The Impact of the Pandemic on the Quality of Colorectal and Anal Cancer Care, and 2-Year Clinical Outcomes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors are commended on this retrospective cohort study comparing outcomes in patients with colon, rectal and anal cancer treated just before and during the COVID-19 pandemic. The investigators found differences in care delivery incl., more frequent 30-day post-surgical readmission, more frequent use of short-course radiotherapy and more frequent interruptions in therapy in the COVID-19 cohort. There were also fewer consultations in the COVID-19 cohort which disproportionally impacted those patients who 30 traditionally experience sociodemographic disparities. However, there was ultimately no difference in 2-year mortality, all-cause mortality, or disease progression between the groups.

The study is important in highlighting challenges during a the pandemic in care delivery and differences based on sociodemographic disparities. Please consider the following comments/questions prior to acceptance.

- Please define what is meant by "appropriate care".

- The differences in age, socioeconomic status, households/dwellings seem marginal and not clinically significant, would qualify this in the text.

- The differences in frequency of surgery is difficult to interpret in the context of multiple primaries, stage at presentation, etc.. Has this been adjusted for?

- Please define "timely" as described in Table 2.

Author Response

Please define what is meant by "appropriate care".

The full definitions of each quality measure are appended in Supplementary Table 1; however, we have amended Table 2 to include select, abbreviated definitions per the reviewer’s comment.  Appropriate oncology consultation was defined as:

• Colon: consultation with a surgical oncologist, with or without a consultation with a medical oncologist
• Rectum: consultation with each of a surgical oncologist AND medical oncologist AND radiation oncologist
• Anus: consultations with each of a medical oncologist AND radiation oncologist

Appropriate treatment was defined for three specific scenarios:

• Colon: proportion of patients who received treatment with surgery for stage I-III disease
• Rectum: proportion of patients who received treatment with systemic therapy for stage II-IV disease
• Anus: proportion of patients who received concurrent chemotherapy and radiation for stage I-III disease

The differences in age, socioeconomic status, households/dwellings seem marginal and not clinically significant, would qualify this in the text.

We have added a qualifier in the Discussion section indicating that observed differences were small.

The differences in frequency of surgery is difficult to interpret in the context of multiple primaries, stage at presentation, etc.. Has this been adjusted for?

We appreciate the reviewer’s feedback.  However, the case mix was essentially the same in both cohorts- while there is different needs based on primary and stage etc, there was no statistically significant differences between the 2 cohorts on relevant baseline factors.  Adjustment was not necessary.

Please define "timely" as described in Table 2.

The full definitions of each quality measure are appended in Supplementary Table 1; however, Table 2 has been edited to include the definitions of timely per the reviewer’s comment:

• Timely Oncologist Consultation: consultation with an oncologist within two weeks of referral
• Timely Receipt of Treatment: initiated treatment within 60 days of the date of diagnosis.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript is well written with the detailed methods section. The main question the research addresses is determining if there was an effect of COVID on the quality of care/ treatment for their newly diagnosed colorectal cancer. This was conducted with the hypothesis that it has impacted the quality delaying treatments and access to care. Similar work has been conducted previously. Unlike some of the other studies, they have examined multiple levels of data ranging from patient- demographics to neighborhood-level information to determine access of care. The author has used multiple sources of data to create an overall dataset and has conducted appropriate analysis. They have also mentioned the details for the study variables and how they derived the study population and the SAP. The authors have clearly stated that the results were not significant but also suggested the need to conduct further research. However, the conclusions were consistent with the aims. The tables are consistent in formatting and include the necessary details. The strengths and limitations of the study have been clearly stated in the discussion section.

Author Response

The manuscript is well written with the detailed methods section. The main question the research addresses is determining if there was an effect of COVID on the quality of care/ treatment for their newly diagnosed colorectal cancer. This was conducted with the hypothesis that it has impacted the quality delaying treatments and access to care. Similar work has been conducted previously. Unlike some of the other studies, they have examined multiple levels of data ranging from patient- demographics to neighborhood-level information to determine access of care. The author has used multiple sources of data to create an overall dataset and has conducted appropriate analysis. They have also mentioned the details for the study variables and how they derived the study population and the SAP. The authors have clearly stated that the results were not significant but also suggested the need to conduct further research. However, the conclusions were consistent with the aims. The tables are consistent in formatting and include the necessary details. The strengths and limitations of the study have been clearly stated in the discussion section.

