Abstract
Background
Despite the variability and complexity of geriatric conditions, few COVID-19 reports of clinical characteristic prognostication provide data specific to oldest-old adults (over age 85), and instead generally report broadly as 65 and older.
Objective
To examine metabolic syndrome criteria in adults across 25 hospitals with variation in chronological age.
Design and Participants
This cohort study examined 39,564 hospitalizations of patients aged 18 or older with COVID-19 who received inpatient care between March 13, 2020, and February 28, 2022.
Exposure
ICU admission and/or in-hospital mortality.
Main Measures
Metabolic syndrome criteria and patient demographics were examined as risk factors. The main outcomes were admission to ICU and hospital mortality.
Key Results
Oldest old patients (≥ 85 years) hospitalized with COVID-19 accounted for 7.0% (2758/39,564) of all adult hospitalizations. They had shorter ICU length of stay, similar overall hospitalization duration, and higher rates of discharge destinations providing healthcare services (i.e., home health, skilled nursing facility) compared to independent care. Chronic conditions varied by age group, with lower proportions of diabetes and uncontrolled diabetes in the oldest-old cohort compared with young-old (65–74 years) and middle-old (75–84 years) groups. Evaluations of the effect of metabolic syndrome and patient demographics (i.e., age, sex, race) on ICU admission demonstrate minimal change in the magnitude of effect for metabolic syndrome on ICU admission across the different models.
Conclusions
Metabolic syndrome measures are important individual predictors of COVID-19 outcomes. Building on prior examinations that metabolic syndrome is associated with death and ARDS across all ages, this analysis supports that metabolic syndrome criteria may be more relevant than chronological age as risk factors for poor outcomes attributed to COVID-19.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11606-024-08744-4.
KEY WORDS: acute care, respiratory diseases, COVID-19 pandemic, older adults, oldest old, intensive care unit, gerontology
Background
The use of chronological age as a central metric for prognosticating outcomes in older adults presents limitations, yet older age is considered a risk factor for poor COVID-19 prognosis.1,2 In line with the concept of immunosenescence, aging can cause immune function variability that influences the efficacy of pathogen recognition and macrophage activation accelerating early-stage COVID-19.3,4 Older age as a risk factor for COVID-19 hospitalization prognosis is often defined as chronological age divided into three life stages: (1) young-old (65 to 74 years old); (2) middle-old (ages 75 to 84 years old); and (3) oldest-old (over age 85). When compared to young adults (age 18–29), the oldest-old age group has a 15-fold increased risk of hospitalization and a 360-fold increase in the risk of death according to Centers for Disease Control and Prevention.1 In contrast, young-old (age 65–74) have a fivefold increased risk of hospitalization and a 60-fold increased risk of death. While these statistics underscore the profound impact of age on COVID-19 survival rates, they are reducing age to a single determinant without acknowledging the influence of biological sex and metabolic syndrome criteria affecting outcomes associated with immune function.5–7
The biological differences between males and females, including hormonal and immune response differences, are considered contributory to poorer outcomes in COVID-19.7 Similarly, the hormonal and immune responses associated with some chronic conditions, including asthma, diabetes, and obesity, may further contribute to individual outcomes associated with COVID-19 infection.8 However, metabolic syndrome criteria may convey a higher impact specificity of risk for adverse outcomes that is tailored to individuals as a measure of vascular aging9 in contrast with a static measure of chronological age. Metabolic syndrome criteria, specifically dyslipidemia, obesity, and diabetes, are associated with worse COVID-19 outcomes and increased odds of respiratory failure requiring oxygen support and/or death while hospitalized.10,11
Despite the variability and complexity of geriatric conditions, few COVID-19 reports of clinical characteristic prognostication provide data specific to oldest-old adults (over age 85), and instead generally report broadly as 65 and older. This is problematic because chronological age as a risk factor is limited as chronological age does not directly measure health.12 Biological differences may be more aptly explained by chronic conditions, with a focus on hormonal and immune responses such as metabolic syndrome. Relevant characteristics related to metabolic syndrome criteria, independent of age, are important to providing a more precise picture beyond the mainstream headlines. Thus, building on prior examinations that metabolic syndrome is associated with death and acute respiratory distress syndrome (ARDS) across all ages,10,11 this analysis examines metabolic syndrome criteria as central to COVID-19 prognostication beyond chronological age.
Methods
Design, Setting, and Participants
This is a prospective, observational cohort study of 39,564 consecutive hospitalizations of patients diagnosed with COVID-19 in 25 hospitals in Texas between March 13, 2020, and February 28, 2022. Data from Baylor Scott & White Health were collected as part of the global Viral Infection and Respiratory Illness Universal Study (VIRUS) registry of patients hospitalized with COVID-19.13 The VIRUS study standard data collection procedures, designed to maximize the quality and fidelity of data elements, were reported previously.14 Data were entered using a Research Electronic Data Capture15 case report form as part of the Society of Critical Care Medicine Discovery VIRUS COVID-19 global registry.14 The registry was approved by the institutional review board at Mayo Clinic (#20–002610) and Baylor Scott & White Research Institute (#020–119). The ClinicalTrials.gov identifier is NCT04323787, and the reporting of this study conforms to the STROBE statement.16
Main Measures
Inclusion criteria included hospitalized, adult (age ≥ 18 years) COVID-19 patients. Diagnostic testing for SARS-CoV-2 included nucleic acid amplification of nasopharyngeal swabs and bronchial aspirate specimens.17 Hospitalizations were analyzed to evaluate risk for ICU admission and in-hospital mortality. Measures for this study included demographics, medications, body mass index (BMI), comorbid conditions to estimate the Charlson Comorbidity Index,18 highest level of oxygen support for each patient’s hospitalization, and last recorded hemoglobin (Hg) A1c. Hyperlipidemia/dyslipidemia, diabetes mellitus, and hypertension were defined using ICD-10 coded comorbidity documentation within the prior 12 months (Supplemental Table 1). Prediabetes was defined as HgA1c 5.7–6.4% without a concurrent diabetes diagnosis,19 and undiagnosed diabetes was defined as HgA1c > 6.4%. Uncontrolled diabetes was defined as HgA1c > 9.0%.20 Obesity was defined as a BMI ≥ 30.
