Table 1: at the end of this file.
2.5. Statistical Analysis
Minitab version 19.0 software was used for statistical analysis. All data was expressed as continuous variables. Continuous data was expressed by median for normally distributed variable i.e. age; whereas absolute numbers were expressed as percentages. The paired t-test was used to compare continuous variables of normal distribution and non-normal distribution, respectively. We used Pearson Coefficient to show association between the variables. Relationship of one response and multiple predictors was examined using linear regression with best fitted model. Mortality was evaluated for bacterial and viral co-infections by using binary logistic expression and expressed as odds ratio. MANOVA (multivariate analysis of variance) was used to analyze mortality in the presence of co-infection and comorbidities and it was expressed as p-value. The patients were grouped by disease severity, comorbidities and co-infection or not. Factors were adjusted for age and gender. A 2-sided α of less than 0.05 was considered statistically significant.
3. Results
We investigated co-infection in 48 COVID-19 patients (including 37 males and 11 females), the male to female ratio was 3:1. Median age of our study population was 52 years (1-92). Fourteen patients (29%) needed admission to intensive care unit (ICU cases). The remaining 34 patients (71%) did not require any admission and were classified as non-ICU cases. We found co-infections in thirty-four (71%) patients. Although severity of disease was negatively correlated (r= -0.09) with presence of a co-infection (p=0.53), it had a positive correlation with co-infecting viruses (r=0.1, p=0.42) by Pearson Coefficient as shown in Figure 1. Furthermore, statistically significant inverse association was observed (r=-0.28, p=0.04) between bacterial co-infection and ICU admission. In other words, this association indicates less likelihood of ICU admission with bacterial co-infection. The most commonly found co-infecting virus was influenza A H1N1 in 17 patients (36%). Chlamydia pneumoniae was the most prevalent co-infecting bacteria found in 13 patients (28 %). Other organisms detected were adenovirus in 10 patients and S. aureus in 4 patients (Figure 1). It was noticed that 4/17 (23.5%) patients with H1N1 had coexisting Chlamydia pneumoniae.