Risks and Benefits of Antibiotics vs. COVID-19 Morbidity and Mortality

Study design

The retrospective statistical analysis study of antibiotic use compared with COVID-19 morbidity and mortality.

Setting and samples

This single-center study was obtained by the institutional ethics review board, and this study was the retrospective cohort study using public data. Data were collected for antibiotic consumption rate and use ratio, and COVID-19 morbidity and mortality.

Procedure

The antibiotic consumption data from the annual epidemiological report for 2019 that contained the consumption of antibacterials for systemic use by EU/EEA countries in 2018 expressed as DDD per 1000 inhabitants per day was used (12). Japan data was obtained from the antimicrobial selling data in 2019 (13). That is, these data represented each antibiotic consumption rate. Each antibacterial use ratio was calculated from these data. COVID-19 morbidity per 0.1 M population and mortality caused by this disease per 0.1 M population in each of the countries were obtained from the WHO coronavirus disease (COVID-19) Dashboard on 25 May 2020. Table 1 showed the population, morbidity per 0.1 M population, mortality per 0.1 M population, the sum of consumption as DDD per 1000 inhabitants per day, and each antibiotic consumption as DDD per 1000 inhabitants per day (tetracyclines: J01A, macrolides: J01F, cephalosporins: J01D, penicillins: J01C, quinolones: J01M, and sulfonamides: J01E) by each of the countries. Table 2 showed morbidity per 0.1 M population with each number, mortality per 0.1 M population with each number, and each antibiotic use ratio by each of the countries.

Statistical analysis

Statistical analyses were carried out using univariate linear regression analysis and Pearson’s correlation coefficient. A p-value of < 0.05 was regarded as significant. Data were analyzed statistical software available through Microsoft Excel 2013 (Microsoft Corporation Redmond, Washington). We analyzed two kinds of relationships. One was the relationship between each antibiotic DDD per 1000 inhabitants per day and COVID-19 morbidity and mortality. The other was the relationship between each antibiotic use ratio and them. These analyses were examined by tetracyclines (J01A), macrolides (J01F), cephalosporins (J01D), penicillins (J01C), quinolones (J01M), and sulfonates (J01E).

The statistical analysis between antibacterial consumption rate and COVID-19 morbidity and mortality

1. Defined Daily Dose (DDD): The sum was a mean±SD: 17.84±5.98.

   The sum of DDD had no relationship with morbidity (Y = 1.64X+161.92; r = 0.05; significance: NS; p = 0.77) and mortality (Y = 0.27X+11.34; r = 0.07; significance: NS; p = 0.69).

2. Macrolides (J01F): The consumption rate was a mean±SD: 2.92±1.42.

   There was no relationship with morbidity (Y = −16.03X+246.25; r = −0.13; significance: NS; p = 0.49) and with mortality (Y = −0.36X+11.41; r = −0.02; significance: NS; p = 0.90).

3. Cephalosporins (J01D): The consumption rate was a mean±SD: 2.23±1.94.

   A negative correlation with morbidity (Y = −38.43X+285.18; r=-0.43; significance: p < 0.05; p = 0.02) and mortality (Y = −3.89X+24.89; r = −0.50; significance: p<0.01; p = 0.004) was seen.

4. Quinolones (J01M): The consumption rate was (a mean±SD: 1.53±1.42).

   There was a negative weak correlation with morbidity (Y = −39.20X+259.20; r = −0.22; significance: NS; p = 0.24) and no relationship with mortality (Y = −2.09X+19.39; r = −0.10; significance: NS; p = 0.61).

5. Sulfonates and trimethoprim (J01E): The consumption rate was a mean±SD: 0.59±0.44. There was a weak positive correlation with morbidity (Y = 95.03X+145.60; r = 0.24; significance: NS; p = 0.20) and no relationship with mortality (Y = −7.02X+20.32; r = −0.14; significance: NS; p = 0.45) was seen.

6. Tetracyclines (J01A): The consumption rate was a mean±SD: 1.93±1.11.

   There was a week positive correlation with morbidity (Y = 49.98X+10.97; r = 0.32; significance: NS; p = 0.08) and no relationship with mortality (Y = 2.48X+11.41; r = −0.13; significance: NS; p = 0.49).

7. Penicillins (J01C): The consumption rate was a mean±SD: 7.41±3.16.

   A week positive correlation with morbidity (Y = 16.97X+73.57; r = 0.31; significance: NS; p = 0.09) and with mortality (Y = 2.34X-1.17; r = 0.34; significance: NS; p = 0.06) was seen.

