Clinical characteristics and predictors of the duration of hospital stay in COVID-19 patients in Jordan [version 1; peer review: 1 approved with reservations, 1 not approved]

Background: On March 11th, 2020, the World Health Organization (WHO) declared coronavirus disease 2019 (COVID-19) as a global pandemic. Healthcare systems in lowand middle-income countries may face serious limitations during a pandemic, for which understanding the predictors of prolonged hospital stay are crucial in decreasing the mortality rate. The aim of this study was to investigate the predictors of increased length of hospitalization among COVID-19 patients. Methods: In this prospective study, we investigated the effect of presenting symptoms and laboratory investigations on the duration of hospitalization of 131 COVID-19 patients at a tertiary hospital in Jordan from March 17th to April 9th, 2020. Results: Patients median age was 24 years [interquartile range (IQR): 8-39], of which 67 (51.15%) were males and 64 (48.85%) were females. Smokers had shorter in-hospital stay (OR: -3.52; 95% CI: -6.73 to -0.32; P=0.03). Taste loss (OR: 5.1; 95% CI: 1.95 to 8.25; P<0.01) and chills or rigors (OR: 4.08; 95% CI: 0.73 to 7.43; P=0.02) were the symptoms significantly associated with increased in-hospital stay, while those who had malaise (OR: -4.98; 95% CI: -8.42 to -1.59; P<0.01) and high white blood cell (WBC) count (OR: -0.74; 95% CI: -1.31 to -0.17; P=0.01) had faster recovery. Conclusions: Our study found that the most common presenting Open Peer Review


Introduction
In December, China reported a group of pneumonia cases of unknown etiology that were later identified to be caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1,2 . On March 11 th , 2020, the World Health Organization (WHO) categorized coronavirus disease 2019 (COVID-19) as a global pandemic 3 . Patients with COVID-19 frequently present with a cluster of different respiratory symptoms, including fever, chills cough, shortness of breath, sore throat, and new loss of taste and/or smell within 2 to 14 days after exposure in most cases 4 .
Coronaviruses belong to viral family Coronaviridae (order Nidovirales) and include viruses approximately 26-32 kilobases in size with a positive-sense single-stranded RNA genome (+ssRNA) 5 . The Coronaviridae family contains four genera, one of which is the Betacoronavirus genus, which SARS-CoV-2 belongs to 6.
Initial assessments of the epidemiologic characteristics and transmission dynamics showed that the basic reproductive number (R 0 ), which is defined as the expected number of additional cases that can directly be generated by one case in a population susceptible to infection on average over the course of its infectious, range from 2.2 to 3.58 7,8 , with the potential for asymptomatic transmission being a major concern for most of previous investigations.
The first identified case of COVID-19 in Jordan was on March 2 nd , 2020, a returning traveler two weeks prior to quarantine procedures 9 . On March 13 th , a wedding ceremony led to a large outbreak of COVID-19 cases in northern Jordan 10 , after which a strict lockdown took place and five tertiary hospitals were selected to provide medical care for patients suspected or diagnosed COVID-19 cases 9,10 .
Healthcare systems in low-and middle-income countries may face serious limitations in capacity and accessibility during a pandemic, leading to worse clinical outcomes and an increase in mortality rate 11 . Therefore, the aim of this study is to investigate the predictors of increased length of hospitalization among COVID-19 patients, in order to provide evidencebased public health outbreak response strategies for COVID-19 and future pandemics.