We used a modified World Health Organization (WHO) criteria21,22 to identify metabolic syndrome, defined as three or more of the following characteristics as present: (1) prediabetes (HgA1c ≥ 5.7%) or history of diabetes mellitus; (2) obesity (BMI ≥ 30); (3) history of hypertension; (4) history of hyperlipidemia or dyslipidemia.
Analysis
The main objective was to determine whether metabolic syndrome characteristics were associated with greater hospital mortality as compared to patients without metabolic syndrome. Student’s t-test was used to compare means of numerical variables in different groups, and Mood’s median test was used for comparison of medians, with an alpha of 0.05 to determine statistical significance. Pearson’s chi-square test was used to compare the difference of distribution in groups for categorical variables.
Building on the analysis approach by Denson et al.,10 to determine the effect of risk increase with each cumulative criteria added, we defined subgroups according to the number of criteria present (e.g., 0, 1, 2, 3, or 4 criteria) and compared ICU admission, hospital mortality, and hospital length of stay between them. The second subgroup analysis compared patients with each individual criteria alone (e.g., patients with obesity but without hypertension, dyslipidemia, and prediabetes/diabetes) to patients with no metabolic syndrome criteria (0 of 4 criteria) to explore which conditions, if any, may be associated with the greatest risk.
For comparison of outcomes, a clustered multivariable logistic regression model (or clustered multivariable linear regression when appropriate) was constructed and clustered by study site and included covariates selected using background knowledge of the causal structure connecting exposure to outcome (i.e., age, self-identified sex, self-identified race and ethnicity, and Charlson Comorbidity Index18). Quantile regression was used to test the difference of median of length of stay and a cumulative logistic model was used for binary outcomes of ICU admission mortality.
Statistical analyses were conducted using SAS Enterprise Guide Version 9.4 (SAS Institute Inc., Cary, NC) software. All statistical tests were two-sided with a statistical significance level set at p-value < 0.05.
Results
The oldest-old cohort (≥ 85 years) accounted for 7.0% (n = 2758) of 39,564 COVID-19 hospitalizations from 25 hospitals during the 2-year study period. Summary of patient characteristics and outcomes by age group is presented in Table 1. Compared to younger groups, the oldest-old cohort had shorter ICU length of stay, similar overall hospitalization duration, and higher rates of discharge destinations providing continued healthcare delivery (i.e., home health, skilled nursing facility).
Table 1.
Characteristics by Age
| Characteristics | 18–64 years (n = 23,057) |
65–74 years (n = 7896) |
75–84 years (n = 5853) |
85 + years (n = 2758) |
p value |
|---|---|---|---|---|---|
| Sex, male | 11,405 (49.5%) | 4258 (53.9%) | 2989 (51.1%) | 1213 (44.0%) | < 0.0011 |
| BMI, median (IQR) | 33.1 (27.9, 41.2) | 30.0 (25.8, 35.9) | 27.6 (23.9, 32.4) | 25.3 (22.1, 29.4) | < 0.0013 |
| Race | < 0.0011 | ||||
| Asian | 247 (1.1%) | 96 (1.2%) | 68 (1.2%) | 28 (1.0%) | |
| Black | 5076 (22.0%) | 1257 (15.9%) | 752 (12.9%) | 276 (10.0%) | |
| White | 14,807 (64.3%) | 5685 (72.0%) | 4520 (77.2%) | 2168 (78.6%) | |
| Other | 2914 (12.6%) | 855 (10.8%) | 512 (8.7%) | 286 (10.4%) | |
| Comorbidities | |||||
| Chronic obstructive pulmonary disease | 1534 (6.7%) | 1468 (18.6%) | 1301 (22.2%) | 514 (18.6%) | < 0.0011 |
| Asthma | 2238 (9.7%) | 568 (7.2%) | 304 (5.2%) | 105 (3.8%) | < 0.0011 |
| Diabetes status* | < 0.0011 | ||||
| No diabetes | 680 (18.4%) | 322 (16.6%) | 256 (18.7%) | 96 (21.3%) | |
| Pre-diabetes | 767 (20.7%) | 492 (25.3%) | 430 (31.5%) | 166 (36.8%) | |
| Diabetes | 1365 (36.9%) | 874 (45.0%) | 569 (41.6%) | 176 (39.0%) | |
| Diabetes, Uncontrolled | 886 (24.0%) | 256 (13.2%) | 112 (8.2%) | 13 (2.9%) | |
| Hypertension | 9983 (43.3%) | 5888 (74.6%) | 4629 (79.1%) | 2178 (79.0%) | < 0.0011 |
| Congestive heart failure | 2600 (11.3%) | 1799 (22.8%) | 1799 (30.7%) | 909 (33.0%) | < 0.0011 |
| Obesity | 8680 (37.6%) | 2553 (32.3%) | 1208 (20.6%) | 268 (9.7%) | < 0.0011 |
| Dyslipidemia/hyperlipidemia | 5803 (25.2%) | 4307 (54.5%) | 3521 (60.2%) | 1479 (53.6%) | < 0.0011 |
| Diagnosis | |||||
| Respiratory failure** | 1825 (7.9%) | 1081 (13.7%) | 739 (12.6%) | 228 (8.3%) | < 0.0011 |
| Stroke | 374 (1.6%) | 195 (2.5%) | 173 (3.0%) | 75 (2.7%) | < 0.0011 |
| Sepsis | 4320 (18.7%) | 1845 (23.4%) | 1307 (22.3%) | 600 (21.8%) | < 0.0011 |
| Oxygenation support, highest level | |||||
| No oxygenation | 1470 (6.4%) | 201 (2.5%) | 79 (1.3%) | 31 (1.1%) | < 0.0011 |
