Severity of COVID-19 Infection in ACEI/ARB Users in Two Saudi Public Specialty Hospitals: Retrospective Cohort Study

Background: The uncertainty about COVID-19 outcomes in angiotensin converting enzyme inhibitors (ACEI)/angiotensin (ARB) users continues with contradictory ndings. The aim of this study was to determine the effect of ACEI/ARB use in patients with severe COVID-19. Methods: This retrospective cohort study was done in two Saudi public specialty hospitals designated as COVID-19 referral facilities. We included 354 patients with conrmed diagnosis of COVID-19 between April and June 2020, of which 146 were ACEI/ARB users and 208 were non-ACEI/ARB users. Controlling for confounders, we conducted a multivariate logistic regression and a sensitivity analyses using propensity score matching (PSM) and Inverse propensity score weighting (IPSW) for high risk patient subsets. Results: Compared to non-ACEI/ARB users, ACEI/ARB users had an eight-fold higher risk of developing critical or severe COVID-19 (OR=8.25, 95%CI=3.32-20.53); a nearly 7-fold higher risk of intensive care unit (ICU) admission (OR=6.76, 95%CI=2.88-15.89) and a nearly 5-fold higher risk of requiring noninvasive ventilation (OR=4.77,95%CI=2.15-10.55). Patients with diabetes, hypertension, and/or renal disease had a ve-fold higher risk of severe COVID-19 disease (OR=5.40,95%CI=2.0-14.54]. These results were conrmed in the PSM and IPSW analyses. Conclusion: In general, but especially among patients with hypertension, diabetes, and/or renal disease, ACEI/ARB use is associated with a signicantly higher risk of severe or critical COVID-19 disease, and ICU care. or critical according to classication and comorbidities hypertension renal disease); both diabetes and hypertension; diabetes only; and (4) hypertension


Introduction
Coronavirus disease 2019 (COVID-19) is pandemic disease caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [1]. It emerged in Wuhan, the largest city in Hubei Province, China and rapidly spread to become a worldwide pandemic [2]. The World Health Organization (WHO) reported as of 31 August 2020 more than 25 million con rmed cases of COVID-19 and more than 800,000 death cases globally [3]. The virus is transmitted through respiratory droplets or aerosols and direct contact. The disease is characterized by a wide range of symptoms, and patients may experience mild, severe, or critical illness [4]. The overall mortality rate in various countries ranges from 2% to 13% [5].
Risk factors associated with the clinical severity of COVID-19 include age (especially >75 years old), obesity, selected medications, and comorbidities such as hypertension, diabetes, and cardiovascular disease [6][7][8][9][10]. Noteworthy is the frequent presence of angiotensin-converting enzyme inhibitors (ACEI) or angiotensin II type-I receptor blockers (ARB) in these patients' pharmacological treatment. It has been hypothesized that exposing patients to ACEI/ARB may increase the expression of ACE2 by a negative feedback mechanism, thus facilitating the binding process of coronavirus to epithelial cells of the kidney, intestine, lung, and blood vessels [11] and increasing the risk of developing severe and fatal COVID-19 disease [12][13][14]. Clinical research evidence to date has been equivocal, especially in terms of identifying differentiating factors that may mediate the ACEI/ARB and COVID-19 severity relationship [15][16][17].
In the case of Saudi Arabia, we have a young developing country with a median age of 30.8 years burdened with comorbid diseases [18]. The Saudi Arabia's healthcare system is mainly funded by the public sector through the Ministry of Health (MOH) which serves 23 million people. Though at different levels, critical care services are provided in primary, secondary, and tertiary/specialty hospitals; with complex cases being referred to tertiary/specialty hospitals that are well equipped with specialized intensive care units (ICU) and highly trained intensivists [19]. Regarding comorbid disease in the Saudi general population and in spite of extensive healthcare resources, the WHO ranks Saudi Arabia 7 th in the world in terms of rates of diabetes with an estimated 7 million people affected, followed by hypertension (26.1%), and rapidly emerging because of its association with both conditions, chronic kidney disease (6%) [20][21][22]. Studies link the development of these diseases to ACE insertion/deletion (I/D) polymorphism in certain populations including Saudis [23][24][25][26].
The uncertainty about COVID-19 outcomes in ACE/ARB users may be a function of how patients with diabetes, hypertension, and/or chronic renal disease are clinically compromised and therefore more susceptible to severe COVID-19 disease. A systematic review concluded that patients with diabetes, hypertension, and chronic renal disease had an increased mortality risk in the setting of Middle Eastern respiratory syndrome (MERS) coronavirus infection (a coronavirus closely related to SARS-CoV-2 that originated in Saudi Arabia in 2012) [27]. We report here on a retrospective real-world clinical study in two public tertiary/specialty hospitals to which severe COVID-19 cases were referred to assess the association of ACEI/ARB and clinical severity of COVID-19. Severity of COVID-19 disease in ACE/ARB users was based on the following variables: ICU admission, non-invasive and invasive mechanical ventilation, in-hospital mortality, and hospital length of stay. In addition to outcomes observed in the overall population of referred patients, we also evaluated severity outcomes for high risk patients with diabetes, hypertension and/or renal disease.

