Reviews/Novel Antibiotics/CRE bacteraemia carries 30 to 45 percent 30-day mortality
Amr Burden · NOVEL-AB-BURDEN-02 · v2
Bloodstream infections caused by carbapenem-resistant Enterobacterales kill a large share of patients, and several studies here put that share around a third to a half, which overlaps the claimed range. But the studies measure death at different moments: some count deaths in hospital or 'until discharge', some count deaths within two weeks, and only a few count deaths strictly at 30 days. The strict 30-day studies pull in opposite directions, one much higher and one lower than the claimed range. Studies focused on the KPC strain, the most common resistant type here, repeatedly report mortality above the upper end of the range. The panel judged the claim roughly right in magnitude but not a clean, settled 30-day figure, because no single pooled study pins the range down and the timeframes do not match.
Broad carbapenem-resistant bloodstream-infection cohorts report mortality near the claimed range: 35% in PANORAMA [10], 34% in Zarkotou [15] and 36.9% in an NDM leukaemia cohort [1]. KPC-specific bacteraemia cohorts run higher, at 48% [17], 57.6% [19] and a strict 30-day figure of 60% [16], above the upper bound. The one large strict-30-day carbapenem-resistant figure drawn across mixed infection sites is lower, at 23.8% [8]. Most in-range figures are in-hospital or infection-related rather than strict 30-day outcomes. The contested point: The division is on timeframe and on the upper bound: the in-range figures [10][15][17] are in-hospital or infection-related rather than 30-day, while the only strict-30-day bacteraemia figure is 60% [16] and KPC cohorts [17][19][16] repeatedly exceed 45%. Against this, the strict-30-day mixed-site figure sits below the range at 23.8% [8].
All 21 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Novel Antibiotics, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
Outcome-timeframe mismatch: the claim specifies 30-day mortality, yet the in-band supporting figures are in-hospital ([10] 35%) or infection-related to discharge/death ([15] 34%, [17] 48%) or within 2 weeks ([19] 57.6%), not strict 30-day. The only strict-30-day CRE-bacteraemia figure, [16], is 60% — above the claimed ceiling.
Upper-bound overreach: KPC-producing bacteraemia — the dominant carbapenemase phenotype in this corpus — repeatedly exceeds 45% ([16] 60%, [19] 57.6%, [17] 48%), so the claimed band understates mortality for a large share of CRE bacteraemia.
Where it remains uncertain. KPC-specific bacteraemia reaches 48% [17], 57.6% [19] and a strict 30-day 60% [16], breaching the upper bound, while the strict 30-day mixed-site figure is 23.8% [8], below it.
Aleatory
Above-band figures rest on small samples such as the 26-patient KPC leukaemia series [19], so chance variation in sick subpopulations drives the spread.
Epistemic
No corpus study provides a pooled, bacteraemia-specific 30-day CRE mortality figure; the range is read off scattered single-centre cohorts [10][15][16][17][19].
Inferential
In-range figures are in-hospital [10], infection-related [15][17] or within two weeks [19], not strict 30-day, so applying them to a 30-day claim is an inference the sources do not directly make.
The load-bearing papers from a corpus of 21, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
NDM-producing K. pneumoniae bacteraemia mortality 36.9%, inside the band
“Patients in the NDM group experienced significantly higher mortality than those in the comparator group (36.9% vs. 12.5%; OR 4.09, 95% CI 2.19-7.65; p < 0.001).”
Panel auditInternal validity: medium · scope of inference: NDM-Kp bloodstream infection in adults with acute myeloid or lymphoblastic leukaemia at one centre, 2020-2025.Key limitations: single centre · acute leukaemia only · timepoint not specified as 30-day · carbapenemase subtype NDM only
Strict 30-day CRE mortality across mixed infection sites 23.8%, below the band; shows bacteraemia-specific figures are higher than all-site
“The 30-day mortality (95% CI) was 23.8% (18.8-29.6), 10.6% (7.2-15.2), and 8.4% (6.5-10.6), respectively.”
Panel auditInternal validity: high · scope of inference: CRE infections (mixed sites) across 50 European hospitals, 2016-2018, attributable mortality.Key limitations: bacteraemia not reported separately · mixed infection sites dilute estimate · aggregate 23.8% below claim band
Large multinational CRE bacteraemia cohort; 35% sits inside the claimed band (in-hospital mortality)
“Crude mortality was 20% (35 of 174 patients) for patients with CSE bloodstream infection and 35% (43 of 123 patients) for patients with CRE bloodstream infection.”
Panel auditInternal validity: high · scope of inference: CRE bloodstream infection in tertiary hospitals across ten low- and middle-income countries, 2014-2015.Key limitations: crude/in-hospital rather than stated 30-day · LMIC tertiary setting · moderate sample
Only strict-30-day CRE-bacteraemia figure in corpus — 60%, above the claimed ceiling
“The overall 30-day mortality rate was 60%; in contrast, 95% of ST16 cases were fatal.”
Panel auditInternal validity: medium · scope of inference: KPC-KP bloodstream infection in a CC258-endemic Brazilian teaching hospital, 2014-2016.Key limitations: endemic high-mortality setting · single centre · high severity scores · clone-specific extremes
4 of 21 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
The claim is directionally correct and overlaps several direct cohorts, but strict-30-day and KPC-specific bacteraemia data breach its upper bound and no pooled estimate pins the range, warranting an Uncertain verdict.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Pooled strict 30-day CRE bacteraemia estimate. The corpus offers no meta-analytic or multi-cohort pooled figure anchored to 30-day all-cause mortality in CRE bloodstream infection; the range is inferred from scattered endpoints. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A systematic review and meta-analysis restricted to CRE bacteraemia with 30-day all-cause mortality as the common endpoint.
Open question: What is the strict 30-day all-cause mortality of CRE bacteraemia across an unselected, multicentre population? Why it is still open: The only strict-30-day bacteraemia figure in the corpus is a single-centre 60% [16]; broad cohorts report other endpoints [10][15].
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19