The study aimed to compare the sociodemographic, comorbidity, and clinical variables associated with coronavirus disease 2019 (COVID-19) mortality across three distinct epidemic waves in Faridabad, India.
MethodsA retrospective analysis of the medical records of patients admitted with COVID-19 was conducted at a tertiary care center at Faridabad, India. COVID-19 epidemic waves were categorized into the first wave (April 2020–January 2021), second wave (March 2021–June 2021), and third wave (December 2021–February 2022). Sociodemographic, comorbidity, and clinical parameters were assessed for their association with mortality in each of the waves by the Chi-square test. The Cochran–Armitage test for trend was used to assess changes in these associations with respect to the mortality rate across the epidemic waves.
ResultsA total of 5217 patient records were assessed, with 4066 in the first wave, 895 in the second wave, and 256 in the third wave. Across all waves, comorbidities (diabetes and hypertension), multimorbidity, severe disease (requiring intensive care unit admission and ventilator support) were consistently associated (p<0.05) with higher mortality. While sociodemographic factors were significant (p<0.05) in the first two waves, their impact diminished in the third. Clinical symptoms, particularly ‘cold and flu’ showed consistent significance (p<0.05) across all waves. COVID-19 mortality trend peaked in the second wave, disproportionately (p<0.05) affecting females, older patients, and those with comorbidities or severe symptoms.
ConclusionsUnderstanding the shifting risk factors across COVID-19 epidemic waves is crucial for targeted interventions. Prioritizing high-risk groups, particularly during peak waves, can optimize resource allocation and minimize mortality.
El estudio tuvo como objetivo comparar las variables sociodemográficas, las comorbilidades y las variables clínicas asociadas con la mortalidad por COVID-19 a través de tres ondas epidémicas en Faridabad, India.
MétodosSe realizó un análisis retrospectivo de los registros médicos de pacientes con COVID-19 en un centro terciario en Faridabad, India. Las ondas epidémicas de COVID-19 fueron categorizadas en la primera ola (abril de 2020-enero de 2021), la segunda ola (marzo-junio de 2021) y la tercera ola (diciembre de 2021-febrero de 2022). Se evaluaron los parámetros sociodemográficos, los clínicos y las comorbilidades para su asociación con la mortalidad en cada ola mediante la prueba de chi-cuadrado. Se utilizó la prueba de Cochran-Armitage para analizar los cambios en estas asociaciones con respecto a la mortalidad a lo largo de las ondas epidémicas.
ResultadosSe evaluaron 5.217 registros de pacientes: 4.066 en la primera ola, 895 en la segunda y 256 en la tercera. Las comorbilidades (diabetes, hipertensión), la multimorbilidad y la enfermedad grave (requiriendo ingreso en la UCI y soporte ventilatorio) se asociaron consistentemente (p<0,05) con mayor mortalidad. Los factores sociodemográficos fueron significativos en las dos primeras ondas epidémicas, pero su impacto disminuyó en la tercera. Los síntomas clínicos, como «resfriado y gripe», mostraron consistencia significativa (p<0,05) en todas las ondas epidémicas. La mortalidad alcanzó su pico en la segunda ola, afectando principalmente a mujeres, pacientes mayores y aquellos con comorbilidades o síntomas graves.
ConclusionesEntender los factores de riesgo cambiantes durante las ondas epidémicas de COVID-19 es clave para intervenciones eficaces. Priorizar grupos de alto riesgo optimiza recursos y reduce la mortalidad.
Coronavirus disease 2019 (COVID-19), caused by SARS-CoV-2, led to a global pandemic that reached over 777 million cases by December 21, 2024.1 India faced a substantial burden, experiencing over 45 million infections and over half a million fatalities during this period.2
India experienced three distinct waves of COVID-19. The first wave (March 2020–January 2021), primarily driven by the original SARS-CoV-2 strain (B.1 lineage), presented with mild-to-moderate symptoms and relatively low mortality. The second wave (March 2021–June 2021), primarily driven by the Delta variant (B.1.617.2), was characterized by a surge in cases, rapid disease progression, and significantly higher mortality. The third wave (December 2021–April 2022), primarily driven by the Omicron variant (B.1.1.529), resulted in milder symptoms and significantly lower mortality.3
Existing research has identified a range of COVID-19 mortality risk factors, encompassing sociodemographic factors such as male sex, older age, smoking status, and obesity, as well as comorbidities like acute kidney injury, chronic obstructive pulmonary disease, diabetes, hypertension, cardiovascular disease, and cancer.4,5 However, these studies often analyze risk factors cumulatively, overlooking the distinct dynamics of individual pandemic waves.6 Furthermore, research specifically examining the Indian context and disaggregating risk factors by wave remains limited.7,8
This study addresses this critical gap by conducting a comparative analysis of mortality risk factors across the first, second, and third COVID-19 waves in India, using a population cohort from a single geographical region. By examining how the influence of these factors shifted between waves, this research aims to provide crucial insights for targeted public health interventions and inform future pandemic preparedness strategies.
