Case history from South Africa

Hello, everyone. My name is Andrew Whitelaw. I’m a clinical microbiologist based in Tygerberg Hospital and Stellenbosch University in Cape Town, South Africa. In Cape Town, we are fortunate enough to have two medical universities attached to three teaching hospitals. The University of Cape Town uses Groote Schuur Hospital and Red Cross Children’s Hospital as its teaching platforms. Stellenbosch University uses Tygerberg Hospital, which has a paediatric unit, as its teaching. But the problem we’ve been facing in many of our hospitals in South Africa, not only the ones in Cape Town, is increasing extended-spectrum beta-lactamase production, particularly amongst Klebsiella, although to a lesser degree amongst E. coli, and resistance to other antimicrobials as well.

And this, is a large part, was driving more and more use of carbapenems in many hospitals in South Africa, including in Cape Town. Although around about 2010, 2011, we weren’t yet seeing carbapenem-resistant Klebsiella and E. coli as problems, we had been seeing carbapenem resistance in Pseudomonas and Acinetobacter, and they were causing major problems on the clinical side. We’d also know from some studies that antimicrobial prescribing was suboptimal, to put it mildly. So a number of patients were prescribed inappropriate antimicrobials, very little de-escalation, inappropriate durations, and so on.

At that stage– so this is roughly 2010, the antimicrobial stewardship in the hospitals in Cape Town that we are serving consisted primarily of ward rounds in the ICUs for varying amounts of time a week. So the microbiologist would go to the ward round once a week with laboratory results and would discuss appropriate antimicrobial use. We were available 24 hours a day to provide advice on appropriate use of antibiotics. And there was a restriction policy in place in most of the hospitals– or most of the academic hospitals in Cape Town at the time. And the restriction policy was quite harsh, if I can put it that way.

So there are a number of agents that required approval from microbiologist or ID to be used for patient. And the system was that the prescribing clinician would phone the on-call microbiologist or infectious disease specialist to get approval to use the antimicrobials. And this is discussed in a couple of other case scenarios that I’ve put up on this learning platform. But this was leading to a lot of work for the on-call person. We would get 30, 40, 50 calls a day to use ceftriaxone, co-amoxiclav, azithromycin, whatever. The clinicians learned how to play the system, if I can put it that way. And they knew what information would allow them to use the antimicrobial of choice.

And many of us got the impression that it was becoming something of a game. The clinicians weren’t taking this particularly seriously, and if they had to phone, they would phone up and sort of not lie, but sort of emphasise the appropriate parts of the case that would allow them to use the agent that they wanted. And there wasn’t much improvement. Antibiotic usage was going up, both at a hospital level as well as in the ICUs. Our resistance rates were staying much the same. This was a study that was done a couple years prior, looking at ceftriaxone use, which had dropped quite drastically, in the purple line.

But the bar graph shows that the ESBL rates of Klebsiellas were staying much the same, and if anything, increasing slightly in the later part of the study. And susceptibility rates of other organisms between 2007 and 2010 didn’t show much change. There was some reduction in resistance in Pseudomonas aeruginosa, possibly in Acinetobacter. Our MRSA rate stayed much the same, and our ESBL production in Cipro-resistant Klebsiella was very much the same as well.

The concerns were the restriction model– concerns that had been raised, at least, with the restriction model were that it stops people’s thinking. And common– well, not a common, but an occasional comment was that when asked why a patient was on a particular antibiotic, was that, oh, the microbiologist said I should use meropenem on this patient. And that doesn’t really help anybody’s cause. There were concerns that it may result in inappropriate choices– when somebody didn’t feel like phoning for release, they would choose something that’s on the nonrestricted list, even though that may not be the most appropriate choice. And there were concerns that it would delay starting antimicrobials if one had to phone and go through a process of getting authorisation.

