Accepted Manuscript Hospitalisation in double-occupancy rooms and the risk of hospital-acquired influenza: a prospective cohort study Elodie Munier, MD, MSc, Thomas Bénet, MD, MPH, Corinne Régis, RN, Bruno Lina, MD, PhD, Florence Morfin, Pharm, PhD, Philippe Vanhems, MD, PhD PII:
S1198-743X(16)00026-4
DOI:
10.1016/j.cmi.2016.01.010
Reference:
CMI 503
To appear in:
Clinical Microbiology and Infection
Received Date: 12 November 2015 Revised Date:
23 December 2015
Accepted Date: 11 January 2016
Please cite this article as: Munier E, Bénet T, Régis C, Lina B, Morfin F, Vanhems P, Hospitalisation in double-occupancy rooms and the risk of hospital-acquired influenza: a prospective cohort study, Clinical Microbiology and Infection (2016), doi: 10.1016/j.cmi.2016.01.010. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACCEPTED MANUSCRIPT Hospitalisation in double-occupancy rooms and the risk of hospital-acquired
2
influenza: a prospective cohort study
3
Elodie Munier, MD, MSc1 ; Thomas Bénet, MD, MPH1,2 ; Corinne Régis, RN1 ; Bruno Lina,
5
MD, PhD3 ; Florence Morfin, Pharm, PhD3 ; Philippe Vanhems, MD, PhD1,2
RI PT
4
6 7
1
8
Lyon, 69003 Lyon, France
Infection Control and Epidemiology Unit, Edouard Herriot Hospital, Hospices Civils de
9 10
2
11
Infectiologie (CIRI) Inserm U1111, CNRS UMR5308, ENS de Lyon, UCBL1, 69007 Lyon,
12
France
M AN U
SC
Emerging Pathogens Laboratory - Fondation Mérieux, Centre International de Recherche en
13 14
3
15
Influenza, Hospices Civils de Lyon, 69500 Bron, France and VIRPATH Laboratory, EA
16
4610, Claude Bernard University of Lyon, René-Théophile-Hyacinthe Laennec School of
17
Medicine, University of Lyon 1, 69003 Lyon, France
18 19 20
Running title: Nosocomial influenza and double-occupancy room
TE D
National Reference Center for Influenzae Viruses, Virology Reference Laboratory of
21
EP
22
Corresponding author: Philippe Vanhems, MD, PhD
24
Service d’Hygiène, Epidémiologie et Prévention, Hôpital Edouard Herriot, Hospices Civils de
25
Lyon, 5 place d'Arsonval, 69437 Lyon cedex 03, France
26
Telephone: +33 4 72 11 07 21
27
Fax: +33 4 72 11 07 26
28
E-mail:
[email protected]
29 30
Text word count: 1 000
AC C
23
1
ACCEPTED MANUSCRIPT
Abstract
32
Hospitalisation in double-occupancy rooms and the risk of hospital-acquired influenza (HAI)
33
were assessed prospectively. The incidence was 2.0 for 100 patient-days in double versus 0.7
34
in single-occupancy rooms (p=0.028). The adjusted HR of HAI was 2.67 (95%CI: 1.05-6.76)
35
in patients hospitalised in double compared to single-occupancy rooms.
RI PT
31
36
Keywords: double-occupancy room; hospital-acquired infection; influenza; epidemiology;
38
cohort
SC
37
AC C
EP
TE D
M AN U
39
2
ACCEPTED MANUSCRIPT
BACKGROUND
41
With attack rates reaching 60%, hospital-acquired influenza (HAI) and influenza-like illness
42
(HA-ILI) are anything but rare [1]. Hospitalised patients exposed to infected cases or
43
healthcare workers (HCWs) are at high risk of HA-ILI [1].
44
Close contacts among patients sharing rooms offer opportunities for pathogen transmission by
45
HCWs, visitors and patients themselves. Indeed, more contacts can be counted among
46
patients hospitalised in double-occupancy rooms than those in single-occupancy rooms [2].
47
Nosocomial bacterial acquisition has been associated with the number of roommate exposures
48
per day [3].
49
The impact of double-occupancy rooms on HAI in nursing homes has been described in
50
retrospective studies [4,5], but this risk has not been quantified in acute care settings
51
prospectively with systematic virological sampling.
52
The objective was to assess the relationship between hospitalisation in double-occupancy
53
rooms and HAI risk in a hospital-based cohort.