MethodsStudy design and settingThis retrospective, cross-sectional study was conducted at a government-run tertiary care teaching hospital located in Faridabad, Haryana, India within the National Capital Region adjoining Delhi. Faridabad is an urban area. The hospital was designated as a dedicated COVID-19 treatment center by the government. By the end of the third wave, the hospital had established a COVID-19 infrastructure, including 24 ventilators, 60 ICU beds, and 510 isolation beds, and in-house testing facilities.
Ethics approvalThe study was conducted after approval of the Institutional Ethics Committee of the institute. This study was performed in lines with the Declaration of Helsinki.
Consent to participate declarationGiven the retrospective nature of this study and the use of anonymized, pre-existing medical records, a waiver of informed consent was granted.
Study populationMedical records of adult patients (≥18 years) admitted to the institute between April 1, 2020, and March 31, 2022, were retrospectively reviewed. Patients with confirmed COVID-19 based on World Health Organization (WHO) interim guidance and positive reverse transcriptase polymerase chain reaction (RT-PCR) nasopharyngeal swab testing were included. Treatment was administered according to national and state guidelines. Discharge was based on clinical improvement and a negative SARS-CoV-2 RT-PCR test.
Data extraction and study variablesData for this study were collected at the Medical Records Department of the institute. A validated, structured proforma was employed to extract information from the medical records of included patients. Incomplete records were excluded from the analysis. Collected variables included sociodemographic data (age, sex, religion, residential area, education, occupation, and total monthly family income) and clinical data (presenting signs and symptoms, comorbidities, physician-assessed disease severity, intensive care unit (ICU) admission, and mechanical ventilation requirement).
Operational definitionsFor this study, the COVID-19 waves in India were categorized as Wave 1 (1 April 2020–31 January 2021), Wave 2 (1 March 2021–30 June 2021), and Wave 3 (1 December 2021–28 February 2022). Patients in the intermediate periods between each wave were excluded to prevent overlap of the waves. Patients were classified as rural or urban based on their reported residence at the time of hospital admission, as defined by the Census of India 2011.9 Educational attainment was categorized into five levels: illiterate, elementary (≤grade 5), secondary (≤grade 10), high school (≤grade 12), and university (undergraduate or postgraduate). Six occupational categories were used: unemployed, unskilled worker, semi-skilled worker, farmer/shop owner, skilled worker, and professional. Hypertension and diabetes were defined as blood pressure≥140/90mmHg and HbA1c≥6.5% or fasting plasma glucose≥126mg/dL, respectively. Cold and flu referred to non-specific upper respiratory tract symptoms, such as cough, runny nose, and congestion, with or without fever. COVID-19 severity was classified as mild, moderate, or severe according to the guidelines established by the Ministry of Health and Family Welfare, Government of India.10
Statistical analysisDescriptive statistics for categorical variables are presented as frequencies and percentages, while continuous variables are presented as mean±standard deviation. First, the association between each potential risk factor and mortality was examined separately for each COVID-19 wave using Chi-square tests. Subsequently, the Cochran–Armitage test for trend was employed to determine whether the strength and direction of these associations with respect to the mortality rate changed significantly across successive waves. All tests were two-sided, and a p-value<0.05 was considered statistically significant. The data was analyzed using IBM SPSS Statistics for Windows, version 23.
ResultsCharacteristics of included patientsAmong the included 5217 patients, the majority were from Wave 1 (n=4066, 77.9%), followed by Wave 2 (n=895, 17.2%) and Wave 3 (n=256, 4.9%). The mean age was 51.54±14.69 years, with the majority aged≤60 years (79.9%), male (65.3%), Hindu (54.0%), urban residents (74.9%), educated (92.4%), employed (86.1%), and earning≥
20,000 (55.1%). Comorbidities were present in 70.0% of patients, with diabetes (42.8%) and hypertension (42.0%) being the most common. Clinical manifestations included cold and flu (81.2%), sore throat (69.7%), and breathlessness (60.1%). Of the patients, 20.8% had severe infection, 54.0% required ICU admission, 26.9% required ventilator support. The mean hospital stay was 9.75±8.46 days. Of these patients, 87.8% (n=4578) survived and 12.2% (n=639) did not (Table 1).
Sociodemographic, comorbidity, and clinical characteristics of the included study population.
