Changes in cancer-related mortality during the COVID-19 pandemic (2020–2021) in Nevada

Abstract

Background

Studies have shown COVID-19 pandemic led to excess mortality globally. However, reports on specific leading causes of death are limited. This study used statewide data to evaluate excess cancer-related death by comparing the expected and observed cancer-related deaths during the first two years of pandemic.

Methods

2015–2021 Nevada statewide death certificates and Nevada State Demographer’s population data were analyzed. We evaluated three outcomes of cancer mortality: cancer as the underlying cause of death, cancer as a contributing cause of death, and a combination of both. Causes of death were determined and provided by the NCHS (ICD-10, 2019 version). Negative binomial regressions were used to model cancer deaths to compare the baseline prior to COVID-19 (2015–2019) with the COVID-19 pandemic period (2020–2021). Observed to expected (O/E) ratios and corresponding confidence intervals were calculated.

Results

During 2020–2021, overall cancer-related deaths, combining underlying and contributing causes of death, were 208 cases lower than expected (O/E = 0.98; 95% CI: 0.96-1.00; p = 0.053), with a notably reduced mortality for respiratory system cancers (O/E = 0.86; 95% CI: 0.83–0.90; adjusted p < 0.001). For cancer as the underlying cause of death, deaths were 550 cases lower than expected (O/E = 0.95; 95% CI: 0.93–0.97; p < 0.001). In contrast, for cancer as a contributing cause of death, deaths were 189 cases higher than expected (O/E = 1.23; 95% CI: 1.20–1.27; p < 0.001).

Conclusion

Cancer-related mortality, combining underlying and contributing cause of death, showed a modest decline during the COVID-19 pandemic (2020–2021), primarily driven by fewer deaths related to respiratory system cancers. The causes for the cancer-related mortality reduction could result from the mortality shifting from cancer to COVID-19, lack of diagnosis, and omissions or misclassification in death certification during the pandemic. Future research incorporating data from subsequent years is needed to better monitor and understand changes in cancer mortality during the pandemic.

Introduction

The COVID-19 pandemic has brought about one of the most significant public health challenges of this century, profoundly affecting mortality patterns and healthcare systems globally, with the United States being one of the hardest-hit countries [1, 2]. The impact of the pandemic extends beyond the direct fatalities due to the virus, manifesting in various indirect consequences that have reshaped healthcare priorities and patient outcomes. The disruption of routine health services, the heightened burden on healthcare infrastructure, and the psychosocial stress induced by prolonged periods of social isolation and economic uncertainty are among the myriad ways the pandemic has altered the health landscape [3, 4].

Cancer, as a prevalent and often life-threatening chronic disease, has been severely affected by these pandemic-related changes [5, 6]. Cancer patients, already burdened by their condition, have faced compounded risks due to COVID-19 [4]. They are not only at increased risk of severe outcomes from COVID-19 due to compromised immune functions but also vulnerable to the repercussions of disrupted cancer care [7]. Delayed screenings, postponed treatments, and altered follow-up schedules have emerged as critical concerns, potentially leading to advanced disease stages at diagnosis and suboptimal management of existing cases [6–8].

Recognizing these challenges and the potential shift in cancer mortality trends during the pandemic is crucial. Despite the fact that it is known that the COVID-19 is an immediate threat to cancer patients, however, studies on its impacts on real world cancer mortality are lacking. The overlapping causes of death during the pandemic, particularly between COVID-19 and pre-existing conditions such as cancer, complicate the interpretation of mortality trends and underscore the need for nuanced analyses. In this study, we aim to investigate the impact of the COVID-19 pandemic on cancer-related mortality in Nevada. By analyzing death certificate records from 2015 to 2021, we seek to unravel how the pandemic has not only influenced the overall number of cancer deaths but also affected specific cancer subtypes. We evaluated cancer mortality through three complementary perspectives: cancer as the underlying cause of death, cancer as a contributing cause of death, and a combined analysis of both. This approach enables us to capture the broader burden of cancer-related mortality, distinguishing between direct effects of cancer and its role as a secondary factor in pandemic-related deaths. This research addresses a critical gap in our understanding of the COVID-19 pandemic’s broader impact on chronic diseases, with a particular focus on cancer mortality. It offers an in-depth analysis of cancer mortality trends during a time of unprecedented health challenges, offering insights for healthcare planning and policy formulation to better navigate similar challenges in the future.

