Factors related to COVID-19 severity and mortality based on comparison between Henan and Zhejiang: an observational study

Background: Coronavirus Disease 2019 (COVID-19) has spread across China and the world, since outbreak at the end of 2019. Although COVID-19 is generally considered as self-limiting, some of cases develop in to fatal respiratory restress, multiple organ failure and even sudden death. According to the data released by the provincial health commissions, there are obvious regional differences in COVID-19 mortality. To explore possible factors of COVID-19 related severity and mortality in laboratory-conrmed cases from Henan and Zhejiang. Methods: We collected the public data released by the government. We analyzed and compared the possible factors, including initial symptoms, gender, age, place of residence, time interval from onset to diagnosis, epidemiological history and the grade of hospital for patient’s treatment in laboratory-conrmed cases. Student’s t-test, Chi-square ( (cid:0) 2) test, Cochran-Mantel-Haenszel (CMH) test and Mann-Whitney U rank sum test were used complied with the applicable conditions. Statistical test was performed with a level of signicance α=0.05. Results: Death cases in Henan were mainly from male, elderly infected persons and those who had a history of contact with patients and underlying basic diseases. Henan had the higher proportion of cases with fever and the lower proportion of asymptomatic infection, and had higher proportions of imported male cases with a contact history, and elderly cases living in rural areas, relative to Zhejiang. The time interval of conrmed cases in rural areas of Henan was 0.8 day longer than that in Zhejiang. Hospitals grade for patients’ treatment in Henan was lower than Zhejiang. Conclusions: Our ndings suggest that initial symptoms, age, place of residence, gender, the time interval from onset to diagnosis and the grade of designated hospital for patient’s treatment are possible factors behind COVID-19 related severity or mortality. This study will provide useful information for public health authorities to develop disease prevention strategies. There is a big urban-rural gap in health resource allocation of public medical institutions in China(10), implying cases from rural areas may have a higher risk of developing severity or death. Our results showed that there was no statistical difference in gender after the stratication (P > 0.05). There were statistical differences in different age groups whatever rural or urban areas (P < 0.05). 21.6% of cases living in urban areas were ≥ 60-year-old in Zhejiang, which was higher than that in Henan (16.7%) with statistically signicant difference (P < 0.05) (Supplementary materials: Table 1). Consistent to early analysis, more cases at the age of 36 ~ 59-year-old or ≥ 60-year-old lived in rural areas in Henan, relative to Zhejiang, and the differences were statistically signicant (P < 0.001). 60.8% of rural cases were imported cases in Henan, which were higher than that in Zhejiang (40.6%) with statistically signicant difference (P < 0.001). Those who lived in rural areas of Henan with Hubei (66.7%) or Wuhan travel history (65.2%) were more than that in Zhejiang (35.0%, 38.7%), and the differences were statistically signicant (P < 0.001). There were 91.9% of cases in urban areas having a contact history, which were higher than that in Zhejiang 85.5% with statistically signicant differences (P < 0.05). The difference of time interval from onset to diagnosis between two provinces had no statistically signicance (P > 0.05), but the time interval in rural areas of Henan was 0.8 day longer than Zhejiang (Table 7). Together, our data indicate that a higher proportion of elderly cases lived in rural areas of Henan, and more young cases with epidemiological history in Henan come from rural areas, relative to Zhejiang.


Introduction
At the end of 2019, an emerging infectious disease named as Coronavirus Disease 2019 (COVID- 19) occurred in several local hospitals of Wuhan, Hubei Province, China (1). Since then, continuous human to human transmissions were observed in Hubei and its outsides leading to more than 80,000 infected persons across mainland China (2). Based on clinical manifestations, blood tests and chest radiographs, this disease was initially considered as viral respiratory illness by clinicians (3). Subsequently, a novel human coronavirus named as severe acute respiratory syndrome (SARS) coronavirus 2 (SARS-CoV-2) was identi ed from the bronchoalveolar lavage uid (BALF) of COVID-19 patients (4,5). COVID-19 is generally considered as a type of self-limiting infectious disease. Most of patients are mild, whereas some of patients develop into fatal respiratory distress posing a great threat to human life (4). To control the epidemic of COVID-19, a variety of interventions, such as containment strategy and emergency response were rapidly started by Chinese government. However, as lack of speci c antiviral protection and vaccines available for COVID-19, more than 3,000 people died from this infectious disease in mainland China until March 09, 2020 (2). applied to analyze strati ed categorical data. Grading variables were analyzed by Mann-Whitney U rank sum test. Statistical test was performed with a level of signi cance α = 0.05.