| Variables | Mean±SD or n (%) (n=5217) |
|---|---|
| Sociodemographic variables | |
| Age (years) | Mean=51.54±14.69 |
| Age distribution (years) n (%) | ≤60: 3855 (79.9) • >60: 1362 (26.1) |
| Gender n (%) | Female: 1809 (34.7) • Male: 3408 (65.3) |
| Religion n (%) | Hinduism: 2815 (54.0) • Others: 2402 (46.0) |
| Area of residence n (%) | Rural: 1308 (25.1) • Urban: 3909 (74.9) |
| Education level n (%) | Illiterate: 397 (7.6) • Educated: 4820 (92.4) |
| Occupation n (%) | Unemployed: 724 (13.9) • Employed: 4493 (86.1) |
| TFMI n (%) | <20,000: 2343 (44.9) • ≥20,000: 2874 (55.1) |
| Comorbidity variables | |
| Comorbidities n (%) | Yes: 3653 (70.0) • No: 1564 (30.0) |
| Number of comorbidities n (%) | 0: 1564 (30.0)1: 1444 (27.7)2: 1078 (20.7)3: 599 (11.5)≥4: 532 (10.2) |
| Diabetes n (%) | Yes: 2231 (42.8) • No: 2986 (57.2) |
| Hypertension n (%) | Yes: 2192 (42.0) • No: 3025 (58.0) |
| Clinical variables | |
| Cold and flu n (%) | Yes: 4238 (81.2) • No: 979 (18.8) |
| Sore throat n (%) | Yes: 3637 (69.7) • No: 1580 (30.3) |
| Diarrhea n (%) | Yes: 1578 (30.2) • No: 3639 (69.8) |
| Pain in abdomen n (%) | Yes: 2689 (51.5) • No: 2528 (48.5) |
| Loss of smell n (%) | Yes: 2987 (57.3) • No: 2230 (42.7) |
| Loss of taste n (%) | Yes: 1823 (34.9) • No: 3394 (65.1) |
| Breathlessness n (%) | Yes: 3133 (60.1) • No: 2084 (39.9) |
| Disease severity n (%) | Mild: 2084 (39.9)Moderate: 2048 (39.3)Severe: 1085 (20.8) |
| ICU admission n (%) | Yes: 2819 (54.0) • No: 2398 (46.0) |
| Ventilator support n (%) | Yes: 1405 (26.9) • No: 3812 (73.1) |
| Duration of admission (days) | Mean=9.75±8.46 |
| Survived n (%) | Yes: 4578 (87.8) • No: 639 (12.2) |
TFMI, total family monthly income; ICU, intensive care unit.
Of the 4066 patients included in the first wave, 3735 survived and 331 died. Sociodemographic factors associated with mortality included age (p<0.001), religion (p=0.005), area of residence (p<0.001), education level (p=0.004), and total family monthly income (p=0.002) (Table 2, Figs. 1, 2A, and 2B). The presence and number of comorbidities were also significantly linked to mortality (p<0.001 for both). Specifically, diabetes (p<0.001) and hypertension (p=0.026) were identified as significant individual comorbidities (Table 3, Figs. 1 and 2B). Among the clinical symptoms examined, cold and flu (p=0.002) were significantly associated with increased mortality. Disease severity (p<0.001), ICU admission (p<0.001), and ventilator support (p=0.023) were also significantly linked to increased mortality (Table 4, Figs. 1 and 2B).
Association of sociodemographic variables with mortality among COVID-19 patients across three waves.
| COVID-19 Wave→ | Wave-1 | Wave 2 | Wave 3 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable↓ | Survived (n=3735) | Dead (n=331) | Total (n=4066) | p value | Survived (n=641) | Dead (n=254) | Total (n=895) | p value | Survived (n=202) | Dead (n=54) | Total (n=256) | p value |
| Sex n (%) | ||||||||||||
| Male | 2453 (91.4) | 230 (8.6) | 2683 (100) | 0.161 | 410 (75.8) | 131(24.2) | 541 (100) | 0.001* | 145 (78.8) | 39 (21.2) | 184 (100) | 0.949 |
| Female | 1282 (92.7) | 101 (7.3) | 1383 (100) | 231 (65.3) | 123 (34.7) | 354 (100) | 57 (79.2) | 15 (20.8) | 72 (100) | |||
| Age distribution, years n (%) | ||||||||||||
| ≤20 | 49 (94.2) | 3 (5.8) | 52 (100) | 0.001* | 6 (66.7) | 3 (33.3) | 9 (100) | 0.014* | 6 (75.0) | 2 (25.0) | 8 (100) | 0.532 |
| 21–30 | 230 (94.3) | 14 (5.7) | 244 (100) | 31 (58.5) | 22 (41.5) | 53 (100) | 10 (76.9) | 3 (23.1) | 13 (100) | |||
| 31–40 | 615 (91.1) | 60 (8.9) | 675 (100) | 95 (76.0) | 30 (24.0) | 125 (100) | 31 (79.5) | 8 (20.5) | 39 (100) | |||
