which I think should not have a .edu top level domain), researchgate.com and the like. 

These articles may be related to the essay writing industry - since it is possible to order a research paper, masters level, of specified number of pages and subject matter, from companies which provide essay writing services.  In this way, a person who has never written an academic article could post articles they purchase in this way, with their name as author in the text of the document - and so build an academic reputation of sorts. 

For instance, https://extraessay.com offers an $11 a page, "plagiarism-free", essay writing service.  Masters level research papers cost $14 a page with a 14 day lead time.  Delivery times as short as 3 hours are offered: USD$49 a page for a masters research paper.

In fact, if anyone looks closely enough at such articles, the ostensible author's reputation will be less than zero.  However, too many people (in the past, myself included) don't look closely enough at the article, cite it, and so add to the author's citation count in Google Scholar and in other systems.  Since Google Scholar can't clearly identify these bogus articles, these articles can cite each other and so build impressive citation counts for the author or purported author.  Algorithms might detect this, but such detection would be made very much harder by legitimate articles citing the bogus ones.

There are numerous dangers to this form of academic fraud.  One is that careless researchers may cite these bogus articles and the data they claim to be based on as a valid way of understanding Nature.  This is an insidious, poisonous, influence on the already difficult business of scientific research, publishing and discussion. 

Another danger is that citation systems, pre-print servers, search engines etc. become clogged up with these articles - just as our email servers have been clogged by spam for decades now. 

A further danger is that fake articles, and their authors, may gain respectability by being cited in genuine articles. 

Vitamin D Supplementation Could Possibly Improve Clinical Outcomes of Patients Infected with Coronavirus-2019 (COVID-2019) Mark Alipio  2020-04-08  (Not peer-reviewed.) Davao Doctors College; University of Southeastern Philippines https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3571484 Posted: 9 Apr 2020 2020-06-30: 6 citations and 9,886 downloads.  This was the 9th most downloaded coronavirus article on SSRN: SSRN-9th-most-popular.png Google Scholar (2020-06-30) supposedly 40 citations, including proper journal articles - but when I checked some, there was no such citation.

2020-07-11 update:


A complete 2020-07-02 image of the profile page is: SSRN-MMA-2020-07-02.png .  From this the stats for this article:



This is article 16 of the 21 listed below: #MA-list.  See also #MA-16-critique below.

I have three PDFs of the article.  There was at least one earlier version.

• AA from Grassrootshealth.net : 151,704 bytes, PDF created 2020-04-15  15:59. local
  
  
• BB on 2020-04-23: 149,907 bytes, PDF created 2020-04-23  16:33. local
  
  
• CC on 2020-06-24: 157,610 bytes, PDF created 2020-05-07 15:10. local
  

AA and BB are both "letters" to an editor.  CC is not, and has an abstract.  BB completes some instances of "(OH)D" to "25(OH)D" and adds some +/- figures to the Serum 25(OH)D line of the table.

BB adds the following text which I comment on below:

All data pertaining to the serum 25(OH)D levels of the cases were extracted from the onset of symptoms. The hospital conducted serum 25(OH)D test, along with other clinical tests, every seven days to monitor the status of patients. For descriptive purposes, mild cases were tested approximately 3 times, slightly lower compared to original cases (4 times), severe cases (6 times), and critical cases (7 times). Pre-admission 25(OH)D measured between 7 and 365 days before hospital admission, was also taken. Mean value for time the latest pre-admission 25(OH)D level was taken, was 12.65 ± 5.32 days. A total of 223 cases were originally extracted in the analysis.  To ascertain no differences between time points, a repeated measure analysis of variance (ANOVA) was used and reported no significant differences in the serum 25(OH)D level of the 212 (95%) cases. Only a small proportion of cases reported significant differences mainly during the course of hospitalization. The 212 cases were used for the final analysis and serum 25(OH)D level taken during the onset of symptoms was considered.

CC adds a header at the top of each page:

Note: This is a pre-print version of the author’s paper before any peer review has taken.

This is technically correct - the article had not been peer-reviewed.  It may give the impression that the article is to be submitted to peer review in some unspecified journal - but as far as I know this not occurred.

When I first read this article I assumed the author was doctor.  This seems reasonable since the first of two affiliations is "Davao Doctors College".  In late June 2020 I looked this up and found the college is attached to Davao Doctors' Hospital, and does not train doctors. 

The main page for the Davao Doctors' College https://www.davaodoctors.edu.ph/  https://en.wikipedia.org/wiki/Davao_Doctors%27_College had (2020-06-28) a scrolling slideshow of award winners, including one for the author who graduated with a Bachelor of Science in Radiologic Technology 2018.   On 2020--08-20 the scrolling images of other students remained, but Mark Alipio's was not amongst them.

The author's profile pages are easy to find with web searches.  Some of these might be automatically generated by the web site in response to his articles:

• https://radiopaedia.org/users/mark-alipio (Sept 2017.)
  https://radiopaedia.org/users/mark-alipio-1 (Dec 2017.)
  
  
• https://publons.com/researcher/3556735/mark-alipio/
  
  
• https://osf.io/8kmh9/ This is the page linked to from article pages such as:
  
  
• https://edarxiv.org/ezwgv/
  
  
• https://scholar.google.com/citations?user=Yb-Xad4AAAAJ
  2020-08-13: this has disappeared.
  
  
• https://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=4083862
  2020-08-20: this has disappeared.
  

• https://www.researchgate.net/profile/Mark_Alipio2
  
  
• https://davaodoctors.academia.edu/MAlipio (29 followers.)
  
  
• https://mpra.ub.uni-muenchen.de/view/people/Alipio=3AMark=3A=3A.html (The four article pages list the author as "Prof. Mark Alipio".)
  

• Twitter (joined June 2020).
  2020-08-13: this has disappeared.
  
  
• https://europepmc.org/authors/0000-0001-8360-0287
  
  
• http://orcid.org/0000-0001-8360-0287
  

These are 2020-06-25 snapshots from Google Scholar of Mark Alipio's articles:


I provide these as a record of my experience, and no-doubt of others, if they searched for this author as part of ascertaining the veracity of his articles.   A quick glance shows that this particular article is part of a wider pattern of publication, albeit - if one actually follows the links - entirely on preprint servers in 2020.  There is an article from previous years with Mark Alipio as co-author which I did not scrutinise.  I am not suggesting that it is part of the pattern I perceive with these 21 articles in March and April 2020.

Since I am explaining my current understanding of this and other articles being invalid, it is important to note that I and other people did not initially think so.  I assumed the author was a medical doctor, probably in his 40s to 60s, conducting vitally needed research on humanity's most novel and pressing health problem.

A Google search finds multiple instances of "Dr Mark Alipio", though nowhere does the author present himself in this way.   Some of these instances are in blogs, comments sections of mass media websites and the like.   A few of them are in peer-reviewed journal articles.  For instance:


cites this article:

Moreover, recently Dr. Alipio has provided substantial information to physicians and health policymakers. Specifically, it concluded that vitamin D supplementation improves the clinical course of patients infected with COVID-19 based on the increased probability of having a mild result when the serum level of vitamin D increases while a serum decrease in vitamin D is associated with worse clinical evolution

This team of seven presumably highly qualified researchers and/or clinicians either didn't read or at least misunderstood the article, which says nothing about observations following vitamin D supplementation.

Articles concerning vitamin D and COVID-19 have their titles and PDF Author text in bold.

