key: cord-1033176-3hf4uo6w
authors: Khan, M.; Khan, M. N.; Mustagir, M. G.; Rana, J.; Islam, M. S.; Kabir, M. I.
title: Effects of pre-existing morbidities on occurrence of death among COVID-19 disease patients: A systematic review and meta-analysis
date: 2020-05-13
journal: nan
DOI: 10.1101/2020.05.08.20095968
sha: a06b3af8f9e0d5446f552e059f7e737536729e37
doc_id: 1033176
cord_uid: 3hf4uo6w

Abstract Background: Coronavirus disease 2019 (COVID-19), most hectic pandemic of the era, is increasing exponentially and taking thousands of lives, worldwide. Objective: This study aimed to assess the prevalence of pre-existing morbidities among COVID-19 infected patients and their mortality risks against each type of pre-existing morbidity categories. Design: Systematic review and meta-analysis Data sources: Medline, Web of Science, Scopus, and CINAHL databases were searched using the following keywords: (COVID-19 or 2019-nCoV or Coronavirus or SARS-CoV-2) AND (Comorbidity or Morbidity) AND (Mortality or Death or Died) up to May 01, 2020. Further searches were conducted using the reference list of the selected studies, renowned pre-print servers (e.g., medRxiv, bioRixv, SSRN), and relevant journal websites. Eligibility criteria: Studies written in the English language included if those were conducted among COVID-19 patients with and without comorbidities and presented survivor vs. non-survivor counts or hazard/odds of deaths or survivors against types of pre-existing morbidities. Methods: Comorbidities reported in the selected studies were grouped into eight categories. The pooled likelihoods of deaths in each category were estimated using a fixed or random-effect model, based on the heterogeneity assessment. Publication bias was assessed by visual inspection of the funnel plot asymmetry and Egger regression test. Trim and Fill method was used if there any publication bias was found. Results: A total of 42 studies included in this study comprised of 39,398 samples. The most common pre-existing morbidities in COVID-19 infected patients were hypertension (36.5%), cardiovascular disease (11.9%), and diabetes (22.0%). The higher likelihood of deaths was found among COVID-19 patients who had pre-existing cardiovascular system diseases (OR: 3.32, 95% CI: 2.79-3.95), immune and metabolic disorders (OR: 2.39, 95% CI: 2.00-2.85), respiratory diseases (OR: 2.02, 95% CI: 1.80-2.26), cerebrovascular system diseases (OR: 4.12, 95% CI: 3.04-5.58), any types of cancers (OR: 2.22, 95% CI: 1.63-3.03), renal (OR: 3.02, 95% CI: 2.60-3.52), and liver system diseases (OR: 1.44, 95% CI: 1.21-1.71). Conclusions: This study provides evidence of a higher likelihood of deaths among COVID-19 patients against morbidity categories. These findings could potentially help healthcare providers to sort out the most endangered COVID-19 patients by comorbidities, take precautionary measures during hospitalization, assess susceptibility to death, and prioritize their treatment, which could potentially reduce the number of fatalities in COVID-19 disease.

Introduction among patients with one or more morbidities could be potential ways to combat its adverse outcomes and severities. Thus, we need to identify possible morbidities that are potentially increasing the risks of mortality, which are still lacking. Studies conducted among patients are highly varied with reported morbidities and the likelihood of mortality [8, 10, 20, 21] . To address these gaps, this study was conducted with two primary aims: (i) to summarize pre-existing morbidities in patients with a secondary disease, COVID-19 and (ii) to estimate the likelihood of mortality from COVID-19 against each category of pre-existing morbidities.

The study findings could help healthcare providers to take appropriate measures to control fatalities from this pandemic.

This systematic review and meta-analysis was conducted by following the Preferred Reporting for Systematic Review and Meta-Analysis (PRISMA) consensus statement [22] .

Studies relevant to the COVID-19 disease among people with pre-existing one or more morbidities were included.

Four databases: Medline, Web of Science, Scopus, and CINAHL were searched, concluded on May 01, 2020, using pre-specified search strategies for each database. The search strategy consists of keywords on COVID-19 disease (COVID-19, 2019-nCoV, Coronavirus, SARS-CoV-2), pre-existing morbidity (comorbidity, morbidity), and patients' survival status (mortality, death, died) combined using the Boolean operators (AND, OR). Details of the search strategies are presented in the supplementary tables (Table 1 -4) . Additional searches were conducted using the reference list of the selected studies, relevant journal websites, and renowned pre-print servers (medRxiv, bioRxiv).