| 41–50 | 948 (92.5) | 77 (7.5) | 1025 (100) | 169 (77.9) | 48 (22.1) | 217 (100) | 52 (82.5) | 11 (17.5) | 63 (100) | |||
| 51–60 | 950 (92.5) | 77 (7.5) | 1027 (100) | 170 (71.1) | 69 (28.9) | 239 (100) | 54 (81.8) | 12 (18.2) | 66 (100) | |||
| 61–70 | 591 (93.2) | 43 (6.8) | 634 (100) | 106 (77.1) | 42 (28.4) | 148 (100) | 34 (77.3) | 10 (22.7) | 44 (100) | |||
| 71–80 | 246 (90.4) | 26 (9.6) | 272 (100) | 44 (66.7) | 22 (33.3) | 66 (100) | 10 (76.9) | 3 (23.1) | 13 (100) | |||
| >80 | 106 (77.4) | 31 (22.6) | 137 (100) | 20 (52.6) | 18 (47.4) | 38 (100) | 5 (50.0) | 5 (50.0) | 10 (100) | |||
| Religion n (%) | ||||||||||||
| Hinduism | 2043 (93.2) | 150 (6.8) | 2193 (100) | 0.005* | 343 (76.4) | 106 (23.6) | 449 (100) | 0.001* | 137 (79.2) | 36 (20.8) | 173 (100) | 0.703 |
| Islam | 774 (89.6) | 86 (10.4) | 830 (100) | 119 (68.0) | 56 (32.0) | 175 (100) | 48 (81.4) | 11 (18.6) | 59 (100) | |||
| Christianity | 413 (90.0) | 46 (10.0) | 459 (100) | 57 (54.3) | 48 (45.7) | 105 (100) | 8 (66.7) | 4 (33.3) | 12 (100) | |||
| Others | 535 (91.6) | 49 (8.4) | 584 (100) | 122 (73.5) | 44 (26.5) | 166 (100) | 9 (75.0) | 3 (25.0) | 12 (100) | |||
| Area n (%) | ||||||||||||
| Rural | 999 (89.3) | 120 (10.7) | 1119 (100) | 0.001* | 109 (79.6) | 28 (20.4) | 137 (100) | 0.025* | 47 (90.4) | 5 (9.6) | 52 (100) | 0.079 |
| Urban | 2736 (92.8) | 211 (7.2) | 2947 (100) | 532 (70.2) | 226 (29.8) | 758 (100) | 163 (79.9) | 41 (20.1) | 204 (100) | |||
| Education n (%) | ||||||||||||
| Illiterate | 261 (92.6) | 21 (7.4) | 282 (100) | 0.004* | 83 (78.3) | 23 (21.7) | 106 (100) | 0.004* | 7 (77.8) | 2 (22.2) | 9 (100) | 0.874 |
| Elementary school | 415 (92.0) | 36 (8.0) | 451 (100) | 65 (74.7) | 22 (25.3) | 87 (100) | 5 (83.3) | 1 (16.7) | 6 (100) | |||
| Secondary school | 549 (89.1) | 67 (10.9) | 616 (100) | 123 (75.9) | 39 (24.1) | 162 (100) | 92 (80.0) | 23 (20.0) | 115 (100) | |||
| High school | 499 (95.4) | 24 (4.6) | 523 (100) | 81 (78.6) | 22 (21.4) | 103 (100) | 39 (73.6) | 14 (26.4) | 53 (100) | |||
| University | 2011 (91.7) | 183 (8.3) | 2194 (100) | 289 (66.1) | 148 (33.9) | 437 (100) | 59 (80.8) | 14 (19.2) | 73 (100) | |||
| Occupation n (%) | ||||||||||||
| Unemployed | 488 (93.7) | 33 (6.3) | 521 (100) | 0.154 | 91 (65.9) | 47 (34.1) | 138 (100) | 0.010* | 52 (80.0) | 13 (20.0) | 65 (100) | 0.337 |
| Unskilled worker | 906 (92.4) | 75 (7.6) | 981 (100) | 101 (70.1) | 43 (29.9) | 144 (100) | 33 (82.5) | 7 (17.5) | 40 (100) | |||
| Semi-skilled worker | 551 (90.8) | 56 (9.2) | 607 (100) | 100 (66.7) | 50 (33.3) | 150 (100) | 19 (63.3) | 11 (36.7) | 30 (100) | |||
| Farmer/shop owner | 703 (90.7) | 72 (9.3) | 775 (100) | 154 (80.6) | 37 (19.4) | 191 (100) | 42 (84.0) | 8 (16.0) | 50 (100) | |||
| Skilled worker | 572 (93.3) | 41 (6.7) | 613 (100) | 108 (70.6) | 45 (29.4) | 153 (100) | 44 (80.0) | 11 (20.0) | 55 (100) | |||
| Professional | 515 (90.5) | 54 (9.5) | 569 (100) | 87 (73.1) | 32 (26.9) | 119 (100) | 12 (84.0) | 8 (16.0) | 50 (100) | |||
| TFMI (INR) n (%) | ||||||||||||
| <20,000 | 1736 (93.1) | 129 (6.9) | 1865 (100) | 0.002* | 274 (74.7) | 93 (25.3) | 367 (100) | 0.032* | 90 (81.1) | 21 (18.9) | 111 (100) | 0.457 |
| 20,001–30,000 | 851 (92.5) | 69 (7.5) | 920 (100) | 139 (69.8) | 60 (30.2) | 199 (100) | 45 (83.3) | 9 (16.7) | 54 (100) | |||
| 30,001–40,000 | 387 (91.9) | 34 (8.1) | 421 (100) | 64 (76.2) | 20 (23.8) | 84 (100) | 24 (80.0) | 6 (20.0) | 30 (100) | |||
| 40001–50000 | 355 (88.1) | 48 (11.9) | 403 (100) | 61 (76.3) | 19 (23.8) | 80 (100) | 30 (71.4) | 12 (28.6) | 42 (100) | |||
| >50,000 | 406 (88.8) | 51 (11.2) | 457 (100) | 103 (62.4) | 62 (37.6) | 165 (100) | 13 (68.4) | 6 (31.6) | 19 (100) | |||
TFMI, total family monthly income; Chi-square test was used.
