COVID -19: Does HCQ Work? The Lancet Study

There has been much debate on whether the antimalarial meds, hydroxychloroquine (HCQ) and chloroquine, help decrease the severity of disease in patient who contract HCQ. Most of the studies have shown a benefit to taking the medications when started early. However, a recent study in the Lancet concluded the opposite.[1]

Results of the Study

The Lancet study was a multinational registry analysis using these drugs and possibly also using certain antibiotics called macrolides, such as Azithromycin. WE will refer to these as the “meds”. Total patients were 96,032 hospitalized patients who had COVID who were treated in the hospital. No out-patient data was collected, I.e. no early users of the meds.

Most of the patients enrolled in the study were from the United States (66%) and Europe (17%). The data was collected when the patients entered the clinics and were admitted to the hospital, then placed in a registry that went for data evaluation. Patients who received the meds within 48 hours of an established diagnosis of COVID by nasal swab. It did not include those who started treatment after 48 hours and who were in ICU.

Their conclusion was those who took the antimalarial drugs with or without the macrolides had higher morbidity and mortality compared to those patients who didn’t take them. However, some of their data actually did not support this conclusion.

The results of this study have rocked the medical institutions and has sent a myriad of questions regarding whether these drugs should or should not be prescribed for patients with COVID-19.

Problems and concerns about the study

The authors admit some limitations of their study:

  1. The data should be interpreted cautiously. It was an observational study and they could not exclude the possibility of many unmeasured confounding factors that could have affected the values.
  2. They admit that a cause-and-effect relationship between drug therapy and survival should not be inferred. Just because there is an association between use of the meds and outcomes does not necessarily mean the meds cause the outcome. For instance, patients with known co-morbidities such as diabetes or obesity may have been given the meds more frequently by the clinic due to the prior studies that showed effectiveness, i.e. unknown variables.
  3. Although they noted that there were more cardiac arrhythmias with the meds group, they admit that there was no link to increased cardiovascular risk due to these EKG changes.

Other concerns about the study

  1. The study was funded by multiple pharmaceutical companies who could have inserted some bias to the outcome.
  2. No references within the study included the different clinics that contributed to the study.
  3. They did not state which clinics around the world they included for evaluation and those they did not include for evaluation and exactly what was their criteria of selection.
  4. It is quite amazing to note that many prior studies, that were smaller, had noted marked benefits of the use of these meds and few further large studies had been published, yet all of a sudden a collection of 671 hospitals located in six continents contributed to this study.
  5. Their data is a bit confusing and did not correlate well with their conclusions. Review of their data in their Apppendix b[2] demonstrates that those with co-morbidities, such as comorbidities such as CVD (heart disease), CHF, arrhythmias, Diabetes, hypertension, smokers and immunocompromised showed very little difference in results whether they took the meds or didn’t take the meds.
  6. We know that the majority of people who die from the disease have these comorbidities, but their data didn’t show this. However, in spite of the lack of difference, their data showed a marked increase risk of mortality from the disease when taking these meds, despite almost similar percent risk of problems from these comorbidities. This does not correlate with what we know about the disease at all, nor does it really make sense.
  7. The data is from patients who went to clinic and were hospitalized, which usually means they were relatively advanced in the COVID process. None were really considered early users of the meds. The patients needed a diagnosis from a nasal swab, that could have taken 3 to 4 days to come back. After diagnosis, they were treated 48 hours after the diagnosis, which means they most likely had had the disease for several days to a week before they got treated.
  8. Their death rate was quite high, a little over 11%, with a survival rate of around 89%. The Latest statistic in the Texas showed a fatality rate of around 2% to 3%. Not even Italy had such high rates. Why the huge disparity is difficult to understand, other than there may have been some incorrect recording of the data.
  9. Only 20% of the total were treated with the meds. They were put on the meds too late since their PO2’s were <94% at the beginning, suggesting too late to treat.
  10. We know that the meds only work if used early; starting late they do not help much, and according to their data, potentially may increase risks.


We cannot make firm conclusions from this study when the meds are used early in the disease. If the data can be confirmed, they may potentially worsen the disease if started later in the disease process. However, their data is rather confusing and has many limitations, making it difficult to draw good conclusions from the study results. They should be confirmed with further studies, especially controlled randomized trials before their conclusions can be believed.

[1] Mehra MR, Desai SS, Ruschitzka FR, P AN. Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis. The Lancet. Journal online. May 22, 2020. DOI:


If you are looking to meet with Dr True, we offer telehealth virtual doctor’s appointments.
Schedule yours here.