An overview of COVID-19 epidemic in Henan and Zhejiang
As shown in Fig. 1, Henan is located in the central China with a population of 106.06 million, and Zhejiang is located in the east China with a population of 53.37 million. Henan borders with Hubei, while Zhejiang is about 200 km from Hubei. On January 21, 2019, the two provinces started to announce information about epidemic situation, and the epidemic trend of these two provinces was very similar after analysis. The cumulative number of laboratory-con rmed cases in these two provinces began to increase rapidly after January 27. The existing number of laboratory-con rmed cases in both provinces had declined since February 7, whereas Zhejiang was less than that in Henan after February 10 (Supplementary material: Figure.1). For new laboratory-con rmed cases ( Figure.2A), the number of cases in both provinces arose rapidly after January 24 and declined after February 3. The epidemic appeared two peaks on January 24 (132 con rmed cases) and February 3 (105 con rmed cases) in Zhejiang and one peak on February 3 (109 con rmed cases) in Henan. As shown in Figure.2B, the cumulative cure rate (%) in Zhejiang was higher than that in Henan. China CDC data (6) showed that COVID-19 related mortality mainly came from severe and critical cases and the case-fatality rate of critical cases was as high as 49%. As shown in Figure.2C, the severe rate of both provinces rapidly increased at the early phase and Henan reached a peak (33.33%) on January 23. Since January 27, the severe rates of both provinces began to decline and uctuated around 10%. Notably, the number of critical cases in Henan was higher than that in Zhejiang since February 1, and the highest number of critical cases in Henan and Zhejiang was 38 and 31, respectively ( Figure.2D). The number of critical cases in Henan and Zhejiang began to decrease after February 12 and 14, respectively.
Taken together, higher severe rate at early epidemic phase, lower cure rate and higher number of critical cases may predict more deaths in Henan.