The dates here are the PDF creation dates, which can be read from the Document Properties or similar function of programs such as PDF-XChange Viewer.  So these are the dates on which a Word file was saved as a PDF which was later uploaded to a preprint server.

Some are available on more than one preprint server.  So if a PDF link doesn't work, please try the Google Scholar link instead.  For quick reference I have PNGs named by the article number I assigned.


2020-08-13: Items with a * have disappeared.

Title and link Google Scholar		Link 1st page PNG	Date
2019-nCOV Scare: Situation Report, Role of Healthcare Professionals and Clinical Findings	GS	PDF * 01.png	3-22
Chest Radiographic Findings of Patients Infected with 2019-nCOV	GS	PDF * 02.png	3-26
Challenges and Strategies for Curbing the 2019-nCOV Pandemic: The Case of the Philippines	GS	PDF * 03.png	3-27
Predicting Academic Performance of College Freshmen in the Philippines using Psychological Variables and Expectancy-Value Beliefs to Outcomes-Based Education: A Path Analysis	GS	PDF 04.png	3-27
A Path Analysis Examining the Relationship Between Access Barriers to Health Services and Healthcare Utilization Among the Publicly Insured: Insights from a Multiprovince Survey in the Philippines	GS	PDF 05.png	3-28
A Framework for Predicting Radiologic Physics Achievement among Radiologic Technology Students	GS	PDF 06.png	3-29
Structural Models of Self-Efficacy of Filipino Radiologic Technology Educators, Current Learners, and Prospective Students in the Senior High School	GS	PDF 07.png	3-30
Adjustment to college and academic performance: Insights from Filipino college freshmen in an allied health science course	GS	PDF 08.png	4-01
Academic Adjustment and Performance among Filipino Freshmen College Students in the Health Sciences: Does Senior High School Strand Matter?	GS	PDF * 09.png	4-01
Public Policy and Program Administration in the Philippines: A Critical Discourse	GS	PDF 10.png	4-02
National Radiologic Technology Licensure Examination Performance: Predicting Success using Discriminant Analysis	GS	PDF 11.png	4-07
Students' Preferences for Medical Schools: A Conjoint Analysis	GS	PDF 12.png	4-07
Academic success as estimated by locus of control and motivation See #MA13-similar.	GS	PDF 13.png	4-07
Epidemiology and Clinical Characteristics of 50 Death Cases with COVID-2019 in the Philippines: A Retrospective Review	GS	PDF * 14a.png 14b.png	4-08
A Structural Model of Organizational Commitment among Higher Education Economics Educators	GS	PDF 15.png	4-08
Vitamin D Supplementation Could Possibly Improve Clinical Outcomes of Patients Infected with Coronavirus-2019 (COVID-19)	GS	PDF * 16a.png 16b.png	4-09
Do latitude and ozone concentration predict Covid-2019 cases in 34 countries?	GS	PDF 17a.png 17b.png	4-10
Do Socio-Economic Indicators Associate with COVID-2019 Cases? Findings from a Philippine Study	GS	PDF * 18a.png 18b.png	4-11
Determinants of tuberculosis incidence in East Asia and Pacific: A panel regression analysis	GS	PDF 19a.png 19b.png	4-15
Revenue Administration in the Philippines:  Significant Collection Reforms, TRAIN Law, Fiscal Incentives, Excise Tax, and Rice Tariffication Law	GS	PDF 20.png	4-15
Education during COVID-19 era: Are learners in a less-economically developed country ready for e-learning?	GS	PDF * 21.png	4-27

Article 16 is the one I am most concerned with - see above: #MA16.  I listed its date as April 09 = 4-09, since this is the date of posting on SSRN .  The PDFs I have of various versions are of later dates, as mentioned above.

Since some of the articles mention journal names, it is important to check to what extent the article is in fact associated with such journals.

2020-08-13: Items with a * have disappeared.

