Factors Associated with Severe Covid-19 Despite Vaccination: A Nationwide, Retrospective Cohort Study

doi:10.21203/rs.3.rs-1654435/v1

Download PDF

Research Article

Factors Associated with Severe Covid-19 Despite Vaccination: A Nationwide, Retrospective Cohort Study

https://doi.org/10.21203/rs.3.rs-1654435/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Importance: With a large proportion of the US adult population vaccinated against SARS-CoV-2, it is important to identify who remains at risk of severe disease in the event of infection.

Objective: To characterize risk factors for severe COVID-19 disease in a vaccinated population.

Design: Nationwide, retrospective cohort study.

Setting: US Department of Veterans Affairs Hospitals.

Participants: US Veteran patients who received a COVID-19 vaccination series and later developed a laboratory-confirmed SARS-CoV-2 infection.

Exposures: Demographics, comorbidities, immunocompromised states, and vaccination-related variables.

Main Outcome and Measures: Development of severe versus non-severe COVID-19 disease. Severe disease was defined as hospitalization within 14 days of a positive SARS-CoV-2 diagnostic test and either SpO2 < 94%, receipt of supplemental oxygen or dexamethasone, mechanical ventilation, or death within 28 days.

Results: Among 110,760 patients with breakthrough infections, 10,612 (9.6%) had severe COVID-19 or died within 28 days of the COVID-19 diagnosis. The strongest risk factor for severe disease despite vaccination was age, which increased with every 5-year increase in age > 50 (approximate adjusted odds ratio (aOR), 1.5). Risk among patients ≥ 80 years old was particularly high (aOR 16.1, CI 13.1 – 19.9, relative to patients 45-50 years old. Immunocompromising conditions, including receipt of different classes of immunosuppressive medications (aORs 1.73 – 2.44, CIs 1.37 – 2.99) or cytotoxic chemotherapy (aOR 2.69, CI 2.25 – 3.21) prior to breakthrough infection, or leukemias/lymphomas (aOR 1.84, CI 1.59 – 2.14) and chronic conditions associated with end-organ disease, such as heart failure, dementia, and chronic kidney disease, were also associated with increased risk. Receipt of an additional (booster) dose of vaccine was associated with reduced odds of severe disease.

Conclusions and Relevance: In this nationwide, retrospective cohort of US Veterans, we identified risk factors for severe disease despite vaccination. Findings could be used to inform outreach efforts for booster vaccinations and to inform clinical decision making about identifying patients who might benefit from pre-exposure prophylaxis and antiviral therapy.

Infectious Diseases

COVID-19

vaccination effectiveness

risk factors for severe disease

breakthrough infections

Question: What are risk factors for severe vaccine breakthrough Covid-19 infections?

Findings: In a nationwide cohort of vaccinated US Veterans, the strongest predictor of severe breakthrough disease was increasing age, steadily above age 50. Immunocompromising conditions and comorbidities indicating chronic heart, lung, kidney, or neurologic damage also increased risk, with a magnitude similar to or less than a 10-year age increase.

Meaning: Identification of the risk factors for severe breakthrough Covid-19 could be used to guide policies and decision-making about preventive measures for those who remain at risk of disease progression despite vaccination.

The two-dose mRNA vaccine series BNT162b2 (Pfizer/BioNTech) and mRNA-1273 (Moderna) and the adenoviral vaccine Ad26.COV2.S (Janssen) reduce symptomatic infection with the original SARS-CoV-2 strain and are effective at reducing risk of severe Covid-19 disease and death in most patients.

“Breakthrough” infections in vaccinated persons have been observed in the US since mid-2021 and continued to increase after omicron predominance. Substantial protection against severe outcomes following vaccination and boosting is maintained for the majority of the population;^1–3 however, some patients remain at risk of severe breakthrough infections, leading to hospitalizations and death.

Identifying patients who remain at risk for severe disease despite vaccination can help inform efforts to promote booster vaccination and guide distribution of antiviral drugs for early treatment or pre-exposure prophylaxis. Thus, the primary aim of this national, retrospective cohort study was to identify demographic and clinical risk factors for severe disease among vaccinated US Veterans who developed breakthrough infections. Secondary aims were to identify and quantify risk factors in groups that may have different risk profiles than the general population (immunocompromised patients and adults age under 50) and to estimate the effectiveness of the primary vaccination series and a single booster dose as functions of time.

Cohort creation:

The cohort consisted of all patients who received an initial vaccination series before 12/1/2021 and subsequently tested positive for SARS-CoV-2 infection (PCR or antigen) within the Veterans Affairs Healthcare System (VA) through 2/28/2022. The study period includes intervals during which the original strain, delta variant (7/1/2021-12/15/2021), and omicron variant 12/16/2021-2/28/2021) were predominant. Full methods are outlined in eMethods.

Data sources and uses:

Data were obtained from the VA COVID-19 shared data resource⁴ and the Corporate Data Warehouse (CDW) that encompasses the clinical electronic health records (EHR) of all VA patients and facilities. See eTable 1 for a complete listing of definitions and data sources.

The primary outcome was severe breakthrough infection, defined by either: death 2–28 days after a positive test; medical/surgical hospitalization within 14 days of the positive test and documented SpO2 < 94%, receipt of any supplemental oxygen, or receipt of dexamethasone; or mechanical ventilation. The comparator was non-severe breakthrough infection: no hospitalization, or hospitalization without meeting criteria for severe disease.