General Features Of Death Cases In Henan Province
Since February 5, deaths continued to be reported in Henan, but no deaths in Zhejiang until February 19. A total of 19 deaths was announced in Henan by February 19, and the general features of 16 death cases (3 deaths with unknown information) were placed in Table 1. The average age of these 16 deaths was (72 ± 12.42) years old with a gender ratio (male/female) of 5:3, and 93.75% (15), 56.25% (9), 31.25% (5), 6.25% (1) of them were ≥ 60-year-old, ≥ 70-year-old, ≥ 80-year-old, < 60-year-old, respectively. The common symptom of 9 cases with known information was fever (6, 66.7%). 6 cases (37.5%) lived in rural areas, and 9 cases (56.25%) lived in urban areas. The time interval from onset to hospitalization was (2.58 ± 2.71) days, and the time interval from onset to diagnosis was (6.77 ± 2.89) days, and time interval from onset to death was (12.08 ± 5.95) days ( Figure.3). There were 3 imported cases (18.75%) coming from Wuhan, Hubei, and 10 (62.50%) non-imported cases including 8 con rmed cases (80.00%) having a history of close contact Page 5/37 with patients. Additionally, underlying basic diseases, such as respiratory disorders, hypertension, cardiovascular diseases, diabetes and cancer were reported in 8 cases with detail information. Presentation of initial symptoms of laboratory-con rmed cases in Henan and Zhejiang Due to individual differences in different patients or factors such as viral virulence and quantity, the initial symptoms of patients after infection often differ. The different initial symptoms may indicate different disease progression and outcomes. Based on the above considerations, we brie y describe and compare the initial symptoms of patients in the two provinces. We nally collected 939 cases in Zhejiang province and 676 cases in Henan province with the data of initial symptoms. As is shown in Figure.4, there were many differences in the initial symptoms of patients between the two provinces. Markedly, the results showed that 87.4% of the patients in Henan had fever in their initial symptoms, compared with Zhejiang Province (75.1%), with signi cant signi cances (P < 0.001). However, 37.8% of patients in Zhejiang Province had more cough than that in Henan Province (17.2%), with signi cant signi cances (P < 0.001). Most other initial symptoms are statistically different (Table 2). It was worth noting that 2.1% of infected persons were asymptomatic in Henan, while 7.2% of patients were asymptomatic in Zhejiang (P < 0.001). However, potential information bias should be considered here. For instance, most of the data reported by Henan didn't include CT imaging performance, and fever and asymptomatic were commonly information reported by both provinces. Herein, the above data indicate that Henan has a higher proportion of patients with fever and lower proportion of patients with asymptomatic. Overall Comparison Of Laboratory-con rmed Cases In Henan And Zhejiang Of total con rmed cases, the proportion of male and female was 50.5%, 49.5% in Zhejiang, and 54.4%, 45.6% in Henan. The proportion of male in Henan was higher, but the difference had no statistically signi cance (P > 0.05). The mean age of the con rmed cases in Zhejiang was (47.25 ± 15.30), which was higher than that in Henan (44.31 ± 16.06) with signi cant difference (P < 0.001). Next, we performed a strati ed analysis of age distribution. The proportion of 19 ~ 35-year-old con rmed cases (28.9%) in Henan was higher than that in Zhejiang (20.7%) (P < 0.001). The proportion of 36 ~ 59-year-old (56.2%) and ≥ 60year-old con rmed cases (20.3%) in Zhejiang was higher than that in Henan (51.8%, 15.1%), with signi cant signi cances (P = 0.033, P = 0.001) (Supplementary materials: Table 1). 79.6% of con rmed cases lived in urban areas of Zhejiang, which was higher than that in Henan (49.7%) with a signi cant difference (P < 0.001). Compared with Zhejiang (46.3%), the proportion of imported cases in Henan (54.4%) was higher, and the difference had a statistically signi cance (P < 0.001). 41.9% and 44.0% of the con rmed cases in Zhejiang had Hubei and Wuhan travel history, which was lower than that in Henan (57.4%, 54.6%). 85.5% of the con rmed cases in Zhejiang had a history of contact with case, which was lower than that in Henan (91.6%). The above differences were statistically signi cant (P < 0.001). The time interval from onset to diagnosis in Henan was (7.35 ± 4.20) days, which was 0.4 day longer than that in Zhejiang (6.95 ± 4.09) days, but the difference was not statistically signi cant (P > 0.05) ( Table 3). These results suggest the Table 3 Comparison of laboratory-con rmed cases in Henan and Zhejiang Comparison of laboratory con rmed cases in Henan and Zhejiang after adjusting for gender Latest studies suggest that the male con rmed cases may have a higher case-fatality rate (6,9).
Nevertheless, the gender is a common confounding factor for other research factors. Than the strati ed χ 2 test was performed. 60.3% of male were imported cases in Henan, which were higher than that in Zhejiang (49.0%) with a signi cant difference (P < 0.001). The proportion of male con rmed cases with a case contact history in Henan (92.5%) was higher than that in Zhejiang (83.1%), with a statistically signi cant difference (P = 0.001). The differences of age in male/female con rmed cases between Henan and Zhejiang had statistically signi cance (P < 0.05). Compared to Zhejiang, the proportion of male/female con rmed cases living in rural areas of Henan was higher with statistically signi cant differences (P < 0.001). The proportion of male con rmed cases with a Hubei or Wuhan travel history was higher than that in Zhejiang, and all differences had statistically signi cances (P < 0.001). The time interval from onset to diagnosis of con rmed cases in Henan was higher than that in Zhejiang whatever male or female, but had no statistically signi cance (P > 0.05) ( Table 4). Above results indicate that Henan has a higher proportion of male cases coming from rural areas, or imported cases with epidemiological history, relative to Zhejiang. Comparison of laboratory-con rmed cases in Henan and Zhejiang after adjusting age Age is an important factor for COVID-19 related severity or deaths, and most of deaths are elderly based on data from Chinese center for disease control and prevention (CDC) (6). Our results showed that the proportions of cases living in rural areas at different ages of Henan were higher than that in Zhejiang, with statistically signi cant differences (P < 0.05). Notably, the proportion of elderly cases living in rural areas in higher than that in Zhejiang and the difference was statistically signi cant (P < 0.001). 