Study Design and Study Setting
This retrospective was conducted in the tertiary specialty referral hospitals known as King Fahad Medical City (KFMC; 1200 beds) and Prince Mohammed Bin Abdulaziz hospital (PMAH; 500 beds), both in Riyadh, Saudi Arabia. The Institutional Review Boards at KFMC and PMAH (IRB 20-200) approved the study. All methods were performed in accordance with the relevant guidelines and regulations. Institutional review boards of both KFMC and PMAH waived informed consent since it was an exempt study conducting a retrospective analysis.

Participant selection
Lists of discharged or dead COVID-19 patients were obtained from the health informatics o cer and categorized into ACEI/ARB users and ACEI/ARB nonusers based on their medication history. Inclusion criteria were age >18 years admitted at KFMC or PMAH from April to June 2020; being treated with ACE/ ARB inhibitors 6 months before and continued during hospital admission and after discharge for any indications such as hypertension, stroke, heart failure, myocardial infarction, diabetes mellitus, chronic kidney disease or nephrotic syndrome (criteria only for ACEI/ARB users); and SARS-CoV-2 infection con rmed by real-time polymerase chain reaction (RT-PCR) from nasopharyngeal swab for inclusion. Patients were excluded if pregnant, incomplete medical records due to patients being transferred recently from other hospitals, or unknown medications history.

Data Collection
Data were extracted manually from electronic health records (Corttex system and HIM system in KFMC, and Cerner Systems in PMAH) by a trained team and included demographic and anthropometric variables such as age, height, weight, body mass index (BMI), class of obesity; medical history of hypertension, diabetes, asthma, cardiovascular disease (including coronary artery disease, heart failure), renal disease (chronic kidney disease or nephrotic syndrome); COVID-19 treatment regimen; use of ACE/ARB; spectrum of illness severity; need for ICU admission; need for non-invasive ventilation (NIV) (e.g., face mask, nasal cannula, nasal mask, or helmet); need for invasive mechanical ventilation (MV); in hospital-death; and length of hospitalization.

COVID-19 Spectrum of Severity de nitions
We de ned the spectrum of disease severity according to the WHO [28]. Mild illness was de ned as uncomplicated upper respiratory tract viral infection and may have non-speci c symptoms such as fever, fatigue, cough, anorexia, malaise, muscle pain, sore throat, dyspnea, nasal congestion, or headache. Moderate illness was the development of non-severe pneumonia that does not require supplemental oxygen. Severe pneumonia was de ned as fever plus symptoms ≥1 of the following: respiratory rate ≥30/min, dyspnea, respiratory distress, SpO2 ≤93% on room air, PaO2/FiO2 ratio <300 or lung in ltrate >50% of lung eld within 24-48hr. Critical illness manifested by symptoms ≥1 of the following: acute respiratory distress syndrome (ARDS), septic shock, altered consciousness, or multi-organ Failure.