Demographic data included age in years at the time of breakthrough infection, biologic sex, race and ethnicity, and US region. Vaccination data included dates of each vaccination dose and the vaccine product. Occurrence of a previous SARS-CoV-2 infection before vaccination was defined as a positive test any time before day 14 after completing the initial vaccination series.

Comorbidities recorded in the patients’ records at the time of infection were assessed using the Chronic Conditions Warehouse.⁵ Body mass index (BMI) in kg/m² was calculated from weight and height, the latter imputed if not available in the EHR.

Definitions of immunocompromised patients:

Drugs known to generally increase risk of infectious diseases were classified into five categories: cytotoxic chemotherapy, glucocorticoids (prednisone or methylprednisolone), cytokine-blocking drugs (e.g., anti-TNF, -IL-6R, -IL-17), leukocyte-inhibiting drugs (e.g., methotrexate, azathioprine, JAK inhibitors, abatacept) and lymphocyte-depleting drugs (e.g., rituximab). Receipt of immunosuppressive drugs was assessed both shortly before vaccination and after vaccination but shortly before breakthrough infection. Different time intervals were used for different drug classes and based on their expected duration of impact (See eTable 2 for a full list of drugs, their classifications and time windows, and the numbers of patients receiving these drugs prior to vaccination and after vaccination (I.e., prior to breakthrough)). The immunocompromised sub-cohort also included patients with leukemia or lymphoma but not patients with HIV, based on the results of the primary analysis.

Time since Vaccination

Interpreting time-since-vaccination as a risk factor for severe disease may be challenging due to high potential for residual confounding and effect modification, due to earlier availability to higher-risk patients, differential use and timing of boosting, and development of the delta and then the less-dangerous omicron variant. Thus, in addition to assessing risk of severe breakthrough infection as a function of time-since-initial-vaccination and time-since-boosting, time-since-vaccination was also assessed in patients who did not receive boosters.

Statistical Analysis:

The primary analysis evaluated exposures associated with development of severe compared to non-severe Covid-19 despite vaccination. Secondary stratified analyses were conducted on subcohorts (Fig. 1), to further explore the impacts of an immunocompromised state, age, and time since vaccination.

Multivariable logistic regression was used to assess the association of the binary outcome (severe vs non-severe breakthrough SARS-CoV-2 infection) with predictor variables. Variables determined a priori to be of interest and those with adjusted odds ratios (aOR) ≥ 1.5 or ≤ 0.67 and statistical significance are presented in the main results. Findings were considered statistically significant at P < 0.05 after adjustment for multiple comparisons using the Benjamini-Hochberg method.

Full Cohort:

During the period from 12/15/2020 to 2/28/2022, 111,151 patients had a breakthrough SARS-CoV-2 infection; severity assessments were available for 110,760 (99.6%). Data for all variables that were analyzed are in eTable 3, and data for variables associated with severity at aOR ≥ 1.5 or aOR ≤ 0.67, and others of clinical interest are shown in Table 1. eTable 4 shows separation of severity into categories, and variables not included in regressions. Demographics of the cohort with breakthrough infections were typical of the VA population: 88% male, 65% white, mean age 61.8, evenly distributed throughout US regions, and with high rates of cardiovascular and pulmonary comorbidities.

27% of breakthrough infections occurred during the delta period (7/1/2021–12/15/2021) and 71% during the omicron period (12/16/2021–2/28/2022), 3.9% of patients had evidence of prior SARS-CoV-2 infection before vaccination, and 27% were boosted prior to diagnosis of breakthrough infection. The 110,760 breakthrough infections were associated with 10,612 severe cases (9.6%) and 1,555 deaths (1.4%). In the delta period, 16% of cases were severe and 2.5% were associated with death, compared to 6.7% and 0.9% during the omicron period.

Whole Cohort:

Multivariable adjusted estimates of risk of severe disease are shown for selected variables in Table 1 and Fig. 2. eTable 3 includes all variables, regardless of magnitude of effect, and results of multivariable logistic regressions. The strongest predictor of severe COVID-19 outcomes was age; risk rose substantially with each 5-year increment at least above age 50, such that aOR was 16.2 (CI 13.1–19.9) for patients above 80. Risk of severe disease was somewhat higher among Native Americans than in white patients. Female patients had lower odds of severe outcomes than male patients. Compared to receipt of the Pfizer vaccine, patients who received the Moderna vaccine had lower odds of severe disease (aOR 0.79, CI 0.76–0.83) and those who received one dose of the Janssen vaccine had higher odds (aOR 1.20, CI 1.30–1.41).

Comorbidities in many different organ systems (cardiac, pulmonary, renal, and neurologic/psychiatric) were associated with a similar magnitude of increased odds (aOR between 1.59–2.83 for the 8 highest-risk conditions). Comorbidities that reflected permanent organ damage (e.g., heart failure, COPD, chronic kidney disease (CKD), dementia) were associated with greater magnitudes of increased risk of disease than predisposing conditions, such as diabetes and hypertension.

