Reply to Chopra and Rizvi

there was an increase in morbidity and length of stay, but there was no signif­ icant increase in mortality due to mul­ tidrug­resistant (MDR) or extremely drug­resistant gram­negative bacterial infection [3]. India is the largest consumer of antibi­ otics, but human use accounts for the mi­ nority of the total antimicrobial products consumed, with the majority of antibiot­ ics being used as growth enhancers in the food animal sector [4]. Although there are strict guidelines on nonutilization of im­ portant human antibiotics, for example, colistin in the food animal sector, there is widespread belief and evidence that these are regularly ignored, resulting in increas­ ing transmission of AMR from farms to hospitals. This has put increasing pressure on regulatory authorities to ban the use of antibiotics in the animal section in the hope of saving them for clinical use [5]. This overutilization has been shown to be responsible for increasing AMR. In India, infections caused by MDR organisms are very frequent, often result­ ing in mortality due to a paucity of treat­ ment options [6]. To address these issues, the Indian Council of Medical Research initiated a nationwide Antimicrobial Resistance Surveillance and Research Initiative (AMRSRI) in 2016. The aim of AMRSRI is to provide reliable and au­ thentic estimates of AMR burden encom­ passing local resistance patterns as well as molecular epidemiology of isolates throughout the country as this data will be invaluable in guiding national policy. AMRSRI has subsequently released an­ nual reports that detail the various resis­ tances encountered in clinical isolates nationally [7]. Collectively, Gandra et al’s report on mortality due to AMR is a welcome step to gauge the effect of AMR on mortality rates in India. This issue is projected to cost the global economy $100 trillion and result in 10 million deaths by 2050 [8]. In order to more accurately represent the effect of AMR on mortality, data from AMRSRI and other initiatives should be collated, along with data from private healthcare providers. Additionally, there should be stringent implementation of antibiotic stewardship and infection con­ trol programs in the healthcare sector and nonutilization of important human antibiotics in the farm and animal hus­ bandry sector as growth promoters. Research aimed at understanding the genetic and molecular basis of AMR in MDR organisms as well as novel drug discovery and development should be declared a national priority as AMR does not respect international boundaries.

there was an increase in morbidity and length of stay, but there was no signif icant increase in mortality due to mul tidrugresistant (MDR) or extremely drugresistant gramnegative bacterial infection [3].
India is the largest consumer of antibi otics, but human use accounts for the mi nority of the total antimicrobial products consumed, with the majority of antibiot ics being used as growth enhancers in the food animal sector [4]. Although there are strict guidelines on nonutilization of im portant human antibiotics, for example, colistin in the food animal sector, there is widespread belief and evidence that these are regularly ignored, resulting in increas ing transmission of AMR from farms to hospitals. This has put increasing pressure on regulatory authorities to ban the use of antibiotics in the animal section in the hope of saving them for clinical use [5]. This overutilization has been shown to be responsible for increasing AMR.
In India, infections caused by MDR organisms are very frequent, often result ing in mortality due to a paucity of treat ment options [6]. To address these issues, the Indian Council of Medical Research initiated a nationwide Antimicrobial Resistance Surveillance and Research Initiative (AMRSRI) in 2016. The aim of AMRSRI is to provide reliable and au thentic estimates of AMR burden encom passing local resistance patterns as well as molecular epidemiology of isolates throughout the country as this data will be invaluable in guiding national policy. AMRSRI has subsequently released an nual reports that detail the various resis tances encountered in clinical isolates nationally [7].
Collectively, Gandra et al's report on mortality due to AMR is a welcome step to gauge the effect of AMR on mortality rates in India. This issue is projected to cost the global economy $100 trillion and result in 10 million deaths by 2050 [8]. In order to more accurately represent the effect of AMR on mortality, data from AMRSRI and other initiatives should be collated, along with data from private healthcare providers. Additionally, there should be stringent implementation of antibiotic stewardship and infection con trol programs in the healthcare sector and nonutilization of important human antibiotics in the farm and animal hus bandry sector as growth promoters. Research aimed at understanding the genetic and molecular basis of AMR in MDR organisms as well as novel drug discovery and development should be declared a national priority as AMR does not respect international boundaries.