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . https://doi.org/10.1101/2020.05.08.20095968 doi: medRxiv preprint

All peer-reviewed and pre-print (not-peer-reviewed) studies met the pre-specified inclusion, and exclusion criteria were included in this study.

Studies met the following inclusion criteria were included if: (i) conducted for the patients infected with COVID-19 with or without pre-existing morbidities, (ii) presented survivor and non-survivor counts following COVID-19 disease among patients with or without preexisting morbidity or presented hazard/risk/odds ratio of deaths or survival following COVID-19 against the types of morbidities, and (iii) published in the English language. Studies without complete information but met our inclusion criteria were included in the narrative review.

Studies excluded if COVID-19 was reported among pregnant women or children (aged <18 years) and written in languages other than English. We also excluded review papers, correspondence, viewpoints, editorials, commentaries, and studies where no information related to the previous morbidity was reported.

A data extraction form was designed, trialled, and modified to extract information from the selected studies. Two authors (MMAK and MGM) used the pre-designed form to extract information independently. The following information was extracted: study location, design, sample size, study population characteristics (e.g., age, gender), and survivor vs. non-survivor counts among COVID-19 patients with or without specific morbidity. If available, the odds/risk/hazard ratio of deaths among COVID-19 patients with comorbidities were extracted . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . https://doi.org/10.1101/2020.05.08.20095968 doi: medRxiv preprint against the types of morbidity. Disagreements reported in data extraction were reviewed and solved by the corresponding and senior authors (MNK and MIK). The modified Newcastle-Ottawa scale, as part of the data extraction strategy, was used to assess the quality of selected studies.

Pre-existing one or more morbidities among COVID-19 disease patients reported in the selected studies were grouped into eight broad categories based on the type of morbidities. These were cardiovascular system diseases (hypertension, cardiovascular disease, arrhythmia, heart failure), immune and metabolic disorders (diabetes, immunosuppression, autoimmune disease, immunodeficiency, metabolic disorder), respiratory system diseases (chronic lung diseases, Chronic Obstructive Pulmonary Disease (COPD), acute respiratory distress syndrome, tuberculosis, etc.), cancer (malignancy, cancer, and tumor), cerebrovascular diseases (cerebrovascular disease, peripheral vascular disease), renal system diseases (chronic kidney disease, urinary disease), liver system diseases (chronic liver disease, cirrhosis, hyperlipidemia, Hepatitis B, etc.), and gastrointestinal system diseases (chronic digestive disorder, gastrointestinal disease). The odds ratios (ORs) of deaths with 95% confidence interval (95% CI) for the people exposed to a particular category of morbidity as compared to people unexposed to any specific morbidity was estimated from the extracted raw data or reported ORs. We first used the Haldane correction (add constant 0.5 to each cell) for the studies in which the sample included in the exposed or unexposed group was zero (such as all exposed patients died or vice versa) [23-25]. We then used either a fixed effect or random effect model to estimate ORs, selected based on heterogeneity assessment. When the test of heterogeneity (I 2 statistics) was moderate (50-74%) or high (≥75%), the pooled estimates of ORs were computed using the random-effects model [26] . Subgroup and meta-regression analyses were conducted for the groups where moderate or higher heterogeneity was . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 13, 2020. 

A total of 247 articles were identified from the databases searched, and the additional 15 articles were identified by checking the reference list of the selected articles and the selected journal's websites ( Figure 1 ). Around 1273 articles were also initially identified from the aforementioned pre-prints servers. Of the selected articles, 1341 articles were excluded after screening titles and abstracts, leaving 114 articles for full-text review for possible inclusion in this study. Of these, 55 articles were excluded based on the inclusion and exclusion criteria for the study sample (e.g., excluded pregnant or children), and 11 articles were excluded for study types (e.g review papers, correspondence, viewpoints, editorials, commentaries), and six articles were excluded for fully incomplete data. A total of 42 articles were finally selected for this study; 36 articles were included in the meta-analysis, and the remaining six articles were synthesized narratively.

A summary of the 42 selected articles is represented in Table 1 . A total of 23 of the selected 42 articles were published in peer-reviewed journals, and 18 articles were published in preprint servers. One of the selected studies was a national report for Australia. The majority of these studies were retrospective in nature (26) along with 7 prospective studies. The sleeted . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 13, 2020. 

Distribution of the type of morbidity presented in Table 2 . Approximately 36.5% of the total COVID-19 disease patients reported that they had hypertension, 22.0% had diabetes, 11.9% had cardiovascular disease, 4.1% had chronic lung disease, 2.3% had COPD, 11.0 % had hyperlipidemia, and 3.0% had chronic kidney disease.

