Comparative analysis of risk factors for COVID-19 mortality before, during and after the vaccination programme in Mexico

Abstract

Objectives

The purpose of this study was to compare case fatality rates (CFRs) and odds for mortality by risk factors of patients with COVID-19 in Mexico, before, during and after the implementation of the national COVID-19 vaccination programme.

Study design

A large database including COVID-19 monitoring cases was used to perform an observational retrospective study.

Methods

The Chi-squared test and multivariate logistic regression analyses were applied to data from COVID-19-positive patients in Mexico. Data were analysed over 3 years, 2020, 2021 and 2022, corresponding with pre-, during and post-vaccination periods. The unadjusted odds ratios and 95% confidence interval were used to estimate the risk factors for COVID-19 mortality in each of the years.

Results

Statistically significant differences in CFR and odds ratio were found in the studied years, favouring postvaccination period. Significant changes in CFR by age, sex and main comorbidities indicated changes in the epidemic dynamics after the implementation of the COVID-19 vaccination campaign. The likelihood of death increased for hospitalised cases and for patients who were middle-aged or older in 2021 and 2022, whereas the odds of death associated with sex and comorbidities remained similar or reduced over the 3 years.

Conclusions

Implementation of the COVID-19 vaccination programme during 2021 showed positive consequences on CFR. The increased odds of dying in hospitalised patients are likely to be due to the unvaccinated proportion of patients.

Introduction

In December 2019, a COVID-19 of the respiratory tract characterised by a severe acute respiratory syndrome, caused by a beta-coronavirus named SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2),^{1 , 2} emerged in Wuhan City, Hubei province in China.³ Despite rigorous global containment and quarantine efforts, the COVID-19 outbreak become a global pandemic. Currently, with a global vaccination strategy in all continents, the disease has resulted in almost 500 million positive cases and around 6.2 million deaths.⁴ In Mexico, the first cases of COVID-19 were registered in January 2020, and official confirmation was published in late February 2020. On 20 April 2020, the Mexican government officially declared that the country had entered phase 3 of the COVID-19 pandemic (i.e. epidemic phase). Following three major peaks of infections, the epidemic has now decreased to an incidence rate of 3.3 cases per 100,000 inhabitants.⁵

Mexico has population of almost 130 million, of whom 52% and 48% are female and male, respectively,⁶ with a high prevalence of comorbidities, such as hypertension, obesity and diabetes, which are considered the risk factors for death.⁷ After a year of implementing the national COVID-19 vaccination programme that formerly prioritised front-line health workers and elderly patients, with or without main comorbidities (i.e. hypertension, diabetes and obesity), it is estimated that >65% of the national population is now fully vaccinated.^{8 , 9}

Although COVID-19 outcomes and risk factors associated with death have now been documented for Mexico,10, 11, 12 there is a lack of national epidemiological analyses after the implementation of the COVID-19 vaccination campaign, including the emergence of new viral variants resulting in important peaks of the disease.^{13 , 14} Hence, the objective of the present study was to compare case fatality rates (CFRs) and the risks of mortality for COVID-19-positive patients during three 1-year periods corresponding with the time before, during and after (i.e. 2020, 2021 and 2022, respectively) the implementation of the national COVID-19 vaccination programme in Mexico.

Methods

Data source

A database including COVID-19 monitoring cases was downloaded from the open data source of the Epidemiologic Surveillance Source of Respiratory Viral Diseases (Sistema de Vigilancia Epidemiológica de Enfermedades Respiratorias Virales) that contained information from 475 monitoring units across the country from the public and private health sectors. Positive cases were extracted and edited. Data from 6,657,667 patients diagnosed as positive for COVID-19, from the first positive case registered on 13 January to 24 october 2022 (database accessed on 25 october 2022) were analysed. The data were divided into three 1-year data sets, as follows: C2020, cases until 23 December 2020, a day before the start date of the national vaccination programme;¹⁵ C2021, cases from 24 December 2020 to 23 December 2021; and C2022, cases from 24 December 2021 to 24 October 2022. All COVID-19-positive cases were diagnosed using real-time polymerase chain reaction and were officially registered by the National Network for Epidemiological Surveillance (Red Nacional de Laboratorios de Vigilancia Epidemiologica), recognised by the Institute of Epidemiological Diagnosis and Reference (Instituto de Diagnóstico y Referencia Epidemiológicos, InDRE).