Methods

Data source

We analyzed Nevada death certificate records from 2015 to 2021 obtained from the National Center for Health Statistics (NCHS) for our analysis. These records encompassed decedents who were residents of Nevada at the time of death. We stratified yearly death counts by age group, sex, and race/ethnicity. The underlying and contributing causes of death were determined by the NCHS using the International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes. These records were de-identified and exempt from Human Subjects Review by the Institutional Review Board. Nevada population data for 2015–2021 was provided by the Nevada State Demographer’s Office, stratified by year, age, sex, and race/ethnicity. The population estimates included individuals in group quarters.

Outcomes and covariates

We evaluated three outcomes for all cancers and specific cancer subtypes: Cancer as the underlying cause of death, Cancer as contributing causes of death, Cancer as either the underlying cause of death or contributing causes of death (any deaths related to cancer).

The analysis of cancer as the underlying cause of death addressed the research question: During the COVID-19 pandemic, has the mortality rate of cancer as the underlying cause of death changed compared to the pre-pandemic period? Specifically, we aimed to evaluate changes in mortality directly attributable to cancer during the pandemic.

The analysis of cancer as a contributing cause of death addressed the research question: During the COVID-19 pandemic, has the mortality rate of cancer as a contributing cause of death changed compared to the pre-pandemic period? In this analysis, we aimed to explore changes in the role of cancer as a secondary cause of death during the pandemic, as well as the proportion of cases where COVID-19 was recorded as the underlying cause of death.

The analysis of all cancer-related deaths, including cases where cancer was recorded as either the underlying or a contributing cause of death, addressed the research question: During the COVID-19 pandemic, has the overall cancer-related mortality rate changed compared to the pre-pandemic period? This analysis aimed to assess changes in the overall burden of cancer-related deaths during the pandemic.

Cancers and their subtypes were classified according to the ICD-10 Version:2019 as follows:

• C00-C97: All cancers.

• C00-C14: Cancers of the lip, oral cavity, and pharynx.

• C15-C26: Cancers of the digestive organs.

• C30-C39: Cancers of the respiratory and intrathoracic organs.

• C40-C41: Cancers of bone and articular cartilage.

• C43-C44: Melanoma and other skin cancers.

• C45-C49: Cancers of mesothelial and soft tissue.

• C50-C50: Breast cancer.

• C51-C58: Cancers of female genital organs.

• C60-C63: Cancers of male genital organs.

• C64-C68: Cancers of the urinary tract.

• C69-C72: Cancers of the eye, brain, and other parts of the central nervous system.

• C73-C75: Cancers of the thyroid and other endocrine glands.

• C76-C80: Cancers of ill-defined, secondary, and unspecified sites.

• C81-C96: Cancers of lymphoid, haematopoietic, and related tissue.

• C97: Cancers of independent (primary) multiple sites.

The basic model contains three demographic covariates: age group, sex, and race/ethnicity. The age distribution of the decedents was segmented into the following categories: <45 years, 45–64 years, 65–79 years, and 80 years and older, in line with the significant age distribution observed in COVID-19 mortality. Sex was categorized into male and female. Race/ethnicity was categorized into Hispanic and non-Hispanic groups, including American Indian or Alaska Native, Asian or Pacific Islander, Black, and White, based on bridged-race categories. In a year-adjusted model, in addition to these three demographic covariates, year of death was included as a continuous variable.

Statistical analysis

Demographic characteristics were described using counts and percentages. To estimate the expected number of deaths during 2020–2021, we employed negative binomial regression models, chosen due to the presence of overdispersion in the data. Our models were constructed using data from 2015 to 2019 as a baseline to represent the pre-pandemic period. The basic models incorporated age group, sex, and race/ethnicity as covariates, and were further adjusted for year to explore temporal trends. An offset was incorporated in our models to account for population counts.

