Statistical Analysis

The prevalence of isolated bacteria was estimated by age groups, sex, and service areas. In addition, these variables were used to determine the prevalence of bacterial resistance and sensitivity. The above was calculated as percentage using the formula: P = (PR/N) x 100, where PR is the number of samples with bacterial resistance and/or sensitivity to antibiotics, N is the total number of samples recorded, and multiplied by one hundred. Difference between Gram-positive and Gram-negative bacteria, and the prevalence of antibiotic resistance and/or sensitivity was compared using the two sided Chi-square test with the package SPSS, 2011, at significant level of p < 0.05.

General prevalence.

An overall bacterial prevalence of 42.8 % from 1945 records within the period December 2016 and January 2018 was estimated. The prevalence of bacterial species is shown in Table 1. Escherichia coli (44.3 %) showed the highest prevalence followed by Klebsiella pneumoniae (7.2 %), Staphylococcus aureus (6.3 %), Staphylococcus epidermidis (6.2 %) and Pseudomonas aeruginosa (5.2 %). On the other hand, Staphylococcus capitis, Staphylococcus lugdunensis, Achromobacter denitrificans, Aeromonas caviae, Cronobacter sakazakii, Raoultella ornithinolytica, Sphingomonas paucimobilis, Morganella spp., Shigella sonnei, Achromobacter xylosoxidans, Plesiomonas shigelloides, Ralstonia pickettii, Streptococcus intermedius, Enterococcus faecium, Bacillus spp., Kocuria rosea, Streptococcus dysgalactiae, Streptococcus parasanguinis and Streptococcus constellatus showed the lowest prevalence (≈ 0.12 %).

Regarding the high prevalence of E. coli, K. pneumoniae, S. aureus, S. epidermidis and P. aeruginosa found in this study, a number of studies in the European Union (19,888 patients) published in 2012 also reported a high prevalence of hospital infections by E. coli (15.2 %), S. aureus (12.1 %), P. aeruginosa (11.2 %), and K. pneumoniae (8.1 %) (^{Zarb et al., 2012}). Despite the level of prevalence and rank order are different to our results, the isolated bacterial species are the same. On the other hand, a surveillance network (PRONARES) in Chile, reported E. coli (35.9 %), S. aureus (20.6 %), and Klebsiella spp. (9.5%) as the predominant isolated species from a total of 5,251 biological samples (^{Trucco et al., 2002}). Similarly, a meta-analysis from 40 different studies published in 2022, reported diarrheagenic E. coli as the most prevalent bacteria in Asia (22.8%) (^{Salleh et al., 2022}). However, identification of type of E. coli in this study was not performed.

Prevalence according to Gram classification

Regarding 834 Gram stains, the prevalence of Gram-negative bacteria was significantly higher (613 isolates; 73.5 %) than Gram-positive bacteria (221 isolates; 26.5 %) (p < 0.05) (Figure 1).

Figure 1 General prevalence (%) of bacterial isolates from 834 clinical samples according to their Gram classification. *Significant p-value by the Chi-square test.  

Figura 1. La prevalencia general (%) de bacterias aisladas de 834 aislamientos de muestras clínicas según su clasificación Gram. *Valor p significativo por Chi-cuadrado prueba. 

In Mexico, reports from Leon Guanajuato and Mexico city in 2012 revealed a higher prevalence of Gram-positive (51.2 %) than Gram-negative bacteria (48.8 %) from 5,117 clinical cultures (^{Duarte-Raya and Granados-Ramírez, 2012}). On the other hand, a previous study carried out at the State Cancer Center of ISSEMYM (Social Security Institute of the State of Mexico and Municipalities) in Mexico, reported 787 (59.9 %) infections associated to Gram-negative bacteria and 401 (30.5 %) to Gram-positive bacteria from a total of 1,313 infections (^{Romero et al., 2013}). Differences between these trends may be attributed to influx of staff, patients, and contact with outside air in waiting rooms, and different cleaning and disinfection procedures by hospital services (^{Zambrano-Gari and Luna-Fontalvo, 2013}).