54
TE D
M AN U
SC
RI PT
40
METHODS
56
Prospective ILI surveillance was implemented between October and April from 2004-2005 to
57
2011-2012 and in 2013-2014 at Edouard Herriot University Hospital (Lyon, France) counting
58
for 1,000 beds. This surveillance has been described elsewhere [1]. ILI was defined as fever
59
>37.8°C (in the absence of antipyretics) and/or cough or sore throat. Eligible patients
60
presented ILI ≥48 hours after admission (incident ILI) with available data on room type,
61
virological results on influenza and length of stay in the unit. No specific infection control
62
measures were implemented because of the study. Standard procedures were followed with
63
droplet precaution in case of ILI.
AC C
EP
55
3
ACCEPTED MANUSCRIPT
The primary endpoint was confirmed HAI defined as virologically-confirmed influenza
65
occurring ≥48h after admission in acute-stay units. The main exposure of interest was
66
hospitalisation in double-occupancy compared with single-occupancy room.
67
The following variables were recorded at ILI diagnosis: admission and discharge dates,
68
underlying diseases, gender, age at admission, influenza vaccination status, room type (single
69
or double-occupancy). Rooms with more than 2 beds did not exist in this hospital.
70
Nasal swabs, tested by the National Reference Center for Influenzae Viruses in Lyon, were
71
collected for each patient included by research nurses. Other respiratory viruses were also
72
tested.
73
Categorical variables were compared by the chi2 test or Fisher’s exact test, and continuous
74
variables, by the Mann-Whitney U-test. Incidence was expressed per 100 patient-days of
75
hospitalisation at risk. Data were censored at patient discharge, death or at 50 days of follow-
76
up. HAI probability according to type of room was described by the Kaplan-Meier method,
77
and survival curves were compared by the log-rank test. Univariate and multivariate Cox
78
proportional hazard models were fitted to assess relationship between type of room and risk of
79
HAI, hazard ratios and 95% confidence interval (95% CI) were calculated. Covariates with
80
p<0.20 after univariate analysis were entered in the multivariate model. All tests were 2-
81
tailed. P<0.05 was considered significant. The data were analysed by Stata software 11.1.
82
The protocol design was approved by the ethic board of the hospital. All patients received
83
study information and gave signed informed consent.
SC
M AN U
TE D
EP
AC C
84
RI PT
64
85
RESULTS
86
Overall, 93 patients, accounting for 1,775 patient-days at risk, were analysed and among
87
them, 20 (21.5%) had confirmed HAI. Among the 73 patients without HAI, 7 had other viral
88
infections: 3 by respiratory syncytial virus, 2 by picornavirus, 1 by metapneumovirus and 1 by
89
herpes simplex virus 1. 4
ACCEPTED MANUSCRIPT
Patients’ characteristics and HAI incidence are presented on Table 1. The mean age of all
91
patients included was 74.2 years. The HAI incidence rate was 0.7 for 100 patient-days in
92
patients hospitalised in single-occupancy rooms (N=9) and 2.0 for 100 patient-days in patients
93
hospitalised in double-occupancy rooms (N=11) (p=0.028). Influenza vaccination coverage
94
was not different between patients in single and double-occupancy rooms (47% vs 62%;
95
p=0.16).
96
The HAI probability was higher among patients hospitalised in double-occupancy rooms
97
compared to those in single-occupancy rooms (Figure 1). Univariate analysis indicated that
98
hospitalisation in double-occupancy rooms increased HAI risk by 2.69 (95% CI: 1.07-6.79).
99
Female gender (crude hazard ratio [cHR]=2.39, 95% CI: 0.86-6.64), age (cHR=1.00 per 1
100
year, 95% CI: 0.98-1.02), being hospitalised outside a geriatric unit (cHR=1.59, 95% CI:
101
0.64-3.93), influenza vaccination (cHR=0.81, 95% CI: 0.32-2.09), cardiac and/or respiratory
102
underlying disease (cHR=0.62, 95% CI: 0.24-1.59) were not significantly associated with
103
HAI risk. After adjustment for gender, the relative risk of HAI in patients was 2.67 (95% CI:
104
1.05-6.76; p=0.039) in case of hospitalisation in double compared to single-occupancy rooms.