The statistical analysis between antibacterial use ratio and COVID-19 morbidity and mortality

1. Macrolides (J01F): The use ratio was a mean±SD: 16.40±1.42.

   A negative weak correlation with morbidity (Y = −9.20X+350.37; r = −0.31; significance: NS; p = 0.09) and with mortality (Y = −0.74X+28.13; r = −0.20; significance: NS; p = 0.27) was seen.

2. Cephalosporins (J01D): The use ratio was a mean±SD: 11.65±8.49.

   A negative correlation with morbidity (Y = −0.30X+319.43; r = −0.50; significance: p< 0.01; p = 0.00) and mortality (Y = −1.12X+29.27; r = −0.44; significance: p<0.05; p = 0.01) was found.

3. Quinolones (J01M): The use ratio was a mean±SD: 8.36±4.21.

   There was a negative weak correlation with morbidity (Y = −13.44X+311.64; r = −0.32; significance NS; p = 0.08) and no relationship with mortality (Y = −0.90X+23.71; r = −0.18; significance: NS; p = 0.35).

4. Sulfonates and trimethoprim (J01E): The use ratio was a mean±SD: 3.58±2.47.

   There was no relationship with morbidity (Y = 8.97X+167.23; r = 0.13; significance: NS; p = 0.50) and with mortality (Y= −1.63X+22.07; r = −0.19; significance: NS; p = 0.32) was seen.

5. Tetracyclines (J01A): The use ratio was a mean±SD: 11.38±6.64.

   There was a week positive correlation with morbidity (Y = 6.99X+119.86; r = 0.28; significance: NS; p = 0.08) and no relationship with mortality (Y = 0.45X+11.07; r = 0.14; significance: NS; p = 0.47).

6. Penicillins (J01C): The use ratio was a mean±SD: 41.53±11.78.

   A weak positive correlation with morbidity (Y = 11.21X+5.07; r = 0.34; significance: NS; p = 0.06) and with mortality (Y = 0.64X-10.26; r = 0.35; significance: NS; p = 0.06) was revealed.

Principle findings

The antibiotic consumption dose for systemic use by EU/EEA countries in 2018 was expressed as DDD per 1000 inhabitants per day. Therefore, the relationship between each DDD per 1000 inhabitants per day, and COVID-19 morbidity and mortality represented whether each antibiotic consumption rate is related to COVID-19. The relationship between the antibiotic use ratio and COVID-19 morbidity and mortality indicated whether the antibiotic use trend affects COVID-19. The sum of DDD suggests no affect morbidity and mortality caused by COVID-19. However, each antibiotic analysis represents various analytical results. Cephalosporins suggest decreasing morbidity and mortality caused by COVID-19. Penicillins may have the potential to accelerate them. Quinolones suggest the potential to reduce morbidity, and the weak potential against mortality. Macrolides may have the potential for reducing morbidity and mortality. Therefore, macrolides and quinolones may suggest working against COVID-19. Sulfonates and trimethoprim have almost the same mortality affect as quinolones. However, they may suggest increasing morbidity. Tetracyclines may increase morbidity, but may not affect mortality. The results of these analyses were similar to those of EU/EEA countries alone. Therefore, cephalosporins, macrolides, and quinolones may be potentially effective against COVID-19 (SARS-CoV-2). And, sulfonates and trimethoprim may have the potential reducing mortality.

Cephalosporins, macrolides, and quinolones are frequently used in Japan, but penicillins are frequently used in European countries (12,13). The different antibiotic use trends may suggest an explanation for less morbidity and mortality caused by COVID-19 in Japan rather than in EU/EEA countries. The antibiotic use trend in Japan is across East Asia. Although the analytical data is from 2018 and is not the latest, the antibiotic consumption rate and use trend from the previous report are not so different. Therefore, the results of this study will appear to fit in with the latest.

Penicillins

The results imply that it is prudent to avoid penicillins use in the current status of the COVID-19 epidemic, as penicillins may suggest increasing morbidity and mortality caused by COVID-19. Penicillins may appear to create an environment for COVID-19 proliferation or may help the virus spread rapidly through cells.

