Corruption and antimicrobial resistance
PLoS . Concerning Collignon et al (2015) ref:_00DU0Ifis._500U0Kmszg

We thank Heister and Kaier for their comments and the extensive time they took to analyse our paper in detail. We have reanalysed our data taking into account their comments and take this “opportunity” to respond to the concerns they raise. Our results continue to show that there is as strong relationship in these countries with variations in the levels of antibiotic resistance seen in blood culture isolates with Governance (GOV, e.g. control of corruption). It appears to be stronger relationship than with documented antibiotic usage (ABU).

We have addressed the issue they raise on inadequate sample sizes, particularly in the first years of data collection in the EU, by rerunning the Pooled OLS and Fixed effects regressions after excluding the years 1998 and 1999. As can be seen in the attached table our results do not change substantially. GOV still remains the most statistically and economically significant explanatory variable. Interestingly the coefficient is much larger and statically more significant in all equations compared to the coefficient of the antibiotic usage variable (ABU). In this FE equation, the coefficients of all variables other than ABU and GOV are not statistically significant simply because the explanatory power of these variables (PHC, TED and AGR, in particular) is partly captured by the country specific fixed effects adjustments. This is an advantage, not a limitation, of the FE estimator.

Removing the first couple of years from the regressions addresses the main concern of Heister and Kaier, which was the number of tested isolates underlying the average resistance in the earliest year, which they believe were insufficient (especially for Estonia and Sweden). It should be noted however that we had used the percentages for resistance as supplied by the European Centre for Disease Prevention and Control (ECDC) on their website. These same issues raised by Heister and Kaier thus pertain to the small minority of these widely accepted results but based only on small numbers as published by this EU agency.

Because of technical issues we could not attach the updated regression results here and so we have uploaded the table on figshare (https://figshare.com/). We also give a brief note below to explain what we did to address the issues raised. The updated regressions results relate to Columns 2,3,5 and 6 in Table 3 of the original paper.

It should be noted that our original System GMM results (which were in any case was not criticized by Heister and Kaier) were based on observations occurring every second year as explained in our paper. These regressions exclude the very first year (1998), because no lagged values are available as instruments for the explanatory variables. These results were, therefore, based on observations occurring in the year 2000, 2002, 2004 and so forth. Thus these results in our paper, in reality had already addressed the Heister and Kaisers’ concern about the inclusion of data for the year 1998, which had low numbers of isolates in some cells.

Finally, we were also criticized for employing an unbalanced data set, a data set which had missing values for a number of countries for some years. The use of unbalanced panel datasets are not uncommon in the empirical analyses, and it is well established that models estimated using unbalanced dataset remain valid in terms of consistency of the estimated coefficients. Excluding some years from the time coverage, just because observations are missing for a few countries leads to a loss of data that could have been fruitfully employed for the analysis. In any case, we would like to note that, to a great extent, we have moved towards a balanced panel dataset by excluding the years 1998 and 1999.

In summary even when we exclude the first 2 years of our data (which is where Heister and Kaier had concerns on some cells being very small re sample size), we have found the same results as when we used the larger data set.
Heister and Kaier seem incredulous that corruption can be related to the levels of antibiotic resistance seen in countries. There are many reasons why antibiotic resistance levels are likely related to the levels of corruption in any country and other examples of poor governance. Indeed, when one considers all the accepted factors that drive the development of resistance and its spread, it would be surprising if such a relationship with poor governance was not seen.

Whenever there are issues with “governance” (in hospitals, in the local community and in a country) then this will most likely result in more resistance being seen. For many healthcare related issues (immunization rates, falls in hospitals, increase in unexpected hospital re-admission), poor governance in countries or institutions will result in worse patients’ outcomes. In the community this includes the use and/or overuse of over-the-counter antibiotics (which can result from poor governance, controls and/or corruption.

Overall, there are only two factors that drive antimicrobial resistance. Both can be controlled. These factors are:
• The volumes of antimicrobials used and
• The ways resistant micro-organisms and genes encoding for resistance spread.

If there is poor governance (of which control of corruption is one variable) there is a high chance that people will more likely bend or break rules. Hence antibiotic usage may be higher than what official figures show, people will use and consume more antibiotics (including more broad spectrum antibiotics) and those in the agriculture sector will more likely use banned, off-label or broad spectrum antibiotics in food animals - even when this is not officially permitted. Factories or hospitals may be more lax about the disposal of antibiotic-contaminated waste into environmental waterways, with subsequent exposure of environmental bacteria to any persistent antibiotics. Poor controls on waste disposal make it also more likely that resistant bacteria will be dumped into the soil and waterways,

Resistant bacteria recycle to people and animals via foods and water. If appropriate standards for food and water are not regulated or enforced, the more resistant bacteria and their resistance genes spread. Within hospitals, lack of control of corruption means some products may not be available or substandard products are only supplied. Infection control will then be more expensive to implement and/or there will less ability of essential supplies to make sure appropriate infection control practices are able to be followed.

Thus there are numerous examples that show how poor governance (of which corruption is one measurable parameter) will likely result in more resistant bacteria developing and being easier to spread to other people.

The general perception of antibiotic resistance is that it is almost entirely related to the amounts of antibiotics used, not only in the broad sense of comparative usage by different countries but also in individuals. However, these two variables are not perfectly correlated at national levels and across countries. Other factors are also likely to very important to account for the variations in resistance observed between regions and countries and may interact with usage volumes and spread. One of these factors is Governance - with control of corruption one important and measurable variable. Other governance measurable variables are of the “rule of law” and “government inefficiency”.

When the Quality of Governance is poor, then there are likely to be less effective controls of antibiotic use (not only in people but in the animal sector). Thus, not only will more antibiotic resistant bacteria develop but the spread of these resistant bacteria will also be easier. This is because there will be less supervision and enforcement of laws that cover issues related to not only human medicine, but also dealing with food and water safety.

Reducing antimicrobial resistance requires a policy mix aimed at lowering antibiotic usage in people and, perhaps even more importantly, improving governance and developing better controls on corruption.

We acknowledge that the issue of antibiotic resistance and how it develops and spreads is complex. Antibiotic resistance involves factors in the human sector, the agricultural sector and the environment. All these factors can’t be examined because adequate data are often lacking. Antibiotic usage in food animals is one very important parameter that hopefully can in the future be examined when adequate data become available for all countries. Another parameter for which there are little or no data available is the usage of the amounts of antibiotics in children. Travel is also a factor that affects the level of resistant bacteria carried by people although in most it tends not to be persistent.

We believe available data suggests that the issue of antimicrobial resistance in human pathogens is not just related to the volume of antibiotic usage in people but other important social factors such as corruption and effective governance. When one thinks about governance issues and the impact of poor governance on the controls on antibiotic usage and on the pathways that resistant bacteria can spread, not unexpectedly, the better any country or institution is with respect to these factors, the lower will be the resistance rates in bacteria causing serious and life-threatening infections.