Patients < 50:

In the sub-cohort limited to patients < 50 years of age, overall risk of severe disease was low (305/25,662 = 1.2%), and death was rare (9/25,662 = 0.04%). The low number of severe cases in this group limits interpretability; however, a small set of severe chronic health problems met criteria for statistical significance (eTable 5 and eResults).

Immunocompromised Cohort:

13,768/110,760 patients with breakthrough infections met criteria for being immunocompromised at the time of vaccination or breakthrough. HIV infection was not included in the definition since the diagnosis was associated with a non-significant and modest increased odds of severe breakthrough Covid-19 (aOR 1.29). Demographic and clinical data for the immunocompromised cohort, stratified by severity, are shown in Table 2 and eTable 6. Immunocompromised patients who developed breakthrough infections had higher absolute unadjusted risk of severe disease (20.9%) or death (3.8%) than patients without immunocompromising conditions or medications (8.0% and 1.1%). Rates of severe disease and death among immunocompromised patients were 28.7% and 5.6% in the delta period and 16.1% and 2.6% in the omicron period. Additional risk factors for severe disease among immunocompromised patients (e.g., age and specific comorbidities) were very similar to those for the whole population (Table 2 and eTable 6).

No specific class of immunosuppressive drugs stood out as conferring higher odds (aOR 1.49–2.19 for use before breakthrough). In multivariable analyses, receipt of immunosuppressive drugs after vaccination (before breakthrough) was a predictor of severe Covid-19, whereas magnitude of risk associated with receipt before vaccination was greatly reduced (Tables 1 and 2, eTables 3 and 6); among 1,949 patients on immunosuppressive drugs only prior to initial vaccination, 272 had severe disease (14.0%), in contrast to 1,740 /7,793 (22.3%) patients immunosuppressed only at the time of the breakthrough infection, and 669/3,114 (21.5%) patients immunosuppressed both before and after vaccination (eTable 7). 1,144/13,768 (8.3%) immunocompromised patients were on at least two different classes of immunosuppressive drugs between vaccination and breakthrough, and these patients appeared to be at additional increased risk of severe disease when compared to those receiving only one immunosuppressive drug class: aOR 1.77 (1.48–2.12) for 2 drugs, aOR 2.34 (1.52–3.62) for 3 drugs (eTable 8).

Impact of Time Since Vaccination or Booster:

Boosting was associated with a decreased risk of severe breakthrough in the main analysis and in all strata (aOR 0.55–0.67) (Fig. 3), as was prior infection with SARS-CoV-2 (aOR 0.55–0.86). In analysis of the whole cohort, time-since-vaccination appeared to be associated with a small but steady increase in risk of severe disease. (Table 1, Figs. 2 and 3, eTables 3, 5, 6, 9 and 10). However, when interpretability was improved by limiting to patients who did not receive boosters and analyzing immunocompromised patients separately, no trends toward decreasing effectiveness were seen, either in patients under or over 65, nor in immunocompromised patients (Fig. 3, eTables 6, 11 and 12).

This nationwide retrospective cohort study of US Veteran patients with documented SARS-CoV-2 infection after vaccination identified clinical and demographic variables associated with risk of severe disease, defined as either death within 28 days or hospitalization with evidence of respiratory failure or hypoxemia. We took a “high-altitude” approach to a broad range of possible risk factors in a large population, rather than attempting to dissect the details of individual comorbidities, immunocompromised states, or factors specific to vaccine products and variants. Risk was reduced among patients who received a booster vaccine or had a history of infection before vaccination and was lower during the omicron period. Increasing age had the most substantial effect on risk, which rose steadily at least above age 50, but a large number of comorbidities remained associated with risk of severe disease in the adjusted model. In light of high rates of vaccination and prior infection, research on these details in breakthrough infection and re-infection, focused on severity and sequelae, has the most potential to inform clinical practice through risk-stratification of individual patients.⁶

Immunization during periods of immunocompromise has been theorized to negatively impact vaccine response rates and predispose to risk of infection and severe outcomes. In this large cohort, an immunocompromised state conferred increased risk of severe outcomes; the impact of the immunosuppressive drugs was stronger if present after vaccination (I.e., producing an immunocompromised state at the time of exposure and subsequent infection) than if present only prior to vaccination. These data suggest that many patients receiving immunosuppressive medications at the time of vaccine develop durable protection, and the causal risk factor is immune status at the time of exposure.

The comorbidities most strongly associated with risk of severe disease include those identified early in the pandemic as the key risk factors for severe disease before vaccines were available.^7–9 Comorbidities indicating pre-existing organ disease (e.g., heart failure, chronic kidney disease, COPD, dementia) or a globally tenuous or frail state (e.g., pressure ulcers, mobility impairments, low BMI) were more important than factors that contribute to future organ dysfunction (e.g., hypertension). However, age was so much stronger as a risk factor for severe outcomes that the magnitudes of risk associated with any source of immunocompromise, or the most important comorbidities were similar to the difference in risk between vaccinated persons aged 60 versus 50. We cannot explain the apparent higher risk among the small number of Native Americans in the cohort (N = 890), but the similar risks among white, Black/African-American, and Hispanic/Latino groups likely reflects the VA providing similar access in most regions.¹⁰

We also identified factors that were protective against severe outcomes; consistent with prior studies, boosters, or a history of infection prior to initial vaccination significantly reduced (to a similar degree) but did not eliminate risk of severe breakthrough in all subcohorts defined by age or presence or absence of immunocompromise. Although time since initial vaccination appeared to be predictive of severe outcomes in the unstratified analyses, different effect estimates in the subcohorts suggested confounding. Among non-immunocompromised patients who did not receive boosters – a lower-risk group overall – there was no trend toward reduction in risk of severe disease. This study does not provide insight into optimal timing for repeated boosting, or if repeated boosting is necessary for all individuals, as only 5 months of data on risk were available after boosters were recommended for non-IC patients with other risk factors, and less for the general population; during this limited follow up period, no clear trends toward reduced benefit were identified.