For each category of cancer mortality (overall cancer deaths, specific cancer subtypes), we compared the observed number of deaths in 2020 and 2021 with those predicted by models. To quantify the deviations, we calculated the observed/expected (O/E) ratios, with ratios greater than 1 indicating excess deaths. We constructed 95% confidence intervals (CI) and performed hypothesis tests for these ratios using Byar’s approximation method [9]. To control the false discovery rate associated with multiple testing in the analysis of cancer subtypes, we applied the Benjamini-Hochberg correction to calculate adjusted p-values [10]. The significance level was set at 0.05.

All statistical analyses were conducted using SAS version 9.4 (SAS Institute Inc.).

Results

Overall cancer mortality

Demographic characteristics of all-cause deaths and deaths from cancers in Nevada during the pandemic years of 2020 and 2021 were presented in Table 1. Nevada’s all-cause mortality was 21% higher than expected (O/E = 1.21; 95%CI: 1.20–1.22; p < 0.001). About 75% of excess deaths were associated with COVID-19. The changes for 2020 (O/E = 1.19; 95%CI: 1.18–1.20; p < 0.001) and 2021 (O/E ratio = 1.24; 95%CI: 1.22–1.25; p < 0.001) were similarly elevated. In contrast, cancer mortality, when assessed as the underlying cause of death, showed a marked decline. Observed cancer deaths totaled 10,312 compared to an expected 10,862, yielding an O/E ratio of 0.95 (95%CI: 0.93–0.97; p < 0.001). However, when analyzing the years 2020 and 2021 separately, only the change in 2021 was statistically significant (O/E = 0.92; 95%CI: 0.89–0.94; p < 0.001), with no statistically significant change observed in 2020 (O/E = 0.98; 95%CI: 0.96–1.01; p = 0.178) (Fig. 1). When considering any death related to cancer (either as the underlying or contributing cause), a similar downward trend was observed for 2021 (O/E = 0.95; 95%CI: 0.93–0.98; p < 0.001) (Fig. 2).

Table 1.

Demographic characteristics of all-cause deaths and deaths from cancers: Nevada, 2020–2021

	All-cause deaths		Deaths from cancers
	(N = 60 871)		(n = 10 312)
	No.	%	No.	%
Age
< 45 years	4693	7.7	325	3.2
45–64 years	12 801	21.0	2278	22.1
65–79 years	22 331	36.7	4797	46.5
80 + years	21 046	34.6	2912	28.2
Sex
Female	26 628	43.7	4820	46.7
Male	34 243	56.3	5492	53.3
Race and ethnicity
White^	42 322	69.5	7611	73.8
Hispanic	7327	12.0	967	9.4
Black^	6223	10.2	893	8.7
Asian or Pacific Islander^	4412	7.3	774	7.5
American Indian or Alaskan Native^	587	1.0	67	0.7

^Non-Hispanic

Fig. 1.

Cancer as the underlying cause of death. A 2020–2021. B 2020. C 2021. “Adj. p” refers to Benjamini-Hochberg adjusted p-values

Fig. 2.

Cancer as either the underlying cause of death or contributing causes of death (any deaths related to cancer). A 2020–2021. B 2020. C 2021. “Adj. p” refers to Benjamini-Hochberg adjusted p-values

Specific cancer subtype mortality

The decline in overall cancer mortality was mainly due to the decline in deaths related to cancers of the respiratory and intrathoracic organs. As one of the leading causes of cancer death, deaths from cancers of the respiratory and intrathoracic organs showed a substantial decrease with an O/E ratio of 0.85 (95%CI: 0.81–0.88; adjusted p < 0.001) for the combined two years, with statistically significant reductions observed in both 2020 (O/E = 0.92; 95%CI: 0.87–0.97; adjusted p = 0.043) and 2021 (O/E = 0.77; 95%CI: 0.73–0.82; adjusted p < 0.001) (Fig. 1). A consistent downward trend was also observed when considering cancers of the respiratory and intrathoracic organs as either the underlying or contributing cause of death (Fig. 2). Conversely, mortality related to cancers of the lip, oral cavity, and pharynx (O/E = 1.30; 95%CI: 1.15–1.46; adjusted p < 0.001), and male genital organs were higher than expected (O/E = 1.16; 95%CI: 1.08–1.24; adjusted p < 0.001) (Figs. 1 and 2).