The pooled ORs of deaths for each category of preexisting morbidities among COVID-19 disease patients, publication bias, and Trim and Fill estimates are presented in Table 3 .

COVID-19 disease patients with preexisting cardiovascular system disease were 3.32 times more likely to die (OR: 3.32, 95% CI: 2.79-3.95; I 2 = 83.8%) than the patients who had no cardiovascular system diseases. The odds of death among COVID-19 disease patients with immune and metabolic disorders were also found to be 239% higher (OR: 2.39, 95% CI: 2.00-2.85; I 2 = 64.5%) than among COVID-19 patients without such disorders. The incidence of COVID-19 disease among people with respiratory system disease increases mortality risk around two times (OR: 2.02, 95% CI: 1.80-2.26; I 2 = 71.2%) than COVID-19 disease patients . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 13, 2020. .

without respiratory system diseases. Similarly, we found higher mortality risk among COVID-19 disease patients who had pre-existing any types of cancers (OR: 2.22, 95% CI:

1.63-3.03, I 2 = 67.7%) and cerebrovascular system diseases (OR: 4.12, 95% CI: 3.04-5.58) than their counterparts. Moreover, the incidence of COVID-19 disease among patients with preexisting renal system disease and chronic liver disease increased mortality risk by about three times (OR: 3.02, 95% CI: 2.60-3.51) and one and half times (OR: 1.44, 95% CI: 1.21-1.71), respectively compared to the COVID-19 disease patients who did not mention such comorbidities.

We found evidence of publication bias for the three categories of preexisting morbidities: any type of cancer, cerebrovascular diseases, and liver system diseases ( Figure 1a to 8a). We then used Trim and Fill methods to impute the number of missing studies, which hypothetically imputed two studies for cardiovascular system diseases, three studies for renal system diseases, and three studies for liver system diseases. The pooled analysis, including these missing studies, showed almost similar results to the summary estimates presented earlier without these missing studies.

Evidence of higher deaths among COVID-19 disease patients with preexisting one or more morbidities were also demonstrated in the narrative review (Supplementary Table 5 ). In two of the three articles reviewed in this study, researchers reported that each of the patients who died following COVID-19 disease had preexisting morbidities, mostly had any types of cardiovascular system diseases and immune and metabolic disorders [29, 30] . Researchers in one study found around 38% of the COVID-19 disease patients with hypertension died [33] and one study reported higher odds for deaths in kidney injury [34] .

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. 

This study aimed to summarize preexisting morbidities among COVID-19 disease patients, which increases their incidence of deaths and their corresponding likelihoods. A total of 42 studies were included that comprised 36,398 samples, and 7,558 (42.5%) of them had preexisting morbidities. The most frequently reported morbidities were hypertension (36.5%), diabetes (22.0%), and cardiovascular disease (11.9%). The likelihood of death was higher among COVID-19 patients who had comorbidities like cardiovascular and cerebrovascular diseases, respiratory diseases, renal diseases, immune and metabolic disorders, hepatic diseases, and cancer. This evidence will guide physicians to take precautionary measures, which could reduce the number of fatalities following secondary infection with COVID-19.

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . Among the total positive COVID-19 disease cases included in this systematic review study, around 43% had preexisting one or more morbidities, mostly cardiovascular system diseases and immune and metabolic disorders. Importantly, patients with these diseases are more likely to have a higher neutrophil-lymphocyte ratio [ is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . adverse consequences on the human body, could, therefore, boost serve clinical events as well as deaths among patients with these pre-existing morbidities.

This study also suggests that patients with cerebrovascular, liver and renal diseases are more vulnerable to mortality following the second incidence of COVID-19 disease than the patient 

This study has several strengths and limitations that should be reported. To our knowledge, this is the first of its kind that summarizes all morbidities among COVID-19 disease patients that lead to death. Moreover, morbidities reported among COVID-19 disease patients were classified into board groups based on their characteristics, and the likelihood of death was estimated separately for each group. This evidence informs healthcare providers about the risk of death among COVID-19 disease patients with different groups of pre-existing morbidities. Thus, they will be able to take precautionary measures early targeting to prevent deaths. However, this study reported the odds of death for COVID-19 disease patients with . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . one preexisting morbidity only. Many COVID-19 disease patients may have multimorbidities (COVID-19 disease with pre-existing two or more morbidities) and a higher risk of death. However, the studies included in this review considered each morbidity separately;