| Nasal cannula | 12,721 (55.2%) | 5451 (69.0%) | 3998 (68.3%) | 1879 (68.1%) | < 0.0011 |
| Face mask | 4504 (19.5%) | 2122 (26.9%) | 1579 (27.0%) | 769 (27.9%) | < 0.0011 |
| High-flow nasal cannula | 5173 (22.4%) | 2457 (31.1%) | 1599 (27.3%) | 722 (26.2%) | < 0.0011 |
| Non-invasive mechanical ventilation | 1586 (6.9%) | 910 (11.5%) | 637 (10.9%) | 223 (8.1%) | < 0.0011 |
| Invasive mechanical ventilation | 2148 (9.3%) | 1123 (14.2%) | 614 (10.5%) | 141 (5.1%) | < 0.0011 |
| Admitted to ICU | 3377 (14.9%) | 1412 (18.1%) | 884 (15.3%) | 328 (11.9%) | < 0.0011 |
| ICU length of stay | 5.0 (1.9, 12.8) | 5.7 (1.9, 13.5) | 4.1 (1.7, 10.0) | 2.8 (1.2, 5.5) | < 0.0013 |
| ICU survival | 2,968 (78.2%) | 1,115 (61.9%) | 688 (59.9%) | 270 (67.8%) | < 0.0011 |
| Hospital length of stay | 4.3 (2.3, 7.9) | 5.9 (3.2, 11.2) | 6.0 (3.4, 10.6) | 5.9 (3.4, 9.1) | < 0.0013 |
| Discharge location | < 0.0011 | ||||
| Home without assistance | 17,997 (82.7%) | 4015 (58.1%) | 1875 (37.3%) | 513 (22.2%) | |
| Home with home health | 1222 (5.6%) | 1088 (15.8%) | 1125 (22.4%) | 473 (20.5%) | |
| Rehab Facility | 507 (2.3%) | 599 (8.7%) | 756 (15.0%) | 384 (16.6%) | |
| Long-term acute care | 436 (2.0%) | 568 (8.2%) | 727 (14.5%) | 565 (24.5%) | |
| Hospice | 39 (0.2%) | 82 (1.2%) | 154 (3.1%) | 213 (9.2%) | |
| Transfer | 916 (4.2%) | 446 (6.5%) | 335 (6.7%) | 143 (6.2%) | |
| Other*** | 640 (2.9%) | 107 (1.5%) | 56 (1.1%) | 18 (0.8%) | |
| Disposition/discharge location | < 0.0011 | ||||
| N-Miss | 312 | 98 | 64 | 43 | |
| In-hospital mortality | 988 (4.3%) | 893 (11.5%) | 761 (13.1%) | 406 (15.0%) | |
| Home without assistant | 17,997 (79.1%) | 4015 (51.5%) | 1875 (32.4%) | 513 (18.9%) | |
| Home with home health | 1222 (5.4%) | 1088 (14.0%) | 1125 (19.4%) | 473 (17.4%) | |
| Rehab facility | 507 (2.2%) | 599 (7.7%) | 756 (13.1%) | 384 (14.1%) | |
*No diabetes classified when A1c is < 5.7% with no diabetes ICD-10 diagnosis. Pre-diabetes classified when A1c 5.7–6.4% without a diabetes diagnosis. Diabetes classified when diabetes ICD-10 diagnosis present and/or A1c 6.5–9.0%. Uncontrolled diabetes when A1c > 9.0% irrespective of diabetes ICD-10 diagnosis
**Acute hypoxic respiratory failure, including acute respiratory distress syndrome (ARDS)
p-value based on: 1Pearson’s chi-square test; 2Student t-test; 3Kruskal-Wallis rank sum test.
***Other discharge disposition includes against medical advice (AMA) and discharge to law enforcement
Chronic conditions varied by age group, with lower proportions of diabetes and uncontrolled diabetes in the oldest-old cohort compared with young-old (65–74 years) and middle-old (75–84 years) groups. Characteristics and outcomes for patients with metabolic syndrome versus those without metabolic syndrome criteria are presented in Table 2. The proportion of patients with metabolic syndrome was significantly higher in males, patients of Black race, those with comorbidities, and those receiving any form of oxygen support during hospitalization. Patients with metabolic syndrome had a significantly higher proportion of admission to the ICU, higher hospital mortality, and longer hospital length of stay.
Table 2.
Patient Characteristics by Metabolic Syndrome Status
| Characteristic | Metabolic syndrome (≥ 3 criteria) (n = 12,274) |
Control (≤ 2 criteria) (n = 27,290) |
p value |
|---|---|---|---|
| Male sex | 6580 (53.6%) | 13,285 (48.7%) | < 0.0011 |
| Age, years—mean (SD) | 65.8 (13.3) | 55.8 (19.7) | < 0.0012 |
| BMI | 33.7 (29.6, 39.9) | 29.1 (25.0, 36.6) | < 0.0013 |
| Race | < 0.0011 | ||
| White | 8318 (67.8%) | 18,862 (69.2%) | |
| Asian | 104 (0.8%) | 335 (1.2%) | |
| Black | 2527 (20.6%) | 4834 (17.7%) | |
| Other | 1324 (10.8%) | 3243 (11.9%) | |
| Comorbidities | |||
| COPD | 2139 (17.4%) | 2678 (9.8%) | < 0.0011 |
| Asthma | 1164 (9.5%) | 2051 (7.5%) | < 0.0011 |
| Diabetes status* | < 0.0011 | ||
| No diabetes | 444 (8.9%) | 910 (36.6%) | |
| Pre-diabetes | 1349 (27.1%) | 506 (20.4%) | |
| Diabetes | 2273 (45.7%) | 711 (28.6%) | |
| Uncontrolled diabetes | 910 (18.3%) | 357 (14.4%) | |
| Hypertension | 11,770 (95.9%) | 10,908 (40.0%) | < 0.0011 |
| Congestive heart failure | 4143 (33.8%) | 2964 (10.9%) | < 0.0011 |
| Obesity | 6450 (52.6%) | 6259 (22.9%) | < 0.0011 |
| Dyslipidemia/hyperlipidemia | 10,335 (84.2%) | 4775 (17.5%) | < 0.0011 |
| Admission diagnoses | |||
| Sepsis | 2949 (24.0%) | 5123 (18.8%) | < 0.0011 |
| Hypoxic respiratory failure** | 1843 (15.0%) | 2030 (7.4%) | < 0.0011 |
| Cardiac arrest | 468 (3.8%) | 565 (2.1%) | < 0.0011 |
| Stroke | 385 (3.1%) | 432 (1.6%) | < 0.0011 |
| Acute MI | 313 (2.6%) | 276 (1.0%) | < 0.0011 |
| Oxygenation support | |||