Time-adjusted cancer mortality

Considering that certain cancer subtypes might naturally fluctuate in mortality over time independent of the COVID-19 pandemic, we incorporated the year of death in our models to further adjust for this natural temporal influence. After testing all cancer subtypes using data from 2015 to 2019, we found that only the mortality related to cancers of the respiratory and intrathoracic organs showed a statistically significant and sustained decrease over time (adjusted p = 0.001). After time-adjustment, the combined two-year O/E ratio for respiratory system cancers was 0.97 (95%CI: 0.94–1.01; p = 0.189), with an O/E ratio of 1.05 (95%CI: 0.99–1.10; p = 0.100) for 2020 and 0.90 (95%CI: 0.85–0.95; p < 0.001) for 2021.

Cancer as contributing cause of death

When examining cancer as a contributing cause of death, there was a general increase across most cancer types. The O/E ratio for overall cancer as a contributing cause was 1.23 (95%CI: 1.20–1.27; p < 0.001), aligning more closely with the excess mortality seen in all-cause deaths. Regarding specific cancer subtypes, statistically significant increases above expected mortality were observed not only in several major cancer types, including cancers of digestive organs (adjusted p = 0.001), respiratory and intrathoracic organs (adjusted p = 0.036), breast cancer (adjusted p < 0.001), and male genital organs (adjusted p < 0.001), but also in cancers of the lip, oral cavity, and pharynx (adjusted p = 0.010), melanoma and skin (adjusted p = 0.036), thyroid and other endocrine glands (adjusted p < 0.001), and lymphoid, hematopoietic, and related tissue (adjusted p < 0.001). Importantly, many of these increases were related to COVID-19 as the underlying cause of death. For the cases where cancer was listed as a contributing cause of death, a substantial portion had COVID-19 as the underlying cause. In 2020, for example, among the excess cases with digestive organs cancer as a contributing cause, 9 out of 23 excess cases (39.1%) listed COVID-19 as the underlying cause. Similarly, for cancers of respiratory and intrathoracic organs, 19 out of 27 excess cases (70.4%) were due to COVID-19 as the primary cause of death; for breast cancer, 11 out of 33 cases (33.3%); for male genital organ cancers, 10 out of 44 cases (22.7%); and for lymphoid, hematopoietic, and related tissue cancers, 18 out of 33 cases (54.5%) were attributed to COVID-19 as the underlying cause (Fig. 3).

Fig. 3.

Cancer as the contributing causes of death. A 2020–2021. B 2020. C 2021. “Adj. p” refers to Benjamini-Hochberg adjusted p-values

Discussion

During the COVID-19 pandemic, cancer patients were considered a particularly vulnerable group due to their weakened immune systems from the disease itself and treatments such as chemotherapy or radiotherapy [11, 12]. Furthermore, cancer patients often require prolonged hospital stays, increasing their risk of exposure to COVID-19 [13]. Compounding this, compared to non-cancer patients, individuals with cancer not only have a higher risk of contracting COVID-19 but also face a greater likelihood of severe clinical events, including death [14, 15]. These factors suggested an expected rise in cancer mortality during the pandemic. However, our real-world study in the state of Nevada, USA, revealed an unexpected finding: from 2020 to 2021, the number of cancers as the underlying cause of death were lower than anticipated (p < 0.001). The first explanation that can be thought of is that the excess deaths caused by COVID-19 lead to the shift in the underlying cause of death from cancer to COVID-19. In our analysis of cancer as a contributing cause, we have already found that the number of deaths where many cancer subtypes were listed as contributing causes was higher than expected. A large part of this excess can be attributed to COVID-19 being the underlying cause. This finding aligns with our existing knowledge that cancer patients have a higher risk of death when infected with COVID-19, directly reflecting the life-threatening impact of the virus on this patient group.

However, the increase in deaths with cancer as a contributing cause cannot fully explain the observed decrease in cancer-related deaths. We propose three possible explanations for this phenomenon: (1) Natural decline in the incidence and mortality of certain major types of cancer, even without the impact of the pandemic; (2) Delays or reductions in cancer diagnosis during the pandemic; (3) Potential omissions or misclassification in death certification during the pandemic.

