Trends in antimicrobial susceptibility patterns in King Fahad Medical City, Riyadh, Saudi Arabia

Discussion

The global action plan to fight AMR adopted by the WHO sets several objectives and recommendations to enhance antibiotic surveillance and research, to strengthen knowledge and improve the awareness regarding AMR, to optimize the use of antimicrobial agents, and to decrease the rate of infections.2 Our study is in line with these recommendations. Gram-negative isolates predominated in our results, with E. coli being the most frequent isolate, followed by P. aeruginosa and K. pneumoniae. Our results were close to what was reported by Amer et al,4 who reported an increasing in resistance rate of gram negative isolates. Our data are in agreement with a review of 45 published articles addressing the AMR prevalence in the gulf corporation countries over 2 decades. Escherichia coli, K. pneumoniae, and P. aeruginosa were the most prevalent agents; however, MRSA rate was higher in our present study at 17.5% compared to 5.4% in theirs.5

A recent survey conducted in a rehabilitation center showed similar results with regard to the most frequently detected organisms, but the susceptibility patterns varied. For example, in the previous study, Pseudomonas was most susceptible to ciprofloxacin and gentamicin, whereas in our study, amikacin and piperacillin-tazobactam had higher susceptibility rates.6 The noted reduction in ceftazidime and meropenem susceptibilities to gram-negative organisms, especially Pseudomonas, is likely because of selection pressure. On the other hand, piperacillin-tazobactam susceptibility remained static, and amikacin had the best susceptibility profile.

Isolates of vancomycin-resistant Enterococcus, carbapenemase-producing K. pneumoniae, and ESBLs tended to increase over the study period; this result is an alarming signal that strict policies for infection control transmission-based precautions and antimicrobial prescriptions need to be followed; moreover, these data need to be taken into consideration when selecting empirical treatment for critically ill patients.

Streptococcus pneumoniae is an important gram-positive organism that causes invasive diseases in children <5 years of age with substantial morbidity and mortality. Before 1967, S. pneumoinae was almost uniformly susceptible to penicillin. The emergence of resistance gradually increased worldwide over time.7 Alteration in the serotype distribution along with the susceptibility patterns has been observed after the introduction of the pneumococcal conjugate vaccines, particularly, post the 7-valent in 2000 and 13-valent in 2010. PCV7 was incorporated into the national immunization program in Saudi Arabia in 2009 followed by PCV13 in 2013. In the results of the present study, there was a significant and steady increase in penicillin susceptibility, with an increase from 66% in 2010 to 100% in 2015. Moreover, S. pneumoniae was the second least detected gram-positive organism at only 2.6%.

In a national prospective surveillance performed in 12 hospitals from 7 different regions in Saudi Arabia between 2007 and 2009, 78% isolates were found to be multidrug resistant. Penicillin susceptibility was observed in only 30% isolates, yet all isolates were sensitive to ceftriaxone.8 In another study published in 2015, recruited data between 2005 and 2010 to assess the antibiotic resistance and serotype differences before and after PCV7 conjugate vaccine administration showed that 66% isolates were resistant to penicillin and 62% were resistant to erythromycin. It was noted that penicillin susceptibility was 54.6% immediately after switching to PCV13.9,10 Same findings were noted in a study conducted at Twam Hospital in the United Arab Emirates over an 8-year period from 2004 to 2011.11

Unfortunately, there are no reports of recent surveillance data in Saudi Arabia to assess the antimicrobial susceptibility trends after PCV13 administration. However, a survey conducted in 4 Gulf and near East countries between 2011 and 2013 showed the penicillin susceptibility was variable, with >85% overall susceptibility and 100% ceftriaxone susceptibility in all countries.12 A similar study conducted in Kuwait to evaluate the impact of PCV7 and PCV13 over a 10-year period documented a drop in penicillin resistance from 67% to 46%.13 A systemic review and meta-analysis study published in 2017 analyzed 68 studies and surveillance data from 2000 to 2015 to define the serotypes causing IPD after PCV vaccination. They found a reduction in IPD caused by vaccine serotypes, but nonvaccine serotypes predominated as the cause of IPD (replacement disease).14