59-year-old (91.5%) who had a case contact history in Henan was higher than that in Zhejiang (84.5%, 82.4%) with statistically signi cant difference (P < 0.05). The time interval (from onset to diagnosis) of cases at 36 ~ 59-year-old (7.86 ± 4.45) days in Henan was longer than that in Zhejiang (7.16 ± 4.12) days, with statistically signi cant difference (P < 0.05) ( Table 5). These results indicate that more older cases lived in rural areas of Henan predicting a low medical level, may be related to high case-fatality rate. Additionally, more young cases linked to an exposure of Hubei or Wuhan in Henan may be migrant workers, resulting in more imported cases at the same age level. The higher proportions of 36 ~ 59-year-old cases with a case contact history or longer interval from onset to diagnosis may be related to high severe rate or deaths in Henan. Comparison of laboratory-con rmed cases in Henan and Zhejiang after adjusting for imported cases Imported cases who have been infected at incubation, subclinical, or onset stage may lead to more severe cases or deaths, especially for the cases from Wuhan or Hubei. Our results showed 60.0% of imported cases in Henan are male, which was higher than that in Zhejiang (53.4%), with statistically signi cant difference (P < 0.05). Among non-imported cases, there was no difference in gender distribution between the two provinces (P > 0.05). There were also differences in age distribution in imported cases between Henan and Zhejiang (P < 0.001), and more imported cases with 19 ~ 35-year-old in Henan (P < 0.001) (Supplementary materials: Table 1). Among imported cases, the proportion of cases having a Hubei travel history in Henan (90.1%) was higher than that in Zhejiang (81.2%) with a statistically signi cant difference (P < 0.001).
Among non-imported cases, there was no statistically signi cant difference in visiting Hubei (P > 0.05). Among imported cases or non-imported cases, higher proportions of cases in Henan came from rural areas, relative to Zhejiang with statistically signi cant differences (P < 0.001). There was no signi cant difference between the two provinces about the imported and non-imported cases having a Wuhan travel history (P > 0.05). Among imported cases, 96.9% of cases having a case contact history in Henan, which were higher than that in Zhejiang (23.2%) with statistically signi cant difference (P < 0.001). Among non-imported cases, 100.0% of cases having a history of contact with cases in Zhejiang, which was higher than that in Henan (89.9%), and the difference was statistically signi cant (P < 0.001) ( Table 6). Together, our results indicate that higher proportions of imported cases are male or have a Hubei travel history, or live in rural areas. As mentioned above, imported cases from Hubei lived in rural areas may be migrant workers. Comparison of laboratory-con rmed cases in Henan and Zhejiang after adjusting place of residence There is a big urban-rural gap in health resource allocation of public medical institutions in China(10), implying cases from rural areas may have a higher risk of developing severity or death. Our results showed that there was no statistical difference in gender after the strati cation (P > 0.05). There were statistical differences in different age groups whatever rural or urban areas (P < 0.05). 21.6% of cases living in urban areas were ≥ 60-year-old in Zhejiang, which was higher than that in Henan (16.7%) with statistically signi cant difference (P < 0.05) (Supplementary materials: Table 1). Consistent to early analysis, more cases at the age of 36 ~ 59-year-old or ≥ 60-year-old lived in rural areas in Henan, relative to Zhejiang, and the differences were statistically signi cant (P < 0.001). 60.8% of rural cases were imported cases in Henan, which were higher than that in Zhejiang (40.6%) with statistically signi cant difference (P < 0.001). Those who lived in rural areas of Henan with Hubei (66.7%) or Wuhan travel history (65.2%) were more than that in Zhejiang (35.0%, 38.7%), and the differences were statistically signi cant (P < 0.001). There were 91.9% of cases in urban areas having a contact history, which were higher than that in Zhejiang 85.5% with statistically signi cant differences (P < 0.05). The difference of time interval from onset to diagnosis between two provinces had no statistically signi cance (P > 0.05), but the time interval in rural areas of Henan was 0.8 day longer than Zhejiang (Table 7). Together, our data indicate that a higher proportion of elderly cases lived in rural areas of Henan, and more young cases with epidemiological history in Henan come from rural areas, relative to Zhejiang.  (25), 3B (32) and 2A (37). In order to further explore whether there is really a difference in the level of treatment for those con rmed cases, we analyzed the medical level of the speci c hospitals where the patients were treated. Together, we found that the patients in Zhejiang were indeed treated in higher-grade hospitals, and this difference was statistically signi cant (P < 0.001) ( Figure.5B and Figure.5C).