Distribution of sociodemographic, comorbidity, and clinical variables among survived and dead patients across the first, second, and third COVID-19 waves. Bars show proportions of patients in each outcome group (survived vs. dead) for each wave, with stacked segments representing different variables, color-coded as per the legend.
(A) Venn diagram showing the overlap of significant risk factors associated with mortality across the first, second, and third COVID-19 waves. The numbers (2, 6, 7) represent the count of factors that are unique to or shared among the three waves. (B) Heatmap showing the strength of associations between risk factors (rows) and mortality in each wave (columns). Darker blue indicates lower p-values (stronger significance), while darker red indicates higher p-values (weaker significance). (C) Heatmap showing mortality rate across various parameters and their subgroups (rows) across different COVID-19 waves (columns). Cells represent mortality rates, with red indicating high rates and yellow indicating low rates.
Association of comorbidity variables with mortality among COVID-19 patients across three waves.
| COVID-19 Wave→ | Wave-1 | Wave 2 | Wave 3 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable↓ | Survived (n=3735) | Dead (n=331) | Total (n=4066) | p value | Survived (n=641) | Dead (n=254) | Total (n=895) | p value | Survived (n=202) | Dead (n=54) | Total (n=256) | p value |
| Comorbidities n (%) | ||||||||||||
| Yes | 2603 (90.7) | 266 (9.3) | 2869 (100) | 0.001* | 437 (75.9) | 139 (24.1) | 576 (100) | 0.001* | 162 (77.9) | 46 (22.1) | 208 (100) | 0.404 |
| No | 1132 (94.6) | 65 (5.4) | 1197 (100) | 204 (63.9) | 115 (36.1) | 319 (100) | 40 (83.3) | 8 (16.7) | 48 (100) | |||
| Number of comorbidities n (%) | ||||||||||||
| No comorbidity | 1132 (94.6) | 65 (5.4) | 1197 (100) | 0.001* | 204 (63.9) | 115 (36.1) | 319 (100) | 0.002* | 40 (83.3) | 8 (16.7) | 48 (100) | 0.001* |
| 1 comorbidity | 1001 (92.3) | 83 (7.7) | 1084 (100) | 199 (79.0) | 53 (21.0) | 252 (100) | 97 (89.8) | 11 (10.2) | 108 (100) | |||
| 2 comorbidities | 838 (89.4) | 99 (10.6) | 937 (100) | 88 (74.6) | 30 (25.4) | 118 (100) | 14 (60.9) | 9 (39.1) | 23 (100) | |||
| 3 comorbidities | 415 (91.4) | 39 (8.6) | 454 (100) | 74 (69.8) | 32 (30.2) | 106 (100) | 27 (69.2) | 12 (30.8) | 39 (100) | |||
| ≥4 comorbidities | 349 (88.6) | 45 (11.4) | 394 (100) | 76 (76.0) | 24 (24.0) | 100 (100) | 24 (63.2) | 14 (36.8) | 38 (100) | |||
| Diabetes n (%) | ||||||||||||
| Yes | 1530 (89.6) | 178 (10.4) | 1708 (100) | <0.001* | 277 (76.9) | 83 (23.1) | 360 (100) | 0.004* | 112 (68.7) | 51 (31.3) | 163 (100) | <0.001* |
| No | 2205 (93.5) | 153 (6.5) | 2358 (100) | 364 (68.0) | 171 (32.0) | 535 (100) | 90 (96.8) | 3 (3.2) | 93 (100) | |||
| Hypertension n (%) | ||||||||||||
| Yes | 1504 (90.7) | 154 (9.3) | 1658 (100) | 0.026* | 283 (73.5) | 91 (24.3) | 374 (100) | 0.023* | 111 (69.4) | 49 (30.6) | 160 (100) | <0.001* |
| No | 2231 (92.6) | 177 (7.4) | 2408 (100) | 358 (68.7) | 163 (31.3) | 521 (100) | 91 (94.8) | 5 (5.2) | 96 (100) | |||
Chi-square test was used.