Start of title	Link 1st page PNG	My terse description of the article
2019-nCOV Scare: Situation Report, Role of Healthcare Professionals and Clinical Findings	PDF * 01.png	General discussion of public data. Header is for the International Journal of Multidisciplinary Health Sciences Research.  There is no such journal.  The two websites I found entitled  International Journal of Multidisciplinary Health Sciences are both defunct.  The issue number 10 exceeds the number of issues per year, which is 4, for the journal which was most recently active. There are 46 SARS-CoV and SARS-CoV-2 references but none of them are mentioned in the text.
Chest Radiographic Findings of Patients Infected with 2019-nCOV	PDF * local 02.png	General discussion of public data. Header as above. There are 53 references, none of which are mentioned in the text.  Most are to COVID-19 articles but 10 are to articles for whom this author is listed as the sole or co-author This looks like an attempt to manipulate Google Scholar and other citation systems.
Challenges and Strategies for Curbing the 2019-nCOV Pandemic: The Case of the Philippines	PDF * local 03.png	General discussion of public data. The final two references are to the above articles, but are not mentioned in the text. This looks like an attempt to manipulate Google Scholar and other citation systems.
Predicting Academic Performance	PDF 04.png	Analysis of results from vaguely defined survey of 12,453 students in 70 higher education institutions.
A Path Analysis Examining	PDF 05.png	Analysis of results from vaguely defined survey of 7,234 Filipino residents using multi-stage cluster sampling.
A Framework for Predicting	PDF 06.png	Analysis of results from vaguely defined survey of 954 Radiology Physics students randomly selected from 12 Radiologic Technology schools in the Philippines.
Structural Models of Self-Efficacy	PDF 07.png	Analysis of results from vaguely defined survey of 256 Radiologic Technology educators and 2,451 students randomly selected from 22 Radiologic Technology schools in the Philippines.  Also 4,263 prospective Radiologic Technology students from the 30 senior high schools of the Philippines.  This seems rather low for a country of 100 million people.
Adjustment to college	PDF 08.png	Analysis of results from survey in June to October 2018 of 132 first year Radiologic Technology students in an unspecified higher education institution in the Philippines, who were chosen through stratified random sampling.  This would require the total number of students to be multiple times larger than this, which seems like a rather large number for any one institution given there are either 12 or 22 Radiologic Technology schools in the Philippines.  I thought this was a lot of such schools, but there are 102. There is no information on the number of students from which the respondents were chosen, how they were assigned to groups for the purpose of stratified random sampling, the statistical goals of the survey, or how Slovin's formula was used to guide these choice of which students to include in the survey. There is no record of the questions or of the answers - just analysis of the answers.
Academic Adjustment and Performance	PDF * 09.png	Analysis of results from vaguely defined survey of 14,062 participants in different health science disciplines at 79 higher education institutions (HEIs) in the Philippines.
Public Policy and Program	PDF 10.png	General discussion.  Some sentences may not have been written by a human: The Identity Crisis should be viewed from the perspective of development administration or from a field now called Development Public Administration. Develop research programs that focuses on the problems in the country, looking into the behavioral insurance of politics and administration, and the accompanying features that open the government to development process. Five irrelevant references to articles by the same author are not mentioned in the text.  This looks like an attempt to manipulate Google and other systems which count citations.
National Radiologic Technology	PDF 11.png	Analysis of results from vaguely defined survey of 2,036 graduates of a baccalaureate Radiologic Technology program in 2016, 2017, and 2018 from 24 higher education institutions (HEIs) in the Philippines. This is the first study to discriminate passing from failing graduates in the national RT licensure examination based on the selected predictor variables and the astounding precision of classifying graduates is a remarkable result for HEIs included in the analysis.
Students' Preferences for Medical	PDF 12.png	Analysis of results from vaguely defined survey of an unspecified number of  respondents selected from 24,263 prospective students in the health sciences and medical courses from the 125 senior high schools in the Philippines. Multi-stage cluster sampling technique was used to recruit the sample. A face to face interview with a set of stimulus cards or through a self-completion questionnaire was used to collect the data and Adaptive Conjoint Analysis (ACA) was performed to analyze students’ preferences.
Academic success as estimated	PDF 13.png	Analysis of results from vaguely defined online survey of 21,012 freshmen. A descriptive-correlational study using online survey questionnaires was  employed to 21,012 respondents who were chosen through simple random sampling and Slovin’s formula. See #MA13-similar.
Epidemiology and clinical characteristics of 50 death cases with Covid-2019 in the Philippines: A retrospective review	PDF * 14a.png 14b.png	Brief analysis of 50 COVID-19 deaths from 2020-01-30 to  2020-04-03.  However, I can't see that the patient details are actually publicly available sites as claimed.  No details of how these 50 cases were chosen from the 135 deaths in this period (https://endcov.ph/dashboard/)
A Structural Model of	PDF 15.png	Analysis of results from vaguely defined survey of 871 higher education Economics educators. The author names of its 28 references all start with A or B.  Except for those attributed to this author, none of them are mentioned in the text.
Vitamin D supplementation could possibly improve clinical outcomes of patients infected with Coronavirus-2019 (COVID-2019)	PDF 16a.png 16b.png	Analysis of results from vaguely defined data extraction from vaguely defined hospitals: Using the database of three hospitals in Southern Asian countries, a retrospective multicentre study of 212 cases with laboratory-confirmed infection of SARS-CoV-2 was conducted. See further notes below.
Do latitude and ozone concentration predict Covid-2019 cases in 34 countries?	PDF 17a.png 17b.png	Analysis of number of cases of COVID-19 in 34 countries with respect to upper atmosphere ozone.  The ozone data is reproduced and looks realistic, though there is little variation between them - I calculated 334 average and SD 47.  The number of cases in the countries is not listed.  The analysis is meaningless because no account is taken of the population of each country.
Do Socio-Economic Indicators Associate with COVID-2019 Cases? Findings from a Philippine Study	PDF * 18a.png 18b.png	Analysis of number of cases of COVID-19 in 17 regions of Philippines.   Case number data is presented but not the 7 items of socioeconomic data for each region.  The analysis is meaningless because no account is taken of the population of each region, which varies by a factor of 8.2.
Determinants of tuberculosis	PDF 19a.png 19b.png	Analysis of TB incidence per 100,000 people and various socioeconomic variables for 23 countries.  The analysis is meaningless because the countries are not listed and no data is presented.
Revenue Administration	PDF 20.png	Cursory discussion of public data.  The first reference is not referred to in the text.  Neither are the other five, which are all to articles in this list of 21 - none of which are relevant to this article.  This looks like an attempt to manipulate Google and other systems which count citations.
Education during COVID-19	PDF * 21.png	2000 students were sent an email survey.  Educational involvement, geographic location and other selection criteria were not defined and they were supposedly initially contacted via Facebook, Twitter and other social networking sites.  However, these sites do not provide email addresses. 440 responded.  This seems like an extraordinarily high rate considering the 27 questions plus demographic details. The 27 questions were adapted from a cited source which allowed responses on a 1 to 5 scale.  In this supposed survey, answers were Yes/No and all 440 respondents responded Yes or No, with none recorded as not responding to a question.  It is impossible to imagine this 100% response rate if they responded via email. The email questionnaire is not reproduced. The questions were shown not in the original order, but in order of decreasing Yes response.  The percentage of Yes answers varied very little between the questions - from 45.5 to 44.0 - so this data is obviously fictitious.

I initially accepted this article 16 as genuine.   It was the first article I knew of which showed what I and many other people expected: a correlation between disease severity and low vitamin D 25OHD levels.  While infection is sometimes thought to lower vitamin D levels, I think the evidence for this is limited.  So, assuming that this could account for only a small part of the association, it is reasonable to conclude that the causation flows the other way: from low vitamin D to worse outcomes.

There are numerous reasons to believe that low vitamin D would cause this directly.  However, other causality pathways also need to be considered, such as low vitamin D causing other conditions, over a long period of time, and those conditions causing worse outcomes.  Also, those conditions might cause both low vitamin D and worse outcomes for COVID-19.

Despite the causation of severe symptoms not necessarily being directly and entirely due to low vitamin D levels, I think it is reasonable to assume that a great deal of causality is by this pathway, which raises the possibility of preventing or reducing these symptoms and resultant damage and death by vitamin D supplementation.

With the world in the early stages of the COVID-19 pandemic, with more and more worrying reports of death, and serious harm to the brain, lungs, heart and other organs, the thought that a few milligrams of inexpensive (USD$2.50 a gram in 1kg lots ex-factory) vitamin D (4000IU a day is 0.1mg) might prevent such tragic and disastrous harm and death motivated me and many others to search for evidence this would be true.

This article, soon joined by the Indonesian article, is what we wanted and needed in our urgent attempts at raising awareness among doctors, researchers and the public.   So, in a hurry (I certainly was) many of us took what we wanted from the article - the numbers of patients in different classes of vitamin D deficiency and symptomatology - and made our own tables and graphs, telling everyone we could about it.

It never occurred to me that someone would write an entirely fictional research article and publish it in any form as if it was a genuine report of real observations.   I assumed the results were from a medical doctor, working in the Philippines, under difficult early stage pandemic conditions.  I didn't scrutinise the article for weaknesses or inconsistencies.

There is no sign of the article being humorous, or a send-up.  It is a quickly written document which shows the author has a non-trivial understanding of vitamin D, the immune system and COVID-19  - well before such notions were more widely discussed.  Indeed, if most doctors had had such an understanding, they would have long been advocating that everyone robustly supplement with vitamin D to raise their 25OHD levels to the 40 to 60ng/ml range which experts have been recommending since 2008:  https://www.grassrootshealth.net/project/our-scientists/

So I think the author knew more about vitamin D than some or many doctors.

These shortcomings of the article are mainly ones I noticed once I began to suspect its veracity.

My initial impression was that this was the work of a doctor somewhere in the Philippines who had access to the records system of hospitals and had taken some precious time to analyse 250HD tests which had already been done.  I was pleasantly surprised that hospitals in this developing nation were suitably aware of vitamin D deficiency in general to be doing these tests.

I remember wondering about the 25OHD tests which were done before COVID-19 infection.  I guessed that these tests were part of ongoing surveillance of outpatients or something like this.

I also wondered about the extensive nature of these tests across three hospitals.

However, I didn't think too much about the details because the data reflected what I suspected was the case - higher proportions of severe symptoms with lower 25OHD levels. 

I should have been wary of this statement:

Using the database of three hospitals in Southern Asian countries,

Why would there be one database for multiple hospitals in multiple countries?   Was this a typo and the author meant "databases"?  I cut the author some slack at the time, being from a developing nation, in a time of crisis - no-doubt working under great difficulties.

Why weren't the countries mentioned?  They should be.  Perhaps the author did not have proper authority to access this data, but did so anyway.  It is a time of crisis after all, and who knows what dodgy security was in place and what bureaucracy which might have stood in the way of this important work.