Study design
In this prospective observational investigation, we reviewed COVID-19 patients who were admitted to the isolation center at Prince Hamza Hospital (PHH), which is a tertiary hospital in Amman, the capital of the Hashemite Kingdom of Jordan. We included Jordanian patients above the age of 18 years who were diagnosed with COVID-19 and admitted to the isolation center of PHH. The diagnosis of COVID-19 was made after the collection of nasopharyngeal swabs using Xpert® sample collection kit (catalog number XPRSARS-COV2-CE-10, Cepheid, Sunnyvale, CA, USA) 12 at the emergency department of five tertiary hospitals in Amman, after which positive cases were transferred to the isolation center at PHH. The data collection took place between March 17 th till April 9 th , 2020, during which all of the 131 patients admitted to PHH with the primary diagnosis of COVID-19 were included in our investigation, representing 40.4% of the total 324 cases diagnosed with COVID-19 in Jordan during the study's timeframe. None of the patients were not eligible, declined to be enrolled, or withdrew from the study, and the study was reported in accordance with the STROBE statement (https://www.strobe-statement. org/). It is noteworthy that, in Jordan, all patients diagnosed with COVID-19 were admitted to hospital during the study's timeframe, regardless of the severity of their illness.

Data collection
Based on semi-structured interviews by anesthesia and intensive care resident physicians, the demographic data, smoking habit, and past medical history of patients were documented directly from the patients during medical history taking. Moreover, we documented the current presenting symptoms, including cough, shortness of breath, chest pain, fever, chills/rigors, sweating, malaise, myalgia, headache, diarrhea, abdominal pain, palpitations, loss of taste, loss of smelling, nasal congestion, and rhinorrhea. Furthermore, baseline vital signs and laboratory investigations were collected, and whether the physician started the patient on hydroxychloroquine as a treatment. All patients were followed-up daily from admission till discharge from the hospital after clinical resolution and having negative test results.

Ethical approval
The study protocol was approved by the Institutional Review Board (IRB) committee of The Hashemite University (No. 1/10/2019-2020). Written informed consent was obtained from all patients prior to participation in the study. All patients were able to withdraw from the investigation at any time without affecting their care. No identifying information were obtained from the patients, and all collected data were used solely for statistical analysis.

Statistical analysis
Statistical analysis was performed using STATA (Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC). The total sample number comprised of 131 patients. The association between the categorical baseline variables and the length of hospitalization was studied using Mann Whitney-U test for the dichotomous variables and Kruskal-Wallis test for the rest of the poly chotomous variables. Data were reported as medians with interquartile range (IQR) as they were nonnormally distributed (P<0.01 for the Shapiro-Wilk test).
A Linear regression analysis was used to examine the factors predicting the length of hospitalization. Baseline characteristics, presenting symptoms, and laboratory results at day of presentation were first analyzed using univariable linear regression analysis. Then, only significant variables were fitted into the final multivariable linear regression analyses. We reported the regression coefficients with their level of significance; p-value and 95% confidence intervals (95% CI). The p-value of statistical tests was two-sided, and statistically significant results were defined as those with a p-value <0.05.