The CDC published a report of N = 189 severe cases among 2,246 breakthrough infections and found that risk factors for hospitalization and death among vaccinated patients included age ≥ 65, immune-compromise, and heart, liver, kidney, neurologic disease and diabetes; all patients who died had multiple comorbidities.¹¹ Most studies of patients in immunocompromised states have focused on laboratory-based studies rather than clinical outcomes.¹² A recent study of breakthrough infection in immunocompromised patients focused on comparing the conditions causing immune dysfunction and did not attempt to determine risk factors for severity.¹³ Multiple studies have shown decline in protection after initial vaccination in the general population, with some incorporating severity^3,14,15 and others showing improvement after boosting.^1,2

This study has important limitations. As with all VA studies, the population was predominantly male and older (mean age 62, and greater than 70 among patients with severe disease), and with a high burden of chronic medical problems. Our use of a nationwide database curated from electronic health records produces additional limitations: uncertainty about the accuracy of Covid-19 as the reason for hospitalization, imperfect algorithms for comorbidities, and missing data, most likely regarding prior infection, boosters, hospitalization at facilities not contracted for reimbursement by the VA and use of monoclonal antibodies. To address these limitations, we focused on questions that would not depend on data that might be missing. The use of a cohort rather than a matched case-control design, which has been commonly used in studies to estimate vaccine effectiveness, with 90% of patients having non-severe disease, is less likely to pick up uncommon risk factors and may be more susceptible to residual confounding considering differences such as younger age and fewer comorbidities. Our pre-specified analysis of patients with immunocompromising conditions addresses a group that has been widely discussed but under-studied with regard to clinical protection via vaccination.^12,13,16 A limitation in that analysis is that these drugs are highly diverse, which necessitated grouping drugs with very different mechanisms of action, e.g. into “leukocyte-inhibiting” or “cytokine-blocking.” In addition, our ability to assess the impact of immunosuppressive medications was limited by our inability to identify systemic use of certain medications (e.g., tacrolimus) and inability to determine glucocorticoid doses.

Strengths of this study start with use of a large, nationwide dataset with granular data on vaccination, clinical and demographic risk factors, and SARS-CoV-2 with an assessment of severity. Inclusion of objective metrics to assess hospitalization severity is a strength compared to studies that do not account for in-hospital screening practices that impact the “hospitalization plus a diagnostic test” definition used by COVID-NET; prior work demonstrates that controlling for hospitalization severity is important for accurately assessing vaccine effectiveness, as vaccinated and boosted patients are differentially represented in non-severe hospitalizations.^17,18 Limitation to patients with documented vaccination and documented breakthrough infection avoids bigger problems with missing data and unmeasurable confounders that can plague population-based studies trying to compare “vaccinated with infection” (our cohort) to either “unvaccinated with infection” or “vaccinated without infection” (including test-negative designs). VA data have high accuracy for distribution of medications, thus assessments of immune status at the time of infection and vaccination are likely to be reliable. Finally, our analyses of immunocompromised patients and patients under 50 provide novel findings in groups that have not been previously studied rigorously on a large scale.

In conclusion, this study provides insight into the demographic, clinical, and vaccination-related risk factors for severe compared to non-severe breakthrough Covid-19 infection. These results could be used to bolster guidelines for use of pre-exposure and post-exposure anti-viral treatments that are in limited supply. Development of models to estimate the probability of a patient progressing to severe disease for individual risk assessment, to guide treatment and prophylaxis planning and outreach, will require more sophisticated machine-learning approaches. The history of the Covid-19 pandemic has indicated that the results of this study will continue to be relevant in future waves in multiple countries with different variants and in populations with different degrees of immunity from vaccination and/or prior infection.

Competing interests and funding:

This study was supported by the VA Cooperative Studies Program. The VA played no role in the design of the study, data collection, analysis, writing, or submission of the article for publication. Views represented are those of the authors and do not necessarily represent those of the VA or the US Federal Government. WBE and PM report receipt of grant funding from Gilead Sciences during the past 3 years; funds to institution. All other authors report no financial conflicts of interest.

Data sharing:

Patient level data cannot be shared due to privacy laws. Code underlying the data will be made available upon request to the authors.

Transparency statement:

The authors affirm that the manuscript is an honest, accurate and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned have been explained.

Patient consent / ethical approval

This study was approved by the VA Boston Research and Development committee as an exempt study prior to data collection and analysis with a waiver of informed consent.