Macrolides

Macrolides have been shown to be active in vitro against RNA viruses. Azithromycin is thought to have antiviral and anti-inflammatory activity and may work synergistically with other antiviral treatments. Azithromycin has demonstrated antiviral activity against Zika virus and against rhinoviruses from in vitro studies. Azithromycin is a macrolide antibiotic that is being investigated as a potential treatment for people with COVID-19 caused by the new coronavirus (SARS-CoV-2) (2,3,4,5). Clarithromycin decreases interferon (IFN)-γ and increases IL-10 levels. Clarithromycin promotes the expansion of immunosuppressive CD11b+Gr-1+ cells essential for the immunomodulatory properties of macrolides (6). Clarithromycin suppresses virus growth (5,6). Ivermectin is a cyclic lactone oral anthelmintic that belongs to the macrolides group. Ivermectin is anti-parasitic previously shown to have broad-spectrum antiviral activity in vitro, and it is an inhibitor of SARS-CoV-2 (14). Indeed, it seems that the mortality rate caused by COVID-19 is low in Onchocerciasis endemic areas of Africa (1). However, from an action point of view, these drugs seem to be effective only in the early stages.

Quinolones

Quinolone derivatives have been shown to inhibit human immunodeficiency virus (HIV) replication at the transcriptional level (7). Richter et al. reported that introducing an aryl group into the piperazine moiety of fluoroquinolone changed its activity from antibacterial to antiviral, with specific effects on human immunodeficiency virus (HIV) (8). Antiviral activity is confirmed by dose-dependently inhibiting viral RNA production (9).

Sulfonates and trimethoprim

Sulfonates and trimethoprim show almost the same mortality effect as quinolones. So, they may work against COVID-19 mortality. Trimethoprim-sulfamethoxazole appears to be effective in suppressing human immunodeficiency virus (HIV) load and antiretroviral therapy (10). It has antimalarial and antibacterial properties. Since trimethoprim-sulfamethoxazole is used in the current state of COVID-19, its function against RNA viruses may be due to its antimalarial properties. Because, in principle, antibacterial agents do not act on viruses. Sulfonates can also act on herpesvirus such as EBV and Kaposi sarcoma herpesvirus belonging to DNA virus (11). Hence, antimalarial properties may act on RNA and DNA viruses.

Cephalosporins

Strangely, in this study, the higher cephalosporins consumption rate and use ratio will make lower morbidity and mortality caused by COVID-19 than other antibacterials. Cephalosporins have not been proven effective against viruses. β-lactam antibiotics including cephalosporins have a bacteriostatic effect. The bacteriostatic action means the action of suppressing the growth of bacteria. β-lactam antibiotics act by inhibiting peptidoglycan synthase, which is an enzyme required for synthesizing bacterial cell walls. However, cephalosporins suggest inhibiting SARS-CoV-2.

There are four effective cases of cephalosporins (Ceftriaxone) combination therapy with minocycline (broad-spectrum tetracycline) in Japan (15). Tetracyclines do not affect mortality caused by COVID-19 from this study. Therefore, the results of this study and these effective cases imply that cephalosporins may work against SARS-CoV-2.

Based on empiric treatment for patients with MERS coronavirus (MERS-CoV)

According to the previous report (16) for patients with MERS coronavirus (MERS-CoV) that is the same coronavirus of SARS-CoV-2, empirical treatment with neuraminidase inhibitors and an association of antibiotics effective against S. pneumoniae and L. pneumophila are the key management of hospitalized patients. Third-generation cephalosporins are effective against S. pneumoniae, and macrolides and quinolones are effective against L. pneumophila. Macrolides (azithromycin or clarithromycin) or quinolones (levofloxacin or moxifloxacin) are the standard treatment for Legionella pneumonia in humans, with levofloxacin being considered first-line with increasing resistance to azithromycin. The results of this study show the similarities in the use of antibiotic agents in the conclusions of the previous report (16). The antibiotic combination of cephalosporins with trimethoprim-sulfamethoxazole (TS) may also be an effective therapy from the perspective of reducing COVID-19 mortality.

Implications for practice

When a physician needs to treat several secondary bacterial infections that two different antibiotic combinations are often used, the combination therapy with cephalosporins and quinolones or macrolides may be an effective treatment under the condition of COVID-19 epidemic period. In general, there are no specific antiviral drugs or vaccines for SARS-CoV-2. All of the drug treatment options come from experience treating SARS, MERS, or some other new influenza virus previously. Active symptomatic support continues to be the key to treatment. This retrospective study suggests that the treatment regimen proposed by Bleibtreu et al. (16) may be effective for COVID-19. However, the efficacy of this regimen needs to be further confirmed.