Model-based evaluation of admission screening strategies for the detection and control of carbapenemase-producing Enterobacterales in the English hospital setting

Pople, Diane; Kypraios, Theodore; Donker, Tjibbe; Stoesser, Nicole; Seale, Anna C.; George, Ryan; Dodgson, Andrew; Freeman, Rachel; Hope, Russell; Walker, Ann Sarah; Hopkins, Susan; Robotham, Julie

doi:10.1186/s12916-023-03007-1

Table 2 Model parameters

From: Model-based evaluation of admission screening strategies for the detection and control of carbapenemase-producing Enterobacterales in the English hospital setting

Process/Parameter		Value				Source
Hospital beds^a
Number of beds in hospital		540				Informed by UK acute trust data [29]
Ward beds
Number of beds on ward		60				Informed by UK study data [6]
Number of admissions^b
Non-local patient	Fixed maximum number of admissions	1				Assumption
Local patient	Probability of maximum admissions = \(\alpha\) (given first admitted with t simulation days remaining)	Fitted values in Additional File 1 §1.a.ii				HES [30]
Geographic source of patient^b Proportion of admissions from each geography		High-prevalence generic hospital		Low-prevalence generic hospital
Local	same region as hospital	0.969263		0.96897		Donker et al/HES [30, 31]
Low-prevalence areas	non-local low-prevalence regions (England)	0.006274		0.00812		Donker et al/HES [30, 31]
High-prevalence areas	non-local high-prevalence regions (England)	0.002251		0.000698		Donker et al/HES [30, 31]
Non-England	non-England geography	0.022212		0.022212		HES [30]
Length of stay^b
Duration of stay in hospital, distribution with median/IQR – given leaving by discharge		2 days / 1 – 5 days				HES [30]
Duration of stay in hospital, distribution with median/IQR – given leaving due to death		8 days / 3 – 18 days				HES [30]
Additional duration of stay if colonised (with or without CPE BSI)		7 days				Knight et al [12]
Probability of death
Probability of leaving hospital due to death – without CPE BSI		2.9%				HES [30] all admission
Probability of leaving hospital due to death – with CPE BSI		48.4%				Xu et al/Budhgram et al/ Hauck et al [32,33,34]
Colonisation on admission^b Prevalence of CPE colonisation upon first admission		High-prevalence generic hospital		Low-prevalence generic hospital		HES/ERS/Donker et al/ European Antimicrobial Resistance Surveillance Network [30, 31, 35, 36]
Local	same region as hospital	0.000532		0.000042
Low-prevalence areas	non-local low-prevalence regions (England)	0.000042		0.000042
High-prevalence areas	non-local high-prevalence regions (England)	0.000532		0.000532
Non-England	non-England geography	0.013		0.013
Infection on admission
Probability of patient colonised with CPE at admission presenting with a CPE BSI infection		310.88 per 100,000 admissions				PHE/HES [30, 37, 38]
Force of infection Transmission parameter \(\beta ={\beta }_{0}+{\beta }_{1}({n}_{C}+{n}_{I})\)		From unknown sources (β₀)		Per infectious patient on ward (β₁)		Estimated from the Bayesian Framework ward models using UK study data [6]
Ward A	Minimum	0.00001308		0.00008105
	Median	0.00040719		0.00036252
	Maximum	0.00188901		0.00075283
Ward B	Minimum	0.00000698		0.00001354
	Median	0.00023081		0.00024106
	Maximum	0.00108142		0.00072196
Ward C	Minimum	0.00000335		0.00002051
	Median	0.00007870		0.00038651
	Maximum	0.00035555		0.00148161
Progression rate
Probability of progression from CPE colonisation to CPE BSI (within hospital)		30.42 per 100,000 bed-days				PHE/ HES [30, 37, 38]
Clearance (in the community)
Probability still CPE colonised, given readmitted after n days		\(p\left(n\right)=\alpha +\left(1-\alpha \right){e}^{-\lambda n}\) \(\alpha\) = 0.11724010, \(\lambda\) = 0.04269721				Estimated from survival analysis of UK study data [6]
Recent positive period
Duration of a Positive CPE status on patient’s record after discharge		365 days				Toolkit [10]
Swab Test
Local CPE testing of rectal swabs from screening and from contact tracing sampling, test methodology used for parameterisation		Parameterised using data for MacConkey agar				Surveys of Acute trusts in England [3, 39]
Confirmation test
Confirmation / reference lab testing of isolates for CPE, test methodology used for parameterisation		Parameterised using data for Cepheid Xpert Carba-R				Surveys of Acute trusts in England [3, 39]
Clinical test
Local blood sample testing for carbapenem sensitivity, performed for clinical purposes, test methodology used for parameterisation		Parameterised using data for VITEK 2				Surveys of Acute trusts in England [3, 39]
Test characteristics		Swab test	Confirmation test		Clinical test
Sensitivity		0.839	0.966		0.786	Meunier et al [40]
Specificity		0.917	0.986		0.820	Meunier et al [40]
Turnaround time (including to/from lab) – positive result		3 days	2 days		2 days	Expert opinion
Turnaround time (including to/from lab) – negative result		2 days	7 days		2 days	Expert opinion
Clinical sample
Probability of selection for carbapenem sensitivity testing (per timestep), given suspected negative patient		1.58 per 1000 bed-days				Linked laboratory/ bed data from UK acute Trust [6]

^abespoke values estimated, from the same source, for individual trust verifications (values here are for generic high- and low-prevalence area hospitals)
^bbespoke values calculated, from the same source, for referral region-typical hospitals and for individual trust verifications (values here are for generic high- and low-prevalence area hospitals)

Back to article page

ISSN: 1741-7015

Contact us

Submission enquiries: bmcmedicineeditorial@biomedcentral.com
General enquiries: info@biomedcentral.com

BMC Medicine

Contact us