Andrews N, Stowe J, Kirsebom F, et al. Effectiveness of COVID-19 booster vaccines against COVID-19-related symptoms, hospitalization and death in England. Nature medicine 2022; 28(4): 831–7.
Bar-On YM, Goldberg Y, Mandel M, et al. Protection of BNT162b2 Vaccine Booster against Covid-19 in Israel. The New England journal of medicine 2021; 385(15): 1393–400.
Feikin DR, Higdon MM, Abu-Raddad LJ, et al. Duration of effectiveness of vaccines against SARS-CoV-2 infection and COVID-19 disease: results of a systematic review and meta-regression. Lancet 2022; 399(10328): 924–44.
US Department of Vetearns Affairs. COVID-19 Shared Data Resource. In: VINCI Infrastructure; 2020.
US Centers for Medicare and Medicaid Services. Chronic Conditions Data Warehouse. 2021. https://www2.ccwdata.org/web/guest/condition-categories (accessed 11/30/2021 2021).
Monach P, Branch-Elliman W. From Pandemic to Endemic. 2022.
Yang J, Zheng Y, Gou X, et al. Prevalence of comorbidities and its effects in patients infected with SARS-CoV-2: a systematic review and meta-analysis. Int J Infect Dis 2020; 94: 91–5.
Garg S, Kim L, Whitaker M, et al. Hospitalization rates and characteristics of patients hospitalized with laboratory-confirmed coronavirus disease 2019—COVID-NET, 14 States, March 1–30, 2020. Morbidity and mortality weekly report 2020; 69(15): 458.
Team CC-R, Team CC-R, Team CC-R, et al. Preliminary estimates of the prevalence of selected underlying health conditions among patients with coronavirus disease 2019—United States, February 12–March 28, 2020. Morbidity and Mortality Weekly Report 2020; 69(13): 382–6.
Fillmore NR, Yellapragada SV, Ifeorah C, et al. With equal access, African American patients have superior survival compared to white patients with multiple myeloma: a VA study. Blood 2019; 133(24): 2615–8.
Yek C, Warner S, Wiltz JL, et al. Risk Factors for Severe COVID-19 Outcomes Among Persons Aged >/=18 Years Who Completed a Primary COVID-19 Vaccination Series – 465 Health Care Facilities, United States, December 2020-October 2021. MMWR Morb Mortal Wkly Rep 2022; 71(1): 19–25.
Galmiche S, Luong Nguyen LB, Tartour E, et al. Immunological and clinical efficacy of COVID-19 vaccines in immunocompromised populations: a systematic review. Clin Microbiol Infect 2022; 28(2): 163–77.
Sun J, Zheng Q, Madhira V, et al. Association Between Immune Dysfunction and COVID-19 Breakthrough Infection After SARS-CoV-2 Vaccination in the US. JAMA Intern Med 2022; 182(2): 153–62.
Goldberg Y, Mandel M, Bar-On YM, et al. Waning Immunity after the BNT162b2 Vaccine in Israel. The New England journal of medicine 2021.
Andrews N, Tessier E, Stowe J, et al. Duration of Protection against Mild and Severe Disease by Covid-19 Vaccines. The New England journal of medicine 2022; 386(4): 340–50.
Bhurwal A, Mutneja H, Bansal V, et al. Effectiveness and safety of SARS-CoV-2 vaccine in Inflammatory Bowel Disease patients: A systematic review, meta-analysis and meta-regression. Aliment Pharmacol Ther 2022.
Fillmore NR, La J, Zheng C, et al. The COVID-19 hospitalization metric in the pre-and post-vaccination eras as a measure of pandemic severity: A retrospective, nationwide cohort study. Infection Control & Hospital Epidemiology 2021: 1–24.
US Centers for Disease Control and Prevention. COVID-Net: COVID-19 Associated Hospitalization Surveillance Network. 11/20/2021 2021. https://gis.cdc.gov/grasp/COVIDNet/COVID19_3.html (accessed 12/1/2021 2021).