Each patient record included information on age, sex, smoking habits, exposure history, comorbidity traits and clinical care management. Sex was recorded as male or female. The following characteristics were recorded as ‘yes’ or ‘no’: smoking habits, hospitalisation, endotracheal intubation, intensive care unit (ICU) admission, hypertension, obesity, cardiopathy, pneumonia, chronic obstructive pulmonary disease (COPD), asthma, immunosuppression, chronic kidney disease (CKD) and other complications.¹⁰ CFR for each year was calculated by dividing the number of deaths from COVID-19 by the number of individuals diagnosed with COVID-19, and the resulting ratio was then multiplied by 100 to be expressed as a percentage. CFRs were calculated for patient characteristics.

Statistical analyses

All statistical analyses were performed using SAS v.9.4 (SAS Institute Inc., Cary, NC, USA). Differences in CFR by year, clinical characteristics and comorbidities were examined by a Chi-squared test using the FREQ procedure, and the differences were confirmed by an exact logistic regression assuming a logit binomial distribution of data using the GENMOD procedure. The unadjusted odds ratio and 95% confidence interval for the different levels of risk factors of COVID-19 were modelled with a multivariate logistic regression model that included the effects of age, sex, smoking habits, patient hospitalisation and comorbidity traits. Comorbidities included hypertension, obesity, pneumonia, COPD, asthma, immunosuppression, CKD and other complications, using the LOGISTIC procedure. For the current analysis, age was classified into 18 groups of 5-year intervals, from 0 to up to >84 years old. Statistical significance was set at <0.01.

Results

The frequencies and CFR for each year group are presented in Table 1 . CFRs were significantly different among the three year groups (P < 0.0001). The C2022 group showed the highest frequency of positive cases; however, the CFR was lower than in the C2020 group. In terms of age groups, the CFR in C2020 was only noticeably different to C2021 for some elderly age groups (>60 years); however, a progressive decrease in CFR was observed over the study period, largely during C2022 among all age groups. CFR across sex, smoking habits and comorbidities from C2020 to C2022 significantly decreased (P < 0.0001). Interestingly, the absolute number of hospitalisations and related traits remained fairly similar during C2020 and C2021, with a slight reduction in C2022, yet their frequencies were statistically different (P < 0.0001).

Table 1.

Frequencies and case fatality rates (CFR) in COVID-19-positive during 2020, 2021 and 2022 in Mexico according to risk factors.