Natural decline in the incidence and mortality rates of lung cancer

Our analysis revealed that the overall decrease in cancer mortality was primarily driven by cancers of the respiratory and intrathoracic organs, which are among the leading causes of cancer death. The majority of this cancer subtype are cases of lung cancer. Our time-adjusted model confirmed that lung cancer mortality in Nevada had been decreasing over time (adjusted p = 0.001), even before the COVID-19 pandemic. National cancer statistics in the United States have shown a steady decline in lung cancer incidence since 2006 [16]. Nevada has observed a similar trend, with a more rapid decrease in both incidence and mortality rates of lung cancer [17]. On one hand, this may be attributed to the implementation of lung cancer screening and advances in lung cancer treatment. Lung cancer screening has been proven to reduce mortality rates in high-risk groups by 16%-24% by detecting asymptomatic malignancies amenable to curative treatment [18, 19], but the actual uptake remains low, especially during the pandemic [20]. On the other hand, public health policies such as smoking restrictions and bans, including the Nevada Clean Indoor Air Act (NCIAA) of 2006 (and subsequent modification in 2011), may also have played a significant role [21]. The NCIAA aims to protect children and adults from cigarette smoke and second-hand smoke in most public places and workplaces. Studies have shown that the incidence and mortality rates of lung cancer in Nevada have rapidly declined in the years following the implementation of the NCIAA and other health policies [17].

Delays or reductions in cancer diagnosis during the pandemic

After adjusting for the natural decline over time, we still observed a statistically significant decrease in mortality for cancers of the respiratory and intrathoracic organs in 2021 (p < 0.001). This additional decrease prompts consideration of several pandemic-related factors that may have contributed to this trend. In March 2020, stay-at-home orders implemented in many countries to combat the COVID-19 pandemic had substantial clinical impacts on healthcare services. A recent study, encompassing about half of the U.S. population, revealed a decline in the incidence of all cancer types beginning in March 2020, coinciding with the implementation of various state and municipal stay-at-home orders and guidelines. This decline reached its nadir in April 2020 and persisted at lower levels, including for lung and prostate cancers. Despite a significant drop in cancer incidence from March to May 2020, there was no compensatory surge in incidence by the end of 2020, failing to return to pre-pandemic levels [22]. There is ample evidence suggesting that the pandemic led to reductions in cancer screening, diagnosis, treatment, and patient visits [23–26]. Changes in healthcare access and delivery during the pandemic could have led to underdiagnosis. Concerns about COVID-19 infection might have caused some patients to defer medical appointments and screenings, potentially leading to delayed cancer diagnoses. Furthermore, the prioritization of COVID-19 care in hospitals may have contributed to reduced cancer screenings and diagnoses.

Omissions or misclassification in death certification during the pandemic period

In addition to the delays in cancer diagnosis or registration, death certificates during the pandemic may have had omissions or misclassification. A study in Michigan revealed that a substantial proportion (about 50%) of reviewed COVID-19 death certificates did not list comorbidities (contributing causes of death) that increase the risk of severe COVID-19 infection or death. Guidelines from the CDC and WHO clearly state that comorbidities should be included on COVID-19 death certificates. Such errors could lead to inappropriate conclusions about the morbidity and mortality associated with COVID-19 [27].

Comparison with other studies

A study based on the underlying cause of death in Mexico found that cancer mortality in 2020–2021 was 5.3% lower than expected based on data from 2015 to 2019, despite a significant excess in all-cause mortality observed during this period [28]. Our study on the underlying cause of death in Nevada showed a very similar result of 5% lower than expected. However, a recent study on national cancer mortality rates in the United States indicated that relying solely on conventional mortality statistics based on the underlying cause may underestimate the impact of the pandemic on cancer patients [29]. The authors of this study found an increase in cancer-related deaths (including both underlying and contributing causes) during the first two years of the pandemic (2020–2021), with a decreasing proportion of deaths attributed to cancer as the underlying cause. Our results largely align with their findings: of 15 cancer subtypes analyzed as the underlying cause of death, only five exceeded expectations, with just one showing a significant increase. When analyzing cancer-related deaths, 10 out of 15 subtypes exceeded expectations, with three showing statistically significant increases. However, it is noteworthy that our data showed an overall modest decrease in cancer-related deaths for 2020–2021, mainly due to a decline in 2021, which contrasts with the national upward trend. In Fedeli et al.’s study, the nationwide decline in respiratory system cancer-related deaths in the United States narrowed in 2021; in contrast, our study observed an expanded decrease in respiratory system cancer-related deaths in Nevada for the same year. This phenomenon may reflect some systemic differences in Nevada compared to other states. In addition to the positive effects of anti-smoking measures previously mentioned, it could also be related to Nevada’s lower lung cancer screening rate. The American Lung Association reported that the screening rate among high-risk population for lung cancer in Nevada is lower than the national level, and there is study pointing out the scarcity of lung cancer screening centers in Nevada [30]. In addition, we adjusted for demographic information in the models to obtain more accurate estimates.