for instance, if COVID-19 patients had both hypertension and diabetes, they were included in both groups. None of the included studies considered COVID-19 disease with two or more morbidities together; therefore, we failed to provide the likelihood of deaths for COVID-19 disease patients with two or more preexisting morbidities. Moreover, the likelihoods presented in this study were mostly unadjusted (31 of the 36 articles included) calculated from the extracted raw data. This may overestimate or underestimate the actual likelihood of deaths in COVID-19 disease patients because age and other socio-demographic characteristics are potential confounders of their deaths, which should be adjusted for getting unbiased estimates. Despite these limitations, this study is unique and beneficial for healthcare providers to handle COVID-19 disease patients with preexisting morbidities.

About 46% of the sample included in this systematic review had one or more preexisting morbidities and get COVID-19 as a secondary infection. The most common preexisting morbidities were hypertension, diabetes, and cardiovascular disease. The likelihood of death was higher among COVID-19 disease patients who had pre-existing cardiovascular and cerebrovascular system diseases, respiratory system diseases, renal diseases, immune and metabolic disorders, liver diseases, and any types of cancer. These findings will help healthcare providers to sort COVID-19 patients by comorbidities, take precautionary measures during hospital admission, assess susceptibility to death, and prioritize their treatment. These could potentially reduce the number of fatalities from secondary infection with COVID-19 disease.

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. 

The authors have no competing interests to declare.

The authors have received no specific funds for this study.

Data related to this study are available upon request to the corresponding author.

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. .

Yang, J., et al. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. .

Egger . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. .

Chen, N., et al. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. .

Lau . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. .

Yao . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . .

It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 13, 2020. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 13, 2020. . https://doi.org/10.1101/2020.05.08.20095968 doi: medRxiv preprint Full-text articles assessed for eligibility, (n = 42)

• 23 peer-reviewed articles, 18 preprints, and 1 national report • 26 studies were retrospective cohort • 7 studies were prospective cohort • 1 study were cross-sectional study Full-text articles excluded, with reasons: • Wrong study design (n=11) • Incomplete data (n=6) Preprints were selected through searching the preprint database of medRxiv, bioRxiv, and SSRN (n = 1273)

Studies included in quantitative synthesis (meta-analysis) (n = 36)

Studies included in qualitative synthesis (n = 6)

Records excluded due to those had different outcomes (n = 55)

. CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 13, 2020. .

A Systematic Review of COVID-19 Epidemiology Based on Current Evidence

A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster

Novel Coronavirus (COVID-19) Situation

COVID-19 Coronavirus Pandemic (Real time database)

Non-pharmaceutical public health measures for mitigating the risk and impact of epidemic and pandemic influenza: annex: report of systematic literature reviews

Impact assessment of non-pharmaceutical interventions against coronavirus disease 2019 and influenza in Hong Kong: an observational study

Comorbidity and its impact on 1590 patients with Covid-19 in China: A Nationwide Analysis

Cardiovascular Disease, Drug Therapy, and Mortality in Covid

Estimation of risk factors for COVID-19 mortality -preliminary results

Identification of a novel coronavirus causing severe pneumonia in human: a descriptive study

Predictors of Mortality for Patients with COVID-19 Pneumonia Caused by SARS-CoV-2: A Prospective Cohort Study

Coronavirus disease 2019 in elderly patients: Characteristics and prognostic factors based on 4-week follow-up

Influence factors of death risk among COVID-19 patients in Wuhan, China: a hospital-based case-cohort study

Clinical Characteristics of Hospitalized Covid-19 Patients in New York City. medRxiv

Epidemiology, clinical course, and outcomes of critically ill adults with COVID-19 in New York City: a prospective cohort study. medRxiv

Clinical Features Predicting Mortality Risk in Older Patients with COVID-19

Dose-Response Association between Risk Factors and Incidence of COVID-19 in 325 Hospitalized Patients: A Multicenter Retrospective Cohort Study

Critically Ill Patients with COVID-19 in China: A Multicenter Retrospective Observational Study

Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement

The authors are grateful to the authors of the paper included in this review.

Patients with more than one comorbidity were missing. Calculated in column percentage format; One study (Caramelo et al.) was excluded from prevalence calculation as the study had not reported frequency of comorbidities; Malnutrition and dementia was skipped from the analysis as we found only one patient (Yang et al.); a Included all types of cardiovascular diseases like coronary heart diseases or artery disease; b Included Cerebral infarction; c anemia, bowel disease, tissue disease, etc.