Table 1

Demographic, clinical, and vaccination-related data for patients with severe and non-severe Covid-19 despite vaccination. Results of multivariable logistic regression are also shown, with statistical significance after adjustment for multiple comparisons noted with asterisks (*). Results for all variables used in the multivariable analysis are shown in eTable 3.
	Total cases	Non-severe	Severe	Multivariable OR (CI)
N	110760	100148	10612
Sex (%)
Male (Referent)	97614 (88.1)	87389 (87.3)	10225 (96.4)	--
Female	13146 (11.9)	22759 (22.7)	387 (3.6)	0.67 (0.60–0.75)*
Age ranges (%)
Less than 40	12703 (11.5)	12600 (12.6)	103 (1.0)	0.57 (0.43–0.76)*
40–45	6317 (5.7)	6226 (6.2)	91 (0.9)	0.94 (0.70–1.25)
45–50 (Referent)	6200 (5.6)	6099 (6.1)	101 (1.0)	--
50–55	9604 (8.7)	9351 (9.3)	253 (2.4)	1.53 (1.21–1.93)*
55–60	10724 (9.7)	10263 (10.2)	461 (4.3)	2.31 (1.85–2.88)*
60–65	12938 (11.7)	12087 (12.1)	851 (8.0)	3.24 (2.62–4.01)*
65–70	12386 (11.2)	11051 (11.0)	1335 (12.6)	4.97 (4.03–6.13)*
70–75	17595 (15.9)	15094 (15.1)	2501 (23.6)	6.74 (5.48–8.30)*
75–80	12585 (11.4)	10441 (10.4)	2144 (20.2)	8.50 (6.90–10.5)*
80 or greater	9708 (8.8)	6936 (6.9)	2772 (26.1)	16.2 (13.1–19.9)*
Period of dominant variant (%)
Pre-delta period (Referent)	2888 (2.6)	2251 (2.2)	637 (6.0)
Delta period	29493 (26.6)	24762 (24.7)	4731 (44.6)	0.95 (0.84–1.09)
Omicron period	78379 (70.8)	73135 (73.0)	5244 (49.4)	0.49 (0.43–0.57)*
History of infection before vaccinated (%)	4373 (3.9)	3959 (4.0)	414 (3.9)	0.69 (0.61–0.78)*
Months since full vax at breakthrough (%)
4 or less (Referent)	12497 (11.3)	10983 (11.0)	1514 (14.3)	--
4–5	7216 (6.5)	6355 (6.3)	861 (8.1)	1.06 (0.95–1.19)
5–6	8148 (7.4)	7039 (7.0)	1109 (10.5)	1.02 (0.91–1.13)
6–7	8902 (8.0)	7763 (7.8)	1139 (10.7)	1.10 (0.99–1.22)
7–8	11739 (10.6)	10665 (10.6)	1074 (10.1)	1.15 (1.03–1.28)*
8–9	19118 (17.3)	17873 (17.8)	1245 (11.7)	1.16 (1.04–1.30)*
9–10	19997 (18.1)	18570 (18.5)	1427 (13.4)	1.18 (1.05–1.32)*
10–11	15731 (14.2)	14252 (14.2)	1479 (13.9)	1.37 (1.21–1.54)*
11–12	6519 (5.9)	5853 (5.8)	666 (6.3)	1.50 (1.31–1.73)*
12 or more	893 (0.8)	795 (0.8)	98 (0.9)	1.59 (1.23–2.06)*
Months since boosted at breakthrough (%)
Not boosted (Referent)	81039 (73.2)	72483 (72.4)	8556 (80.6)	--
1 or less	5755 (5.2)	5401 (5.4)	354 (3.3)	0.56 (0.49–0.63)*
1–2	7732 (7.0)	7341 (7.3)	391 (3.7)	0.43 (0.38–0.48)*
2–3	8550 (7.7)	7996 (8.0)	554 (5.2)	0.46 (0.42–0.51)*
3–4	4633 (4.2)	4197 (4.2)	436 (4.1)	0.56 (0.50–0.63)*
4–5	1749 (1.6)	1543 (1.5)	206 (1.9)	0.63 (0.54–0.75)*
5–6	370 (0.3)	320 (0.3)	50 (0.5)	0.71 (0.51–1.0)
6 or more	932 (0.8)	867 (0.9)	65 (0.6)	0.78 (0.59–1.02)
IS medications before breakthrough (%)
Before breakthrough Chemotherapy	1006 (0.9)	696 (0.7)	310 (2.9)	2.69 (2.25–3.21)*
Before breakthrough Cytokine-blocking	1601 (1.4)	1401 (1.4)	200 (1.9)	1.73 (1.37–2.18)*
Before breakthrough Glucocorticoids	7544 (6.8)	5723 (5.7)	1821 (17.2)	2.41 (2.25–2.58)*
Before breakthrough Leukocyte-inhibitory	1458 (1.3)	1163 (1.2)	295 (2.8)	2.44 (1.98–2.99)*
Before breakthrough Lymphocyte-depleting	585 (0.5)	406 (0.4)	179 (1.7)	2.12 (1.61–2.79)*
IS medications before initial vaccination (%)
Before vaccine Chemotherapy	430 (0.4)	316 (0.3)	114 (1.1)	0.85 (0.64–1.14)
Before vaccine Cytokine-blocking	1211 (1.1)	1096 (1.1)	115 (1.1)	0.62 (0.46–0.83)*
Before vaccine Glucocorticoids	2754 (2.5)	2124 (2.1)	630 (5.9)	1.45 (1.29–1.62)*
Before vaccine Leukocyte-inhibitory	1110 (1.0)	942 (0.9)	168 (1.6)	0.61 (0.47–0.79)*
Before vaccine Lymphocyte-depleting	223 (0.2)	153 (0.2)	70 (0.7)	1.09 (0.70–1.67)
BMI Class (%)
Underweight	677 (0.6)	496 (0.5)	181 (1.7)	1.51 (1.23–1.86)*
Normal (Referent)	14394 (13.0)	12299 (12.3)	2095 (19.7)	--
Overweight	32707 (29.5)	29792 (29.7)	2915 (27.5)	0.71 (0.66–0.76)*
Obesity I	31156 (28.1)	28557 (28.5)	2599 (24.5)	0.77 (0.72–0.83)*
Obesity II	16862 (15.2)	15335 (15.3)	1527 (14.4)	0.92 (0.85–1.00)
Severe Obesity	10641 (9.6)	9510 (9.5)	1131 (10.7)	1.21 (1.10–1.32)*
Unknown	4323 (3.9)	4159 (4.2)	164 (1.5)	0.88 (0.73–1.05)
Comorbidities (%)
Alzheimers disease and related disorders or senile dementia	3050 (2.8)	1915 (1.9)	1135 (10.7)	2.01 (1.84–2.21)*
Chronic kidney disease	11832 (10.7)	9071 (9.1)	2761 (26.0)	1.61 (1.51–1.72)*
Chronic obstructive pulmonary disease and bronchiectasis	8337 (7.5)	6103 (6.1)	2234 (21.1)	1.63 (1.53–1.74)*
Diabetes	26083 (23.5)	21919 (21.9)	4164 (39.2)	1.25 (1.19–1.32)*
Heart failure	5444 (4.9)	3681 (3.7)	1763 (16.6)	1.73 (1.60–1.88)*
Leukemias and lymphomas	1336 (1.2)	993 (1.0)	343 (3.2)	1.84 (1.59–2.14)*
Lung cancer	824 (0.7)	573 (0.6)	251 (2.4)	1.59 (1.34–1.89)*
Mobility impairments	1030 (0.9)	728 (0.7)	302 (2.8)	1.90 (1.62–2.24)*
Multiple sclerosis and transverse myelitis	454 (0.4)	362 (0.4)	92 (0.9)	2.83 (2.14–3.73)*
Pressure and chronic ulcers	1347 (1.2)	872 (0.9)	475 (4.5)	1.58 (1.37–1.81)*
Schizophrenia and other psychotic disorders	2755 (2.5)	2320 (2.3)	435 (4.1)	1.69 (1.49–1.91)*
BMI = body mass index (for criteria, see Supplementary File 1). IS = immune-suppressive. OR = odds ratio. CI = 95% confidence interval.

