Bacteria are cleverer than one can imagine. They often mutate their DNA to cope up with antibiotics, they are treated with. Further, they can horizontally transfer genetic information to other bacteria in the clan by plasmid exchange. These abilities make them invincible dwellers of our bodies and environment. Development of drug resistance in this manner accumulate and expand over time, to make more and more infections hard to tackle. However, it is not that all bacterial species/strains are equally poised to infect a person. They choose their hosts (humans) depending on host-specific features. For instance, neonatal sepsis is frequently caused by infection due to group-B streptococcus and Escherichia coli. Here, the host age-group plays a perceivable role.
Ghosh et al., for the first time, reported a comprehensive analysis of the association between infecting bacterial species and patient factors, including patient demographics, comorbidities, and certain healthcare-specific features. Their study unraveled a plethora of unknown and emerging associations in the critical care settings of a prominent tertiary care hospital in the US. Many of these deserve the attention of ID specialists.
Below we quote a relevant part of the reported findings by the authors.
“… We observed that Enterococcus faecium and interval of more than 5 days had the highest positive association with mean coefficient value ∼ 1.84. Next, we noticed that Beta-hemolytic Streptococcus, group B and Escherichia coli are more likely to cause infections in patients of age 0−15 years as compared to patients belonging to other age groups with mean coefficients 1.75 and 1.74 respectively. The bacteria being Enterococcus faecalis, ‘diseases of the respiratory system’ had a positive coefficient ∼1.17 on an average. Furthermore, 3−5 days after a patient’s admission into hospital, Serratia marcescens was found to have a higher infection rate, represented by the strong positive mean coefficient value of 1.04. We also noted that incidents of Pseudomonas aeruginosa were positively correlated with interval greater than 5 days and ‘endocrine, nutritional and metabolic diseases, and immunity disorders’ with mean coefficients 1.02 and 0.93 respectively. Staphylococcus also had positive association (mean coefficient value ∼ 0.92) with interval more than 5 days. On the other hand, Staphylococcus had strong negative correlation (mean coefficient value −0.66) with‘ diseases of the digestive system’. Staphylococcus epidermidis infections were much less frequent in females compared to males supported by mean coefficient value −0.63….“
While the behaviour of bacteria is unpredictable, systematic patient record mining can generate informed leads that can be tapped for proactive and dynamic calibration of hospital-wide infection management policies. The associations derived can be used by antibiotic stewardship teams to implement hospital-specific guidelines to curb the spread of infection within the hospital against present alarming situations. This in turn can help ASP teams to cut down the additional load of antibiotics on patients occurring due to hospital acquired infections.