We thank the reviewer for their thoughtful comments.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The Covid 19 pandemic has had a relevant impact on chronic disease management. This study evaluates healthcare in newly diagnosed patients with colorectal and anal cancer during 2020 compared to 2019.

The main concern is related to the timing of comparison and the analyses are more than confusing. However, the innovation of this study is that they analyse the relationship with mortality.

- The shutdown conducted during 2020 affected all health systems and was difficult to access. It is essential to assess these differences during and after the lockout.

- Patients who had a new diagnosis between February 2019-December 2019 and February 2020 and December 2020 were included. However, it would be essential to know the follow-up time of these patients and also to adjust for this time.

- The authors compare mortality during the following two years, so the comparison cohort was also evaluated during the COVID period. Perhaps this is the reason why there was no differences in mortality.

- The use of SD is not justified given the sample size included in both cohorts and is difficult to interpret in this study. In addition, the authors use the p-value in the third table, making the analyses confusing.

Author Response

The shutdown conducted during 2020 affected all health systems and was difficult to access. It is essential to assess these differences during and after the lockout.

We agree with the reviewer’s comment, hence we compared patients diagnosed prior to COVID to those after.

Patients who had a new diagnosis between February 2019-December 2019 and February 2020 and December 2020 were included. However, it would be essential to know the follow-up time of these patients and also to adjust for this time.

Data was collected on the 6 months following the patient’s first consultation at PM to evaluate performance on quality measures.  This information has been added to the Data Collection section of the Methods.  Data on progression and death was captured for 2 years following date of diagnosis; this information is detailed in the Statistical Analysis section of the Methods.

The authors compare mortality during the following two years, so the comparison cohort was also evaluated during the COVID period. Perhaps this is the reason why there was no differences in mortality.

The time period of the comparator cohort is acknowledged as a limitation of the study design in the Discussion. However, the fact that the follow-up period for patients in the comparator cohort overlaps with the pandemic helps to account for confounding of mortality rates by COVID-related deaths, since cause of death was not always recorded in the chart data.  Delays in treatment initiation have been most closely tied to poorer 5-year clinical outcomes, including progression and mortality, in the published literature, and there have been few if any reports of delays or deferrals of cancer care in those patients who were already receiving treatment when the pandemic happened.  We are starting to see divergence in outcomes at the end of the two-year follow-up, whereby cumulative incidence rates of both disease progression and death are higher in the COVID cohort albeit not yet statistically significant.  Given the trend, it is likely that this difference would become significant given more follow-up time, and a repeat analysis at 5-years post-diagnosis is warranted.

The use of SD is not justified given the sample size included in both cohorts and is difficult to interpret in this study. In addition, the authors use the p-value in the third table, making the analyses confusing.

P-values are capturing the likelihood that an observed difference is due to chance, which is dependent on the sample size. The larger the sample size, the smaller the p-value, and the smaller the sample size the larger the p-value.  Additionally, P-values are computed to test the hypothesis whether the observed values for two samples are different, but it does not confer information on effect size and answer the question “how different are they?” In the case of a small sample size (as with this study), understanding the magnitude of the difference (effect size) is more relevant than testing whether the two values are different, which is why standardized difference was chosen for comparisons in Tables 1 and 2 on the recommendation of our study statistician.  The reason for reporting standardized differences is included in the Statistical Analysis section of the methods. For Table 3, which focuses on time to disease progression and mortality outcomes, hazard ratios are a measure of effect size, so it is relevant to report p-values.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

No further revisions warranted.

Author Response

Thank you for your comments.

Reviewer 3 Report

Comments and Suggestions for Authors

The main concern of this stufdy is that both cohorts compare similar periods of time (during COVID pandemia) and thus, it is difficult to assess differences in the outcomes.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf