That’s what the WHO stated this week, and it was based on a study, in Lancet Planetary Health. In most news items that I saw animal antibiotic use was directly linked to human infections caused by antibiotic resistant bacteria. A journalist even asked if eating meat was safe. Although most of us (including me) support reduction of unnecessary antibiotic use, it’s worth reading this excellent meta-analysis, initiated by WHO. Did this study answer the burning research question “to what extent does animal antibiotic use influence infections in humans?“
The study aim was to do a “systematic review and meta-analysis to summarise the effect that interventions to reduce antibiotic use in food-producing animals have on the presence of antibiotic-resistant bacteria in animals and in humans.” That less antibiotic use in animals reduces resistance in animals has been demonstrated in many studies. But does that effect translate into measurable effects in humans? Twelve (!) studies were selected to analyse the association between reduced animal antibiotic use and AMR in humans. The pooled estimate was a 24% reduction (95% CI 6-42%). Yet, as correctly stated by the authors, these studies had large clinical heterogeneity (other word for comparing apples and pears) and incredibly large statistical heterogeneity (I2=97.4%, other word for findings from individual studies go in all directions). What to do with such results was discussed recently, here.
The authors are fully aware of this caveat and limitation, nevertheless provided 3 reasons for pooling data: “First, we anticipated a priori that there would be clinical heterogeneity across studies, given the wide variety of settings studied and interventions described and tested. Second, it would have been surprising to not find statistical heterogeneity, given the large number of samples in the constituent studies that would have resulted in high power when statistical testing for heterogeneity. Finally, even with this expected clinical and statistical heterogeneity, the overall findings were surprisingly homogeneous in their ultimate findings, with a consistent effect that spanned different populations, interventions, and bacterial groups.” This sounds like my 5-year old daughter caught by surprise on a high chair with her hand in the candy box, immediately realizing that she’s doing something not allowed, then telling me: “I was only looking at it”. The additional meta-regression failed to explain the heterogeneity, although effects were larger in studies with poorly described interventions.
These 12 twelve studies (mostly from Scandinavia and the Netherlands) deserve a closer look: From 1 (Klare, 1999) I couldn’t find anything and of the remaining 11, 9 studied carriage in farmers, 1 included carriage (VRE) of healthy subjects and 1 included clinical isolates (VRE) from patients; 6 studied VRE, 4 live-stock associated MRSA (LA-MRSA) and 1 LA-MRSA and ESBL. The 24% estimate resulted from 3 studies in particular (2 LA-MRSA and 1 VRE); in 1 (Cuny) I couldn’t find a control population, in 1 (Rinski) I only found slides from a conference presentation from which I could not derive a 75% reduction. All studies on VRE were published around 2000, and more recent studies have clearly demonstrated the differences between VRE found in animals and infected humans (but may the plasmid may be shared). So, any interpretation from these studies for the general population should be made very, very carefully (if at all).
And that is what the authors did: “Interventions that restrict antibiotic use in food-producing animals are associated with a reduction in the presence of antibiotic-resistant bacteria in these animals. A smaller body of evidence suggests a similar association in the studied human populations, particularly those with direct exposure to food-producing animals. The implications for the general human population are less clear, given the low number of studies.” But this got more or less lost in “scientific” translation.
That leaves us with a still unanswered research question: “to what extent does animal antibiotic use influence infections in humans?“ The largest natural experiment thinkable is currently executed in the Netherlands, where animal antibiotic use has declined with almost 70% in the last 7 years. Yet, bloodstream infection rates with MRSA and ESBL-producers remained fairly stable, and VRE increased (a little). So, this may suggest that the animal reservoir has no effect, or that without the intervention rates would have gone up. Who knows? Only WHO, I’m afraid.
3 thoughts on “No more antibiotics for animals”
Nice post. It’s interesting that folks think that antibiotic exposure in animals would have an impact on rates of infections (human) caused by pathogens containing already prevalent resistance genes. Why would we even think that when antibiotic use rates are already so high in humans? I’ve always thought that global antibiotic exposure in animals/environment would only drive emergence of new or rare genes. This is the paper that was most influential, but there are others. Maybe you no longer believe in mathematical models? (humor)
Thanks for highlighting this article – fascinating that it was written in 2002! Especially given it’s modelling of transient gut presence becoming colonisation becoming high level carriage with abx use and subsequent transfer.
It’s such an eminently sensible conclusion, that once AR is out in a big way in the human population, it’s human behaviour that governs it’s prevalence, but the timing of the initial rare imports are hastened by exposure to resistant organisms created, among other things, by agricultural antibiotic use.
Complete game changer in how I think about it.
Don’t worry, I still believe in math models, and I think that there is sufficient evidence of direct transfer of AR bacteria from animals to caretakers. Yet, from there on the R-value is probably very low, and further spread unlikely (and certainly not creating endemic levels in a population without animal contacts). The same may hold for transmission through contaminated meat. With low transmission parameters for these 2 routes the current lack of evidence may well be explained by math models. Yet, in countries with different transmission risks (where animals and caretakers and family live in closer proximity, where meat is eaten raw or where sewage from animals directly reaches humans) things may be differerent, but so far (as in this meta-analysis) only data from rich countries where animals and humans are strictly separated.