Table 2

Demographic, clinical, and vaccination-related data, and results of multivariable logistic regression for immunocompromised patients with severe and non-severe Covid-19 despite vaccination. Statistical significance after adjustment for multiple comparisons is noted with asterisks (*). Results for all variables used in the analysis are shown in eTable 6.
	Total cases	Non-severe	Severe	Multivariable OR (CI)
N	13768	10888	2880
Sex
Male (Referent)	12274 (89.1)	9495 (87.2)	2779 (96.5)	--
Female	1494 (10.9)	1393 (12.8)	101 (3.5)	0.58 (0.46–0.73)*
Age ranges (%)
Less than 40	841 (6.1)	822 (7.5)	19 (0.7)	0.67 (0.36–1.26)
40–45	517 (3.8)	500 (4.6)	17 (0.6)	0.92 (0.48–1.77)
45–50 (Referent)	629 (4.6)	606 (5.6)	23 (0.8)	--
50–55	979 (7.1)	908 (8.3)	71 (2.5)	1.86 (1.14–3.04)*
55–60	1201 (8.7)	1073 (9.9)	128 (4.4)	2.41 (1.51–3.84)*
60–65	1705 (12.4)	1451 (13.3)	254 (8.8)	3.33 (2.12–5.22)*
65–70	1837 (13.3)	1428 (13.1)	409 (14.2)	4.83 (3.09–7.55)*
70–75	2885 (21.0)	2075 (19.1)	810 (28.1)	6.62 (4.25–10.3)*
75–80	1811 (13.2)	1253 (11.5)	558 (19.4)	7.53 (4.81–11.8)*
80 or greater	1363 (9.9)	772 (7.1)	591 (20.5)	12.2 (7.71–19.1)*
Period of dominant variant (Pre-Delta, Referent)
Delta period	4353 (31.6)	3104 (28.5)	1249 (43.4)	0.73 (0.56–0.95)*
Omicron period	8928 (64.8)	7494 (68.8)	1434 (49.8)	0.43 (0.32–0.57)*
History of infection before vaccinated (%)	629 (4.6)	508 (4.7)	121 (4.2)	0.63 (0.49–0.80)*
Months since full vax at breakthrough (%)
4 or less (Referent)	1647 (12.0)	1209 (11.1)	438 (15.2)	--
4–5	956 (6.9)	729 (6.7)	227 (7.9)	1.08 (0.86–1.37)
5–6	1145 (8.3)	824 (7.6)	321 (11.1)	1.13 (0.90–1.40)
6–7	1125 (8.2)	836 (7.7)	289 (10.0)	1.03 (0.83–1.29)
7–8	1315 (9.6)	1067 (9.8)	248 (8.6)	0.92 (0.73–1.16)
8–9	2040 (14.8)	1738 (16.0)	302 (10.5)	0.98 (0.78–1.23)
9–10	2436 (17.7)	2032 (18.7)	404 (14.0)	1.11 (0.88–1.41)
10–11	2104 (15.3)	1669 (15.3)	435 (15.1)	1.32 (1.03–1.69)
11–12	894 (6.5)	698 (6.4)	196 (6.8)	1.36 (1.02–1.82)
12 or more	106 (0.8)	86 (0.8)	20 (0.7)	0.95 (0.53–1.69)
Months since boosted at breakthrough (%)
Not boosted (Referent)	9291 (67.5)	7161 (65.8)	2130 (74.0)	--
1 or less	650 (4.7)	530 (4.9)	120 (4.2)	0.80 (0.64–1.01)
1–2	864 (6.3)	724 (6.6)	140 (4.9)	0.64 (0.51–0.79)*
2–3	1085 (7.9)	890 (8.2)	195 (6.8)	0.68 (0.56–0.83)*
3–4	995 (7.2)	851 (7.8)	144 (5.0)	0.52 (0.41–0.64)*
4–5	635 (4.6)	534 (4.9)	101 (3.5)	0.59 (0.45–0.76)*
5–6	135 (1.0)	104 (1.0)	31 (1.1)	0.86 (0.54–1.37)
6 or more	113 (0.8)	94 (0.9)	19 (0.7)	0.84 (0.49–1.45)
IS medications before breakthrough
Before breakthrough Chemotherapy (%)	1006 (7.3)	696 (6.4)	310 (10.8)	2.19 (1.81–2.65)*
Before breakthrough Cytokine-blocking (%)	1601 (11.6)	1401 (12.9)	200 (6.9)	1.47 (1.16–1.86)*
Before breakthrough Glucocorticoids (%)	7544 (54.8)	5723 (52.6)	1821 (63.2)	1.98 (1.75–2.24)*