Study Outcomes
The primary outcome was COVID-19 severity classi ed as mild, moderate, severe, or critical according to the WHO classi cation in ACEI/ARB users as compared to non-ACEI/ARB users. Secondarily, we evaluated the need of ICU admission, NIV and MV, in-hospital death, and length of hospital stay in patients on ACEI/ARB as compared to non-ACEI/ARB users. We also assessed COVID-19 severity (severe or critical) in patients with (1) one of the three comorbidities of interest (diabetes, hypertension or renal disease); (2) both diabetes and hypertension; (3) diabetes only; and (4) hypertension only.

Statistical Analysis
Using R Core Team (2020) software (R Foundation for Statistical Computing, Version 4.0.1, Vienna, Austria), continuous data were expressed as mean with standard deviation (±SD) or median with interquartile range (IQR). If the normality assumption was met using normal Q-Q plot and the Shapiro-Wilk test, we used the Student's t-test for group comparisons; if not met, we used the Mann-Whitney. Categorical data were reported as frequencies and percentages and analyzed either using the Chi-square test for nxm tables or Fisher's exact test for 2x2 tables group comparisons. To obtain odds ratios, we performed a multivariable logistic regression model adjusting for the confounders (either P<0.3 in a univariable logistic regression model or clinically important confounders) of age, sex, BMI, diabetes, hypertension, renal disease, number of comorbidities (diabetes, hypertension, cardiovascular disease [heart failure/coronary artery disease], stroke, renal disease, asthma and obesity) and COVID-19 treatment regimen. All statistical inferences were drawn with 95% con dence intervals with P<0.05. Bonferroni adjustments were applied to control for multiplicity [29].

Sensitivity analyses
Propensity score-matched analysis Propensity score-matched (PSM) procedure was implemented to the dataset. Using MatchIt package in R [30], the treatment assignment was modelled with a multivariable logistic regression with the following speci cations: nearest neighbor 1:3 with replacement and a caliper of 0.1. Multiple models were tted and compared using the "cobalt" package to plot propensity score distributional mirror diagrams and standardized mean difference (SMD) before and after matching. The SMD check for data balance is an indicated statistical procedure to judge the quality of matched data rather than means or standard deviations of the matched subjects [31]. In our case we considered values of SMD <0.1 to be indicative of adequate balance and fruitful matching. We carefully examined models in which we evaluated different interaction terms. The nal model was then selected based on the above criterion. The following covariates were included in nal model: age, BMI, diabetes, hypertension, renal disease and number of comorbidities. We illustrated the propensity score distribution (prior matching was also evaluated using mirror diagram and the love plot for SMD distribution) (see Supplementary Fig. S1 and Fig. S2).
Inverse propensity score weighting (IPSW) In addition to PSM, we considered a secondary sensitivity analysis for high risk patients (diabetes, hypertension or renal disease). We limited the analysis to the severity outcome to avoid a multiplicity problem. In this technique, we used "WeightIt" and "ipw" packages in R to calculate inverse propensity weights [32] [33]. The following covariates were considered in the IPSW analysis: age, gender, diabetes, hypertension, renal disease, cardiovascular disease, body mass index, and asthma and number of comorbidities. For speci cs, see Fig S3 on covariates selected and balance distribution. Again, various models were tted and toughly tested before selecting the best model for each data subsets. Following the recommendation from Desai et al [34], the estimand was the average treatment effect. Stabilized weights were utilized in our analysis and then we examined weights with plots and summary statistics to prevent variance in ation from extreme weights. Love plots for SMD distribution presented in supplementary. Lastly, we tted a structural causal model using "Survey" package to obtain robust variance estimation (sandwich type).