Risk factor	2020				2021				2022
	Non-survivors	Survivors	Total	CFR (%)	Non-survivors	Survivors	Total	CFR (%)	Non-survivors	Survivors	Total	CFR (%)
n	138,789	1,258,207	1,396,996	9.9	140,973	2,171,003	2,311,976	6.1	23,841	2,924,854	2,948,695	0.8
Age (years)
0–4	262	7127	7389	3.5	219	20,479	20,698	1.1	181	36,613	36,794	0.5
5–9	48	7571	7619	0.6	55	30,085	30,140	0.2	41	49,835	49,876	0.1
10–14	67	14,411	14,478	0.5	101	58,700	58,801	0.2	52	78,243	78,295	0.1
15–19	166	34,475	34,641	0.5	271	113,351	113,622	0.2	91	101,414	101,505	0.1
20–24	501	90,270	90,771	0.6	732	231,006	231,738	0.3	129	288,240	288,369	0.0
25–29	1096	146,100	147,196	0.7	1832	294,145	295,977	0.6	231	379,891	380,122	0.1
30–34	2101	153,417	155,518	1.4	2908	264,796	267,704	1.1	319	376,921	377,240	0.1
35–39	3347	147,909	151,256	2.2	5031	244,947	249,978	2.0	422	334,645	335,067	0.1
40–44	5754	137,690	143,444	4.0	6484	207,504	213,988	3.0	513	305,582	306,095	0.2
45–49	9539	138,085	147,624	6.5	9977	200,450	210,427	4.7	871	285,337	286,208	0.3
50–54	12,772	116,093	128,865	9.9	11,869	153,231	165,100	7.2	1352	238,709	240,061	0.6
55–59	16,743	92,739	109,482	15.3	15,268	117,612	132,880	11.5	1810	172,482	174,292	1.0
60–64	19,392	64,878	84,270	23.0	17,860	84,580	102,440	14.9	2227	107,617	109,844	2.0
65–69	19,544	43,796	63,340	30.9	18,584	59,079	77,663	23.0	2669	67,801	70,470	3.8
70–74	17,465	28,390	45,855	38.1	16,855	39,361	56,216	33.1	2863	43,005	45,868	6.2
75–79	13,590	17,820	31,410	43.3	14,019	25,076	39,095	43.1	3025	27,690	30,715	9.8
80–84	9332	10,020	19,352	48.2	9816	14,677	24,493	57.2	2987	16,437	19,424	15.4
>84	7070	7416	14,486	48.8	9092	11,924	21,016	46.7	4058	14,392	18,450	22.0
Sex
Female	5,0928	639,709	690,637	7.0	56,963	1,112,215	1,169,178	5.0	9538	1,669,679	1,679,217	0.6
Male	87,861	618,468	706,329	12.0	84,010	1,058,788	1,142,798	7.0	14,303	1,255,175	1,269,478	1.0
Smoking status				0.0				0.0				0.0
Smoking	11,035	93,776	104,811	11.0	9753	126,674	136,427	7.0	1817	118,192	120,009	2.0
Non-smoking	127,754	1,164,431	1,292,185	10.0	131,220	2,044,329	2,175,549	6.0	21,938	2,795,433	2,817,371	0.8
Hospitalisation
Non-hospitalised	11,916	1,097,953	1,109,869	1.1	5922	2,013,746	2,019,668	0.3	1903	2,865,234	2,867,137	0.1
Hospitalised	126,873	160,254	287,127	44.0	135,051	157,257	292,308	46.0	23,839	59,620	83,459	29.0
ICU	14,336	10,378	24,714	58.0	13,302	8829	22,131	60.0	1878	2493	4371	43.0
ETI	36,300	7058	43,358	84.0	25,557	4737	30294	84.0	3556	1231	4787	74.0
Pneumonia status
With pneumonia	102,157	116,828	218,985	47.0	100,646	112,602	213,248	47.0	14,725	38,899	53,624	27.0
Without pneumonia	36,632	1,141,379	1,178,011	3.0	40,327	2,058,401	2,098,728	2.0	9115	2,873,150	2,882,265	0.3
Comorbidity				0.0				0.0				0.0
Hypertension	63,582	190,117	253,699	25.0	60,532	226,258	286,790	21.0	11,251	245,935	257,186	4.0
Obesity	32,441	190,259	222,700	15.0	28,628	208,150	236,778	12.0	2785	190,749	193,534	1.0
Diabetes	53,089	143,606	196,695	27.0	49,738	169,299	219,037	23.0	9036	164,673	173,709	5.0
Cardiopathy	7374	16,814	24,188	30.0	6329	15,778	22,107	29.0	1749	18,303	20,052	9.0
COPD	6487	10,815	17,302	37.0	5478	10,780	16,258	34.0	1505	10,518	12,023	13.0
Asthma	2640	30,922	33,562	8.0	2326	37,691	40,017	6.0	363	50,909	51,272	0.7
Immunosuppressed	3267	9813	13,080	25.0	2809	9009	11,818	24.0	946	11,637	12,583	8.0
CKD	9781	13,225	23,006	43.0	9158	13,640	22,798	40.0	2998	14,779	17,777	17.0
Other complication	7209	22,377	29,586	24.0	6557	28,341	34,898	19.0	1815	34,149	35,964	5.0

CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; ETI, endotracheal intubation; ICU, intensive care unit.

Fatality ratio (FR) were estimated by trait.

Multivariate odds ratios of mortality were estimated for age, sex, smoking habits, hospitalisation admission characteristics and comorbidities (Table 2 ). The results revealed an age-dependent increase in the odds of mortality for individuals middle-aged and older during C2021 and C2022, where the likelihood of death was twice as high compared with C2020 (Table 2). With a significant and slight reduction over the 3 years, male patients had higher risk of death than female patients.

Table 2.

Multivariate odds ratios (±95% confidence interval) for risk factors of case fatality rate in patients positive to COVID-19 in Mexico during 2020, 2021 and 2022 in Mexico.

