Some critical notes on a COVID-19 TCM paper

The recent COVID-19 outbreak has led to an enormous amount of preprints and rapidly-approved papers of variable quality. A recently published paper in Pharmacological Research called “Traditional Chinese Medicine for COVID-19 Treatment” caught my eye. The title suggested that Traditional Chinese Medicine (TCM) could be used to treat patients that had fallen ill with the viral disease, but a quick read showed that the paper promised much more than it delivered. Here is a critical review.

As is true for most COVID-19 papers and news articles, it is a free download. The paper was received on March 1st, and apparently accepted and published online within a whopping three days. It is a “Pre-proof”, which according to the ScienceDirect Support Center means “Articles in Press that have been peer reviewed and accepted for publication by the Editorial Board of this publication.”

The paper starts off with a description of the COVID-19 outbreak and how Traditional Chinese Medicine (TCM) might bring new hope to treat the disease. It describes the successful treatment of a COVID-19 patient with plant-based mixture called qingfei paidu decoction (QPD). At first glance, this might be a welcome alternative treatment for a novel disease that is quickly growing to be a pandemic, and for which there are no good treatments or vaccines available. But with great claims, we need to see great data, and this is where the paper does not deliver at all.

Some text on this blog post has also been used in my PubPeer review of this paper, which you can read here.

Great claims, but no data

The introduction states “In 102 cases of mild symptoms treated with TCM, the clinical symptom disappearance time was shortened by 2 days, the recovery time of body temperature was shortened by 1.7 days, the average length of stay in hospital was shortened by 2.2 days, the improvement rate of CT image was increased by 22%, the clinical cure rate was increased by 33%, 27.4% reduction in the rate of common to severe cases and 70% increase in lymphocyte.[3]” 

Reference [3] is a link to a press release in Chinese, but the link does not seem to work, at least from the US. There is no clear description of what the control group was. For example, what was the size of the control group? If 102 cases were treated with TCM, how many controls were not treated? Where can we see that shortening of hospital stay or the cure rate? Fantastic results like these need to be accompanied by fantastic data, and I don’t see any data.

Another sentence from the introduction states: “Among the 701 confirmed cases treated by QPD, 130 cases were cured and discharged, clinical symptoms of 51 cases disappeared, 268 cases of symptoms improved, and 212 cases of stable symptoms without aggravation.[3]” No control group is mentioned here. What were the curation rates for people who were not treated with QPD? Again, without a working link, this statement is not verifiable.

“The effective cure rate of QPD against COVID-19 is over 90%.” What is the “effective cure rate” of COVID-19 cases without QPD treatment? Again, without a control group, this statement does not bear much weight.

Only 1 patient cured, but did he have COVID-19?

After several statements about the high success rates for TCM to treat hundreds of COVID-19 patients, this paper only describes 1 patient. What happened to the hundreds of other patients?

Unfortunately, although the patient was described as “highly suspected” for COVID-19, the paper also casually mentions that the nucleic acid test was negative. Therefore, it is not confirmed that this was a COVID-19 patient.

With only one patient, how sure are the authors that this patient’s illness would have improved without the TCM treatment? There is no good medical history. The paper just mentions he had fever and cough, but how severe his symptoms were is not clear. Maybe they were mild. Maybe the patient would have gotten better without the TCM treatment. With just one patient, this is not clear at all.

A vague description of a miracle drug

Let’s assume that authors indeed have found a great treatment for COVID-19. Wouldn’t they want to share their miracle cure with the rest of the world, so that every doctor can start treating patients and saving lives? So let’s see what the amazing mixture consists of. Here is the description of the QPD mix:

QPD … consisted of Ephedrae Herba, Glycyrrhizae Radix et Rhizoma Praeprata cum Melle, Armeniacae Semen Amarum, Gypsum Fibrosum, Cinnamomi Ramulus, Alismatis Rhizoma, Polyporus, Atractylodis Macrocephalae Rhizoma, Poria, Bupleuri Radix, Scutellariae Radix, Pinelliae Rhizoma Praepratum cum Zingibere et Alumine, Zingiberis Rhizoma Recens, Asteris Radix et Rhizoma, Farfarae Flos, Belamcandae Rhizoma, Asari Radix et Rhizoma, Dioscoreae Rhizoma, Aurantii Fructus Immaturus, Citri Reticulatae Pericarpium, and Pogostemonis Herba

From: Ren et al., Pharmacological Research (2020), doi: 10.1016/j.phrs.2020.104743.

This pseudo-Latin description of plant roots and fruits is less precise than Severus Snape’s potion recipes in the Harry Potter books. Which exact plants do these ingredients come from and in which proportions should they be mixed? With this vague description, it appears impossible to exactly replicate the miracle mixture and actually save a patient. 

The authors are affiliated with the “Engineering Research Center of Efficacy Evaluation and Industrial Development of TCM Classic Formulae “. This might raise the possibility that a patent or industrial application is planned with this TCM mixture, something that could potentially lead to some monetary gain for the authors or their institute. The authors declare they have no conflicts of interest, but the fact that they do not share the exact recipe of the miracle mixture might lead other people to disagree with that.

Four CT scans – one with a mystery watermark

Looking at the four sets of CT scans, scan d) of Feb 4th, where the patient’s lungs look much clearer than in the previous week, looks a bit different than that of the other three dates.

Not only do the lungs look much cleaner (as expected), the surrounding tissue looks different than on the other three sets of images. For example, there is much less dorsal light-grey tissue than in the other three images. The rounding of the lung cavity is also slightly different between the panels in a-c vs. that in d. I am obviously not very skilled in lung CT scans, so this might be easily explained by e.g. a different cross section or a different machine, but it makes me wonder if this is really the same patient.

Of particular interest, panels d) have a green water mark that is not present in the other three sets. It resembles that of certain pre-proof manuscripts, but it looks different than the water mark for the rest of the paper.

Most of the text above has been used in my critical review of this paper on PubPeer, which you can read here. You can leave additional comments below.