Before breakthrough Leukocyte-inhibitory (%)	1458 (10.6)	1163 (10.7)	295 (10.2)	2.17 (1.76–2.67)*
Before breakthrough Lymphocyte-depleting (%)	585 (4.2)	406 (3.7)	179 (6.2)	2.00 (1.52–2.63)*
IS medications before initial vaccination
Before vaccine Chemotherapy (%)	430 (3.1)	316 (2.9)	114 (4.0)	0.86 (0.65–1.13)
Before vaccine Cytokine-blocking (%)	1211 (8.8)	1096 (10.1)	115 (4.0)	0.62 (0.46–0.82)*
Before vaccine Glucocorticoids (%)	2754 (20.0)	2124 (19.5)	630 (21.9)	1.28 (1.13–1.46)*
Before vaccine Leukocyte-inhibitory (%)	1110 (8.1)	942 (8.7)	168 (5.8)	0.60 (0.47–0.77)*
Before vaccine Lymphocyte-depleting (%)	223 (1.6)	153 (1.4)	70 (2.4)	1.13 (0.74–1.72)
BMI Class (%)
Underweight	116 (0.8)	58 (0.5)	58 (2.0)	1.65 (1.09–2.51)*
Normal (Referent)	1952 (14.2)	1345 (12.4)	607 (21.1)	--
Overweight	4008 (29.1)	3216 (29.5)	792 (27.5)	0.65 (0.57–0.75)*
Obesity I	3808 (27.7)	3103 (28.5)	705 (24.5)	0.69 (0.60–0.80)*
Obesity II	2167 (15.7)	1772 (16.3)	395 (13.7)	0.75 (0.64–0.89)*
Severe Obesity	1550 (11.3)	1241 (11.4)	309 (10.7)	0.92 (0.76–1.11)
Unknown	167 (1.2)	153 (1.4)	14 (0.5)	0.56 (0.31–1.02)
Comorbidities
Alzheimers disease and related disorders or senile dementia	447 (3.2)	225 (2.1)	222 (7.7)	1.54 (1.23–1.92)*
Chronic kidney disease	2174 (15.8)	1369 (12.6)	805 (28.0)	1.58 (1.39–1.80)*
Chronic obstructive pulmonary disease and bronchiectasis	2475 (18.0)	1531 (14.1)	944 (32.8)	1.68 (1.50–1.89)*
Diabetes	3814 (27.7)	2723 (25.0)	1091 (37.9)	1.23 (1.10–1.38)*
Heart failure	1172 (8.5)	655 (6.0)	517 (18.0)	1.49 (1.27–1.75)*
Leukemias and lymphomas	1336 (9.7)	993 (9.1)	343 (11.9)	1.56 (1.31–1.86)*
Lung cancer	286 (2.1)	171 (1.6)	115 (4.0)	1.36 (1.04–1.79)*
Mobility impairments	164 (1.2)	93 (0.9)	71 (2.5)	1.72 (1.18–2.49)*
Multiple sclerosis and transverse myelitis	207 (1.5)	166 (1.5)	41 (1.4)	2.07 (1.34–3.19)*
Pressure and chronic ulcers	227 (1.6)	113 (1.0)	114 (4.0)	1.68 (1.23–2.30)*
Schizophrenia and other psychotic disorders	322 (2.3)	231 (2.1)	91 (3.2)	1.46 (1.09–1.95)*
BMI = body mass index (for criteria, see Supplementary File 1). IS = immune-suppressive. OR = odds ratio. CI = 95% confidence interval.

RiskFactorsAllSuppsPDF.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Factors Associated with Severe Covid-19 Despite Vaccination: A Nationwide, Retrospective Cohort Study

Status:

Version 1

Abstract

Figures

Key Points

Introduction

Methods

Cohort creation:

Data sources and uses:

Definitions of immunocompromised patients:

Time since Vaccination

Statistical Analysis:

Results

Full Cohort:

Whole Cohort:

Patients < 50:

Immunocompromised Cohort:

Impact of Time Since Vaccination or Booster:

Discussion

Declarations

Competing interests and funding:

Data sharing:

Transparency statement:

Patient consent / ethical approval

References

Tables

Supplementary Files

Status:

Version 1