Risk factor	2020		2021		2022
	Multivariate odds ratioa	P-value	Multivariate odds ratiob	P-value	Multivariate odds ratioc	P-value
Age (years)
0–4	1.0		1.0		1.0
5–9	0.488 (0.354–0.672)	<0.0001	0.548 (0.403–0.746)	0.0001	0.602 (0.426–0.851)	0.0040
10–14	0.430 (0.325–0.570)	<0.0001	0.692 (0.541–0.884)	0.0033	0.892 (0.648–1.227)	0.4832
15–19	0.505 (0.411–0.621)	<0.0001	0.961 (0.797–1.159)	0.6780	1.419 (1.093–1.842)	0.0086
20–24	0.570 (0.486–0.670)	<0.0001	1.218 (1.039–1.428)	0.0151	1.304 (1.032–1.647)	0.0263
25–29	0.608 (0.525–0.704)	<0.0001	1.754 (1.513–2.035)	<0.0001	1.686 (1.375–2.067)	<0.0001
30–34	0.867 (0.754–0.998)	0.0472	2.345 (2.028–2.712)	<0.0001	2.208 (1.824–2.673)	<0.0001
35–39	1.127 (0.982–1.294)	0.0887	3.290 (2.850–3.797)	<0.0001	3.169 (2.637–3.808)	<0.0001
40–44	1.598 (1.395–1.831)	<0.0001	4.082 (3.539–4.708)	<0.0001	3.663 (3.062–4.383)	<0.0001
45–49	2.108 (1.842–2.413)	<0.0001	5.043 (4.376–5.811)	<0.0001	5.369 (4.529–6.365)	<0.0001
50–54	2.691 (2.352–3.078)	<0.0001	6.100 (5.294–7.029)	<0.0001	6.902 (5.851–8.141)	<0.0001
55–59	3.634 (3.178–4.157)	<0.0001	7.766 (6.741–8.946)	<0.0001	8.082 (6.869–9.509)	<0.0001
60–64	4.970 (4.345–5.684)	<0.0001	9.972 (8.657–11.488)	<0.0001	9.693 (8.248–11.392)	<0.0001
65–69	6.341 (5.543–7.254)	<0.0001	12.239 (10.623–14.102)	<0.0001	11.127 (9.479–13.062)	<0.0001
70–74	8.000 (6.990–9.156)	<0.0001	14.620 (12.684–16.851)	<0.0001	12.883 (10.975–15.123)	<0.0001
75–79	9.493 (8.287–10.875)	<0.0001	16.642 (14.429–19.194)	<0.0001	14.476 (12.334–16.991)	<0.0001
80–84	11.197 (9.757–12.850)	<0.0001	18.502 (16.018–21.370)	<0.0001	16.700 (14.220–19.611)	<0.0001
>84	12.684 (11.033–14.583)	<0.0001	21.204 (18.347–24.508)	<0.0001	19.081 (16.279–22.366)	<0.0001
Sex
Female	1.0		1.0		1.0
Male	1.479 (1.455–1.502)	<0.0001	1.346 (1.326–1.368)	<0.0001	1.386 (1.340–1.439)	<0.0001
Smoking status
Non-smoking	1.0		1.0		1.0
Smoking	0.889 (0.863–0.915)	<0.0001	0.853 (0.827–0.879)	<0.0001	0.945 (0.885–1.010)	0.0942
Hospitalisation
Non-hospitalised	1.0		1.0		1.0
Hospitalised	17.438 (17.042–17.843)	<0.0001	82.699 (80.305–85.165)	<0.0001	146.073 (136.574–156.232)	<0.0001
ICU	0.698 (0.674–0.723)	<0.0001	0.788 (0.760–0.816)	<0.0001	0.739 (0.681–0.802)	<0.0001
ETI	9.976 (9.384–9.976)	<0.0001	8.272 (7.974–8.580)	<0.0001	7.751 (7.145–8.409)	<0.0001
Pneumonia status
Without pneumonia	1.0		1.0		1.0
With pneumonia	3.128 (3.074–3.182)	<0.0001	2.237 (2.200–2.275)	<0.0001	2.488 (2.406–2.573)	<0.0001
Comorbidity
Not present	1.0		1.0
Hypertension	1.176 (1.156–1.197)	<0.0001	1.131 (1.111–1.151)	<0.0001	1.051 (1.011–1.092)	0.0126
Obesity	1.292 (1.269–1.317)	<0.0001	1.335 (1.308–1.362)	<0.0001	1.020 (0.967–1.076)	0.4722
Diabetes	1.234 (1.213–1.256)	<0.0001	1.131 (1.111–1.152)	<0.0001	1.090 (1.048–1.133)	<0.0001
Cardiopathy	0.959 (0.922–0.996)	0.0319	0.911 (0.874–0.949)	<0.0001	0.799 (0.748–0.854)	<0.0001
COPD	1.261 (1.150–1.383)	<0.0001	0.992 (0.948–1.037)	0.7172	0.908 (0.844–0.976)	0.0093
Asthma	1.096 (1.051–1.143)	<0.0009	0.917 (0.865–0.972)	0.0036	0.852 (0.747–0.971)	0.0160
Immunosuppressed	1.273 (1.206–1.345)	<0.0001	1.347 (1.267–1.431)	<0.0001	1.224 (1.120–1.337)	<0.0001
CKD	2.086 (2.011–2.163)	<0.0001	1.763 (1.669–1.831)	<0.0001	1.450 (1.371–1.533)	<0.0001
Other complication	1.317 (1.269–1.368)	<0.0001	1.242 (1.194–1.291)	<0.0001	1.404 (1.314–1.501)	<0.0001

