The retrospective analysis of the variables influencing the acquisition of CRE and their association with mortality is studied.

Univariate analysis of age and gender vs mortality, age had significant impact on mortality (p-value <0.001*). Similar to our study findings, there are reports of median age of patients with CRE-66 years (interquartile range of 57 to 80), with upto 45.2% being males (70 out of 155 patients).⁸

The distribution of CRE across diverse clinical sample types highlights the importance of monitoring. Increased rate of CRE isolation was from skin and soft tissue infection (36.7%) and respiratory infection (34.29%). Univariate analysis of clinical sample type with mortality was not statistically significant (p-value = 0.051). The result is in par with a study by Wang et al, in which no statistically significant variance in infection types was observed amongst CRE groups.⁹

Klebsiella pneumoniae was the commonest CRE isolated (n = 194, 69.2%), but there was no significant association of the organism isolated with mortality rate. (p-value = 0.053). Sekar et al. study and a study from South India highlighted an increase in resistance to carbapenem, notably in Klebsiella species.¹⁰ Research focusing on the pediatric demographic in rural areas of Southern India reported notably high rates of carbapenem-resistant Klebsiella spp.¹¹ PUO, skin and soft tissue infection and respiratory infections account for majority of cases (over 80%) in our setup. The rising occurrence of CRE have been reported from skin and soft tissue infections.¹² In a study by Rebold et al., respiratory infections accounted for 38%, followed by urinary infections at 20%, intra-abdominal at 16%, and primary bacteremia at 14% of the CRE infections.¹³

Research conducted by Guarnera et al. revealed that the colonization and subsequent infection of multidrug-resistant organisms commonly occurs and poses a challenging obstacle to eliminate in patients undergoing various stages of treatment for acute myeloid leukemia and/or Fever of unknown origin. This presence significantly impacts long-term outcomes.¹⁴

The type of morbidity had a significant association with mortality as shown in Table 2. (p-value <0.05). As per review article by Qin Hu et al, mortalities due to CRE infection ranged from 10% to 59% in 12 studies with combination therapies and from 8.3% to 64.7% in 11 studies with monotherapies.¹⁵ A study by Mariappan et al. found that comparing infections caused by susceptible strains of the same species to those caused by multiple antibiotic-resistant bacteria reveals significantly more severe consequences. These include extended hospitalization and heightened rates of both morbidity and mortality. This study did not isolate any specific factor as an independent risk for acquiring carbapenemase-producing enterobacterales (CPE) infection. However, through multivariate analysis, it became apparent that patients with CPE infections face an elevated risk of mortality when they require ventilation and rely on indwelling medical devices for support.^16,17

Our study showed that risk factors such as prior Hospitalization/Surgery during preceding 12 months, prior use of Multiple antibiotics/carbapenems, Chemotherapy, COVID status, Duration of ICU stay, Mechanical ventilation and presence of indwelling devices were significantly associated with mortality (p-value <0.05). These findings are on par with the findings by Kong et al. in China, in which the independent predictors of in-hospital death included age over 65 years (OR: 3.19), prior exposure to carbapenems (OR: 3.54), and the presence of a central venous catheter (OR: 4.19).¹⁸ A retrospective cross-sectional analysis conducted in Penang over a 3-year period (January 2021-December 2023), demonstrated a strong positive association between carbapenem utilization and the incidence of carbapenem-resistant Enterobacterales (CRE).¹⁹

Multivariate analysis was done to analyze the relationship between mortality and multiple independent risk factors. The Table 4 depicts information about the significance levels, odds ratios and confidence intervals for the predictors in the model. Using the logistic regression model, the factors: age, gender, >7 days-duration of hospital stay, and COVID status were considered as covariates, while the factors that were statistically significant on univariate analysis: infection type, prior hospitalization/surgery in 12 months, prior use of multiple antibiotics/carbapenems/Beta lactams, chemotherapy, mechanical ventilation, indwelling device were considered as variates.

The type of infection, clinical sample, organism and prior carbapenem use, maintained statistical significance on multivariate analysis. (p-value <0.05). Prior use of carbapenems was statistically significant (p-value of 0.048) with a high odds ratio of 6.844, suggesting a substantial impact on the outcome. The significant p-value suggests that these predictors included in the final model are important for explaining the variation in mortality.

