Enterobacter resistance shifted unevenly across 12 years in southern Saudi Arabia
Laboratory surveillance across a tertiary-care hospital in southern Saudi Arabia found a resistance pattern in Enterobacter species that changed unevenly over time, with some drugs losing activity while others remained comparatively stable. The 12-year review, spanning August 2012 through August 2024, captured 1,319 non-duplicate adult isolates and traced non-susceptibility trends across multiple antibiotic classes using contemporaneous CLSI criteria. The most important signal was not a single uniform rise in resistance, but a fragmented pattern: aminoglycoside and cefepime non-susceptibility fell, carbapenem non-susceptibility climbed, and fluoroquinolone and piperacillin–tazobactam trends required caution because breakpoint changes complicated interpretation.
The cohort was dominated by the Enterobacter cloacae complex, which accounted for 91.8% of isolates. In practical microbiology terms, that matters because the complex contains closely related species with clinically meaningful differences in susceptibility patterns, yet routine phenotypic methods do not always separate them with precision. In everyday hospital practice, this leaves stewardship teams working with a group of organisms that can behave like a moving target: one year apparently manageable with a cephalosporin-based regimen, another year far less predictable because of AmpC-mediated resistance and additional acquired mechanisms.
Resistance did not move in one direction across all drugs.
Adjusted annual percent change showed declining odds of non-susceptibility for amikacin at −6.1%, gentamicin at −9.4%, and cefepime at −9.1%. Those falling lines suggest that at least some local susceptibility remained preserved over time, preserving the possibility of targeted therapy when laboratory results support it. By contrast, carbapenem non-susceptibility increased at a rate between +13.1% and +15.8%, a shift that has immediate clinical weight because carbapenems are often reserved as a dependable backstop for serious infections caused by Enterobacterales. When that backstop weakens, empiric treatment becomes harder to trust and definitive treatment narrows quickly.
Carbapenem non-susceptibility rose early while DTR stayed less common
The study separated carbapenem-resistant Enterobacterales from difficult-to-treat resistance, and that distinction exposed an important dissociation. CRE rose early and remained common, reaching 28.7% overall, while DTR appeared later and stayed infrequent at 7.5% of eligible isolates. DTR increased at an annual percent change of +16.4% and reached a highest annual value of 20.2% in 2023, but the timing and frequency never mirrored the broader carbapenem-resistant signal. Clinically, this means a hospital can face a serious carbapenem problem without yet confronting the more extreme endpoint of resistance to all first-line agents.
The biology fits the pattern. Carbapenem non-susceptibility often emerges through mechanisms that do not automatically eliminate every other treatment option, whereas DTR requires a broader collapse in antimicrobial activity across multiple classes. In this dataset, the two phenotypes moved on different timelines, arguing against the false assumption that every carbapenem-resistant isolate is already resistant to nearly everything useful. For infection control and bedside decision-making, that difference changes how urgently clinicians need to escalate therapy, seek susceptibility data, and avoid broad empiric overreach.
Inpatient isolates carried a heavier resistance burden than outpatient isolates
Setting mattered throughout the analysis. Inpatient isolates were consistently more non-susceptible than outpatient isolates by roughly 16 to 39 percentage points across several agents, including gentamicin, piperacillin–tazobactam, cefotaxime, cefepime, ceftazidime, and ceftriaxone. The contrasts were large: gentamicin non-susceptibility was 31.7% in inpatient isolates versus 14.8% in outpatient isolates; piperacillin–tazobactam was 38.7% versus 23.1%; cefotaxime was 63.9% versus 41.5%; cefepime was 45.7% versus 28.8%; ceftazidime was 61.4% versus 35.3%; and ceftriaxone was 57.4% versus 18.8%. All comparisons met statistical significance at p ≤ 0.003.
Those gaps are exactly what one would expect in a tertiary-care setting where device-related infection, prolonged hospitalization, prior antibiotic exposure, and high-acuity illness favor selection of resistant strains. The hospital environment amplifies selective pressure: repeated exposure to broad-spectrum agents rewards organisms carrying AmpC overexpression, porin changes, or carbapenemase-associated pathways. Outpatient isolates, by contrast, more often reflect less intense antimicrobial selection, which helps explain the lower resistance burden in that setting.
Breakpoint changes complicate the levofloxacin and piperacillin–tazobactam signals
Two apparent upward trends deserve restraint. Levofloxacin increased at +6.3% and piperacillin–tazobactam at +7.0% in the main adjusted analysis, but those increases did not persist in breakpoint-comparable analyses. The authors explicitly cautioned that these were not unequivocal biological increases. For clinical interpretation, this is an important boundary between true epidemiologic drift and changes created by evolving interpretive standards. A lab trend can look like worsening resistance simply because the rules used to classify susceptibility have changed.
Ciprofloxacin stayed stable over time, which further supports the idea that not every fluoroquinolone-related signal moved in the same direction. In antimicrobial surveillance, consistency across closely related drugs can be informative, but inconsistency often points to a mixture of biology, prescribing pressure, and breakpoint methodology rather than a single monotonic resistance event.
Stewardship needs antibiotic-specific surveillance, not one broad resistance label
The most useful lesson from this 12-year series is methodological as much as microbiologic. Enterobacter resistance could not be summarized cleanly with one headline number, because the pattern differed by drug, by ward setting, and by resistance phenotype. The authors’ conclusion follows directly from the data: antibiotic-specific trend analysis should sit alongside clinically meaningful phenotypes such as CRE and DTR when hospitals report surveillance results and shape stewardship policy.
For clinicians, that approach has direct bedside consequences. A cephalosporin may remain useful in one subset of isolates even while carbapenem non-susceptibility is worsening. An isolate that meets CRE criteria may still differ sharply from a truly difficult-to-treat strain. And when breakpoint revisions distort long-term trends, surveillance systems need analytic safeguards before they translate those signals into policy.
The dataset itself was complete and de-identified, and the analysis adjusted for care setting, specimen source, age, and sex. Even so, the largest unresolved issue is not whether resistance was present—it clearly was—but how best to interpret long-term surveillance when taxonomic complexity, local ecology, and evolving breakpoints all shape the numbers at the same time.