COPD, chronic obstructive pulmonary disease; CRD, chronic kidney disease; ETI, endotracheal intubation; ICU, intensive care unit.

Odds for ICU and ETI estimated from hospitalised data.

^an = 1,387,777.

^bn = 2,289,037.

^cn = 2,918,646.

A significant risk was associated with smoking habits, where smokers showed slightly lower odds of mortality than non-smokers. In terms of comorbidities, a slight reduction from C2020 to partial 2022 year was observed; however, the presence of comorbidities significantly increased the odds of mortality in all years analysed, with the exception of COPD, which did not influence the odds during C2021, and obesity in C2022 (Table 2).

The most significant change observed in the years analysed was the substantial increased likelihood of death associated with hospitalisation – from 17.4 to 82.7 times higher among inpatients compared with non-hospitalised COVID-19-positive patients in C2020 and C2021, respectively. The odds of mortality further increased to 146.1 in C2022 in COVID-19 hospitalised patients, an eight-fold increase compared with C2020.

Discussion

In this study, three 1-year groups were analysed from a large data set of registered COVID-19-positive cases (N = 6,657,667) in Mexico from 13 January 2020 to 24 October 2022. The odds of death related to age, sex and comorbidities in COVID-19 patients from Mexico have been previously discussed.10, 11, 12 In general, most of the estimated odds for mortality remained similar among the year groups, and previous risk factors associated with COVID-19 were confirmed in the present analysis.¹⁰ For instance, comorbidities previously reported as important by their odds of mortality, such as hypertension, diabetes, obesity and CKD,¹⁰ maintained similar and significant odds in the year groups assessed in the present study, highlighting the relevance of continuous monitoring of vulnerable populations during vaccination campaigns. However, the comparative analysis of year groups revealed the following three main findings that are relevant to discuss: (1) a reduction of total CFR over the 3-year study period; (2) increasing odds of mortality in older ages during the second- and third-year group; and (3) the dramatic and sustained increase in odds of mortality in hospitalised patients.

Constant reduction in CFRs in the last three years

In Mexico, the results from the first epidemiological analyses of COVID-19 were uncertain because of the reported and unprecedented increasing CFR for some vulnerable population groups.¹⁰ The implementation of the national COVID-19 vaccination programme, initiated on 24 December 2020, was believed to be a relief to the constant increasing number of cases and mortality. The vaccination strategy was implemented in stages, with the aim to vaccinate the majority of the Mexican population by the end of 2021. The strategy prioritised, step-by-step, (1) front-line healthcare personnel, (2) the elderly population (≥60 years old), (3) those aged 50–59 years and pregnant women, (4) those aged 40–49 years and, finally (5) the rest of the population.¹⁵ However, some amendments were necessary to cover comorbidities in vulnerable individuals and school personnel. As the programme developed, it was evident the vaccination coverage was dependent on laborious government vaccine acquisition and reduced widespread vaccine availability that delayed vaccine uptake.¹⁴ By the end of 2021, approximately 63% of the Mexican population was vaccinated, with 56% fully vaccinated and 7.1% partially vaccinated (i.e. had only received one dose).⁹ Following the fourth peak of COVID-19 infections during the early months of 2022, it was estimated that around 66% of the Mexican population had been vaccinated.⁹