To examine the impact of PCV on antibiotic resistance, several studies have been conducted in different countries, and their results were encouraging.15-17 One study was conducted to compare susceptibility between pre-vaccination and post-vaccination eras (in the 2007–2009 and 2010–2014 periods) for S. pneumoniae isolates from patients with otitis media. They showed significant improvement in penicillin susceptibility, from 37% in the pre-PCV13 vaccination period to 51% in the transitional period and to 100% in the post-PCV13 vaccination period. Ceftriaxone susceptibility also improved from 95% to 100%.15

In line with these studies, our results showed improved penicillin susceptibility. This finding reflects the effect of the post-vaccination era, as it is highly documented that vaccination leads to a dramatic decrease in the nasopharyngeal carriage rate and IPD by vaccine serotypes in children, in addition to the development of herd immunity in adults.14 Moreover, these findings might lead to major changes in empirical antimicrobial selection when coverage of S. pneumoniae is needed. Physicians may reassess the needs for the previously used practice of broad-spectrum coverage to a narrower-spectrum antibiotic. This practice is noted in the recently published guideline for community-acquired pneumonia endorsed by the Saudi Pediatric Infectious Disease Society, when experts decided to choose ampicillin as initial therapy for treating uncomplicated pneumonia.18

The exact prevalence of MRSA, either community-acquired (CA-MRSA) or hospital-acquired (HA-MRSA) infections, in Saudi Arabia is unknown despite extensive studies. In one review, it was estimated to be 35.6% with great variations among regions. Another study reported an MRSA rate of 23.2% among S. aureus carriers.19,20 However, it is estimated by the Center of Disease Control’s report on MRSA tracking that 33% of people are typically colonized with Staphylococcus and 2 in 100 carry MRSA.21

In our present data, S. aureus was the second most common gram-positive isolate (27.5%), among which 17.5% isolates were MRSA (defined as oxacillin resistant S. aureus (MIC >4 in Mueller-Hinton agar). The calculated rate of MRSA among S. aureus was 39%. It is difficult to extrapolate the CA-MRSA rate based on antibiograms alone, because it is distinctively different from HA-MRSA both genetically and phenotypically. Infection control (IC) policies have a fundamental role in decreasing HA-MRSA rates, although the effectiveness of active surveillance in HA-MRSA prevention is still controversial.22-24 A recent study documented that early identification of MRSA using rapid diagnostic technologies (PCR) along with timely implementation of infection control strategies have a great financial impact.25

At KFMC, the IC department designed an MRSA-prevention program, which includes 2 components, preemptive screening of at-risk patients and implementation of strict transmission-based precautions for the tested positive patients. There was a reduction in HA-MRSA, from 0.17 case per 1000 to 0.03 case per 1000 between years 2007 to 2009.26 Wider surveillance is needed in the future to determine the rates of CA-MRSA and HA-MRSA, because it has a direct effect on the empirical therapy of skin and soft-tissue infections as well as osteoarticular infections.

The infection rates of ESBLs, such as E. coli, K. pneumoniae, and P. mirabilis are increasing. These infections are usually associated with longer hospital stays, increased costs, and worse outcomes if they are not anticipated early and treated, especially in neonates. Therefore, we aimed to assess their trends during the study period. The overall percentage of total ESBLs of the 3 most common isolates (E. coli, K. pneumoniae, and P. mirabilis) including bloodstream and urine infections, was estimated to be 20.8%, with a steadily increasing frequency over the years from 32% to 41% per year for E. coli.

In comparison, a systemic review and meta-analysis of bloodstream infections caused by ESBLs showed the overall percentage was 9%, with a 3% annual increase. The higher prevalence and mortality rates were observed in neonates. In Saudi Arabia, it was noted to range between 6% and 38.5% in different regions.27,28 The risk of increasing rates of ESBLs is significant and alarming; it needs to be considered in the empirical coverage of bloodstream and urinary tract infections, especially in critically ill patients.

In light of the above mentioned findings, particularly, with regard to S. pneumoniae susceptibility, lack of exact prevalence of certain organisms such as MRSA, and the rising trends of the MDRO, further research and multi-center surveillance studies are extremely needed to tackle the antimicrobial resistance.