Other predictors, such as Gender, Surgical intervention, presence of indwelling device, and most of the medical history-related variables as mentioned in the Table 4, do not seem to significantly impact mortality since it is not statistically significant in predicting the outcome. This suggests that these variables may not be strongly associated with the outcome.

The established risk factors for carbapenem-resistant Enterobacteriaceae (CRE) infection included advanced age, the use of ventilation devices, and prior exposure to antimicrobials (multiple and carbapenems). The findings of this study, thus have implications for clinical practice, particularly in the areas of infection control and antimicrobial stewardship.

Elderly patients are generally at higher risk due to age-related immunosenescence and the presence of multiple comorbidities, necessitating closer monitoring and timely intervention. The association between ventilation devices and CRE necessitates the need for strict adherence to infection prevention protocols, such as proper hand hygiene, regular equipment disinfection, and care of devices.

Prior antimicrobial exposure, especially to broad-spectrum antibiotics, was found to be a significant contributor to the development of CRE. This highlights the critical importance of antimicrobial stewardship programs to ensure the judicious use of antibiotics, reduce unnecessary prescriptions, and promote the use of targeted therapies based on culture and sensitivity results.

The interventions to mitigate the risks associate with CRE may include early screening for CRE colonization upon admission, especially in intensive care units, and implementation of isolation protocols for colonized or infected individuals to prevent spread in healthcare settings.²⁰ Additionally, strengthening antimicrobial stewardship programs to limit the unnecessary use of broad-spectrum antibiotics is critical. Tailoring empirical therapy based on local resistance patterns and enforcing de-escalation strategies can also reduce selection pressure and curb resistance. Collectively, these interventions can help healthcare facilities control CRE transmission and improve patient outcomes. By addressing these modifiable risk factors, healthcare providers can reduce the incidence of CRE infections and enhance patient outcomes, while also contributing to the global effort against antimicrobial resistance.²¹

A six-year study by Lin et al. revealed a 7%-13% decline in susceptibility rates to Beta lactams/Beta-lactam inhibitor combinations among initial CRE isolates. This trend may be attributed to higher antibiotic consumption, the acquisition of resistance determinants, or shifts in the prevalence of carbapenemase types within the Southeast region.²²

The antibiotic susceptibility pattern of our study of 280 cases of CRE in the decreasing order of frequency is as shown in Figure 2. The antibiotics Cotrimoxazole (n = 125, 45%), aminoglycosides {Amikacin (n = 142, 51%) and gentamicin (n = 136, 49%)}, tigecycline (n = 121, 43%) were sensitive in a substantial portion of CRE cases. Over 70% of the total CRE cases were sensitive to either of these antibiotics as shown in Figure 2. Hence, these antibiotics which showed lower rates of resistance, may be relied upon for empirical treatment in the present setup in contrast to the to 3^rd generation cephalosporins, beta-lactam inhibitor drugs and carbapenems.

The antibiotics cefaperazone sulbactam, imipenem and meropenem were not as sensitive as aminoglycosides/cotrimoxazole/tigecycline, but they were still sensitive in 86 cases (31%), 73 cases (26%) and 66 cases (24%) of CRE respectively. Antibiotics piperacillin tazobactam (n = 58, 21%), ertapenem (n = 58, 21%), cefepime (n = 56, 20%), 3^rd gen cephalosporins (n = 44, 16%) were least sensitive against CRE. Hence, they may not be relied upon as first-line treatments, and their use should be carefully considered in case of a suspected CRE infection. Our research outcomes align with Mohamed et al. findings, which investigated the antibiogram of CRE isolates and observed varying levels of resistance to different antibiotics. Specifically, concerning cephalosporins, the resistance rates for CRE were notably high, ranging from 69% to 80% for 3^rd generation cephalosporins, 89% for cefepime, 97% for imipenem and 97% for meropenem.^23,24