| No oxygenation | 186 (1.5%) | 1595 (5.8%) | < 0.0011 |
| Nasal cannula | 8821 (71.9%) | 15,228 (55.8%) | < 0.0011 |
| Face mask | 3287 (26.8%) | 5687 (20.8%) | < 0.0011 |
| HFNC | 3671 (29.9%) | 6280 (23.0%) | < 0.0011 |
| Non-invasive | 1641 (13.4%) | 1715 (6.3%) | < 0.0011 |
| Invasive | 1767 (14.4%) | 2259 (8.3%) | < 0.0011 |
| ECMO | 31 (0.3%) | 40 (0.1%) | 0.0211 |
| Admitted to ICU | 2325 (18.9%) | 3676 (13.8%) | < 0.0011 |
| ICU length of stay | 5.5 (1.9, 13.4) | 4.5 (1.7, 11.0) | < 0.0013 |
| ICU discharge status | < 0.0011 | ||
| ICU mortality | 960 (34.2%) | 1143 (26.3%) | |
| ICU survival | 1844 (65.8%) | 3197 (73.7%) | |
| Hospital length of stay | 6.0 (3.6, 11.0) | 4.5 (2.4, 8.2) | < 0.0013 |
| Hospital status | |||
| Missing | 1 | 516 | |
| Hospitalization survival | 10,960 (89.3%) | 25,039 (93.5%) | < 0.0011 |
| Disposition/discharge location | < 0.0011 | ||
| Missing | 1 | 516 | |
| Expired | 1313 (10.7%) | 1735 (6.5%) | |
| Home without assistant | 6252 (50.9%) | 18,148 (67.8%) | |
| Home with home health | 1748 (14.2%) | 2160 (8.1%) | |
| Rehab Facility | 983 (8.0%) | 1263 (4.7%) | |
| Long-term acute care | 885 (7.2%) | 1411 (5.3%) | |
| Hospice | 132 (1.1%) | 356 (1.3%) | |
| Transfer | 746 (6.1%) | 1094 (4.1%) | |
| Other*** | 214 (1.7%) | 607 (2.3%) | |
| Discharge location | < 0.0011 | ||
| Home without assistant | 6252 (57.0%) | 18,148 (72.5%) | |
| Home with home health | 1748 (15.9%) | 2160 (8.6%) | |
| Rehab facility | 983 (9.0%) | 1263 (5.0%) | |
P-value based on: 1Pearson’s chi-square test; 2Student t-test; 3Kruskal-Wallis rank sum test
*Normal A1c when A1c is < 5.7% with no diabetes ICD-10 diagnosis. Pre-diabetes when A1c 5.7–6.4% with no diabetes diagnosis. Diabetes when diabetes ICD-10 diagnosis present and/or A1c < 9.1%. Uncontrolled diabetes when A1c > 9.0% irrespective of diabetes ICD-10 diagnosis.
**Acute hypoxic respiratory failure, including Acute Respiratory Distress Syndrome (ARDS)
***Other discharge disposition includes against medical advice (AMA) and discharge to law enforcement
Evaluations of the effect of metabolic syndrome and patient demographics (i.e., age, sex, race) on ICU admission (Table 3) demonstrate minimal change in the magnitude of effect for metabolic syndrome on ICU admission across the different models. Thus, the effect of metabolic syndrome and patient demographics on ICU admission was affected with chronological age as compared to models without age. In contrast, the effect of metabolic syndrome and patient demographics on hospital mortality was significantly modified in the model with age as compared to models without age (OR [95% CI] = 1.44 [1.33,1.56] vs 1.71 [1.58,1.84]) (Table 4). There was a stepwise progression in odds of ICU admission (Table 5) and hospital mortality (Table 6) when modeled by the number metabolic syndrome criteria, demonstrating stepwise increases in the odds of ICU admission and hospital mortality associated with increases in the number of metabolic syndrome criteria. Propensity score matching was conducted to compare results with logistic regressions, with similar findings (Supplemental Table 2; Supplemental Table 3; Supplemental Fig. 1). Post hoc analyses were also conducted to inspect for differences associated with comorbidities, changes in clinical practice over time, or COVID-19 variants (Supplemental Table 4; Supplemental Table 5). The association of metabolic syndrome on outcomes weakened when additional comorbidities were used, and adjustments for differences over time (e.g., hospitalization in 2020 vs 2021) did not change the magnitude of effect.
Table 3.
Logistic Regression Modeling for ICU Admission
| Model with metabolic syndrome only | Adjusted model including age | Adjusted model excluding age | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristic | OR | 95% CI | p | OR | 95% CI | p | OR | 95% CI | p |
| Metabolic syndrome | |||||||||
| Control | 1.00 | — | 1.00 | — | 1.00 | — | |||
| Metabolic syndrome | 1.46 | 1.38, 1.55 | < 0.001 | 1.42 | 1.34, 1.50 | < 0.001 | 1.44 | 1.36, 1.53 | < 0.001 |
| Sex | |||||||||
| Female | 1.00 | — | 1.00 | — | |||||
| Male | 1.52 | 1.44, 1.61 | < 0.001 | 1.53 | 1.45, 1.62 | < 0.001 | |||
| Race | |||||||||
| White/Caucasian | 1.00 | — | 1.00 | — | |||||
| Asian | 0.79 | 0.59, 1.05 | 0.12 | 0.80 | 0.59, 1.05 | 0.13 | |||
| Black/African American | 0.95 | 0.88, 1.02 | 0.15 | 0.95 | 0.88, 1.03 | 0.20 | |||
| Other | 1.05 | 0.96, 1.14 | 0.31 | 1.05 | 0.96, 1.14 | 0.27 | |||
| Age group | |||||||||
| 18–64 years | 1.00 | — | |||||||
| 65–74 years | 1.15 | 1.07, 1.23 | < 0.001 | ||||||
| 75–84 years | 0.96 | 0.89, 1.04 | 0.35 | ||||||
| ≥ 85 years | 0.78 | 0.69, 0.87 | < 0.001 | ||||||
OR odds ratio, CI confidence interval
Table 4.