Prevalence of bacteria per service unit

Table 2 shows the bacterial prevalence from 834 positive records per service unit. The service with the highest overall prevalence of bacterial species was the internal medicine (IM) with 241 isolates (28.8 %), followed by external consultation (EC) with 133 isolates (15.9 %); gynecology (GYNE) with 82 isolates (9.8 %); neonatal special care unit (NSCU) with 81 isolates (9.7 %); pediatrics (PED) and emergency department (ED) with 75 isolates each one (8.9 %); surgery (SUR) with 64 (7.6 %); and pediatric emergencies (PEDE) with 52 (6.2 %). The intensive care unit (ICU) service showed the lowest prevalence (15 isolates, 1.7 %).

The area with the highest reported infections was the IM (28.8 %) in this study. A study carried out in 2013 with 126 patients at the Mexican Social Security Institute in Salamanca, Guanajuato, and Mexico City, reported a 9.5 % prevalence of hospital infections from which the highest prevalence (16.2 %) was observed at the IM area, followed by the SUR (12.5 %), the traumatology and orthopedics (5 %), and the GYNE and obstetrics (3.1 %). The high prevalence of infections at the IM in this study, is probably associated to the location of the urology service at the IM, where most urine culture analysis are performed (^{Castañeda-Martínez and Valdespino-Padilla, 2015}).

In relation to EC, the second service unit with the major number of isolates, it showed the highest demand for laboratory requests. Therefore, this area can be regarded as an indicator of frequency and type of bacterial infections affecting a community.

On the other hand, a 2014 study of 360 samples at the Fernando Troconis University Hospital in Colombia, reported S. haemolyticus as the most prevalent bacterium at the ICU service (81.1 %) (^{Vélez-Pereira and Caicedo, 2014}). In this study, this bacterium also showed the highest prevalence at the ICU service (20 %). It should be noted that this bacterium can be easily transmitted between the potential sources of infection and hosts, such as patients, medical personal, and routine equipment.

Apparently, the uncontrolled application of broad-spectrum antibiotics in ICU patients promotes the settlement of bacterial infections by opportunistic pathogens. This may be due to unawareness of the site of infection and the etiological agent, or prolonged use of catheters and often probes, as well as the contamination provoked by the personnel in charge of patient care (^{Russotto et al., 2015}).

Prevalence by age and sex groups

Table 3 shows the prevalence of the main bacterial species isolated by age groups from 834 isolates. E. coli was predominant in all age groups, and increased with age, being those over 60 years (52 %) the most susceptible age group. A study carried out at two cities of Honduras in 2014, analyzed 602 bacterial isolates, of which 84 (14 %) were from the 0-10 years group, 30 (5 %) from the 11 - 20 years group, 260 (43 %) from the 21 - 60 years group, and 205 (34 %) from people over 61 years old. In addition, authors observed a higher prevalence of hospital infections in people over 18 years of age, but particularly in the age over 60 years (^{Zúniga-Moya et al., 2016}). It was proposed that older people are more sensitive to various stressors associated to a progressive loss of cellular homeostasis which may lead to immunosenescence and greater exposure to different ailments (^{Ciaglia et al., 2022}).

On the other hand, Table 4 shows E. coli as the most isolated bacteria from male and female patients. A higher prevalence was observed in female (50.8 %), with 251 isolates, than in male (35 %) patients with 119 isolates (p < 0.05). Furthermore, male had a higher prevalence of infections by P. aeruginosa (9.4 %), S. aureus (8.2 %) and E. cloacae (7.3 %).

The disparity in infection prevalence between male and female patients may be explained by the anatomical and physiological differences which may promote a greater predisposition to urinary tract infections, particularly in female (^{Alós, 2005}). Likewise, other studies have reported that female are more susceptible to infections, but a clear physiological explanation for this event has not yet been proposed (^{Combes et al., 2009}).