I think I first read the BB version of the article, not knowing there had been an earlier version without the following text:

All data pertaining to the serum 25(OH)D levels of the cases were extracted from the onset of symptoms. The hospital conducted serum 25(OH)D test, along with other clinical tests, every seven days to monitor the status of patients. For descriptive purposes, mild cases were tested approximately 3 times, slightly lower compared to original cases (4 times), severe cases (6 times), and critical cases (7 times). Pre-admission 25(OH)D measured between 7 and 365 days before hospital admission, was also taken. Mean value for time the latest pre-admission 25(OH)D level was taken, was 12.65 ± 5.32 days. A total of 223 cases were originally extracted in the analysis.  To ascertain no differences between time points, a repeated measure analysis of variance (ANOVA) was used and reported no significant differences in the serum 25(OH)D level of the 212 (95%) cases. Only a small proportion of cases reported significant differences mainly during the course of hospitalization. The 212 cases were used for the final analysis and serum 25(OH)D level taken during the onset of symptoms was considered.

There are lots of clangers here.  If the version I read then had this text, I should have paid much more attention.

Why would patients be tested fixed numbers of times depending on severity.  Severity varies over time and some die or are discharged before others.  7 weekly measurements for severe symptom patients makes no sense, because some would have died by then, and because this would have had to be in February and March, when there were few cases outside China.

I remember wondering why there were one or more 25OHD tests taken before the onset of COVID-19, up to a year before.  I glossed over this concern - imagining some kind of outpatient monitoring arrangement, which I thought both unlikely and actually quite impressive, considering the lack of interest many doctors in first world countries have in vitamin D.   I really should have questioned the veracity of the whole article for this and other reasons.

Then there is the question of three separate hospitals doing similar 25OHD tests on outpatients when they were not even in the same country.  I probably thought about this for a few seconds before moving on, provisionally impressed by the foresight of these doctors in countries far less rich than my own.

This reminds me of the pattern of spam from African countries - or at least ostensibly from them.   No-one would believe an email from a previously unheard-of person from Germany, the UK, Japan or the USA who suddenly was considering donating a few million dollars , or was urgently in need of a place to park his or her vast fortune due to some impending coup, assassination plot or whatever.   But one can imagine such quasi-noble characters bumbling through their troubles, burdened by their immense and probably ill-gotten wealth, struggling with English, legality and morality if they came from some distant African country with immense oil or mineral wealth.  The effect is to cut the spammer some extra slack due to their non-English-speaking and generally third-world background.

I should have recognised that the whole article was made up, with the extreme precision of the following, and there being no reason at all for the pre-admission 25OHD tests so soon before:


This would require doctors in all three hospitals, in different countries, to either be doing these frequent 25OHD tests to a vast number of people who were not ill at the time, or only for the smaller number of people who would soon get COVID-19 - without a time machine to predict this.

What were the significant differences in these multiple pre-illness 25OHD levels?  What was the criteria for exclusion from the subset chosen for the main analysis?

If I had known that these paragraphs had been added after the initial version of the article, I might - should at least - have been more suspicious.  However, I had no idea some people were writing entirely fictional research articles.


Never having studied statistics, I usually gloss over these parts of research articles - but what does this actually mean in this context?  I still have no idea.

Also, I didn't notice this:


If I had recognised that the data was open-access, I would have expected its URL to be mentioned.  I would probably have written to the author, though I was sure he was very busy, to request access to such data so I could scrutinise it in greater detail than he had reported it.

My eyes went straight to the 25OHD levels in table 1.  Not so much the averages but the number of patients in each of the three bands of values for the four levels of severity.

I wrote this up as a table to make it really easy for people to see the great dependence of severity on vitamin D levels:


While I think the real author is to be congratulated for having an early sense of the degree to which low 25OHD levels probably cause COVID-19 disease severity, I think he was straining credulity somewhat with the sharp dichotomy between severity above and below 30ng/ml.   My best guess, in July 2020, is that the relationship is not as sharp as this - since we are dealing with biological systems with numerous dimensions of variation from one individual to another.

I took no interest in the predictions the author had supposedly discovered - in part because I lack the statistical knowledge to understand what was supposedly being achieved there.  I could have educated myself on this quick-smart if I was interested - but what I really wanted was the observational data.

Until now I didn't notice the final sentence, which is a clanger:


The immune system is generally not dependent on the serum 1,25OHD levels.  Anyway, these are generally reasonably stable, for a reasonably wide range of D3 inputs and so 25OHD levels.  What the immune system requires for proper operation is robust levels of 25OHD - I now know that 40 to 60ng/ml is recommended by many researchers and a few doctors, and this is directly affected by supplementation levels.

Still, I could imagine someone collecting these statistics while not having a very detailed understanding of how vitamin D affects the immune system.

I don't recall using Google to search for the author.  I pictured someone in their 40s to 60s, with a wealth of experience, and a busy caseload, taking some time to access and analyse some crucial and so-far unreported data.

I tended to think of him treating COVID-19 patients, and was vaguely expecting further articles on the effectiveness of the vitamin D supplements suggested in this article.

Important update 2020-07-27:  See the following article in which three Indonesian MDs find no trace of the purported authors of the above article.  This includes searching the Indonesian Medical Council database and contacting the hospital mentioned in the article:

This is the first of two fake articles which fall into the pattern established by Mark Alipio's articles, with which they share many commonalities.  The other is the "Glicio" article mentioned below: #G-critique

Patterns of COVID-19 Mortality and Vitamin D: An Indonesian Study Prabowo Raharusuna, Sadiah Priambada, Cahni Budiarti, Erdie Agung, Cipta Budi 2020-04-30  (Not peer-reviewed.) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3585561 (Deleted mid-to-late June 2020.) It is available at: https://archive.org/details/ssrn-id-3585561 and, with an additional page 1 signed by Lorenz Borsche and Dr. Bernd Glauner, June 2020, from https://borsche.de/res/...COVID_19_Mortality_and_Vitamin_D_An_Indonesian_Study.pdf

"Raharusuna" (in both PDF versions I have seen) seems to be a typo since Google only finds it in pages which mention this article. 

The article was posted to the SSRN preprint system at this URL.  The page where it was listed disappeared sometime around mid-to-late June 2020.  I don't have a snapshot of what the page looked like, but I found an image EbEe6w_U0AAzSI1.png apparently of this page, in this tweet .  It shows the initial date for the article is 30 April.

I have a PDF I saved from this page on 2020-05-02.


This page, which seems to be associated with SSRN, tracks websites, social media etc. which mention the article:


Here is a screenshot of that page from 2020-06-30: plu~mx~ssrn~a~ssrn_id~3585561.png .  You can see the first author's surname is spelt Raharusuna while the article was listed with this name spelt Raharusun. SSRN is owned by Elsevier so this page indicates the article is published by Elsevier.

The date in the text of the file is April 26, 2020.

Here are the versions I am aware of.  "RR", "SS" and "TT" are my names.

• RR I downloaded from SSRN.  148,019 bytes, PDF creation date 2020-04-26 19:12.  This would be the original version.  It was the only one I was familiar with until 2020-05-30 when I searched for PDFs of the article and found:
  
  
• SS I found at: https://emerginnova.com/patterns-of-covid19-mortality-and-vitamin-d-an-indonesian-study/
  
  and downloaded on 2020-06-30.  211,568 bytes, PDF creation date 2020-05-06 10:59.
  