Association of clinical variables with mortality among COVID-19 patients across three waves.
| COVID-19 Wave→ | Wave-1 | Wave 2 | Wave 3 | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable↓ | Survived (n=3735) | Dead (n=331) | Total (n=4066) | p value | Survived (n=641) | Dead (n=254) | Total (n=895) | p value | Survived (n=202) | Dead (n=54) | Total (n=256) | p value |
| Cold and flu n (%) | ||||||||||||
| Yes | 2981 (91.2) | 287 (8.8) | 3268 (100) | 0.002* | 522 (69.9) | 225 (30.1) | 747 (100) | 0.009* | 181 (81.2) | 42 (18.8) | 223 (100) | 0.021* |
| No | 754 (94.5) | 44 (5.5) | 798 (100) | 119 (80.4) | 29 (19.6) | 148 (100) | 21 (63.6) | 12 (36.4) | 33 (100) | |||
| Sore throat n (%) | ||||||||||||
| Yes | 2619 (91.8) | 234 (8.2) | 2853 (100) | 0.827 | 410 (68.8) | 186 (31.2) | 596 (100) | 0.008* | 146 (77.7) | 42 (22.3) | 188 (100) | 0.416 |
| No | 1116 (92.0) | 97 (8.0) | 1213 (100) | 231 (77.3) | 68 (22.7) | 299 (100) | 56 (82.4) | 12 (17.6) | 68 (100) | |||
| Diarrhea n (%) | ||||||||||||
| Yes | 1115 (91.8) | 99 (8.2) | 1214 (100) | 0.983 | 223 (76.6) | 68 (23.4) | 291 (100) | 0.021* | 51 (69.9) | 22 (30.1) | 73 (100) | 0.025* |
| No | 2620 (91.9) | 232 (8.1) | 2852 (100) | 418 (69.2) | 186 (30.8) | 604 (100) | 151 (82.5) | 32 (17.5) | 183 (100) | |||
| Pain in abdomen n (%) | ||||||||||||
| Yes | 1973 (91.6) | 180 (8.4) | 2153 (100) | 0.587 | 272 (66.8) | 135 (33.2) | 407 (100) | 0.004* | 103 (79.8) | 26 (20.2) | 129 (100) | 0.711 |
| No | 1763 (92.1) | 150 (7.9) | 1913 (100) | 369 (75.6) | 119 (24.4) | 488 (100) | 99 (78.0) | 28 (22.0) | 127 (100) | |||
| Loss of smell n (%) | ||||||||||||
| Yes | 2171 (91.6) | 200 (8.4) | 2371 (100) | 0.417 | 312 (68.0) | 147 (32.0) | 459 (100) | 0.013* | 126 (80.3) | 31 (19.7) | 157 (100) | 0.505 |
| No | 1564 (92.3) | 131 (7.7) | 1695 (100) | 329 (75.5) | 107 (24.5) | 436 (100) | 76 (76.8) | 23 (23.2) | 99 (100) | |||
| Loss of taste n (%) | ||||||||||||
| Yes | 1313 (91.6) | 120 (8.4) | 1433 (100) | 0.688 | 190 (65.7) | 99 (34.3) | 289 (100) | 0.007* | 70 (69.3) | 31 (30.7) | 101 (100) | 0.002* |
| No | 2422 (92.0) | 211 (8.0) | 2633 (100) | 451 (74.4) | 155 (25.6) | 606 (100) | 132 (85.2) | 23 (14.8) | 155 (100) | |||
| Breathlessness n (%) | ||||||||||||
| Yes | 2300 (91.5) | 215 (8.5) | 2515 (100) | 0.226 | 333 (67.8) | 158 (32.2) | 491 (100) | 0.005* | 92 (72.4) | 35 (27.6) | 127 (100) | 0.012* |
| No | 1435 (92.5) | 116 (7.5) | 1551 (100) | 308 (76.2) | 96 (23.8) | 404 (100) | 110 (85.3) | 19 (14.7) | 129 (100) | |||
| Disease severity n (%) | ||||||||||||
| Mild | 1592 (94.3) | 96 (5.7) | 1688 (100) | 0.001* | 194 (66.2) | 99 (33.8) | 293 (100) | 0.007* | 75 (72.8) | 28 (27.2) | 103 (100) | 0.008* |
| Moderate | 1435 (90.8) | 145 (9.2) | 1580 (100) | 256 (71.3) | 103 (28.7) | 359 (100) | 96 (88.1) | 13 (11.9) | 109 (100) | |||
| Severe | 708 (88.7) | 90 (11.3) | 1022 (100) | 191 (78.6) | 52 (21.4) | 243 (100) | 31 (70.5) | 13 (29.5) | 44 (100) | |||
| ICU admission n (%) | ||||||||||||
| Yes | 1870 (90.2) | 204 (9.8) | 2074 (100) | 0.001* | 417 (74.6) | 142 (25.4) | 559 (100) | 0.011* | 139 (74.7) | 47 (25.3) | 186 (100) | 0.008* |
| No | 1865 (93.6) | 127 (6.4) | 1992 (100) | 224 (66.7) | 112 (33.3) | 336 (100) | 63 (90.0) | 7 (10.0) | 70 (100) | |||
| Ventilator support n (%) | ||||||||||||
| Yes | 803 (90.0) | 89 (10.0) | 892 (100) | 0.023* | 263 (76.2) | 82 (23.8) | 345 (100) | 0.015* | 126 (75.0) | 42 (25.0) | 168 (100) | 0.034* |
| No | 2932 (92.4) | 242 (7.6) | 3174 (100) | 378 (68.7) | 172 (31.3) | 550 (100) | 76 (86.4) | 12 (13.6) | 88 (100) | |||
ICU, intensive care unit; Chi-square test was used.