Hence, considering the emergence of the SARS-CoV-2 Delta variant and the triggering of the third wave of COVID-19 in May 2021,^{13 , 16} the current analysis strongly suggests that vaccination against COVID-19 in Mexico was an effective way to reduce CFR in 2021 and 2022. Similarly, a study in the United Kingdom, the first country that implemented a COVID-19 vaccination programme, reported that COVID-19 vaccination was effective against symptomatic disease with the Delta variant, especially after the application of two doses.¹⁷ Moreover, Liang et al.¹⁸ analysed data from 90 countries from November 2020 to April 2021 and estimated that a 10% increase in vaccine coverage resulted in a 7.6% reduction in COVID-19 CFR. The gradual decrease in CFR observed in the present study from 2021 to 2022 following the implementation of the national COVID-19 vaccination programme at the end of 2020 is in line with a report from China indicating that vaccination decreases the risk of developing severe COVID-19.¹⁹ The latter study also found that in an unvaccinated population, in general, more severe cases were seen with Delta infections than with Omicron infections,¹⁹ which would have increased levels of CFR.

Few studies have assessed the effects of COVID-19 vaccination in Mexico. A study with 312 health workers concluded that the BNT162b2 COVID-19 vaccine was 100% effective against severe illness; however, only 22 individuals were vaccinated in the trial,²⁰ so the results should be interpreted with caution. In addition, in a small study (n = 53), Galán-Huerta et al.¹⁴ found that in vaccinated individuals, mainly with the Cansino vaccine, patients with complete vaccination were less likely to develop severe COVID-19 disease requiring hospitalisation compared with those who received incomplete immunisation. Moreover, a preprint study comparing 793,487 vaccinated individuals with 4,792,338 unvaccinated individuals from December 2020 to September 2021 suggests that vaccination can decrease hospitalisation and death for adults (aged ≥18 years) in Mexico.²¹

Increasing odds of mortality in middle-aged or older individuals and for hospitalised patients during 2021–2022

The odds of COVID-19 mortality in patients who were hospitalised increased from 2020 to 2022. This result can be explained by a relatively low incidence of cases that required hospitalisation from the total cases observed during 2021 and 2022; nonetheless, a higher proportion of these cases had an adverse outcome (96% and 94% for 2021 and 2022, respectively), increasing the odds of mortality in hospitalised patients. Distribution of cases between age groups suggests that a greater number of the patients in these groups required hospitalisation. The results, as previously discussed, might also suggest that these inpatients were not vaccinated.

The years 2021 and 2022 were characterised by the emergence of new SARS-CoV-2 variants (i.e. Delta and Omicron)¹³ but also by the implementation of the national COVID-19 vaccination programme. As observed in the CFR, most of the positive cases that did not require hospitalisation during 2021 and 2022 had higher probabilities of survival than non-hospitalised cases in 2020. Several studies support the effectiveness of vaccination protocols in Mexico and other countries.^{9 ,} 16, 17, 18, 19, 20, 21, 22, 23, 24 For example, Mhawish et al.²⁴ indicated that most COVID-19 patients admitted to ICU in Saudi Arabia were non-immunised patients. Acuti Martelluci et al.²² analysed 313,068 unvaccinated and 966,626 vaccinated residents in Italy and found that patients receiving two or three vaccine doses showed 80–90% lower risk of COVID-19 hospitalisation or death compared with unvaccinated patients. Similarly, Muthukrishnan et al.²³ examined the effectiveness of the COVISHIELD vaccine (ChAdOx1 nCoV-19) in 1168 patients in India, showing that fully vaccinated patients who required hospitalisation had a higher likelihood of survival than unvaccinated inpatients. One of the limitations of the present study is that vaccination status of patients was not included in the analysis because of the unavailability of information in the data sets. Nevertheless, although a significant increase in the odds of mortality was observed for hospitalised cases during 2021 and 2022, the CFR decreased during those years, suggesting that hospitalisation could have been related to unvaccinated patients or those with incomplete immunisation. This theory is supported by evidence highlighting the reduction in hospitalisation of vaccinated COVID-19 patients.^{14 , 22 , 24}