Our study findings would guide the clinicians to make use of the local hospital and ward level susceptibility data to help their antibiotic treatment decisions when their patients are at risk of being infected by a carbapenem-resistant pathogen. Out of the total 280 patients, 147 patients (52.5%) received empirical antibiotics, and among them, 112 patients (76.19%) survived. This highlights the importance of early diagnosis and quick start of empirical antibiotic therapy in improving the patient outcome. Interestingly, patients who did not receive any antibiotic therapy had a lower mortality rate (4.51%) compared to those who received. This could be attributed to the good practice of restricted antibiotic use, antibiotic stewardship and subsequent culture driven specific antibiotic therapy. Ohnuma
et al.’s research demonstrated that the use of suitable initial empirical therapy differed based on pathogens, often being less common for resistant organisms. Surprisingly, multiple studies have shown that inadequate empirical antibiotic therapy did not correlate with increased mortality. Consequently, the selection of appropriate empirical therapy continues to pose a significant challenge.²⁵

Beta-lactams/Beta-lactam inhibitor monotherapy was the most commonly administered empiric antibiotic (n = 70) as shown in Table 5. The combination of 3^rd generation cephalosporins + Clindamycin + Aminoglycosides category had a lower mortality rate (11.11%) compared to other categories, probably due to the fact that amikacin and gentamicin were sensitive in majority of CRE isolates (n = 142, 51% and n = 136, 49%) respectively, while the combination of third generation cephalosporins + metronidazole had a relatively higher mortality rate (40%). Other empirical antibiotics had varying success rates in treating patients as shown in the Table 5. This study thus helps to assess the effectiveness of these empirical antibiotic treatments in managing patient outcomes, so as to guide in making informed decisions regarding empirical therapy.

Higher doses of Tigecycline, Carbapenem, Colistin are found to be associated with better clinical outcomes and another strategy is to combine these drugs though evidence for the latter is more obscure. In our study, as shown in Table 6, the effectiveness of different antibiotic treatments in terms of patient survival and mortality when dealing with CRE infections is described. As shown in Table 6, a majority (n = 214,76.4%) of CRE infected patients received combination of Beta-lactam inhibitor (Piperacillin-tazobactam, Cefaperazone-sulbactum, ceftazidime-avibactum) with colistin/tigecycline/carbapenem, out of which 87.85% survived.

Among the total of 34 patients who received carbapenem with tigecycline, 67.65% survived and among the 15 patients who received carbapenem with colistin, 73.33% survived. Thus, our study supports the role of combination therapy vs monotherapy. As indicated by Sheu et al. study, the use of combination therapy is associated with reduced mortality rates compared to the use of monotherapy.²⁶

None of the patients in the categories who received beta-lactam inhibitor with aminoglycosides, carbapenem with Polymyxin B, Cotrimoxazole with aminoglycoside and Cotrimoxazole with Polymyxin B, expired (0% mortality). This difference in mortality maybe attributed to the fact that cotrimoxazole (45%) amikacin and gentamicin were sensitive in majority of CRE isolates (45%, 51% and 49% respectively). According to Sahitya et al. study on CRE infections, the effectiveness in addressing CRE has been established through the utilization of combination therapies that encompass carbapenems, polymyxin, tigecycline, aminoglycoside, and Fosfomycin.²⁷ Conversely, a recently developed antibiotic involves ceftazidime avibactam alone or avibactam, often combined with carbapenem-containing regimens, showcasing a newer addition to the treatment options available for addressing CRE infections.²⁸ Current evidence indicates that ceftazidime-avibactam demonstrates superior efficacy and safety over polymyxins in the management of infections caused by carbapenem-resistant Enterobacteriaceae.²⁹ Determining drug MICs through phenotypic susceptibility testing is key to tailoring treatment, particularly in refractory cases or infections at difficult-to-access sites. When carbapenem-resistant Gram-negative bacteria are strongly suspected such as in severe sepsis, septic shock, or immunocompromised hosts empirical combination therapy with two or more active agents is recommended, with antibiotic selection guided by local resistance trends and antibiogram data.³⁰

In countries with limited resources, obstacles to enhancing the management of CRE infections include the absence of advanced molecular diagnostics capable of offering a deeper insight into resistance patterns related to CRE infections. Additionally, the scarcity of new antibiotics poses a challenge, hindering the potential for improved outcomes and decreased mortality among critically ill patients suffering from CRE infections.³¹ The study’s constraint was the absence of statistical analysis correlating empirical and definitive treatment methods with mortality rates. Despite investigating the various risk factors and mortality, the study did not capture patient outcomes regarding average duration of hospitalization and reduced functional status at discharge.