Analysis of Pneumonia, Hospitalization, and Fatality Among COVID-19 Cases by Mexican States in Women Under 19 Years: An Ecological Study

1Department of Research and Technological Development, Teaching and Research Directorate, Institute of Public Health from Guanajuato State.Victor Cervera Pacheco 14 Primer Piso Plaza Santa Fé Guanajuato, Gto. Mexico CP36250 2Department of Medicine and Nutrition, Division of Health Sciences, Campus Leon, University of Guanajuato, León, Guanajuato, México CP37670 3Directorate of Health Services, Institute of Public Health from Guanajuato State, Guanajuato, Guanajuato, Mexico CP36000 *Corresponding Author E-mail: npadillar@guanajuato.gob.mx

Since the first report of pneumonia cases in Wuhan, China, in December 2019 1 , and until February 6, 2020, only nine corresponded to children under one year 2 . In contrast to Mexican population up to 50 years old, children have a lower frequency and less severe COVID-19 3 . According to the Centers for Disease Control and Prevention (CDC) in the United States, children under one year of age are more prone to severe COVID-19 4 .
The severity of COVID-19 may be due to the following reasons: endothelium damage by increasing age, changes in coagulation function and the distribution of angiotensin-converting enzyme 2 (ACE-2) receptors, antibodies pre-existing anti-SARS-CoV-2, effects of chronic cytomegalovirus infection, presence of co-morbidities considered high risk for severe and death by COVID-19, and low levels of vitamin D. Most of these factors occur with higher frequency in adulthood that in childhood 3 .
The access level to health services in Mexico varies across states due to its high social, cultural, and geographic diversity, being the industrialization difference between the north and south of the country one of the main reasons 7,8 . Hence, the interest arises in analyze women under 19 years, affected by COVID-19 in each Mexican state, from an ecological design. Another reason is the fear declaration that women have less access to health services, and a gloomy outlook is present for girls and women in the world 9 .
In the national context, the General Directorate of Epidemiology from the Mexican National Health Office publishes an anonymized version of the surveillance system database at a personal level 10 , which, on the gender lens view, is a great advantage to understand the COVID-19 impact on focused groups 11 .
With the above mentioned, this study aims to analyze pneumonia, hospitalizations, and fatality of confirmed cases of COVID-19 in women under 19 years of age in Mexico. For this purpose, we used the National Epidemiological Surveillance System database up to March 31, 2021 10 .

Material and Methods
The study design is quantitative, ecological, comparative, and retrolective.
The universe was the registries of suspected and confirmed cases of COVID-19 in the National Epidemiological Surveillance System from the General Directorate of Epidemiology (NESS/GDE) up to March 31, 2021 10 .
The study population was female patients under 19 years whose data was available from the open NESS database.
The sampling scheme was by availability. All the records included up to the end of March 2021 were considered.
The registries for the analysis were the ones from female patients under 19 years of age with clinical data suggesting COVID-19, according to the operational definition of a viral respiratory disease suspected case: fever, cough, headache, or dyspnea (seriousness), accompanied by at least one of the following: myalgias, arthralgias, odynophagia, chills, chest pain, rhinorrhea, anosmia, dysgeusia, or conjunctivitis 12 , and who underwent the test in Real-time Polymerase Chain Reaction (rRT-PCR) or SARS-CoV-2 antigen determination. The excluded cases were the ones that remained as suspected (registries without rRT-PCR results or SARS-CoV-2 antigen). The deleted entries were those with missing data.
The variables considered for the analysis were: age, state that registered the case, type of patient (who was classified as an outpatient or hospitalized), date of onset of symptoms, date of registration, date of death, if occurred, and diagnosis of pneumonia.
After the approval by the Ethics Committee from the Salamanca General Hospital on April 28, 2021, the database was accessed 10 .
Statistical analysis was descriptive for all variables. For inferential analysis, for each state and at the national level, the proportion of pneumonia cases, hospitalized cases, and Case Fatality Ratio (RFC) were calculated among confirmed cases. The state with the lowest proportion of pneumonia (Durango), hospitalizations (Mexico City), and fatality of cases (Mexico City) was used as baseline group to calculate Odds Ratio and Attributable Fraction in exposed and in the population.
Pearson's r calculation, linear regression, t-test, the respective p values computation, and 95% CI were performed between cases of pneumonia and hospitalizations with CFR. In all cases, to assess the result's statistical significance, the alpha level was set to .05.