General description
From April to June 2020, a total of 532 con rmed cases of COVID-19 disease were admitted to the KFMC and PMAH hospitals. Of these, 178 patients were excluded from analysis as they did not meet eligibility criteria. A total of 146 patients on ACEI/ARB and 208 were non-ACEI/ARB users included in multivariable logistic regression analysis. Propensity score matching yielded 145 ACEI/ARB users and 63 patients' non-ACEI/ARB users ( Figure 1). The mean (±SD) age of the ACEI/ARB group was 56.3 (11.9), BMI was 28.8 kg/m 2 (6.3) and 78.1% were men, compared with 40.2 years (11.5), BMI of 27.4 kg/m 2 (7.4) and 81.7% were men in non-ACEI/ARB group ( Table 1). Rates of diabetes of 41.6%, hypertension of 34.5%, and renal disease of 7.3% were observed in the total sample. The propensity score matching controlling for the selected covariates was balanced [Table S1; see also Supplementary Figure S1 and Figure S2 (love plot) for propensity scores distribution and balance of covariates, respectively]. Covariate balance for IPSW analysis presented in supplementary ( Figure S3)   Table 4.   [35][36]. Whether this is a general trend among ACEI/ARB users with COVID-19 or limited to some high-risk substrata is unclear and here our ndings may prove relevant.
Many of the patients in our study had comorbid hypertension, diabetes, and renal disease and therefore were being treated with ACEI/ARB agents. It is in these high-risk patients that we observed proportionately more of the severe COVID-19-associated outcomes, which is physiologically plausible. The renin angiotensin aldosterone system (RAAS) is reported to play a major role in the pathogenesis of type 2 diabetes and hypertension [37]. Over 30 different variants of the ACE2 gene reported in the Lieden open variation database (LOVD) [38]. Several studies showed a potential link between certain ACE2 variants and the development of cardiovascular diseases such as hypertension, while other reports nd the evidence is inconclusive [41][42][43]. The high rates of comorbidities in our population may be linked to the existence of ACE2 polymorphism. In particular, the association between ACE insertion/deletion (I/D) polymorphism and the development of type 2 diabetes and hypertension was reported by Alsaikhan et al in 2017 in Saudi Arabian population [23]. Moreover, studies in Malaysian, Taiwanese, Iranian and Turkish populations reported similar ndings [24][25][26] [42]. The evidence that ACEI/ARB use in COVID-19 patients is associated with more severe COVID-19 disease is growing but still insu cient to be conclusive yet biologically plausible. One hypothesis is that SARS-CoV-2 uses ACE2 receptors for cellular entry and that, therefore, ACEI/ARB use can block the receptor and thus prevent viral attachment, entry and multiplication. In this hypothesis, ACEI/ARB use would ultimately improve patients' outcomes. Conversely, and the hypothesis being evaluated more intensively, that continuous use of these medications may lead to over-expression of ACE2 receptors by a negative feedback mechanism and thus increase coronaviruses binding to target cells [11] [43]. Whether the existence of genetic susceptibility to ACE2 polymorphism and ACEI/ARB use has an effect on COVID-19 severity or not is yet to be determined [44].
Despite the fact that the de nitions of severity of COVID-19 disease need to be harmonized, the contradictory ndings reported in different populations with regards to the severity of COVID-19 may suggest population differences [16][17] [48]. Further, there could be an ethnicity signal certain populations are more susceptible to severe COVID-19 disease in association with ACEI/ARB use. In addition to our study in a high-risk Saudi population treated in tertiary referral hospitals, a large UK cohort study found that Black Africans on ACEI/ARB had a higher risk of COVID-19 disease [50]. Our population may be considered similar to a study conducted in Turkey [51] that found that ACEI/ARB use in hypertensive patients was independently associated with ICU admission, mechanical ventilation and in-hospital mortality.
Our study has several strengths and provides additional insights into the association of ACEI/ARB use and severity of COVID-19 disease as potentially mediated by ethnicity. First, to our knowledge, this is the rst retrospective study conducted in two public specialty hospitals in the Gulf Cooperation Council (GCC) region designated speci cally as national referral hospitals for complex COVID-19 patients that other hospitals in the country cannot manage. The setting lends external validity to the study as these hospitals accommodate patients from all social backgrounds in Saudi Arabia. Secondly, we used multiple de nitions to capture the severity of the disease, all of which were signi cant except for the need for mechanical ventilation, length of stay, and in-hospital death. Thirdly, unlike other studies, we conducted three models to adjust