  
• TT is a modified version, not from SSRN, with an extra page at the start, apparently created by Lorenz Borsche and Dr. Bernd Glauner, June 2020.  This is available from: https://borsche.de/res/Patterns_of_COVID_19_Mortality_and_Vitamin_D_An_Indonesian_Study.pdf
  
  I downloaded a copy on 2020-06-30. It is 629,563 bytes.  Except for the new first page, it is identical to SS.  It has the same PDF creation date and PDF Author.  The PDF Producer field has been changed from the original "Microsoft ® Word 2019" to something which indicates the file was created by software from https://www.pdf-tools.com .  I will not discuss the first page in this section, but in the next section I discuss a graph from this first page.
  

So two versions concern us: RR is the original and SS is an update 10 days later.  SS has two extra elements.  Firstly, at the end of the Descriptive Statistics section, there is a new paragraph:

A total of 179 cases had Vitamin D deficiency (Vitamin D < 20 ng/ml), mean level of serum 25(OH)D and mean age for this group were 18.2 ± 0.6 ng/ml and 66.9 ± 13.8 years, respectively (Table 2). 213 cases had Vitamin D insufficiency (Vitamin D 20-30 ng/ml), mean level of serum 25(OH)D and mean age for this group were 26.7 ± 1.3 ng/ml and 62.9 ± 14.7 years, respectively. 388 cases had normal Vitamin D levels (Vitamin D > 30 ng/ml), mean level of serum 25(OH)D and mean age for this group were 32.2 ± 1.2 ng/ml and 46.6 ± 12.6 years, respectively. 80.0% of Vitamin D deficient cases had pre-existing conditions (comorbidity). 73.8% of Vitamin D insufficient cases had pre-existing conditions (comorbidity). 18.8% of cases with normal Vitamin D levels had pre-existing conditions (comorbidity). 98.9% of Vitamin D deficient cases died while only 1.1% of them were active cases. 87.8% of Vitamin D insufficient cases died while only 12.2% of them were active cases. Only 4.1% of cases with normal Vitamin D levels died while 95.9% of them were active cases.

Secondly, there is an additional table, though the text references to the prior tables 2 and 3 were not bumped to 3 and 4:

This page no longer contains anything about my attempts to find out about the authors.   However, here are some notes about the hospital mentioned in the paper.  I only pursued this in late June, once I was evaluating my hypothesis that the article does not report on real research.

Before that, I assumed that one, some or all authors worked at a hospital somewhere in Indonesia - and I was aware that it was not on Java or Sumatra, but on the less densely populated Kalimantan district, which is part of Borneo.  The address is:

RSUD Kabupaten Sukamara translates to Sukamara District Hospital. 

2020-08-13 update:  Some of the text below is now in grey since I wrote it on the assumption that the website I was looking at was the hospital's website.  As explained in the #indon-govt section below, I now know that website was fake.  The hospital does not and never had a website.

I initially assumed it was like a major public hospital in Australia or other developed countries - hundreds of beds, well connected as part of a larger presumably government network and with all aspects of its operation fully certified.  I didn't look up the location on Google maps until about 10th July (link).  It is a small hospital, in a semi-rural location, 3km south of the centre of Sukamura.  (In 2019, 63.5 percent of 2,813 hospitals in Indonesia were run by private organizations.) Sukamara is the capital of Sukamara Regency [W] which in 2014 had a population of 48,134 in its 3,827 square kilometres.  There are no Google streetview pictures, which is not surprising.  The two-storey hospital is located in some streets which are not part of Google Maps' database.  Its opening hours are 6AM to 6PM Monday to Saturday.

Its website http://rsudsukamarakab.net (on 2020-07-11) has its two most recent updates in April.  Searching for web pages which mention or link to rsudsukamarakab.net (with English and Indonesian selected in Google's settings), I found only pages which list domain names and a handful of Facebook pages.  I found it surprising that any hospital, even a small one in a rural area, would have so few links to its website.  However, on 2020-08-13, this makes sense, because the website I was viewing, which disappeared sometime around mid-July, was fake.

According to 2012 Worldbank data cited in Wikipedia there were 0.2 physicians and 1.2 nurses per 1000 population.  I couldn't find a list of hospitals at https://www.kemkes.go.id .  The hospital is presumably accredited, though in early 2019 it apparently was not .

If I had looked up this hospital in April or early May when I first read and cited this article, I would have found it to be very different to what I imagined.  Then, I might have paid greater attention to potential weaknesses in the article. 

Looking up the institution(s) the authors are affiliated with is an important part of assessing the veracity of an article.   I just assumed that hospitals in remote areas of Indonesia were operating on the same scale and with the same level of expertise as the larger hospitals here in Australia - although I guess they might be smaller. 

It is possible that a doctor living in this area could communicate with large government hospitals in the rest of the country, working with three colleagues, to obtain all the required data, analyse it and collectively write the Word file from which the PDF file was created, in about a day and a half from the end of data collection - as I found in July 2020 would have to be the case given the article's date range for data collection and the PDF creation date less than two days after this.

This article fits the pattern I observed in many of the Mark Alipio articles: a vaguely defined survey or set of sourced data, with no actual raw data - just some analysis and discussion. 

The first version PDF was made at 7:12PM on 26th April, yet the article states that the researchers used electronic medical record data from March 2 to April 24.  I can't imagine this being realistic.  Blood tests etc. take time to perform and to enter into data systems.  How could the researchers perform their analysis and write their article in a day or so?  This is a clanger I should have noticed.

Again we see vague descriptions of the data sources: "Indonesian government hospitals".  "To ensure anonymity, all names were preserved throughout the analysis." makes no sense.

I should have paid more attention to problems such as this:

The pre-admission serum 25(OH)D levels were considered for the analysis. Serum 25(OH)D level was checked by two physicians based on the available clinical data of the patients.

Why had all these 780 patients been tested for vitamin D levels before being admitted to hospital?  "considered"?  Why would two physicians be needed to establish a patient's 25OHD level if it had been recorded?  Who were these physicians - and how did they do this in a fraction of a day, for a data dump of 780 patient records?  What about incomplete data?

There are a number of similarities between the Mark Alipio article 16 (A16) and this Indonesian article (R1).  Both have a unique (in the entire world, according to Google) header at the top of each page.

A16	R1
Note: This is a pre-print version of the author’s paper before any peer review has taken.	Disclaimer: This is a preliminary study for early dissemination of results. Data are subject to changes
Header was not present on the BB version PDF date 2020-04-23 but was added with the CC version, PDF date 2020-05-07.	Header was not present on the RR version PDF date 2020-04-26 but was added with the SS version, PDF date 2020-05-06.

I don't understand what the first phrase in the following table means - it is quite rare.

A16	R1
Frequency and percentage were used for categorical variables.  (2,000 Google hits.)	frequency and percentage were employed for categorical variables. (2 Google hits.)

I don't have time to pick through the minutiae of the data and analysis.

One thing which did strike me as curious when I first read it, and which remains a problem, is the extreme sharpness of the relationship between 25OHD level and death.  I wrote it up as:

25OHD  Death rate
ng/ml

> 30     4.2%    
 
20-30   49.1%  

< 20    46.7%    

There is further analysis, supposedly based on the data of this article, at: https://borsche.de/res/Indo_3_EN.pdf which I don't have the time or inclination to check.   This analysis is stated to have been done by or in collaboration with the statistician member of the author team.   According to the next PDF, this collaboration resulted in the graph presented in the new first page of: https://borsche.de/res/Patterns_of_COVID_19_Mortality_and_Vitamin_D_An_Indonesian_Study.pdf


 
This is so sharp.  I should have paid more attention to this when I first reproduced the graph on my site.  A 20% drop in 25OHD levels leads to the death rate going from a few percent to 80%??  This is surely not real data.  Even if a population of humans responded in this dramatic and clear-cut manner to a slight change in 25OHD levels, the per-sample errors in the 25OHD measurements would preclude such a sharp relationship appearing in the data.  I really should have thought more about this before taking this data seriously.