Discussion
Currently, COVID-19 still remains a big challenge for China and global public health. Even though most of cases are asymptomatic or mild, a large number of severe cases have cost a lot of medical resources and pose a tremendous pressure on clinical treatment. Nowadays, few of cases continue to die every day which cause serious panic to society. Outside Hubei, the case-fatality rate of most provinces such as Zhejiang keeps an extremely low level, while some of provinces such as Henan and Heilongjiang far exceed national average level. Recent studies have indicated that gender (male), age (elder), and physical health (underlying basic disorders) are closely related to the adverse outcomes of the disease (4,6). To deep understand factors behind COVID-19 related severity or mortality, we compared variables of initial symptoms, gender, age, place of residence, epidemiological history, time interval from onset to diagnosis and designated hospitals for COVID-19 treatment in Henan and Zhejiang.
First, we found that higher severe rate at early epidemic phase, lower cure rate and higher number of critical cases were found in Henan, relative to Zhejiang. It has been reported that most of deaths come from severe and critical cases (7), and lower cure rate may associate with medical level of Henan. Features of COVID-19 related deaths suggest that most of deaths come from elderly and male cases and those who have underlying basic diseases or a history of contact with cases, or live in rural areas. Previous study also indicated that elderly patients and those who have potential chronic diseases such as respiratory diseases, cardiovascular diseases, diabetes, hypertension and cancer (7,11). Management of elderly patients, especially those with chronic diseases should be concerned in clinical practice (12). SARS-CoV-2, the pathogen of COVID-19, hijacks human angiotensin converting enzyme 2 (ACE2) receptor for cell entry, the expression of ACE2 involves with the severity of COVID-19 (13,14). Male persons with smoking showed higher ACE2 expression in lungs (9), which may explain the higher proportion of deaths in male population. Fever is an important predictor for the severity of infectious diseases (15,16). Compare to Zhejiang, Henan has a higher proportion of cases with fever, which may support the higher severe rate of Henan in the early epidemic phase. Several studies have reported that asymptomatic patients can shed virus, but whether asymptomatic infection is contagious remains controversial (6,17,18). It is now generally accepted that asymptomatic infection is weak, indicating Zhejiang with a higher proportion of asymptomatic infection, may have more mild patients compared to Henan.
Next, we conducted strati ed analysis of the data from these two provinces. Our ndings suggest that most of imported cases in rural areas of Henan come from male young persons who have a Hubei or Wuhan travel history, or directly contact with cases. Henan is the largest labor export province in China, and the labor output is nearly 13 million in 2018 (19). Therefore, we speculate this group of infected persons are migrant workers and return to Henan before the Spring Festival. The medical level and health literacy in rural areas are always lower than urban areas (20,21). These group of infected persons predict to cause their family members infection, which is di cult to control. Therefore, Chinese governments at all levels adopted containment strategies at early phase of epidemic, and had been demonstrated to be effective.
Overall, the average age of cases in Zhejiang was higher than that in Henan, and most cases are middleaged and elderly infected persons, because aging population in Zhejiang is more serious than in Henan (22,23). However, the case-fatality rate was 0.0% until February 19, which was contrary to our original inference. Our further analysis indicates that most of elderly cases in Zhejiang live in urban areas, while the higher proportion of elderly cases in Henan live in rural areas with a longer time interval, and the time interval in rural areas of Henan is also longer relative to Zhejiang. The level of urbanization in Zhejiang is 68.9%, while Henan is only 51.71% (24,25). There is a big urban-rural gap in health resource allocation of public medical institutions in China (10), Zhejiang has higher-grade of hospitals for COVID-19 treatment, which may explain a higher cure rate and lower severe rate compared to Henan. In the previous battles against SARS and H7N9 (26,27), advanced treatment strategy of severe respiratory diseases was also well established in Zhejiang, and a lower number of critical cases was observed in Zhejiang. The longer interval from onset to diagnosis can also affect the development of disease in elderly infected persons in Henan. As for above reasons, we think elderly cases living in rural areas may be responsible for the higher severe rate and number of critical cases, which further cause more deaths in Henan.

Conclusion
In summary, our study found that higher proportions of patients with fever, imported male cases with a contact history, and elderly cases with a longer time interval from onset to diagnosis who live in rural areas, and the lower proportion of asymptomatic infection and the lower-grade of hospitals for COVID-19 treatment might be related to COVID-19 severity and mortality in Henan. Thus, our ndings suggest that initial symptoms, age, place of residence, gender, the time interval from onset to diagnosis and the grade of designated hospital for patient's treatment may be potential factors behind COVID-19 related severity or mortality. At last, as lack of detail information of the reported cases from Henan and Zhejiang, there are certain limitations to our study.

Declarations
Ethics approval and Consent to participate The data used in this study was completely from public data released by the government, so the ethics approval and consent to participate were not applicable.

Consent for publication
Not applicable.

Availability of supporting data
All data generated and analyzed during this study are included in this published article and its supplementary information les.The details are available from the following o cial websites: National Figure 1 Geographic location of Henan Province and Zhejiang Province More information is available at http://www.henan.gov.cn/jchn/ and http://www.zj.gov.cn/col/col1544731/index.html. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.  The time points of onset, admission, diagnosis and death in 16 death cases. The time points of onset, admission, diagnosis and death in 16 death cases with known information in Henan. Supplementary Files