Reply to Chopra and Rizvi
To the Editor-Rising rates of antimicro bial resistance (AMR) are a global public health crisis that threatens to abrogate gains made in improving healthcare outcomes over the last 60 years. As our report [1] on more than 5000 patients in India makes clear, Gramnegative infections are signifi cantly associated with mortality, and these infections are spreading rapidly around the globe in healthcare facilities as well as in the community [2]. Because our study was conducted in a private hospital setting and the vast majority of patients in India receive care in government hospitals that typically have less advanced care options, our re sult that patients with multidrug resistant Gramnegative infections were associated with 2-3 times higher mortality may ac tually understate the problem in India, as Chopra et al note [3]. Therefore, we echo the call that more data from government hospitals in India are urgently needed to better understand the burden of AMR in India.
We believe that the Antimicrobial Resistance Surveillance Network initiated by the Indian Council of Medical Research [4], which includes both public and private hospitals, is an excellent start at complying with the Indian National Action Plan on AMR [5]. However, increased investment to strengthen laboratory capacity is urgently needed in many public and private hospitals to both improve surveillance as well as patient outcomes. Enhancements in surveillance will allow for better understanding of how disease severity and comorbid conditions impact mortality outcomes in patients with AMR infections. For instance, the contrast in mortality risk between our findings on nonintensive care unit patients [6] may be due to differences in infection location and di sease severity. Overall mortality in the study of 116 patients by Naim et al. was only 4.3% (in contrast to 13.1% in our study), and none of their patients had bacteremia or lower res piratory tract infections (nearly 90% were from wounds or urine). In our study, the odds of mortality were significantly higher among patients with lower respiratory tract and bloodstream infections when compared with urinary tract and wound infections. Understanding these differences was only possible because of high data availability. Yet, national data are needed to understand the frequency of these infections and to develop nationalscale plans to combat the problem of resistance.
While the scale of the problem is national and even global in nature, the outcomes are local. Thus, we further endorse the call by Chopra et al that there needs to be strin gent implementation of antimicrobial stewardship and infection control activ ities to improve patient outcomes; how ever, greater investments are necessary on a countrywide level [7]. Large investments in training are needed as there are few in fectious diseases physicians in India. These investments would likely bring significant returns as interventions by infectious dis eases physicians in the United States are associated with improved outcomes and lower costs [8]. Furthermore, because re sistance can spread rapidly around the world, greater investments are needed in AMR hot spots, such as India, to help con tain and reduce the spread of resistance. In addition, increased investments in novel drug discovery is of significant importance as currently available therapeutic options are not effective against the common re sistance mechanisms encountered in ex tremely drugresistant Gramnegative bacteria in India.

Cerebral herniation after lumbar puncture
To the Editor-In their recent article, Costerus and colleagues [1] conclude that cerebral herniation after lumbar puncture (LP) is a rare event, giving a frequency range of 0.1-3%. However, in the context of adverse drug reactions to medicinal products European Commission guid ance considers this level of risk to be un common (0.1% to 1%) or even common (1% to 10%); rare would typically be de fined as 0.01% to 0.1% [2]. Furthermore, the severity of the adverse reaction, in this case death, needs to be considered. Most doctors and patients would be con cerned by a risk of death of up to 3 in 100 procedures and would likely choose not to proceed.
Although LP is an important investiga tion in patients with suspected bacterial meningitis, given the risk of undertaking an LP in this patient group it is valuable to consider the role of other less invasive diagnostic tests that may complement data from LP or confirm a microbio logical diagnosis in patients where LP is contraindicated.
UK and European guidelines on bac terial meningitis stress the importance of blood cultures, which can be positive in up to 74% of patients if samples are taken prior to commencing antibiotics [3,4]. Too often, this simple procedure is not performed, and this represents a missed opportunity to confirm the bacterial eti ology: Shallcross and colleagues found that of 4357 patients attending the emer gency department who received par enteral antibiotics, less than a third had blood cultures taken [5].
Of newer molecular tests, polymerase chain reaction of peripheral blood for Neisseria meningitidis has been shown to have high sensitivity and specificity for detection of meningococcal infec tion [6] and remains positive for up to 5 days after initiation of parenteral anti biotics [7]. Urinary pneumococcal an tigen has been assessed for the diagnosis of pneumococcal meningitis. Although