I find this data unbelievable.  However, I have no formal qualifications in any field, and I hope other people will make up their own minds about this, and the other aspects of the article.

Important update 2020-08-03: I have been corresponding with Dr Raymond Pranata, who is one of three authors of the article previously mentioned in which they report finding no trace of the 'Raharusun" article's authors, and list several critiques of its veracity:


Dr Pranata kindly authorised me to reproduce here three further reasons why he and his coauthors doubt the veracity of the "Raharusun" article:


Also, despite the Campaign Owner (Lorenz Borsche) being fully aware of these criticisms and the article by the three doctors - as discussed in acrimonious email correspondence I have been a party to since 2020-07-28 - this fundraiser for COVID-19 research, in honour of the late "Dr Prabowo Raharusun", to be administered by Mark Alipio and "Dr Glicio" (below), remained active until it stopped accepting donations on 2020-08-16:


Here is a composite screenshot of the fundraiser page on 2020-08-05.  It remained in much the same state, with the image of the supposed gravestone, until 2022-08-16 when two more small donations had been made and no more were accepted. 

Update 2020-08-20: There was a major change to the fundraiser page.  The gravestone photo was removed and the state of the page was as shown in this PDF, pages 6 to 9 of which reflected the state before this update:


It seems that Lorenz Borsch is contemplating several scenarios in which the whole affair was a scam (and so in which he was a dupe, and unreasonably denied the veracity of my efforts, and those of Dr Pranata and colleagues, to help him).  Yet he does not seem to really believe any of these scenarios, since he refers to us as "attackers from AU and Inonesia".  He notes that the "graveplate" image might have been photoshopped, offers to repay all donations and notes that this will not put him in financial trouble.  The graph is still there. 

There were a number of people - mainly in the UK and Germany, but with some in the USA - who had engaged in extensive email correspondence and at least one phone call with persons they believed to "Dr Raharusun" and at least one of his co-authors.   Monies were solicited from and donated by some of these people for supposed hospital expenses.  Offers were made of vitamin D3 supplements to be sent to Indonesia to further research, but these were not accepted.  Further data was requested from the supposed Raharusun study but only one person received anything in this regard - or at least correspondence regarding further analysis of such purported data - and that person did not share it with anyone else.  Then there was the fundraiser mentioned above.

Further data was also solicited from Mark Alipio and at least one larger, different, "data" set was provided.  I haven't seen this, but some people who received it quickly decided it was fake, due to gaps in the distribution of data points which would never occur in real clinical observations.

Some of these people (I know this from email correspondence with multiple former believers) became suspicious in the early stages of affair, I think around May.  Others retained their belief in the veracity of the "Raharusun" article and in the existence of its purported authors, and likewise of the veracity of Mark Alipio's best-known article, MA16 above until they became aware of the critiques on this webpage.   As far as I know, most of the remaining believers have now (2020-08-20) - 7 weeks after this page was first first established and made known to them - totally or almost totally abandoned their former beliefs.  The first news I had of Lorenz Borsch having any doubts about the veracity of the "Raharusun" article was on 2020-08-20, as described above.

With the three most important articles now withdrawn from their preprint servers - ("Raharusun", then "Glicio" and in recent days Mark Alipio's most widely cited MA16 article (see list above) - there should be absolutely no reason to cite these articles in any website, journal article etc. except to note that these fake articles have been withdrawn, and ideally to point to this page, the Henrina et al, article cited above, and the hospital director's statement  mentioned below, as evidence that these articles were fraudulently created.

However, these articles and the derivative work (graph above) supposedly based on analysis of the supposed data of the "Raharusun" article, have been cited in legitimate journal articles and probably will continue to be cited in the future.   These articles are a cancerous development, and to the extent that author C cites them because he or she saw them cited in an article by author A or B, without checking that the article is still available, and is not fake, then these articles arguably constitute a viral infection or some other pathological species of self-propagating academic fraud.

Ideally I would list all such articles here, and write to all their authors.  I don't have time for this.  If someone would search for such articles on Google Scholar and do this for at least some of them and let me know, I would really appreciate it.

Here, I want to point out a prominent and in many ways promising new website regarding vitamin D which cites the bogus graph developed by Lorenz Borsche and Dr Glauner, supposedly from the "Raharusun data" as mentioned in the previous section.

The site is:


It is produced by the Organic & Natural Health Association  https://organicandnatural.org/about-us/mission-and-vision/ 
whose membership consists of some nutrition organisations (and one concerned with grassfeed meat certification) and quite a number of nutrition supplement companies.

There's an image of a cheery and apparently very healthy brown-skinned young woman with the caption "My level is 55! What's yours?".   Assuming this is ng/ml (nmol/L would be only 22ng/ml) then this is a perfectly healthy 25OHD level.  I am very glad to see vitamin D supplementation promoted, including with the prominent mention of Ken Redcross MD https://www.drkenredcross.com , who also has a line of nutritional supplements, including this one, which combines a gram of EPA/DHA omega 3 fatty acids (which would normally only be found in several grams of fish oil) with 0.05mg (2000IU) vitamin D3 and 250mg vitamin C:


I think this is an unusually good dietary supplement and I think the website is an excellent development, with its focus on people with pigmented skin and the elderly who are most at risk of vitamin D deficiency.

Here is Dr Ken Redcross tweeting (2020-08-18) about his new website PowerOfD.org :


For the avoidance of doubt: I totally support the aims of this website.  My criticisms of their use of the bogus data are not directed at Dr Ken Redcross or any of the other MDs and other specialists who are behind this website:



I salute these energetic people for developing this website.  

The other reason I mention them is to demonstrate that good, professional, people can easily be swept into the process of believing, making healthcare decisions based upon, citing and so promoting and further propagating bogus research.  It has probably not occurred to them that anyone fake such research articles.  It didn't occur to me until late June, and that was only prompted by the suspicious supposed death of a supposed author.  I had been citing these articles with enthusiasm and even gratitude (to Dr Alipio in the Philippines for doing much-needed vitamin D research in the midst of the COVID-19 pandemic), simply because it had no occurred to me anyone would fake research articles like this and because the article presented data which confirmed my belief that low vitamin D was driving COVID-19 severity and because there was a lack of such research. 

The main page of PowerOfD.org has 10 boxes concerning vitamin D and depression, sleep, healthier children etc. each with a link to another website and a research article.  I assume these are all good.

Then there is a link to a page with a video of Joseph Mercola interviewing Ken Redcross.  More on that in a moment.   Below this are more links to research articles and then this black white and orange graphic, boldly captioned:
      STATS DON'T LIE


Actually, statistics need to be interpreted very carefully at the best of times.  In this case, everything in the graph is fake.  See the previous section for the image from which this was adapted and why the data is not just fake, but wildly unrealistic.  Please also see all the research linked to from https://aminotheory.com/cv19/ especially the Iranian research linked to and discussed at: https://aminotheory.com/cv19/#2020-Maghbooli where 25OHD levels above 30ng/ml are shown to be protective against COVID-19 severe symptoms, but where some patients with levels of 40ng/ml and higher still die.