Strengths

This study has several notable strengths. First, we utilized Nevada state-level mortality data from the NCHS, which offers high reliability and provides a comprehensive representation of cancer mortality trends during the early phase of the COVID-19 pandemic in Nevada. Second, we analyzed cancer mortality from three distinct perspectives: as an underlying cause, a contributing cause, and a combination of both. This approach offers a multifaceted understanding of the role of cancer in mortality under different classifications. Third, we accounted for potential confounders by adjusting for age, gender, race, population size, and year of death, ensuring robust and reliable results. Lastly, to control the false discovery rate and reduce the risk of false-positive findings, we applied the Benjamini-Hochberg correction in the analysis of cancer subtypes.

Limitations

Despite its strengths, this study has several limitations. First, this study relies solely on death certificate records, which are influenced by the certifying physician’s subjective judgment. As mentioned in the discussion, contributing causes of death were often overlooked during the pandemic, and we lack detailed clinical data to determine whether COVID-19 directly worsened cancer or whether cancer-related deaths involved prior SARS-CoV-2 infection. Second, collider bias may have influenced the observed relationships due to the inherent nature of mortality data, which only captures cases of death and excludes those who survived. The exposure variable (the COVID-19 pandemic) and the outcome variable (cancer-related mortality) are both linked through the collider—mortality. During the pandemic, the recording of COVID-19 as the underlying cause of death while overlooking contributing causes may have led to an underestimation of the cancer-related mortality burden. This selective reporting in death records could potentially distort the observed changes in cancer-related mortality. Third, this study only covers data from 2015 to 2021, focusing on the early phase of the COVID-19 pandemic. Future studies incorporating data from subsequent years are needed to validate the findings presented here and provide insights into post-pandemic trends. Fourth, the data used in this study are derived from death certificates, which only capture information at the time of death. Therefore, the findings should be interpreted as associations between the pandemic and cancer-related mortality rather than causal inferences.

Conclusions

In conclusion, our analysis of death records in Nevada reveals the changes in cancer-related deaths during the first two years of the COVID-19 pandemic (2020–2021). Overall, cancer-related deaths were slightly lower than expected, but the trends varied among different cancer subtypes. Most subtypes cancers either saw higher-than-expected death or no statistically significant changes during COVID-19 pandemics. The substantial decrease in lung cancer-related deaths drove the overall reduction in cancer-related mortality. This is partly due to the continuous decline in lung cancer incidence in Nevada in recent years. On the other hand, it might be due to delayed cancer diagnoses or omissions of cancer as contributing causes on death certificates during the pandemic. Moreover, these indirect effects of the pandemic likely led to a systemic underestimation of the overall cancer-related death burden, not just for lung cancer. The contrasting trends of decreased cancer deaths as the underlying cause and increased cancer deaths as the contributing cause highlight a potential shift in the role of cancer in mortality during the pandemic. Continued monitoring of cancer-related deaths is necessary to further assess the impact of the COVID-19 pandemic on cancer patients. At the same time, actions should be taken to increase participation rates in recommended cancer screenings among high-risk populations.

Author contributions

W.Y. provided data and software.L.W. and K.S. conducted the statistical analysis.L.W. wrote the main manuscript text and prepared figures and tables. W.Y. and K.S. edited the main manuscript text.W.Y. supervised the whole project.All authors reviewed the manuscript.

Funding

Not applicable.

Data availability

The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request. Data are located in controlled access data storage at University of Nevada, Reno.

Code availability

The codes used during the present study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.