results and discussion
In the open database of the NESS/GDE, 48,091 records corresponded to women under 19 years of age throughout Mexico with positive rRT-PCR 10 . Figure 1 shows the Mexican states by Attributable Fraction in those exposed. The comparison was made against Durango state, which showed the lowest prevalence of pneumonia (0.71%) ( Table 1). Table 1 shows the results of confirmed COVID-19 cases in women under 19 years who developed pneumonia. The national prevalence of pneumonia was 2.81%; the highest prevalence was observed in Veracruz with 13.77%, followed by Nayarit with 12.66% and Quintana Roo with 12.08%, and the lowest in Durango with 0.71%. The OR and attributable fractions were calculated using the Durango state as the baseline group. The ORs are statistically significant for most states, and the attributable fractions in the exposed are above 75%. For the Attributable Fractions in the population, the majority were higher in most of the states, around 60%.  Figure 2 shows the correlation and linear relationship between cases of pneumonia and CFR in women under 19 years of age; there is a good correlation and a significant linear relationship, confirming a direct linear relationship between pneumonia and dying from COVID-19. Figure 3 shows the AFe at the national level and in each Mexican state. The highest AFes are present in states from the North of Mexico (Chihuahua, Nuevo León) and the coastlines from the Pacific Ocean and the Gulf of Mexico. Figure 3 shows the AFe at the national level and in each Mexican state. The highest AFes are present in states from the North of Mexico (Chihuahua, Nuevo León) and the coastlines from the Pacific Ocean and the Gulf of Mexico.  Figure 4 shows a good correlation (r=0.82) between the proportion of hospitalized cases and CFR. Also, there is a linear relationship (P<.05) between the variables.
The fatality of cases was 0.60% nationwide. The states with the highest CFRs were Chiapas (4.81%), Aguascalientes (3.99%), and Nayarit (3.80%). For the same states, the ORs were 86.03, 68.59, 65.20; the AFes 99%, 99%, and 98%; and the AFps 5%, 4%, and 4% (Table 3). Table 3 shows the CFRs. The national one was 0.60%, the highest was in Chiapas with 4.81%, and the lowest was in Mexico City with 0.06%. The ORs are high, showing the effect of each state on the CFR, which corroborates the high AFe. On the other hand, the AFps are low in most states.
In most Mexican states, the peaks of pneumonia and those hospitalized by COVID-19 In general, the prevalence curves for pneumonia, hospitalizations, and CFR coincide in the states with the highest prevalence of pneumonia and hospitalizations ( Figure 6).
The girls under 19 years of age positive for SARS-CoV-2, registered in the NESS/GDE up to March 31, 2021, were 48,091. For pneumonia, the states of central and southern Mexico had the highest AFe (> 0.9) (Fig 1), and the lowest prevalence (0.71%) was detected in Durango state, with only 3 cases, while the highest was in the state of Veracruz with 13.77% (Table 2). Also, it is detected that the higher the prevalence of pneumonia, the higher the CFR (r = 0.79, t = 0.68, P 0.0001) (Fig 2).
For the prevalence of hospitalizations due to COVID-19, it was higher in the southern states of Mexico and some northern states such as Baja California, Baja California Sur, Chihuahua, and Nuevo León (Fig 3); the lowest prevalence was in Mexico City with 1.43% and the highest in Nayarit with 30.38% (Table 2). A likely reason for these differences is the socioeconomic inequality across the Mexican states 7,8 . A good correlation and linear relationship were detected between number of hospitalizations and CFR (r = 0.82; t = 7.77, P = .0001) (Fig 4).
Regarding the CFR, the highest was in Aguascalientes state with 3.99%, and the lowest in Mexico City with 0.06%. (Table 3). Meanwhile, the AFes were higher than 0.9 for all states but Durango. Oualha et al. (13) reported that among 27 children hospitalized for COVID-19, pneumonia was present in 17 (62.96%). Women predominated with 63%, and from five deaths, three fatalities from pneumonia were in women.