Logistic Regression Modeling for in-hospital Mortality
| Model with metabolic syndrome only | Adjusted model including age | Adjusted model excluding age | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristic | OR | 95% CI | p | OR | 95% CI | p | OR | 95% CI | p |
| Metabolic syndrome | |||||||||
| Control | 1.00 | — | 1.00 | — | 1.00 | — | |||
| Metabolic syndrome | 1.73 | 1.60, 1.86 | < 0.001 | 1.44 | 1.33, 1.56 | < 0.001 | 1.71 | 1.58, 1.84 | < 0.001 |
| Sex | |||||||||
| Female | 1.00 | — | 1.00 | — | |||||
| Male | 1.55 | 1.43, 1.67 | < 0.001 | 1.51 | 1.40, 1.63 | < 0.001 | |||
| Race | |||||||||
| White/Caucasian | 1.00 | — | 1.00 | — | |||||
| Asian | 0.97 | 0.67, 1.36 | 0.86 | 0.95 | 0.66, 1.33 | 0.78 | |||
| Black/African American | 0.76 | 0.68, 0.84 | < 0.001 | 0.63 | 0.57, 0.70 | < 0.001 | |||
| Other | 0.65 | 0.56, 0.74 | 0.001 | 0.60 | 0.52, 0.68 | < 0.001 | |||
| Age group | |||||||||
| 18–64 years | 1.00 | — | |||||||
| 65–74 years | 2.54 | 2.30, 2.79 | < 0.001 | ||||||
| 75–84 years | 3.01 | 2.72, 3.33 | < 0.001 | ||||||
| ≥ 85 years | 3.79 | 3.34, 4.29 | < 0.001 | ||||||
OR odds ratio, CI confidence interval
Table 5.
Modeling of ICU Admission by Number of Metabolic Syndrome Measures
| Overall | |||
|---|---|---|---|
| Characteristic | OR | 95% CI | p |
| No. of conditions for metabolic syndrome | |||
| 0 | 1.00 | — | |
| 1 | 0.95 | 0.87, 1.05 | 0.33 |
| 2 | 1.35 | 1.23, 1.49 | < 0.001 |
| 3 | 1.59 | 1.45, 1.75 | < 0.001 |
| 4 | 1.67 | 1.50, 1.86 | < 0.001 |
OR odds ratio, CI confidence interval
Table 6.
Modeling of Mortality by Number of Metabolic Syndrome Measures
| Overall | |||
|---|---|---|---|
| Characteristic | OR | 95% CI | p |
| No of conditions for metabolic syndrome | |||
| 0 | 1.00 | — | |
| 1 | 1.23 | 1.06, 1.44 | 0.008 |
| 2 | 2.19 | 1.90, 2.54 | < 0.001 |
| 3 | 2.61 | 2.26, 3.03 | < 0.001 |
| 4 | 2.70 | 2.30, 3.17 | < 0.001 |
OR odds ratio, CI confidence interval
Discussion
Delineating accurate prognostic indicators for COVID-19 outcomes are important for clinical decision-making and public health planning. Based on our prospective, observational cohort study of 39,564 consecutive hospitalizations of patients diagnosed with COVID-19, we identified that patients with metabolic syndrome had a significantly higher proportion of admission to ICU, greater hospital mortality, and longer hospital length of stay. Importantly, metabolic syndrome was a better predictor of hospitalization outcomes than chronological age alone. During COVID-19 hospitalization, we noted a progressive increase in the risk of both ICU admission and hospital mortality as the number of metabolic syndrome criteria incrementally increased. Thus, we illustrate that metabolic syndrome measures are more clinically relevant for estimating adverse outcomes of COVID-19 hospitalizations compared with chronological age.
Analysis of oldest-old adults (80 and older) in the SEMI-COVID-19 registry of 150 Spanish hospitals conveyed that chronological age, male sex, and poor pre-admission functional status independently contribute to higher odds of in-hospital mortality, while comorbidities showed no significant association.23 The study used a comorbidity measure instead of physiological criteria to predict adverse outcomes during hospitalization. This approach underscores the differentiation between risk factors associated with chronological age and comorbidity compared to those associated with the more direct physiologic markers of metabolic syndrome. Similarly, Lohia (2021)11 and Denson (2021)10 demonstrate the predictive power of metabolic syndrome criteria irrespective of chronological age. When individuals had high cholesterol, hypertension, mild obesity, and pre-diabetes or diabetes and are hospitalized with COVID-19, they had a 1 in 4 chance of developing acute respiratory failure, and were almost 20% more likely to die during hospitalization.
Inclusion of covariates and measures like mechanical ventilation, ICU-acquired delirium, evidence-based bundle performance, and length of stay are important to consider for advancing epidemiologic, health services, and patient outcomes research across the lifespan. Intra-hospitalization characteristics of daily oxygen support types illustrate that basic (i.e., nasal cannula) and intermediate interventions (i.e., high-flow nasal cannula) interventions were adequate for older adults at high risk of ICU admission and mortality.24 In a secondary analysis matching hospitalized oldest-old adults (80 and older) with and without COVID-19, Guidet (2022)25 suggests that older adults with COVID-19 were more likely to experience withholding of life-sustaining treatment compared with their non-COVID ARDS counterparts. Overall, the characteristics of their COVID-19 cohort exhibited lower frailty levels and acuity scores, highlighting the potential interplay between level of care and physiologic condition. In a similar investigation spanning all Dutch ICUs, the ICU and hospital mortality rates of older adults (70 and older) with COVID-19 were substantively higher compared with non-COVID pneumonia diagnoses, both with and without adjustments for acuity and occupancy rates.26 The applicability of metabolic syndrome criteria enables a more holistic approach by prioritizing physiological conditions rather than age when considering the risk of mortality or poor outcomes from COVID-19.