In the second wave, of the included 895 patients, 641 survived and 254 died. Male sex was associated with a higher mortality rate (p<0.001). Age (p=0.014), religion (p<0.001), area of residence (p=0.025), education level (p=0.004), occupation (p=0.010), and total family monthly income (p=0.032) were all significantly associated with mortality (Table 2, Fig. 1). Similar to the first wave, the presence and number of comorbidities were significantly linked to mortality (p<0.001 and p=0.002, respectively). Diabetes (p=0.004) and hypertension (p=0.023) continued to be significant individual comorbidities (Table 3, Figs. 1, 2A and 2B). Clinical symptoms showing significant associations included cold and flu (p=0.009), sore throat (p=0.008), diarrhea (p=0.021), abdominal pain (p=0.004), loss of smell (p=0.013) and taste (p=0.007), and breathlessness (p=0.005). Disease severity (p=0.007), ICU admission (p=0.011), and ventilator support (p=0.015) remained significantly associated with increased mortality (Table 4, Figs. 1, 2A and 2B).
Third wave associationsIn the third wave, of the 256 patients included, 202 survived, and 54 died. None of the socio-demographic parameters evaluated showed significant associations with mortality (Table 2, Figs. 1, 2A and 2B). While the presence of any comorbidity was not a significant factor (p=0.404), the number of comorbidities was significantly associated with an increased risk of death (p=0.001). Specifically, having diabetes (p<0.001) or hypertension (p<0.001) significantly increased the risk of death (Table 3, Figs. 1 and 2B). Clinical symptoms significantly associated with mortality included cold and flu (p=0.021), diarrhea (p=0.025), loss of taste (p=0.002), and breathlessness (p=0.012). Disease severity (p=0.008), ICU admission (p=0.008), and the need for ventilator support (p=0.034) were also significantly associated with a higher risk of death (Table 4, Figs. 1 and 2B).
Cochran–Armitage test for trendAll the sociodemographic, comorbidity, and clinical parameters assessed by the Cochran–Armitage test for trend in mortality rates showed statistical significance (p<0.05) across the three waves (Fig. 2C). Mortality rates were highest in the second wave, especially among females (34.7%) the older age groups (>80 years, 47.4%), patients from urban areas (29.8%), university-educated patients (33.9%) and unemployed patients (34.1%). Mortality then declined in the third wave, particularly in females (20.8%), younger patients (≤20 years, 25%), patients from rural areas (9.6%) and those with higher education (university, 19.2%), and income levels (>50,000 INR, 31.6%). Patients with comorbidities or multimorbidity had the highest mortality in the second and third waves. In the third wave, the mortality declined, especially in patients with no or fewer comorbidities (no comorbidity, 16.7%; no diabetes, 3.2%; no hypertension 5.2%). Based on the clinical symptoms, the mortality increased from the first wave to the second wave, particularly in patients with cold and flu (30.1%), sore throat (31.2%), pain in abdomen (33.2%), and breathlessness (32.2%). Also, patients with severe disease needing ICU admission (25.4%) and ventilator support (23.8%) showed a significant (p<0.05) trend toward a higher mortality in the second wave. In the third wave, the mortality declined, particularly in patients who did not require ICU admission (8.6%) or ventilator support (13.6%) (Fig. 2C).
DiscussionIn this study, we assessed the mortality risk factors across the first, second, and third COVID-19 waves in Faridabad, India, examining sociodemographic, comorbidity, and clinical factors associated with mortality. The findings revealed that multiple factors exhibited associations with mortality, with varying patterns across the waves.
Regarding sociodemographic factors we observed that, in the first wave, mortality was primarily associated with male sex, older age, rural residence, higher education (university), and professional occupation. In the second wave, higher mortality was associated with female sex, increased age, urban residence, higher education, and higher income groups. Strikingly, in the third wave, no sociodemographic variables, were associated with increased mortality, suggesting a potential equalization of the risk across different demographic groups. These evolving patterns underscore the importance of considering the specific context of each wave. Our findings were largely consistent with previous research evaluating risk factors in the first,11 second,12 and third waves.13 Further investigation is needed to fully understand the underlaying reasons for these shifts.