In Mexico, Rivas-Ruiz et al. (14) reported that from 1,443 confirmed cases of COVID-19, 141 (9.8%) developed pneumonia. In this analysis, pneumonia presented a RR of 53.1 (18. In Mexico, according to the suspected case of respiratory disease operational definition 9 , only the symptomatic patients were tested for the presence of SARS-CoV-2 infection. It partially explains that the mortality in hospitalized girls in Mexico was 5.59%, while the fatality of COVID-19, at the national level, was 0.60%. Also, we see a big difference between the percentages of hospitalized in the states, from 1.46% in Mexico City to 21.93% in Aguascalientes ( Table 2).
The differences between CFRs, AFes, and AFps, may be explained by the fact that each state has a health agency which, although linked to the Federal Health Secretariat, develops independent plans to improve health care for its inhabitants 16 . For example, Guanajuato state was the first entity to have a hospital exclusively for COVID-19 cases in March 2020. Also, socioeconomic conditions and educational level vary from state to state.
It is also shown in Figure 4 that there is a correlation (r = 0.82) and a linear relationship (P <.05) between the proportion of hospitalized girls with the CFR. A likely reason is that the ones  [10] considered for hospitalization were those with dyspnea or chest pain, altered blood pressure, high heart or respiratory rate, high body temperature, or O2 saturation, which were referred urgently to hospitals COVID 17 . It follows that many of the fatal cases would return home for outpatient treatment and days later would become complicated and hospitalized in worse conditions. The national CFR among COVID-19 cases corresponding to women under 19 years was 0.60%, and it varied among Mexican states. The lowest CFR being in Mexico City with 0.06%, meanwhile Chiapas showed the highest with 4.81% (Table 3). The overall CFR worldwide was 2.2%, and for Mexico, it was 9.07% 18 .
The high ORs evidences the effect of dying by COVID-19 for living in each state. Nevertheless, the corresponding 95% confidence intervals are wide, which tells us the low precision of the data. The AFes show this effect, which all are above 90% but few (Table 3). On the other hand, the AFps show the impact of living in each state in death, compared with Mexico City. They were small, noting the effect of the state of residence on mortality (Table 3).
In general, the prevalence curves for pneumonia, hospitalizations, and CFR coincide in  the states with the highest prevalence of pneumonia and hospitalizations. strengths The sample size is large. Therefore, the results are reliable. limitations As it is a database analysis, the results and data quality depend on the correct collection. At the state level, each team collects information and works together with the National Epidemiological Surveillance System.
The AFp could be subject to bias due to variations in the frequency of exposure in the population.
conclusion Level and quality of attention vary across states, which was observed through the values of Attributable Fractions and ORs. Although women under 19 years seem to be mildly affected by COVID-19 in clinical regard, the socioeconomic effects of the pandemic in this sector must be studied and addressed. The future in COVID-19 research could be in social aspects by the Mexican state. The sharing of strategies among states may benefit the attention of the COVID-19 emergency.

Conflict of Interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. author contributions GFV, designed the protocol, analyzed the data, and wrote the first draft of manuscript; NPR, participated in designing the protocol, analyzing the data and participated in writing the first draft of manuscript; ENO, participated in clearing the database and participated in the analysis of the data; MJGL, obtained the database and checked that registries had all data, and participated in writing the first draft of the manuscript; ELL, participated in the analysis of data and writing the first draft; FJMV, searched the literature and made critical review of the protocol and the first draft of manuscript: DADM, had the idea, participated in the analysis of data and writing the first draft of manuscript.

Funding
No funding for this research. data availability statement The datasets obtained from General Directorate of Epidemiology analyzed for this study can be found in Open Science Framework [Padilla-Raygoza N.