Overall, despite the relevance of care intensity requirements at the critical stage of hospital discharge, post-acute destination outcomes are not broadly reported in analyses of COVID-19 outcomes stratified by older age.25–29 In our cohort, while approximately 80% of adults < 65 years were discharged home without home health, the reverse occurred for the oldest-old (≥ 85 years) with approximately 80% relying on formal care for support from home health, hospice, or institutional care (e.g., inpatient rehabilitation, skilled nursing facility). In comparison, in a large cohort of oldest-old adults ≥ 80 years hospitalized with non-COVID-19 critical illness in Australia and New Zealand, almost two-thirds (65%) were discharged to home.30 Similarly, in a single-site cohort study of 2872 patients in the USA, more than half (56%) of oldest-old adults (≥ 80 years) were discharged home after hospitalization for COVID-19, and rates of discharge to a rehabilitation facility were comparable to national rates for any type of hospitalization.31 These findings are tempered by the absence of the pre-admission location of care (e.g., home, skilled nursing facility), because care requirements after COVID-19 hospitalization are typically higher.32,33 In the USA, the demand for home health care after COVID-19 is considered comparable to the intensity of home health services used by sepsis survivors, with both groups characterized with poor symptom and functional profiles that benefit from interventions to address pain, dyspnea, cognition, anxiety, and physical functioning.33 As life expectancy increases globally, the oldest-old population is projected to triple between 2019 and 2050, as the number of people above 80 years outpaces people 65 years and older. In this large cohort, hospitalization discharge destination characteristics suggest an accelerated and compounded need for post-hospitalization rehabilitation attributed to COVID-19 across the lifespan.
Strengths and Limitations
A major strength of our study is the novel combination of data sources to describe diabetes status using both HgA1c levels and validated ICD-10 codes. While ICD-10 classifications for diabetes provide a binary distinction (presence or absence of a diagnosis), there are risks for misclassification or incomplete data. We supplemented ICD-10 classifications by including HgA1c levels when available, to both mitigate potential ICD-10 missingness and generate glycemic control categories (e.g., pre-diabetes, uncontrolled diabetes). By considering both criteria, we provide a more robust description of diabetes status. The crosswalk format to combine ICD-10 codes for diabetes with HgA1c values is provided (Supplemental Table 1). Further, our use of metabolic syndrome criteria demonstrates a reliance on fewer variables, resulting in improved simplicity and interpretability. Despite the modest improvement in prediction by including additional comorbid conditions (Supplemental Table 4), our use of fewer measures guards against overfitting, multicollinearity, and spurious correlations. Furthermore, the selection of clinically relevant and available measures is better suited to real-world translation.
COVID-19 databases, including the National COVID Cohort Collaborative (N3C) and COVID-NET cohorts, are population-based data repositories of electronic health record and claims data designed to accelerate research and response efforts to the pandemic.34 The N3C database alone contains data drawn from more than 21 million patients in the USA,35 with published analyses representing up to 137,870 hospitalizations drawn from 43 health systems.36 This breadth enables analyses spanning broad geographic regions, although challenges inherent to all databases are data quality and standardization. For example, the specificity of chronic disease measures in the N3C database are not comprehensive enough to evaluate the association of metabolic syndrome characteristics with interventions or outcomes. In contrast, data extracted from a single integrated health system using the VIRUS registry measures can be used, although not all patient history elements are included (i.e., familial hypercholesterolemia). Additional strengths of the Baylor Scott & White Health cohort of the VIRUS dataset are that data was drawn from a common data model, enabling important data of interest, including intra-hospitalization process measures, such as type of oxygen delivery required on each day of hospitalization.24
A limitation of this observational study design is the potential residual confounding by unmeasured variables that limits causal inferences. Second, while our population-based approach is drawn from a large integrated health system in the Southwest region of the USA, it does not reflect characteristics of other regions; thus, analyses from other health systems and regions are needed to confirm findings. Third, excluding hospitalization admission source (e.g., home, nursing home) could be viewed as a weakness given the impact of nursing homes as significant sources of COVID-19 transmission, with questions of non-random negative outcomes associated with variability in timeliness of patient transitions from nursing homes to hospitals, while also serving as a strength because bypassing complex nested analyses favors a parsimonious analysis centered on chronological age compared with metabolic syndrome criteria. Similarly, excluding COVID-19 pharmacologic treatments (i.e., remdesivir, dexamethasone) may affect outcomes, though influence is likely restricted to non-mortality measures. For example, there is no evidence suggesting that the efficacy of remdesivir for COVID-19 in hospital settings varies by chronological age.37
Metabolic syndrome criteria are particularly important in the USA because of the markedly increased prevalence compared with other countries.10 Future research to evaluate whether well-controlled diabetes and/or hyperlipidemia remain salient risk factors is warranted, because current risk factor analyses do not differentiate beyond metabolic syndrome criterion presence and absence. Exploring the impact of sustained glycemic control and optimal lipid profiles could contribute to identifying distinct health trajectories and outcomes, ultimately guiding more targeted and personalized prognostication and interventional decisions.
Conclusions
Metabolic syndrome measures are important individual predictors of COVID-19 outcomes. Building on prior examinations that metabolic syndrome is associated with death and ARDS across all ages, this analysis examines metabolic syndrome criteria independent of age. Biological age measures, including vascular aging measured using pragmatic measures of metabolic syndrome, may be more relevant than chronological age as risk factors for poor outcomes attributed to COVID-19. The applicability of metabolic syndrome criteria with clinically relevant outcomes is positioned for enhancing clinical relevance and aiding stakeholders at the bedside.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary file1 (DOCX 43 KB) Supplemental Table 1. Diabetes classification methodology
Supplementary file2 (DOCX 45 KB) Supplemental Table 2. Propensity score matched cohort on ICU admission and mortality
Supplementary file3 (DOCX 49 KB) Supplemental Table 3. Comparison of propensity score matched cohort characteristics
Supplementary file4 (DOCX 47 KB) Supplemental Table 4. Covariate adjustments expanded to most common chronic conditions
Supplementary file5 (DOCX 48 KB) Supplemental Table 5. Covariate adjustments expanded to chronic conditions and year of admission
Supplementary file6 (SVG 228 KB) Supplemental Figure 1. Absolute standardized mean difference of pre-matched (all) and matched cohort
Author Contribution:
Valerie Danesh, PhD RN: conceptualization; investigation; methodology; project administration; resources; supervision; validation; first draft; review and editing
Alaina Tellson, PhD RN: review and editing
Alan B Stevens, PhD: methodology; review and editing
Leanne M Boehm, PhD RN, ACNS-BC: review and editing
Gerald O Ogola, PhD: methodology; analysis; first draft; review and editing
Anisha Shrestha, MS: methodology; validation; review and editing
Jinmyoung Cho, PhD: review and editing
Edgar J Jimenez, MD: review and editing
Alejandro C Arroliga, MD MSc: conceptualization; project administration; supervision; first draft; review and editing
Funding
Dr. Danesh received grant funding from NIH/NIA (#R21AG080339). Dr. Boehm received grant funding from NIH/NIA (#R01AG077644). Dr. Stevens received grant funding from NIH/NIA (#R01AG061973).