Vaccine hesitancy is multifactorial, but a survey in Mexico indicated that middle-aged and older individuals were significantly more likely to refuse any COVID-19 vaccine, regardless of its effectiveness.²⁵ Similarly, another study in Mexico reported that a young adult population (18–34 years) was most likely to get vaccinated.²⁶ This could partially explain the increase in odds of mortality in middle-aged and older individuals observed in the present study following the implementation of the national COVID-19 vaccination campaign. Distrust in federal government recommendations also seems to play a significant role in COVID-19 vaccine refusal in the Mexican population,²⁵ but this is known to be a global problem.²⁷ Other factors for COVID-19 vaccine refusal in Mexico include a perception of adverse effects, conspiracy theories (e.g. the virus was created by the government) and anti-vaccine feedback from social media, friends and/or family.^{26 , 28 , 29} Indeed, misinformation from social media has played a pivotal role in the development of vaccine hesitancy.^{30 , 31}

Implications

The present investigation is an observational study based on retrospective analysis of a large data set (6,657,667 COVID-19-positive patients) of recorded information on COVID-19 cases. The findings suggest that the implementation of the COVID-19 vaccination programme had a noticeable positive effect, despite the significant increase in the number of positive cases during 2021 and 2022. The study also highlights the relevance for constant surveillance of COVID-19 vaccine effectiveness and the need to improve the information collected in free data sets provided by the government (e.g. background clinical information, vaccination status and adverse effects following vaccination). This information could provide confidence in the effectiveness of COVID-19 vaccination and improve vaccination acceptance.

An important limitation from previous reports and this study includes the lack of specific information available to determine clinical conditions; for example, disease severity or lack of specific information on comorbidity and clinical tests, which may reveal more evidence on the final cause of deaths. As revealed by Parra-Bracamonte et al.,¹⁰ hospitalised patients might have more accurate and accumulative data regarding comorbidities leading to bias in their relationship to death, evidenced by the accumulated proportion of specific clinical conditions (i.e. pneumonia). Furthermore, interactions between comorbidities may bias disease outcomes, as suggested for CKD, where patients showed an increasingly very high risk for death, also in association with major comorbidities (diabetes, hypertension and obesity).¹¹ In addition, the size of the country, regional diversity and other patient habits are factors that need to be considered when analysing observed variations so that better management of the course of infections and strategies to improve prognosis in patients can be implemented.¹⁰ This is particularly true in observed odds of mortality, where endotracheal intubation and non-hospitalised patients have a greater risk for death than those admitted to the ICU. This observation has been previously explained as a possible overburdening of the healthcare system capabilities, where patients are not receiving appropriate escalation of care because of limited resources;^{10 , 32} this, although not properly documented, was evident, especially during the peaks of pandemic.

Finally, the differences in CFR among the studied years could also be explained by changes in the characteristics of SARS-CoV-2 variants, particularly during C2022, where the emergence of the Omicron variant was characterised by a rapid spread in the human population, lower fusogenicity and attenuated pathogenicity.³³

Conclusions

The present COVID-19 analysis highlighted the lack of complementary information to fully understand and interpret the disease dynamic in Mexico, despite the availability of a large registered database. Regardless of this limitation, the results are based on a large data set and indirectly support the beneficial effects of implementing COVID-19 vaccination protocols, leading to a significant reduction of CFR from 10% during 2020 to less than 1% during 2022 in Mexico. However, an increase in the odds of mortality in hospitalised patients was seen during 2021 and 2022, which could be related to the unvaccinated proportion of the infected population following vaccination hesitancy. Some important comorbidities that were previously associated with increased odds of mortality became non-significant as the vaccination programme was implemented and a large proportion of the population were vaccinated.

Author statements

Acknowledgements

The authors thank the Secretariat of Health, Mexican Government, for free access to the database used in this study.

Ethical approval

None required.

Funding

None declared.

Competing interests

None declared.