105
TE D
M AN U
SC
RI PT
90
DISCUSSION
107
Higher HAI risk was detected among patients hospitalised in double-occupancy rooms
108
compared to those in single-occupancy rooms, confirming that double-occupancy rooms
109
contribute to the transmission of influenza, in acute-care hospitals. A previous study analysing
110
contact patterns between individuals in Edouard Herriot Hospital [6] reported more frequent
111
and longer contacts for patients hospitalised in double-occupancy rooms. This association was
112
noted in other settings and for other micro-organisms, mostly transmitted by contact and
113
relative risks of bacterial HA infections have been reported to be reduced by 58 to 82%
114
among patients in single-occupancy rooms compared to those in multiple-occupancy rooms
115
[7]. The added values of the study were a standardised prospective design carried out during
AC C
EP
106
5
ACCEPTED MANUSCRIPT
9 influenza seasons and virological analysis conducted by the National Reference Center for
117
Influenzae Viruses. The primary end-point was very specific to prevent misclassification bias.
118
Moreover, HAI risk was ascertained by survival analysis that took duration of exposure into
119
account [8].
120
A lack of a standardised HAI definition might be discussed [9,10]. We undertook sensitivity
121
analysis with different time interval between admission and symptoms onset. The association
122
remained consistent for time interval of 24 and 72 hours, respectively (data not shown).
123
Study limitations included possible residual confounding due to limited sample size and
124
misclassification bias related to possible intra-unit room transfers preceding onset. Linkage
125
between hospital architecture (including bed management) and the HA infection rate is not
126
obvious because of their multifactorial nature [11]. Then, environmental factors including
127
patient repartition in the unit should be considered.
128
We observed higher HAI risk among patients hospitalised in double-occupancy rooms than
129
those in single-occupancy rooms. In addition to being more appropriate for global patient care
130
regarding privacy, rest, familial support, costs and confidentiality, single-occupancy rooms
131
are reducing the risk of HA infections, including HA influenza [12,13].
AC C
EP
TE D
M AN U
SC
RI PT
116
6
ACCEPTED MANUSCRIPT
Author contributions
133
Dr Vanhems had full access to all the data in the study and takes responsibility for the
134
integrity of the data and the accuracy of the data analysis. Study concept and design:
135
Vanhems, Lina. Acquisition of data: Munier, Regis, Morfin, Lina. Analysis and interpretation
136
of data: Munier, Bénet, Vanhems. Drafting of the manuscript: Munier, Bénet, Vanhems.
137
Critical revision of the manuscript for important intellectual content: Munier, Bénet, Regis,
138
Lina, Morfin, Vanhems. Statistical analysis: Munier, Bénet. Obtained funding: Vanhems,
139
Lina. Administrative, technical, and material support: Vanhems, Regis. Study supervision:
140
Vanhems.
SC
RI PT
132
M AN U
141
Potential conflicts of interest
143
Dr Vanhems received research grants from Sanofi MSD, Biomerieux, and ANIOS and
144
consulting fees from GSK, and Biomerieux. Dr Lina belongs to the scientific council of the
145
French Influenza Clinical Information Network (Groupe d’Expertise et d’Information sur la
146
Grippe) and the European Scientific Working group on Influenza, he received congress and
147
travel fees from Biomerieux. Munier, Bénet and Morfin had no conflict of interest.
TE D
142
EP
148
Funding
150
This work was supported by a “Programme Hospitalier de Recherche Clinique (PHRC)
151
Regional” from the French Ministry of Health and Sanofi Pasteur.
152
AC C
149
153
Acknowledgements
154
We would like to thank the healthcare workers from the participating units of Edouard Herriot
155
Hospital.
7
156
ACCEPTED MANUSCRIPT
References
157 158
1.
Vanhems P, Voirin N, Roche S, et al. Risk of influenza-like illness in an acute health care setting during community influenza epidemics in 2004-2005, 2005-2006, and 2006-
160
2007: A prospective study. Arch Intern Med. 2011; 171(2):151–157.
161
2.
RI PT
159
Benet T, Comte B, Vaillant L, Vanhems P. Contacts entre soignants, visiteurs et patients âgés hospitalisés pendant la période de haute incidence des syndromes grippaux : une
163
étude observationnelle. Rev Gériatrie. 2007; 32(10):727–733. 3.
for health care–associated infection. Am J Infect Control. 2010; 38(3):173–181.
165 166
Hamel M, Zoutman D, O’Callaghan C. Exposure to hospital roommates as a risk factor
4.
M AN U
164
SC
162
Drinka PJ, Krause P, Nest L, Goodman BM, Gravenstein S. Risk of acquiring influenza
167
A in a nursing home from a culture-positive roommate. Infect Control Hosp Epidemiol
168
Off J Soc Hosp Epidemiol Am. 2003; 24(11):872–874.
