Also available as an audio podcast here. This is a plain language summary of a paper we (myself and colleagues from the Centre for Clinical Microbiology at UCL, and colleagues from HerpeZ in Lusaka, Zambia and the Institute for Endemic Diseases (IEND) in Khartoum, Sudan) wrote about how well prepared countries in sub-Saharan Africa are when it comes to tackling antimicrobial resistance (AMR). If you want the link to the original paper, you can find it here, it’s open access!
Background to the study
As a bit of background to the project, during the COVID-19 pandemic in 2020, I was asked for help by my infectious disease clinical colleagues at the Royal Free Hospital in North London. They’d noticed that on the COVID-19 wards (and by that time, it was pretty much the whole hospital!) patients were catching a different set of bacterial infections, which had different resistances to antibiotics, than they saw before the pandemic. If you want to learn more about what antimicrobial resistance (AMR) is, check out my other blog! The aim of the study was to identify, using sequencing, whether the bacteria were related, and whether they were sharing AMR genes.
As I was still working with my international colleagues, we decided to see if the same thing was happening in our partner sites. Along with my colleagues Dr John Tembo (HerpeZ) and Dr Muzamil Mahdi (IEND) and their teams, I submitted a grant proposal to the British Society for Antimicrobial Chemotherapy and we won $65,000 (my first grant as principle investigator, so I was thrilled). Below is a summary of the work we did and what we found:
Background
Across the world, patients who develop severe illness due to COVID-19 (i.e. patients who get sick enough to require ventilators to help them breathe) are more likely to be admitted to hospital and acquire bacterial infections whilst they were in hospital. Because they are more at risk of catching bacterial infections, the WHO recommends rapid treatment with antibiotics, which can mean they are treated before the tests have been done to confirm what organism is causing their illness (if you want to learn more about the drugs we use to treat pathogens, check out my other blog!).
At the time there were few reports from resource constrained settings that looked at how COVID-19 patients were looked after affected whether they might catch a bacterial infection on the ward, and whether it would affect what drugs the bacteria would respond to. Our study aimed to identify whether being admitted to a COVID-19 ward (being sick with COVID-19) meant you were more likely to get a bacterial infection whilst in hospital. We then compared these COVID-19 positive patients with COVID-19 negative patients on a different, COVID-19 negative ward. You can read our introductory blog on the BSAC website here.
Methods
We enrolled patients from COVID-19 positive wards and COVID-19 negative wards if their doctors thought they might have a hospital acquired bacterial infection. The patients were swabbed and their samples sent to the microbiology laboratory to identify if they did have an infection, and if so, what species it was. At the same time, we got hold of Infection prevention and control guidelines (protocols that guide doctors, nurses and cleaning staff on the best ways to keep wards clean and patients safe) and analysed them to see if there were differences between the different wards.
We also did whole genome sequencing on some of the bacterial cultures that we collected. Whilst our Zambian site already had an Oxford Nanopore sequencing device, our Sudanese site didn’t. Usually in this situation, I’d have gone out to the sites to help train them on how to sequence the samples, but of course during the pandemic, there was no international travel! We decided to try and do the training online, using Zoom. We recorded the training sessions, so that they could refer back to them, but after the success of the YouTube tutorial video we made on diagnosing SARS-CoV-2 using PCR, we decided to upload the sequencing tutorials too. As a result, we recorded the sequencing tutorials and uploaded them to the PANDORA YouTube channel and also to The Global Health Network sequencing hub pages. These pages have since been viewed loads of times and we’ve had messages saying how useful they’ve been for other scientists!
Results
We collected 109 patient samples from Sudan and 66 from Zambia. We found that patients who had COVID-19 were significantly more likely to catch a multi-drug resistant bacterial infection. That result was found in both countries. The total number of patients who caught a bacterial infection whilst in hospital (both treatable with drugs (susceptible) and resistant to some antibiotics) increased significantly on COVID-19 wards in Sudan. We found the opposite in Zambia though, COVID-19 patients were less likely to catch a bacterial infection in hospital.
When we looked at the genome of the bacterial samples using whole genome sequencing, we found significantly more β-lactamase genes per isolate on COVID-19 wards in both countries. If a bacteria carries β-lactamase genes, it will be able to inactivate antibiotics within the β-lactam family, including penicillins, cephalosporins, monobactams and carbapenems (that covers a large number of the routine antibiotics that we use to treat bacterial infections). Bacteria with β-lactamase genes can be really difficult to treat, which means the patient is likely to be sicker for longer.
We also looked at the infection prevention and control policies on COVID-19 positive and COVID-19 negative wards. We found that there weren’t too many differences, the main changes were that the policies were more strongly emphasized on COVID-19 wards (e.g. making sure nurses and doctors wash their hands and use the correct personal protective equipment). You can read our BSAC news and views article with a summary of our results here.
Conclusions
When we analysed our results, we decided that the changes in bacterial infections and AMR patterns we saw in COVID-19 patients were likely a result of lots of different factors. These were likely to include patient factors (such as underlying health conditions such as diabetes and heart conditions), the types of drugs that the patients were treated with and also differing emphases on infection prevention and control in the wards. As a result of this study, the hospitals in both countries adapted their infection prevention and control policies in an attempt to protect patients from catching AMR bacterial infections (as well as ensuring that nurses and doctors were still protected from catching COVID-19 whilst caring for their patients). Definitely a good result!
Notes from a small scientist 