Comorbidities, particularly diabetes and hypertension, and multimorbidity played a consistent role in mortality across all three waves. In both the first and second waves, patients with comorbidities had markedly higher mortality, consistent with previous reports.11,12,14,15 In the third wave, while the presence of any comorbidity did not significantly affect mortality, multimorbidity, diabetes, and hypertension remained significant predictors of death, aligning with previous reports.13,15 Notably, even in the third wave characterized by Omicron's lower virulence, multimorbidity remained a significant predictor of death, underscoring its cumulative burden in exacerbating the impact of COVID-19.3,15
The clinical symptoms of COVID-19 showed varying associations with mortality across all the three waves. In all three waves, the symptoms of cold and flu were significantly linked with higher mortality. Additionally, symptoms such as sore throat, diarrhea, abdominal pain, loss of smell and taste, and breathlessness showed significant associations with mortality in the second and/or third waves. Disease severity, including ICU admission and ventilator support, were also strong predictors of mortality throughout all three waves. These findings are consistent with the clinical course of COVID-19, where respiratory failure and severe disease results in worse outcomes.6 The consistent associations of these clinical markers across multiple waves emphasises their utility in identifying high-risk patients and guiding clinical management strategies.16,17 Consistent with global reports, mortality was much lower in the third wave.3,15
The Cochran–Armitage test for trend demonstrated significant changes in mortality rates across the three COVID-19 waves in India, reflecting the dynamic nature of the pandemic. The first wave in India saw a lower incidence and mortality, likely due to early lockdowns, limited transmission, and relatively lower virulence of the strain. Limited testing capacity also masked the true extent of the incidence and mortality.18 Consistent with global reports,19,20 the second wave, driven by the Delta variant, witnessed the highest mortality rates across most subgroups. This aligns with the higher virulence of the Delta variant, which was associated with faster disease progression and increased complications, coupled with healthcare system strain.6,20,21 Mortality rates subsequently declined in the third wave, particularly among patients with fewer comorbidities, and those not requiring ICU admission and ventilator support. The lower mortality in the third wave was likely due to a combination of factors, including reduced virulence of the Omicron variant, increased population immunity from prior infections and vaccinations, and improved management strategies.22
India's experience with three distinct COVID-19 waves reveals a dynamic interplay of policy decisions, viral strain evolution, and vaccination efforts. During the first wave, India adopted strict mitigation strategies, including a nationwide lockdown, travel bans, and quarantine measures.23 This zero-COVID approach delayed the first wave's peak. During the second wave, the approach shifted toward relaxation of travel and social restrictions, localized lockdowns, increased testing, and a ramped-up vaccination drive. This approach continued during the third wave as well.24 India's COVID-19 vaccination campaign was launched in January 2021 i.e. between the first and second wave. Initially, vulnerable groups and frontline workers were prioritized. By the time of the third wave, a significant portion of the population had received at least one dose of the vaccine.25 Data reveals a strong correlation between increased vaccination rates and decreased hospitalization and mortality.24–26 Further research is needed to disentangle the precise impact of mitigation policies and vaccination on COVID-19 mortality.26
The study's strengths include its large sample size across three COVID-19 waves, diverse variables offering a detailed mortality perspective. The focus on a specific region within India, allows for a nuanced understanding of the pandemic's impact on a particular population, providing valuable insights for targeted health interventions.
There are some limitations to this study. First, the data was collected from one healthcare center catering to a specific region of India, potentially limiting the generalizability of the findings. Second, the causal relationships between different variables and mortality cannot be established due to the study's observational design. Third, unmeasured confounding factors may have influenced some of the observations. Fourth, the varying durations of each of the waves (6 months for the first, 4 months for the second, and 3 months for the third), combined with exclusion of patients between waves, may have skewed the observed distribution of patient admissions across the pandemic period. Specifically, the longer first wave duration has likely inflated patient counts compared to the shorter second and third waves, as observed in an earlier study.8 Additionally, limited hospital capacity, including only 24 ventilators, restricted admissions despite high community caseloads during the second wave, particularly among patients seeking ICU-level care with ventilator support. This constraint may have potentially led to a selection bias. Changing admission practices (admitting mild cases in the first wave versus selective admissions during the overwhelmed second wave) further complicate comparisons.27 Finally, the lack of multivariate analysis limits our ability to determine independent risk factors and fully adjust potential confounders. These limitations should be considered when interpreting the study's mortality comparisons across the three COVID-19 waves.
Future studies should investigate causal relationships using multivariate analysis, assess intervention impacts on subgroups, incorporate detailed clinical data, and expand research to diverse populations.
ConclusionsThis comparative analysis across three COVID-19 waves in Faridabad, India highlights key risk factors associated with mortality, emphasizing the evolving dynamics of the pandemic. While disease severity and comorbidities consistently posed significant risks, the influence of specific factors varied across waves, underscoring the need for adaptive public health strategies. Further research exploring causal relationships and evaluating targeted interventions is crucial to inform future pandemic preparedness and response.
Authors’ contributionPrasad J and Vig SL contributed to the concept and design of the study and supervised the project. Parashar L was the PhD Research Scholar responsible for collection of the data and doing the statistical analysis. Meshram GG drafted the manuscript and prepared the graphs and tables. All authors approved the final version of the manuscript.
Informed consentWritten informed consent was waived because of the secondary use of medical data.
Ethics approvalThe study was conducted after approval of the Institutional Ethics Committee of ESIC-MCH (Date: 9-4-2021/F.No.134/A/11/16/Academic/MC/2016/196). This study was performed in lines with the Declaration of Helsinki.
FundingNo funding was received for conducting this study.
Conflict of interestsThe authors have no relevant financial or non-financial interests to disclose.
We gratefully acknowledge Dr. Asim Das, Dean of ESIC-MCH, Faridabad, during the COVID-19 waves, for his support and guidance in facilitating this research.