Discussion
In spite of the great progress in understanding COVID-19, a therapeutic or preventive solution is yet to be achieved 14 . Consequently, better management of medical facilities during the next phase of this pandemic is crucial in improving the outcomes in these patients 15 . In the present study, smoker, as well as patients presenting with malaise and elevated WBCs at presentation had shorter hospital stay, while loss of taste and chills or rigors at presentation were associated with lengthier in-hospital stay.
Several previous studies investigated the clinical manifestations of COVID-19. Most common presenting symptoms in most of these studies were fever, cough, dyspnea, malaise and myalgia 16,17 . Interestingly, it has been suggested that anosmia (loss of smelling sense) and ageusia (loss of taste function) can represent the first or only symptomatology 18 , while an investigation from Italy found that 64% mildly symptomatic patients had impaired olfaction 19 .
Remarkably, even though multivariable linear regression analyses in the current study did not show significant correlation between demographic factors and the length of in-hospital stay except for smoking habit, an investigation from Germany revealed that those with preexisting respiratory diseases, obese patients, and those with persistently elevated inflammatory markers are at increased risk of developing acute respiratory distress syndrome (ARDS), which will prolong their hospitalization period 20 . Moreover, a study conducted in China found that severe cases more frequently had dyspnea, lymphopenia, and hypoalbuminemia, with higher levels of c-reactive protein, d-dimer, lactate dehydrogenase, alanine aminotransferase, ferritin, IL-2R, IL-6, IL-10, and TNF-α 21 . In the current study, patients with elevated WBCs count had shorter in-hospital stay. Immunocompetent WBCs play a significant role in systemic inflammatory response to infection, with neutrophil-lymphocyte count ratio being significantly higher in mortality cases of community-acquired pneumonia 22 .
An interesting finding in our study is that smokers had shorter in-hospital stay, moreover, only 22% of hospitalized COVID-19 patients were smokers. There is a controversy in the current literature about the role of smoking and nicotine in COVID-19 23,24 . It is postulated that smokers are at higher risk to get respiratory tract infections and develop more severe illness, due to the preexisting bronchopulmonary damage, reduced mucociliary clearance, and the local inflammatory status 23 . Some reports, however, showed lower COVID-19 related mortality and morbidity in smokers 24 . Nicotine, as a cholinergic agonist, inhibits pro-inflammatory cytokines such as TNF, IL-1, IL-6 by binding acetylcholine receptors (nAChR). These cytokines, among the others, might result in the notorious feature of this illness, the cytokine storm. Therefore, it was hypothesized that COVID-19 is a disease of nicotinic cholinergic system. Another supporting observation is that patients with ageusia in our sample had longer in-hospital stay; this sensory disturbance is mainly related to the cholinergic system of the brain, and it might indicate that such patients are having more extensive disease, or technically, nicotinic cholinergic dysfunction 23,24 . Although it is not wise or acceptable to advice people to smoke, this might shed the light on the promising role of nicotine in preventing and treating COVID-19 infection.
The rapid upsurge in the number of confirmed cases makes the control of the spread of COVID-19 and its treatment challenging 25 . Several ongoing clinical trials will soon confirm or disprove the usefulness of several candidate medications in treating COVID-19 26 . Studies investigating the use of hydroxychloroquine were unable to confirm its benefit on in-hospital outcomes of COVID-19 patients 27 . On the other hand, dexamethasone has shown a decrease in the 28-day mortality among those who were receiving respiratory support, with both of dexamethasone and methylprednisolone being equally effective in treating moderate to severe COVID-19 cases 28,29 . With no definitive method of prevention or treatment being available to date, precautions should be made in order to control the spread of COVID-19, including proper hand hygiene, isolation of infected or suspected persons in properly ventilated hospitals, social distancing, discouraging large gatherings, and avoiding direct contact with suspected animal reservoir hosts 30 .
The main limitation of this study is that it did not investigate the predictors of in-hospital mortality of COVID-19 patients.  Previous studies suggested that older age, high Sequential Organ Failure Assessment (SOFA) score, and elevated d-dimer are associated with poor prognosis 31,32 . However, upon plotting mortality against the incidence of COVID-19, a significant positive correlation was found, suggesting that mortality is associated with heavier healthcare burden 33 . Hence, understanding the predictors of prolonged hospital stay and prognostic factors are crucial in decreasing the mortality of any pandemic. Moreover, we recommend future studies to take into consideration duration and severity of each symptom in order to develop a better understanding of the clinical course of COVID-19.
The main strength of our study is that, in Jordan, all COVID-19 patients were hospitalized during the study's timeframe, regardless of the severity of their illness. Therefore, our sample gives a good opportunity to observe the natural progression of COVID-19 under controlled conditions.

Conclusion
Our study found that the most common presenting symptoms of COVID-19 are cough, malaise, and headache. beginning of the Corona crisis; therefore, the conclusions are premature and cannot reflect the logical and the expected conclusions regarding COVID-19. Thus, the study should be revised by including data of a larger sample size to be more representative and provide evidence-based conclusions. Having said that, the study rationale is good and interesting; but when supported with robust design, it will be of more interest to the scientific community and will better reflect the real situation.
Is the work clearly and accurately presented and does it cite the current literature? Partly

If applicable, is the statistical analysis and its interpretation appropriate? Partly
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Partly