Patients with 30ng/ml or more 25OHD had lower severity and death rates than those with less than this.

This what we need to understand.  The sharp relationship between deaths and 25OHD levels depicted in the bogus Prabowo Raharusun Lorenz Borsche and Dr Bernd Glauner graph above gives a completely false picture of the relationship.  The bogus graph, if believed, gives us a false sense of security that if we could just get everyone above 32 or 35ng/ml there would be few, if any, deaths from COVID-19.  

I am sure that if we could get everyone in a given country to 30ng/ml 25OHD, that symptoms severity would be greatly improved.  Also, I think the amount of virus shedding would be greatly reduced so the infection rates would drop, rather than keep growing, as has happened over summer in the UK and the whole EU, with infection rates now rising as the Sun heads south again and vitamin D levels drop: twitter.com/Covid19Crusher/
 
However, it would be better still if we got everyone to at least 40ng/ml.  For some people, there would be little or no difference in general health, all through their life, between 30ng/ml and 40ng/ml.  However, for some - especially those with particular genetic characteristics, or obesity, or when they are 70+, 40ng/ml will be better for them than 30ng/ml.   Within that set, some people, will be better with 50ng/ml than 40ng/ml and so on.  

For some of those infected, 30ng/ml will help them a lot more than 20ng/ml, but it is not a get-out-of-trouble-free card, as the Iranian data shows.  

I proposed a long downward-sloping  25OHD vs risk of COVID-19 harm curve (red line in second graph) on 2020-07-28: https://aminotheory.com/cv19/#hazard .   I was happy to see a similar hazard curve (also in red) in this article which was published on 2020-08-18.  I am yet to write about this article, which concerns pre-COVID-19 respiratory tract infections:

I added the ng/ml figures for the horizontal scale.



Also from the same article:

 


Now to the 2020-08-01 page with the interview video (linked to from the main page of https://www.powerofd.org) , which links to a transcript:


The text of this page has many good elements.   Dr Redcross advocates 40 to 60ng/ml (100 to 150nmol/L) 25OHD levels which accords with the best research.  He ensures all his patients supplement with vitamin D3 adequately and would much rather do this then prescribe drugs.  He is particularly concerned about low vitamin D levels in people with pigmented skin.  He advocates at least 0.125mg (5000IU) a day D3 supplementation and is disappointed to see mainly 0.025mg and 0.05mg (1000IU and 2000IU) capsules on sale.

Unfortunately, a central part of this page - and apparently of Dr Redcross's current understanding - is directly based on the 100% fake Raharusun and MA16 articles mentioned above.  The above page begins with three dot points, which together with other similar sentences, repeats much of the material in these two later paragraphs:


 
Ref [1] is to a site with a copy of the bogus Raharusun article.  Ref [2] is to Ilie et al.'s country-wide 25OHD COVID-19 mortality article, which is very broad brush.  I don't regard it as concrete evidence for the risk of harm or death virtually disappearing with 25OHD levels above 30ng/ml - see the Iranian chart above.  Ref [3] is to a 2020-06-22 commentary by Damien Downing which cites, first and second, Mark Alipio's bogus article MA16 above and the Raharusun article.  (I might have done the same on that date.  It was only about then that I started looking at these articles critically.)

So here we see great work by Dr Redcross being partly corrupted by quite a high degree of reliance on 100% bogus data.   Furthermore, one of the articles he cites is based on the same fake data.

The same applies to Dr Mercola, who in the interview says:


This page links to a detailed 46 page report Stealth Strategies to Stop COVID Cold:


in which (PDF date 2020-08-11) the Raharusun and Alipio fake articles are also cited.  However, these are just two of 185 references.

After updating this web page, I will write to all the PowerOfD website people pointing them to this critique.   I look forward to updating this account of academic rot to be in the past tense.

Even if all the authors listed for this "Raharusun" article were real people, with good publication histories, we can see from the critiques above that the article cannot be accepted as a reliable account of real research.  Furthermore, the article was only a preprint, has not been submitted to any journal and the preprint has been withdrawn.

Therefore, I suggest, the only reason for citing this article anywhere is to mention that according to all the available evidence it should not be considered a reliable account of actual research.   This is a polite, minimal statement about an article which is a deliberate fake.

Following the publication of the three MDs' article I was party to a number of email exchanges involving one of the three MDs and a number of people who doubted the veracity of their article: that is, they still tentatively or absolutely believed in the existence (and recent death) of "Dr Prabowo Raharusun" and so in the veracity of the article Patterns of COVID-19 Mortality and Vitamin D: An Indonesian Study.

I thought such positions to be entirely at odds with all the evidence.   I am not sure that they had taken the time to read this web page.  The wider situation is that all three articles discussed here, and sometimes one or more of the other Mark Alipio articles concerning vitamin D and COVID-19, were still being cited in numerous peer-reviewed (or in the process of being reviewed) articles as if they were legitimate research reports.

This integration of bogus articles into the fabric of proper research is an appalling development.  Some authors continued to cite the articles even though they had been withdrawn, without any note to this effect.

In order to try to reduce these corrosive practices, the three Indonesian MDs went to some further trouble.  They contacted the director of the hospital cited in the "Raharusun" article and obtained signed statements from her in both Indonesian and English regarding this supposed "Dr Prabowo Raharusun".  Then they made this document available at an Indonesian government website as evidence of the veracity of the director's statement.

In what follows, I have assumed that the statement by the hospital director is entirely true, and that this corroborates the article of the three doctors, as well as my own inability to find any sign that this so-called "Dr Prabowo Raharusun" ever existed.

From this, a number of important things can be reliably ascertained:

1. The three MDs did contact the hospital before completing their article.
   
   
2. The hospital has no record of a doctor of this name.
   
   
3. The hospital has had no involvement with any research project as described in the "Raharusun" article.
   
   
4. The hospital does not have and never has had a website.
   
   
5. The website I and other people referred to, assuming it was the hospital's website, was a deliberate fake.
   

Points 4 and 5 came as a surprise to me.  I discuss the fake website and its contact details in the next section.

The document begins with the three MD's article, followed by the RR initial version of the "Raharusun" article.  Pages 21 and 22 are the Indonesian and the English versions respectively of the statement by the director of the hospital.

The PDF document's name is:



The website it is located on is: http://ppid.sukamarakab.go.id .  However (on 2020-08-13 at least) this server only responds to http and https requests (and pings) from IP addresses deemed to be in Indonesia.   While Google's search engine spider has accessed it in the past, as far as I know, it cannot be accessed directly from most or all computers outside Indonesia.  In order to access it from outside Indonesia, a proxy server or VPN router in Indonesia is required.  I know one person who used another proxy server in Indonesia without success.   This one suffices:


If you use your browser to access this proxy server site, and then in the "Enter URL" box there, type or copy and paste the following URL:

you will see the home page of this server, a screenshot (2020-08-13) of which is here, showing a link to the document we are interested in under the heading Informasi Terbaru (New Information).  The heading of the link is as follows and in Google translation to English:


This link doesn't work (at least for me) in this proxy system, and neither does the proxy's URL bar, so to reach the document page, reload the proxy's home page https://proxy.web.id and type or copy and paste this URL into the proxy-server's "Enter URL" bar, and then click the "Go" button or press the "Enter" key on your keyboard:


Here is a screenshot of what you should see.   Now, you should be able to click the "Download File" button and read and save the PDF.  Here is a screenshot of page 22: the English version of the letter from the director of the hospital, RSUD Sukamara, to one of the three MDs:

During at times acrimonious email discussions with various people in early July 2020, all those involved, including me, were looking at a website which disappeared in mid to late July.