Prevalence of resistance and sensitivity

Table 5 shows the resistance and sensitivity prevalence by the most isolated bacteria from 834 clinical samples. The highest resistance was shown by K. pneumoniae to ampicillin (AMP) (100 %) and nitrofurantoin (NIT) (50.0 %); P. aeruginosa to trimethoprim/sulfamethoxazole (SXT) (100.0 %), AMP (97.6 %), and ampicillin/sulbactam (SAM) (97.6 %); Enterobacter cloacae to cefazolin (CZO) (100 %) and NIT (69.4 %), S. aureus (90.3 %) and S. epidermidis (97.7 %) to benzylpenicillin (PEN); and E. coli to AMP (78.3 %) and SAM (68.4 %).

On the other hand, the species which showed the highest sensitivity were E. coli to meropenem (MEM) (99.4 %) and tigecycline (TGC) (99.1 %); K. pneumoniae (98.2 %) to MEM, ertapenem (ETP), and amikacin (AMK); S. aureus (98.0 %) to TGC, vancomycin (VAN), SXT, NIT, Quinupristin/Dalfopristin (QDA), Linezolid (LNZ), and Rifampicin; S. epidermidis (100 %) to TGC, QDA, NIT, and LNZ; P. aeruginosa (72.09 %)to gentamicin (GEN) and Piperacillin/Tazobactam (TZP); and E. cloacae (88.89 %)to MEM, AMK, and TGC.

Among the few published Mexican studies on bacterial resistance, the State Oncology Center of the Mexican Institute of Social Security of the State of Mexico, and Municipalities (ISSEMYM) in 2012, analyzed 4,652 samples and reported a high resistance of E. coli to AMP (86.7 %), ciprofloxacin (86.1 %), levofloxacin (85.3 %); and from 1,313 records P. aeruginosa (100 %) to tetracycline, CZO, and ceftriaxone (CRO). In addition, S. aureus, showed a high resistance to PEN (92.2 %), erythromycin (62.9 %), and clindamycin (62.4 %), and S. epidermidis with a similar resistance trend to the same antibiotics (^{Romero et al., 2013}). On the other hand, it has been recognized that the most common resistant bacteria associated to hospital infections are E. coli, K. pneumoniae, and P. aeruginosa (^{Rello et al., 2019}). In addition, an increase in resistance has been observed against CRO and amoxicillin (^{Rolain et al., 2016}). It is probably that the number of bacteria capable of generating extended-spectrum beta-lactamases (ESBL) has increased worldwide (^{Hermann et al., 2006}).

Antimicrobial resistance mechanisms can be categorized into four primary groups: (1) restricting drug uptake, (2) altering drug targets, (3) deactivating drugs, and (4) actively expelling drugs. Intrinsic resistance may involve limited drug uptake, drug deactivation, and drug efflux; while acquired resistance mechanisms may include modifying drug targets, drug deactivation, and drug efflux. Variations in structures lead to differences in the mechanisms used by gram-negative and gram-positive bacteria. Gram-negative bacteria use all four main mechanisms, while gram-positive bacteria use less frequently the limiting drug uptake (due to the absence of an LPS outer membrane) and lack the capacity for certain types of drug efflux mechanisms (^{Reygaert, 2018}).

Finally, a high prevalence of A. baumanni (8.61 %), E. coli (38.03 %), K. pneumoniae (37.41 %) and P. aeruginosa (15.94%) have been reported by some hospitals in Saudi Arabia (^{Aloraifi et al., 2023}). A systematic review, from 2011 to 2021, reported that these bacteria are the most common multridrug-resistant specifically to aztreonam, ceftazidime, cefotaxime, related cxyimino-β-lactams, cephalosporins, and penicillins (^{Borgio et al., 2021}; ^{Pishtiwan and Khadija, 2019}). Apparently, the trends of bacterial resistance based on previous few published Mexican study data and those from this study remains unchanged, and probably is expected that this health problem increases in Mexico.