Significant Decrease in Annual Cancer Diagnoses in Spain during the COVID-19 Pandemic: A Real-Data Study
, Esperanza Torres 1, Manuel Cobo 1, Laura Galvez 1, Antonio Rueda 1, Emilio Alba 1 and Nuria Ribelles 1,*
Round 1
Reviewer 1 Report
Ruiz-Medina et al investigated the impact of COVID-19 pandemic on the histopathological diagnosis of cancer at two University Hospitals in Malaga (E) with a cross-sectional study which recorded the number of diagnoses of cancer made between March 13th 2020 and March 13th 2012 and compared them with the number of diagnoses made in the same periods of 2019, 2018, and 2017. Compared with 2019, the number of diagnoses of cancer had a 17.2% decrease. The most affected malignancies were breast cancer (-26.1%), colorectal neoplasms (-16.9%), and head and neck tumors (-19.8%). A comparison was also made concerning the distribution by stages (I-II lumped as early-stage; III-IV lumped as advanced-stage) at diagnosis; no relevant difference was found concerning these parameters as recorded in the investigated periods.
Even if the study design seems to largely repeat what was previously reported on this topic, I think that this kind of data collection is worth being done in order to guide resource reallocation in Health Systems after the pandemic. The reported data are, however, redundant, even if many of them provided as a supplementary material. I don't think that a comparison among data recorded 2017, 2018, and 2019 is truly necessary.
It would be great to see a comparative of the monthly percent diagnostic decrease and the monthly incidence of COVID-19 cases.
The previous studies on this topic can be summarized in a table.
As a minor query, acronyms (GIST, CUP, CNS) should be spelled out.
Overall, I am strongly convinced that a slimming diet will give us a much better readable manuscript.
Author Response
Ruiz-Medina et al investigated the impact of COVID-19 pandemic on the histopathological diagnosis of cancer at two University Hospitals in Malaga (E) with a cross-sectional study which recorded the number of diagnoses of cancer made between March 13th 2020 and March 13th 2012 and compared them with the number of diagnoses made in the same periods of 2019, 2018, and 2017. Compared with 2019, the number of diagnoses of cancer had a 17.2% decrease. The most affected malignancies were breast cancer (-26.1%), colorectal neoplasms (-16.9%), and head and neck tumors (-19.8%). A comparison was also made concerning the distribution by stages (I-II lumped as early-stage; III-IV lumped as advanced-stage) at diagnosis; no relevant difference was found concerning these parameters as recorded in the investigated periods.
Point 1: Even if the study design seems to largely repeat what was previously reported on this topic, I think that this kind of data collection is worth being done in order to guide resource reallocation in Health Systems after the pandemic. The reported data are, however, redundant, even if many of them provided as a supplementary material. I don't think that a comparison among data recorded 2017, 2018, and 2019 is truly necessary.
Response 1: Regarding the redundancy of data indicated by the reviewer, we consider that it is of crucial importance to try to establish, with the greatest possible evidence, that the decrease in the number of cancer diagnosed during 2020 is attributable to the only extraordinary and differential event that has occurred during that year: the COVID pandemic. And for this, we believe it necessary to show that the incidence of cases in our area has remained clearly stable in previous years.
Point 2: It would be great to see a comparative of the monthly percent diagnostic decrease and the monthly incidence of COVID-19 cases.
Response 2: We have added a new column in Table S1 with the monthly percent diagnostic decrease.
Point 3: The previous studies on this topic can be summarized in a table.
Response 3: The issue indicated by the reviewer is a good idea, but the characteristics of the data provided in previous studies are somewhat heterogeneous, and therefore difficult to group in a table.
Point 4: As a minor query, acronyms (GIST, CUP, CNS) should be spelled out.
Response 4: The acronyms have been spelled out as in Tables 2, S4 and S5
Point 5: Overall, I am strongly convinced that a slimming diet will give us a much better readable manuscript.
Response 5: Maybe the reviewer is right, but the only way to shorten the length of the manuscript would be by eliminating the information referring to the years 2017 and 2018, and as we have already indicated previously, we believe that maintaining these data adds strength to our results.
Reviewer 2 Report
From our point of view, because the study is conducted in two university-affiliated hospitals, you should obtain the ethical comittee approval, that must be specified in the article body.
In the abstract, it would be better if you specify the IMRAD structure, in order to be easier to read.
At the lines 193 and 208, you should review the paragraphs because the readers cannot understand the periods you talked about; you should be more specific and the lines should be better delimited regarding the period of time.
In the tables, it would be better if you specify the percent of patients, not only the absolute number, in order to highlight the difference between the periods of study.
Author Response
Point 1: From our point of view, because the study is conducted in two university-affiliated hospitals, you should obtain the ethical committee approval, that must be specified in the article body.
Response 1: We have considered that it is not necessary to obtain the ethical committee approval since the data of the patients have been analyzed in a grouped way and had been deidentified. Furthermore, our results do not imply any change in clinical practice.
Point 2: In the abstract, it would be better if you specify the IMRAD structure, in order to be easier to read.
Response 2: We have followed the IMRAD structure in the abstract, but we have not added the headers as indicated in the Cancers template
Point 3: At the lines 193 and 208, you should review the paragraphs because the readers cannot understand the periods you talked about; you should be more specific and the lines should be better delimited regarding the period of time.
Response 3: We have modified the paragraphs and we hope to have clarified the information.
Point 4: In the tables, it would be better if you specify the percent of patients, not only the absolute number, in order to highlight the difference between the periods of study.
Response 4: We have added the percentage of patients in the tables.
Reviewer 3 Report
The paper by Sofía Ruiz-Medina1 and colleagues have presented a cross-sectional study to evaluate the impact of COVID 19 on the health system at both primary and hospital care level. Authors are looking for a potential correlation between the decrease in the number of patients diagnosed with cancer in two university hospitals (Spain, 106 Hospital Regional and Hospital Virgen de la Victoria) and COVID 19. Collected data during the first year of the pandemic (from March 13th, 2020, to March 13th, 2021) were compared to those collected one year before (from March 13th, 2019, to March 13th, 2020) within the same institutions. The impact of COVID 19 on the regular performance of the medical care system was significant with a decrease of 17.2% of the diagnosed patient compared to the pre COVID. The paper is well organized and bring up quantitative information about the healthcare system in Málaga which is the trend in Spain as a country and generally in the world.
However, some minor points should be considered.
- The link given in the Reference [1] does not exist anymore. It might be better to refer to a published paper.
- Is there any reason behind using nonparametric test instead of the commonly used parametric tests.
- Authors might display collected data in a contingency table for the clarity of the analysis.
Author Response
Point 1: The link given in the Reference [1] does not exist anymore. It might be better to refer to a published paper.
Response 1: We have rechecked the link of the Reference [1] and it is still active (accessed Jun 20, 2021)
Point 2: Is there any reason behind using nonparametric test instead of the commonly used parametric tests.
Response 2: Parametric tests make the assumption that the data comes from a normal distribution. We verified this by applying the Shapiro-Wilk test to verify if the population is normally distributed. However the p-value was below 0.05, thus we rejected the null hypothesis and have evidence that the data is not normally distributed. Therefore, we applied non-parametric tests (sometimes referred to as distribution free tests) instead of the commonly used parametric tests, since the data did not meet all the assumptions required to apply parametric tests.
Point 3: Authors might display collected data in a contingency table for the clarity of the analysis.
Response 3: We believe that the data already appears in the tables presented in the manuscript and it is possible that the inclusion of new tables would not add new information.
Round 2
Reviewer 1 Report
None