169
5.
Drinka PJ, Krause PF, Nest LJ, Goodman BM, Gravenstein S. Risk of acquiring influenza B in a nursing home from a culture-positive roommate. J Am Geriatr Soc.
171
2005; 53(8):1437. 6.
Voirin N, Payet C, Barrat A, et al. Combining high-resolution contact data with
EP
172
TE D
170
virological data to investigate influenza transmission in a tertiary care hospital. Infect
174
Control Hosp Epidemiol. 2015; 36(3):254–260.
175
7.
AC C
173
Bracco D, Dubois M-J, Bouali R, Eggimann P. Single rooms may help to prevent
176
nosocomial bloodstream infection and cross-transmission of methicillin-resistant
177
Staphylococcus aureus in intensive care units. Intensive Care Med. 2007; 33(5):836–
178
840.
179
8.
Schumacher M, Allignol A, Beyersmann J, Binder N, Wolkewitz M. Hospital-acquired
180
infections--appropriate statistical treatment is urgently needed! Int J Epidemiol. 2013;
181
42(5):1502–1508. 8
182
9.
ACCEPTED MANUSCRIPT
Voirin N, Barret B, Metzger M-H, Vanhems P. Hospital-acquired influenza: a synthesis
183
using the Outbreak Reports and Intervention Studies of Nosocomial Infection (ORION)
184
statement. J Hosp Infect. 2009; 71(1):1–14.
186 187
10. Vanhems P, Landelle C, Bénet T. Toward a standardized definition of healthcareassociated influenza? Infect Control Hosp Epidemiol. 2014; 35(8):1074–1075.
RI PT
185
11. Dettenkofer M, Seegers S, Antes G, Motschall E, Schumacher M, Daschner FD. Does the architecture of hospital facilities influence nosocomial infection rates? A systematic
189
review. Infect Control Hosp Epidemiol. 2004; 25(1):21–25.
M AN U
193
13. Glind I van de, Roode S de, Goossensen A. Do patients in hospitals benefit from single rooms? A literature review. Health Policy Amst Neth. 2007; 84(2-3):153–161.
TE D
192
2008; 300(8):954–956.
EP
191
12. Detsky ME, Etchells E. Single-patient rooms for safe patient-centered hospitals. JAMA.
AC C
190
SC
188
9
ACCEPTED MANUSCRIPT
194
Table 1. Patients’ characteristics and HAI incidence by stay in a single or double-
195
occupancy room (2004-2014), Edouard Herriot University Hospital (Lyon, France) Patients in single
Patients in double
Total
room (N=51)
room (N=42)
32 (64%)
25 (60%)
0.66
57 (62%)
73.5 (21.8)
76.1 (17.7)
0.97
74.2 (19.9)
P (N=93)
Gender, female
RI PT
Patients’ characteristics
Medium (standard deviation)
0.57
31 (61%)
Medical
15 (29%)
10 (24%)
25 (27%)
Surgical
5 (10%)
7 (17%)
12 (13%)
36 (71%)
28 (67%)
0.68
64 (69%)
21 (47%)
23 (62%)
0.16
44 (54%)
9
11
20
1 227
548
1 775
0.7 (0.4-1.4)
2.0 (1.1-3.6)
Cardiac and/or respiratory underlying diseases Influenza vaccination Confirmed HAIa
EP
Number
25 (60%)
M AN U
Geriatric
TE D
Unit type
SC
Age, years
Cumulative patient-days at risk
56 (60%)
AC C
Incidence rate (95% CI), for
0.028
1.1 (0.7-1.7)
100 patient-days at risk 196
HAI: Hospital-acquired influenza
197
a
Onset ≥48h and virologic confirmation
10
ACCEPTED MANUSCRIPT
198
Figure Legend.
199
Figure 1. Probability of hospital-acquired influenza at Edouard Herriot University Hospital
200
(Lyon, France) by stay in a single or double-occupancy room (2004-2014) (Kaplan-Meier
201
survival curves).
RI PT
202
HAI: Hospital-acquired Influenza
204
At admission (T0), no patient has influenza. The event is influenza acquisition confirmed
205
virologically during hospital stay. Data are censored at patient discharge, death or at 50 days
206
of follow-up. Survival curves were compared by the log-rank test.
AC C
EP
TE D
M AN U
SC
203
11
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT