Charted: Acinetobacter Carbapenem Resistance Hits 54.3% in WHO GLASS Bloodstream Isolates

Bloodstream infections are the severity end of bacterial disease — and the place where antibiotic failure shows up first in global surveillance. WHO’s Global Antimicrobial Resistance and Use Surveillance System (GLASS) now publishes adjusted 2023 estimates across more than 23 million bacteriologically confirmed infections from 104 reporting countries. This desk cut asks one question: which pathogen–drug cells, and which WHO regions, post the highest share of bloodstream isolates resistant to last-line and first-choice drugs?

The interactive dashboard above is built as a resistant-isolate share lens. Toggle Peak cells, Regions, Trends, and US vs GLASS. On peaks, filter last-line carbapenems versus first-choice agents. On regions, flip among Acinetobacter carbapenem, K. pneumoniae carbapenem, E. coli carbapenem, K. pneumoniae 3GC, E. coli 3GC, and MRSA. The punchline is multi-cell, not a single slogan. Globally, Acinetobacter · imipenem sits at 54.3% resistant (95% CrI 49.3–59.2). African Region E. coli third-generation cephalosporin resistance hits 70.7%. South-East Asia K. pneumoniae carbapenem resistance reaches 41.2%. Eastern Mediterranean MRSA bloodstream share is 50.3%. Same surveillance system — four different failure tips.

The resistant-isolate scoreboard

Read the table as a family of peak cells, not one global average. Carbapenem rows are the last-line story: Acinetobacter already fails more than half the time globally, and regional peaks push into the high sixties and seventies. Third-generation cephalosporin rows are the first-choice story: E. coli and K. pneumoniae lose the drugs clinicians reach for first, especially in Africa and South-East Asia. MRSA is a separate Gram-positive lane — lower globally at 27.1%, but still one in two bloodstream isolates in the Eastern Mediterranean Region.

Peak cells: where last-line and first-choice break

Open Peak cells. The ranked bars put African K. pneumoniae cefotaxime first at 77.8%, then African Acinetobacter imipenem at 71.0%, African E. coli 3GC at 70.7%, and SE Asia E. coli cefotaxime at 67.3%. Filter Last-line ★ and the ladder collapses to carbapenem peaks: Africa Acinetobacter, EMR Acinetobacter, SE Asia K. pneumoniae, SE Asia E. coli. Filter First-choice and the 3GC / MRSA cells dominate.

The companion global scoreboard panel shows adjusted bloodstream shares for the same pathogen–drug pairs without the regional tip: K. pneumoniae 3GC 55.2%, Acinetobacter imipenem 54.3%, E. coli 3GC 44.8%, MRSA 27.1%, Salmonella ciprofloxacin 18.0%, K. pneumoniae imipenem 16.7%, E. coli imipenem 2.4%. Global medians hide the regional extremes — which is why the peak-cell view exists.

Regions: the map disagrees by drug

Toggle Regions. Flip the metric control. On Acinetobacter · carbapenem, Africa and the Eastern Mediterranean lead; Europe sits far lower. On K. pneumoniae · carbapenem, South-East Asia leads at 41.2% while Europe and the Western Pacific stay in single digits to low teens. On E. coli · 3GC, Africa’s 70.7% dwarfs Europe’s 19.9%. On MRSA, the Eastern Mediterranean’s 50.3% sits against Europe’s 9.7%.

The radar panel overlays four regions across six metrics at once. Africa and South-East Asia inflate the carbapenem and 3GC axes; Europe compresses almost every spoke. That is the geographic content of GLASS bloodstream resistance in one view: failure is thickest where health-system capacity is thinnest, and the correlation WHO reports between median bloodstream AMR and the UHC service coverage index is strongly inverse (r = −0.77). Surveillance coverage itself correlates inversely with reported median resistance (r = −0.74) — a reminder that sparse, tertiary-heavy sampling can inflate apparent prevalence even as true burden remains high.

Trends: level is not slope

Open Trends. The scatter plots 2023 resistance level against median annual percentage change from 2018–2023. Carbapenem tips for K. pneumoniae and E. coli sit mid/low on the level axis but high on slope: +15.3% and +12.5% per year respectively. Acinetobacter imipenem already sits high on level (54.3%) and still climbs (+5.3%/yr). E. coli 3GC is high and roughly stable (+1.3%/yr). MRSA trends slightly downward on the global median (−2.5%/yr) while remaining a severe regional problem in EMR.

This is the desk distinction that matters for procurement and stewardship. A drug that is already failing half the time (Acinetobacter carbapenems) is a stock crisis. A drug that is still rare globally but rising double-digits annually (E. coli imipenem at 2.4% with +12.5%/yr) is a flow crisis. Treating only today’s level as the signal underweights the carbapenem trajectory in Enterobacterales.

US comparator: different frame, useful contrast

Toggle US vs GLASS. CDC EIP invasive E. coli surveillance is not a GLASS cell — different sampling frame, different geography, different clinical mix. It still helps desks avoid treating WHO global medians as if they described US hospital wards. Invasive E. coli ESBL share sits near 14%; ceftriaxone resistance exceeds 15%; fluoroquinolone resistance lands near 26–27%. Among carbapenemase-producing E. coli in EIP samples, blaNDM has become the dominant carbapenemase (~73% of CP E. coli in 2023 samples). Beside that, WHO’s global E. coli 3GC bloodstream share of 44.8% is a different universe — useful as a scale check, not as a pooled average.

The companion panel shows GLASS reporting coverage by WHO region in 2023: South-East Asia 90.9% of Member States reporting, Eastern Mediterranean 76.2%, Europe 58.5%, Africa 57.4%, Western Pacific 37.0%, Americas 20.0%. High resistance and incomplete participation often coincide. Desks should read thick African and EMR peaks with that coverage caveat attached — not as a reason to dismiss the signal, but as a reason to demand denser surveillance.

AWaRe use and the syndemic frame

GLASS sits beside WHO’s antimicrobial-use ledger. Access-group antibiotics — the preferred first-choice set — made up only 52.7% of global use in 2022 against a UN political-declaration target of 70% Access by 2030. Watch-group agents accounted for 45.3%, exceeding 70% of consumption in nearly one third of countries. Reserve agents remain rare at 0.3%. When 3GC drugs fail, clinicians escalate to carbapenems; when carbapenems fail, options collapse into scarce Reserve agents that are costly and unevenly available in LMICs.

That escalation ladder is why bloodstream resistance shares are not a microbiology curiosity. They are a treatability map. Carbapenem-resistant K. pneumoniae and Acinetobacter are associated with fatality exceeding 30% in published burden work WHO cites. Third-generation cephalosporin-resistant E. coli and MRSA drive high incidence and disability even when last-line agents still work. The syndemic WHO describes — high resistance where UHC coverage and diagnostic capacity are weak — means the places least able to confirm and treat resistant bacteraemia are often the places reporting the thickest resistant-isolate shares.

Caveats desks should keep on the page

• Adjusted ≠ crude. GLASS 2025 bloodstream percentages are Bayesian, coverage- and population-weighted estimates with credible intervalsnot simple lab medians.
• Sparse surveillance inflates. Where few tertiary hospitals report, case-mix bias can push resistance up; the inverse coverage–AMR correlation is part of the signal, not noise to ignore.
• Regional peaks ≠ every country. National maps in the WHO annex vary widely inside each region; this post ranks pathogen–drug–region cells, not a country league table.
• US EIP ≠ GLASS. Do not average CDC invasive E. coli shares with WHO regional percentages into one “global US-adjusted” rate.
• Trend ≠ inevitability. Annual percentage changes are modelled medians over 2018–2023; stewardship, infection prevention, and Access-share targets can bend slopes.
• Infection type matters. This cut is bloodstream only. Urinary and gastrointestinal resistance patterns differ and should not be pasted onto sepsis empiric protocols.

Bottom line for desks

WHO GLASS bloodstream surveillance in 2023 puts last-line failure in plain view: Acinetobacter carbapenem resistance at 54.3% globally, with African and EMR regional peaks above 65%. First-choice failure is thicker still for Enterobacterales 3GC drugs — 70.7% for African E. coli, 77.8% for African K. pneumoniae cefotaxime. The carbapenem slope on E. coli and K. pneumoniae (+12.5% and +15.3% per year) is the forward-looking alarm even where today’s levels remain lower than Acinetobacter. Read resistant-isolate share as a treatability meter by pathogen, drug, and region — and keep CDC US figures in a separate column.

Primary sources: WHO Global antibiotic resistance surveillance report 2025; WHO news briefing (13 Oct 2025); CDC AMR facts & surveillance.