This study received financial support from the Cardiovascular Research Review Committee of the Baylor Healthcare System Foundation, the Society of Critical Care Medicine, and Gordon and Betty Moore Foundation.
Data Availability
The dataset analyzed for the current study is available from the corresponding author on reasonable request and the execution of an institutional Data Use Agreement.
Declarations
Conflict of Interest
The authors report no conflicts of interest.
Footnotes
Context
Question
Do metabolic syndrome criteria have more explanatory power for risk prediction for COVID-19 than chronological age?
Findings
In this cohort study of 39,564 hospital admissions in 25 hospitals in Texas, we identified that patients with metabolic syndrome had a significantly higher proportion of admission to ICU, greater hospital mortality, and longer hospital length of stay.
Meaning
These findings suggest that metabolic syndrome measures may be a more direct measure of individual heterogeneity than chronological age.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.CDC. Risk for COVID-19 infection, hospitalization, and death by age group. CDC. COVID-19 Web site. https://www.cdc.gov/coronavirus/2019-ncov/covid-data/investigations-discovery/hospitalization-death-by-age.html. Published 2023. Accessed 2023–05–30.
- 2.Taylor CA, Patel K, Patton ME, et al. COVID-19-Associated Hospitalizations Among U.S. Adults Aged >/=65 Years - COVID-NET, 13 States, January-August 2023. MMWR Morb Mortal Wkly Rep. 2023;72(40):1089-1094. 10.15585/mmwr.mm7240a3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Mueller AL, McNamara MS, Sinclair DA. Why does COVID-19 disproportionately affect older people? Aging (Albany NY). 2020;12(10):9959-9981. 10.18632/aging.103344 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Napoli C, Coscioni E, Trama U, Strozziero MG, Benincasa G. An evidence-based debate on epigenetics and immunosenescence in COVID-19. Curr Res Immunol. 2023;4:100069. 10.1016/j.crimmu.2023.100069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Singh AK, Gillies CL, Singh R, et al. Prevalence of co-morbidities and their association with mortality in patients with COVID-19: A systematic review and meta-analysis. Diabetes Obes Metab. 2020;22(10):1915-1924. 10.1111/dom.14124 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Vo AD, La J, Wu JT-Y, et al. Factors Associated With Severe COVID-19 Among Vaccinated Adults Treated in US Veterans Affairs Hospitals. JAMA Network Open. 2022;5(10):e2240037. 10.1001/jamanetworkopen.2022.40037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wiersinga WJ, Rhodes A, Cheng AC, Peacock SJ, Prescott HC. Pathophysiology, transmission, diagnosis, and treatment of coronavirus disease 2019 (COVID-19): A review. JAMA. 2020;324(8):782-793. [DOI] [PubMed]
- 8.Thakur B, Dubey P, Benitez J, et al. A systematic review and meta-analysis of geographic differences in comorbidities and associated severity and mortality among individuals with COVID-19. Sci Rep. 2021;11(1):8562. 10.1038/s41598-021-88130-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hamczyk MR, Nevado RM, Barettino A, Fuster V, Andres V. Biological Versus Chronological Aging: JACC Focus Seminar. J Am Coll Cardiol. 2020;75(8):919-930. 10.1016/j.jacc.2019.11.062 [DOI] [PubMed] [Google Scholar]
- 10.Denson JL, Gillet AS, Zu Y, et al. Metabolic Syndrome and Acute Respiratory Distress Syndrome in Hospitalized Patients With COVID-19. JAMA Network Open. 2021;4(12):e2140568. 10.1001/jamanetworkopen.2021.40568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lohia P, Kapur S, Benjaram S, Pandey A, Mir T, Seyoum B. Metabolic syndrome and clinical outcomes in patients infected with COVID-19: Does age, sex, and race of the patient with metabolic syndrome matter? J Diabetes. 2021;13(5):420-429. 10.1111/1753-0407.13157 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Burch JB, Augustine AD, Frieden LA, et al. Advances in geroscience: impact on healthspan and chronic disease. J Gerontol A Biol Sci Med Sci. 2014;69 Suppl 1:S1-3. 10.1093/gerona/glu041 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Domecq JP, Lal A, Sheldrick CR, et al. Outcomes of patients with coronavirus disease 2019 receiving organ support therapies: The international viral infection and respiratory illness universal study registry. Crit Care Med. 2021;49(3):437-448. 10.1097/CCM.0000000000004879 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Walkey AJ, Kumar VK, Harhay MO, et al. The Viral Infection and Respiratory Illness Universal Study (VIRUS): An International Registry of Coronavirus 2019-Related Critical Illness. Crit Care Explor. 2020;2(4):e0113. 10.1097/CCE.0000000000000113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377-381. 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. 10.1016/S0140-6736(07)61602-X [DOI] [PubMed] [Google Scholar]
- 17.Mutnal MB, Johnson S, Mohamed N, et al. Surveillance genome sequencing reveals multiple SARS-CoV-2 variants circulating in central Texas, USA, with a predominance of delta variant and review of vaccine breakthrough cases. J Med Virol. 2022;94(3):937-944. 10.1002/jmv.27373 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Quan H, Sundararajan V, Halfon P, et al. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005;43(11):1130-1139. 10.1097/01.mlr.0000182534.19832.83 [DOI] [PubMed] [Google Scholar]
- 19.CDC. All about your A1c. Centers for Disease Control (CDC). Living with diabetes Web site. https://www.cdc.gov/diabetes/managing/managing-blood-sugar/a1c.html#print. Published 2022. Accessed 2023–03–29, 2023.