Its URL was: http://rsudsukamarakab.net .   Here is how you can view a screenshot from the homepage or that site, from 2020-07-10.   Click the following link, which is not to any file, and then delete the Z from the end of the file address, and get your browser to load that, the correct URL.  (I have done this to avoid search engine spiders finding this image.)


The site was running Wordpress and evidently had pages added as early as 2017.  I found a few links to this site, such as from Facebook pages, in 2017 or 2018, so evidently this site as we saw it in June and July 2020 had been in existence since 2017, though we can't be sure exactly what it looked like then.  It is not archived in https://archive.org , which as far as I know only occurs if the site owner writes to archive.org, proves they own the site to them and requests it not be archived. 

On 2020-08-14 I could still find Google's search engine caching pages and images from this site: Search-1.pngZ and Search-2.pngZ (Again, click the link, delete the 'Z' and hit "Enter".)

The website only appeared because there was a server, with all the files which made the website AND because whoever controlled the domain rsudsukamarakab.net configured the DNS (Domain Name Service) zone file for that domain to point to that web server.  On 2020-08-14, looking up the full WHOIS response, such as via tools.dnsstuff.com
(enter rsudsukamarakab.net into the WHOIS/IPSHOIS Lookup box and click the arrow there - then click the "Detailed WHOIS Response" arrow on the results page) there is an email address listed as being for the owner of this domain, and it is not any of the addresses listed below.  Screenshot: zyxwvut.pngZ .

From my 2020-07-10 screenshot of the website which appeared to me and others as being of the hospital, here are the email and phone details:


Until 2020-08-13 I assumed this website was of the hospital.  Then I read - as you can see above - that the hospital has no website.

The 26752 number matches that on the hospital letterhead.  The (fake) hospital website displayed three email addresses:


none of which match the one email address on the letterhead:


The dot and the underscore make these two totally different Yahoo email addresses.

During email discussions about the veracity of the "Raharusun" article one person and later a second reported that they had written to the hospital email address inquiring about the late "Dr Raharusun" who they had been told (by emails from his colleagues and others who knew him, including Mark Alipio) had died.  The replies they received were officious but sad confirmation that he had worked at the hospital and that he had died.  One of these replies was forwarded to me.  So I wrote to the admin address about this and received a similar reply.  I didn't believe it and hypothesised that the hospital had somehow lost control of its domain and that whoever then controlled it were part of, or were cooperating with, whichever one or more individuals was encouraging some people to sincerely believe that "Dr Prabowo Raharusun" person really had existed, and had since died.  I didn't pursue this any further, since my aim was to explain to people my reasons for considering all these articles fake - and my analysis of the text of the articles provided more than sufficient evidence to establish this.

On 2020-08-13 and subsequently, because of the PDF mentioned in the previous section, I believe that the domain and its website, and so the three email addresses above, were all the work of some persons who were not employed by the hospital.  The site certainly looked to me like an official hospital site (given this was a small hospital in a remote rural part of a developing country).  The reply I received was evidently written to give the impression that it was sent by an authorised member of the hospital staff.

I have no idea why anyone would create a fake site like this - and apparently do so for three years or more.   It cannot be a friendly gesture, since the email addresses do not match that of the hospital.  Nor do I understand why a hospital would not have a website or take action to shut down an unauthorised site such as this.   However, I have no experience of places such as rural Kalimantan.

Dr Pranata also directed my attention to this Indonesian Ministry of Health listing for the hospital:



This has the same email address and phone/fax numbers as on the hospital letterhead.

This is one of 9 hospitals in the area listed at:


some others of which list no email address or website.

There is another article on the SSRN preprint server which I had previously looked at only briefly.  I decided to evaluate its veracity in late June 2020.

I took this snapshot of the page (2020-06-30), in case it disappeared.  The background image states UNDER REVIEW BY SSRN.  The PDF has the full, four MD, author list:

Vitamin D Level of Mild and Severe Elderly Cases of COVID-19: A Preliminary Report El James Glicio, MD; Siddharth Neelam, MD; Rajeev Rashi, MD; Deepak Ramya, MD No date on the PDF, but 2020-05-05 on the SSRN page. (Not peer-reviewed.) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3593258 PDF creation date is 2020-05-05 19:28.

Update 2020-07-06:  This article was withdrawn in early July.   I wrote to SSRN enquiring as to why, and will report any answer I get here.

This article follows the pattern established by many of the Mark Alipio articles and by the Indonesian article.  

A vaguely defined source of clinical data - the two hospitals are not named, but are in "South Asia".   176 patients had vitamin D levels recorded pre-hospital, from an initial set of 672.  (This seems highly unlikely to me.)  There are no details on the selection criteria of the initial 672, or what sorts of wards they were in.  There is no information on the range of dates in which data was collected.

General discussion with references, including the Mark Alipio article 16 and the Indonesian article.  Analysis of the data . . .   There seems to be rather sharp upper and lower cutoffs in the scatter plot of BMI in Figure 9 - so this does not look like real data to me.

I can find no reason to believe this article represents real research.

On 2020-08-13 I went through some links above to find out what had disappeared on various websites since the page was established on 2020-07-01.  All these items concern Mark Alipio:

• Google Scholar citations URL.
  
  
• Twitter profile.
  
  
• PDFs of articles 01, 02 and 03 at Academia.edu.
  
  
• PDFs of articles 14, 18 and 21 at SSRN.

2020-08-20:  Mark Alipio's most cited fake article number 16 disappeared from the SSRN site a day or two ago, as has his profile page.  

2020-07-01  Initial version.  I updated this several times.

2020-07-13  Revised, simplified version focusing on the text and diagrams of the articles, but also looking at the publication history of Mark Alipio, the content of his articles, and on what I later found out about the Indonesian hospital, and how different this was from what I had initially assumed, with little or no thought.

2020-07-28  Linked to the British Journal of Nutrition article.

2020-08-03  Added three points from Dr Pranata at: #R-critique .

2020-08-10  Added notes #MA13-similar about how to determine whether a journal is predatory or a legitimate open-access journal, with regard to the International Journal of Engineering Technology Research & Management predatory journal in an article similar to Mark Alipio's 13th article appears.

2020-08-13  Added two new sections concerning the "Raharusun" article: #indon-govt and #R-hospital-website .   A new section directly above lists items which this page's links point to, but where the item no longer appears.

2020-08-16  Added publication histories for the three Indonesian MDs.  Noted that the fundraiser had finished, but that the page still exists.  There has been a flurry of email correspondence in the last few days, including with the Davao Doctor's College, who attest that Mark Alipio was not associated with them after November 2019.  "Our school is not party to nor does it support any of the supposed undertakings of Mr. Alipio."

2020-08-20  Mark Alipio's best known article MA16 has disappeared from the SSRN preprint server as has his profile page.   His image is no longer on rotation in the "Topnotcher's Corner" of the Davao Doctors' College homepage.

Lorenz Borsche contemplates that this whole affair may have been a scam, and updates the fundraiser page: #fundraiser-2020-08-20 .

I added the #academic-rot section which primarily discusses the new and promising https://www.powerofd.org site's mistaken reliance on the fake Raharusun and Alipio articles and the Raharusun-Borsche graph.