for potential confounders with adjustment for multiplicity problem [47] [52][53][54][55]. The three analytical approaches con rm the severity outcome. It is noteworthy, the research question of ACEI/ARB versus non-ACEI/ARB users and COVID-19 severity presents a statistical challenge due to the fact that ACEI/ARB medications are indicated for particular set of patients that differs from control subjects. This will eventually lead to non-overlapping propensity score distributions. The PSM will model will try to nd close matches of the control and the treatment based on our speci cations and propensity scores, but in this process many cases from the control/or treatment will be discarded leading to decrease sample size and imprecise estimates. The IPSW circumvents this problem by utilizing all the available sample by assigning weights to observations.
The balance can be achieved by creating pseudo-population in which treatment allocation is independent of observed covariates. In this analysis, extreme weights can be problematic and running weighted logistic regression may lead to variance in ation due to the fact that pseudo-population size may be in ated or de ated. In our analysis, we used stabilized weights and weights trimming was not necessary. The robust variance estimator from the "Survey" package gave us precise con dence intervals. We limited the IPSW analysis to a subset of the high population to which we think it's safe to assume that this population will have an indication for ACEI/ARB use. All of the three analyses point towards higher odds of severity among ACEI/ARB users; especially in high risk patients. Due to the small sample size, the COVID-19 severity was not found in the diabetes and hypertension subset of patients. Both PSM and IPSW are very powerful tools for observational studies as means to reduce bias.
We recognize that the retrospective nature of this study is a limitation, though due largely because it was virtually impossible to mount a study when the hospitals were being overwhelmed with complex and severe cases of COVID-19. We could not stratify patients according to their chronicity of ACEI/ARB use because this information was not available. Notwithstanding the multiple models to adjust for confounding variables, we may have overlooked other variables that may explain the severity of COVID-19 disease. Some con dence intervals are wide and may limit the interpretability of some of the results. We could not assess the duration of non-invasive ventilation in patients on ACEI/ARB as compared to non-ACEI/ARB users. Lastly, our study did not differentiate between the two classes of antihypertensive agents (ACEI or ARB).
Clinically, a temporary ACEI/ARB withdrawal in COVID-19 patients with frequent blood pressure monitoring and management with other appropriate therapies maybe a prudent action especially for ICU patients where the risk is higher than the (short-term) bene t in the short term, until randomized controlled trials evaluate the impact of interim ACEI/ARB cessation in COVID-19 patients in the hospital setting. This may be viewed by some as a strong statement that goes against the mainstream recommendation to continue these medications [56]; however, our clinical experience is that a short-term ACEI/ARB withdrawal (≤2 weeks) strategy does not seem to increase the risk of cardiovascular disease deterioration in low-risk patients. Indirect evidence may be inferred from a randomized trial comparing ACEI (quinapril) continuation or withdrawal (by giving placebo) for 16 weeks in heart failure patients, which showed that heart failure symptoms worsened only after 4-6 weeks following ACEI withdrawal [57]. Conclusions ACEI/ARB use was found to be associated with adverse effects on the severity of COVID-19 disease in Saudi Arabian patients, most of them with the comorbidities, singularly or concomitantly, of hypertension, diabetes, and renal disease. As compared to non-ACE/ARB users, ACEI/ARB users had higher disease severity and ICU admission and non-invasive ventilation. Notably, patients on ACEI/ARB with diabetes, hypertension and renal disease were at higher risk of developing severe COVID-19 disease compared to non-ACEI/ARB.

Declarations
Ethics approval and consent to participate The Institutional Review Boards at KFMC and PMAH (IRB 20-200) approved the study. All methods were performed in accordance with the relevant guidelines and regulations. Informed consent was waived since it was an exempt study conducting a retrospective analysis.

Consent for publication
Not applicable Availability of data and materials All data generated or analyzed during this study are included in this published article (and its Supplementary Information les).

Competing interests
The authors declare that they have no competing interests.

Funding
The authors would like to thank the Research Center at King Fahd Medical City, Riyadh, for their nancial support provided for the manuscript.