- 20.Medicaid. Comprehensive diabetes care: Hemoglobin A1c (HbA1c) poor control (>9.0%): Ages 18 to 75. State Overviews Web site. https://www.medicaid.gov/state-overviews/scorecard/comprehensive-diabetes-care/index.html. Published 2023. Accessed September 11, 2023.
- 21.Grundy SM, Brewer HBJ, Cleeman JI, et al. Definition of metabolic syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation. 2004;109(3):433-438. 10.1161/01.CIR.0000111245.75752.C6 [DOI] [PubMed] [Google Scholar]
- 22.Xie J, Zu Y, Alkhatib A, et al. Metabolic Syndrome and COVID-19 Mortality Among Adult Black Patients in New Orleans. Diabetes Care. 2020;44(1):188-193. 10.2337/dc20-1714 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ramos-Rincon JM, Buonaiuto V, Ricci M, et al. Clinical Characteristics and Risk Factors for Mortality in Very Old Patients Hospitalized with COVID-19 in Spain. J Gerontol A Biol Sci Med Sci. 2020;76(3):e28–e37. 10.1093/gerona/glaa243. [DOI] [PMC free article] [PubMed]
- 24.Danesh V, White HD, Tecson KM, et al. Daily oxygenation support for patients hospitalized with SARS-CoV-2 in an integrated health system. Respir Care. 2023;68(4):497-504. [DOI] [PMC free article] [PubMed]
- 25.Guidet B, Jung C, Flaatten H, et al. Increased 30-day mortality in very old ICU patients with COVID-19 compared to patients with respiratory failure without COVID-19. Intensive Care Med. 2022;48(4):435-447. 10.1007/s00134-022-06642-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Haas LEM, Termorshuizen F, Den Uil CA, De Keizer NF, De Lange DW. Increased mortality in ICU patients >=70 years old with COVID-19 compared to patients with other pneumonias. J Am Geriatr Soc. 2023;71(5):1440-1451. 10.1111/jgs.18220. [DOI] [PubMed]
- 27.Lozano-Montoya I, Quezada-Feijoo M, Jaramillo-Hidalgo J, Garmendia-Prieto B, Lisette-Carrillo P, Gomez-Pavon FJ. Mortality risk factors in a Spanish cohort of oldest-old patients hospitalized with COVID-19 in an acute geriatric unit: the OCTA-COVID study. Eur Geriatr Med. 2021;12(6):1169-1180. 10.1007/s41999-021-00541-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Oba S, Altinay M, Salkaya A, Turk HS. Evaluation of the effect of clinical characteristics and intensive care treatment methods on the mortality of covid-19 patients aged 80 years and older. BMC Anesthesiol. 2021;21(1):8. 10.1186/s12871-021-01511-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhu X, Yuan W, Shao J, et al. Risk factors for mortality in patients over 70 years old with COVID-19 in Wuhan at the early break: retrospective case series. BMC Infect Dis. 2021;21(1):821. 10.1186/s12879-021-06450-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Rai S, Brace C, Ross P, et al. Characteristics and outcomes of very elderly patients admitted to intensive care: A retrospective multicenter cohort analysis. Crit Care Med. 2023;51(10):1328-1338. 10.1097/CCM.0000000000005943 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lewis CW, Gray E, Dreyer S, Goodman D, Jayabalan P. The Relationship Between Patient-Specific Factors and Discharge Destination After COVID-19 Hospitalization. Am J Phys Med Rehabil. 2023;102(7):611-618. 10.1097/PHM.0000000000002159 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Kapur S, Cantor Z, Bishop P, Benjaram S, Lohia P. Need for inpatient rehabilitation and skilled nursing services in COVID-19 patients post hospitalization: A retrospective cohort study. Ann Palliat Med. 2023;12(1):90-102. 10.21037/apm-22-858 [DOI] [PubMed] [Google Scholar]
- 33.Bowles KH, McDonald M, Barron Y, Kennedy E, O'Connor M, Mikkelsen M. Surviving COVID-19 after hospital discharge: Symptom, functional, and adverse outcomes of home health recipients. Ann Intern Med. 2021;174(3):316-325. 10.7326/M20-5206 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.O'Halloran A, Whitaker M, Patel K, et al. Developing a sampling methodology for timely reporting of population‐based COVID‐19‐associated hospitalization surveillance in the United States, COVID‐NET 2020–2021. Influenza Other Respir Viruses. 2023;17(1):e13089. 10.1111/irv.13089. [DOI] [PMC free article] [PubMed]
- 35.National COVID Cohort Collaborative (N3C). Comorbidity and demographics breakdown of all patients in the N3C enclave. National Institutes of Health. National COVID Cohort Collaborative (N3C) Web site. https://covid.cd2h.org/dashboard/public-health/enclave-health/3. Published 2024. Accessed 2024.
- 36.Mehta HB, An H, Andersen KM, et al. Use of Hydroxychloroquine, Remdesivir, and Dexamethasone Among Adults Hospitalized With COVID-19 in the United States. Ann Intern Med. 2021;174(10):1395-1403. 10.7326/M21-0857 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Singh S, Khera D, Chugh A, Khera PS, Chugh VK. Efficacy and safety of remdesivir in COVID-19 caused by SARS-CoV-2: a systematic review and meta-analysis. BMJ Open. 2021;11(6):e048416. 10.1136/bmjopen-2020-048416 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary file1 (DOCX 43 KB) Supplemental Table 1. Diabetes classification methodology
Supplementary file2 (DOCX 45 KB) Supplemental Table 2. Propensity score matched cohort on ICU admission and mortality
Supplementary file3 (DOCX 49 KB) Supplemental Table 3. Comparison of propensity score matched cohort characteristics
Supplementary file4 (DOCX 47 KB) Supplemental Table 4. Covariate adjustments expanded to most common chronic conditions
Supplementary file5 (DOCX 48 KB) Supplemental Table 5. Covariate adjustments expanded to chronic conditions and year of admission
Supplementary file6 (SVG 228 KB) Supplemental Figure 1. Absolute standardized mean difference of pre-matched (all) and matched cohort
Data Availability Statement
The dataset analyzed for the current study is available from the corresponding author on reasonable request and the execution of an institutional Data Use Agreement.