The education model, if anything, is even more labour-intensive. It needs a lot of buy-in from everyone concerned. But studies had shown that it was– even if it may take longer to implement, became a more sustainable solution in the long term. So we slashed our restriction list radically and limited it to a small number of agents based on whether the patient was in the ICU or the general wards. We introduced prescription charts, stewardship ward rounds, and emphasised education. The stewardship chart– this is the first draft of our stewardship chart. It’s gone through a number of iterations since then, but I think this illustrates the principle. A dedicated prescribing chart for antibiotics only.

It was divided into infection episodes, that one could quite easily see what a particular infection was being treated with. So if the patient had pneumonia, was started on antibiotic X, escalated or de-escalated to antibiotic Y, and you could see how long that was carried on for. And the intention of this was primarily to focus the mind on the antibiotic that was being prescribed– why is it being prescribed? What is the indication? Is it community hospital acquired? And a reminder that cultures need to be sent before starting antibiotics.

A reminder to review the antibiotics at least a day– And then separate sections for Staph antibiotics and some information on the back page around appropriate cultures to be sent; which antibiotics didn’t need to be given, to avoid double-cover or overlapping activity and so on. And we formed an antibiotic stewardship task team, which consisted, at that stage, of a ward pharmacist, an infection control specialist, a consultant microbiologist, and an ID specialist. And this team would then go to a particular ward and look at all the patients on antibiotics in that ward and review and provide advice on whether it was appropriate or inappropriate, how it would be optimised and changed if necessary.

This was first done as a nine-week pilot study, followed by weekly rounds in two medical wards over a 12-month period. And we compared antibiotic use and laboratory tests pre and post– intervention The errors identified on the ward rounds were not totally surprising. The most common were lack of indication for antibiotics and no indication had been documented in the notes– the patient would be on a third-generation cephalosporin with no clear reason why– incorrect duration, incorrect dose, and failure to de-escalate them. Those four together made up nearly 3/4 of the errors found on the stewardship.

After the 12-month period comparing the pre and post usage data, there was an approximately 20% reduction in antibiotic volume overall, and particularly, if you look at the graph on the top left, reductions in intravenous antimicrobials with increases in– or also reduction in ciprofloxacin IV, with an increase in oral ciprofloxacin. So talking to increased attention on IV to oral switch. A massive reduction in IV ceftriaxone, which in many cases was being used inappropriately. And similar reductions in many other intravenous antimicrobials as well. Small increases in either oral agents or narrow-spectrum intravenous agents. The flip side of this was that we also found that laboratory costs increased by 22% over the time period as well.

And you see– this small table here just shows the increase in laboratory investigations, particularly related to C-reactive protein and procalcitonin probably went up inappropriately higher. So full blood count stayed much the same. White blood cell count went up 22%. CRP & PC went up 100% and 500%, respectively. And that was an inappropriate or unwanted side effect of the stewardship programme, and further interventions were designed to address that particular issue as well. But even with the increasing laboratory cost, there was a total cost saving of the antimicrobial costs of in South African rands, nearly a quarter of a million grand, which was substantial for one programme in two wards over one year.

So after this, we then started rolling out the same model at other tertiary care hospitals, focusing in the ICUs, but then moving to paediatric settings and adult wards, and after that moving to the secondary level hospitals as well. We started developing provincial support to strengthen antimicrobial stewardship and infection control. I think one key element is that we’ve always tried to link up antimicrobial stewardship and IPC programmes together. Without IPC, antimicrobial stewardship is substantially weakened. Pharmacists have long been seen as a key resource to promote antimicrobial stewardship in South Africa.

And we started empowering pharmacists through training programmes and developed the National South African Antibiotic Stewardship Programme, which facilitated workshops and advocacy and engagement with National Department of Health to drive the National Antimicrobial Stewardship agenda. Just in the Western Cape more briefly, we set up teams to visit various hospitals in the Western Cape, consisting of infectious diseases, microbiology, infection control, pharmacy, public health, and assess the structures and activities related to antimicrobial stewardship at these hospitals and made recommendations on how to implement an AMS Programme in these hospitals. And this led to the formation of the Provincial Antimicrobial Stewardship Committee, which is still in place and reviews policies and progress and liaises with the National Committee on Antimicrobial Resistance.