Exposure to extrinsic risk factors in prevalence surveys of hospital-acquired infections: a methodological approach
Introduction
The best approach to take when studying the risk factors for hospital-acquired infection (HAI) within the framework of cross-sectional surveys has not been established. In one-day, point-prevalence surveys, it is recognized that the temporal association between a risk factor and a disease may not be causal, since it may be impossible to determine whether exposure preceded the disease if both are present simultaneously. Despite this, the majority of studies collect data related to the main risk factors for HAI and carry out analyses upon the data. If interhospital comparisons or comparisons of successive surveys for the same hospital are to be made, it is necessary to adjust for variables that are understood to modify infection risk. Due to the relevance of collecting this information and the fact that withdrawal of the risk factor may also be a therapeutic measure,1, 2, 3 the literature contains a number of methodological variants relating to risk factors in cross-sectional surveys. For example, some study protocols treat the elapsed time between the day of admission and the day on which the survey is conducted as the period of exposure, so that any patient who has been exposed to the risk factor at any given moment during this time period is also considered to have been exposed to illness.4, 5 Other protocols have used a window of six to seven days before the date of the survey as a guide for considering exposure for those patients not exposed on the day of the survey.6, 7, 8 Other protocols only include exposure on the day of the survey,9 or else limit data collection to those patients who have been infected,10 and there are even some studies that have not collected any data regarding risk factors whatsoever.11 Given this variety of methodological approaches, the authors decided: (1) to analyse the effect of obtaining concurrent and retrospective data concerning exposure to extrinsic risk factors on the association between the risk factors and the prevalence of HAI; and (2) to study the influence of the prevalence of risk factors in the different retrospective data gathering schemes, and to control for duration of exposure.
Section snippets
Materials and methods
For the first part of this study, the EPINE database for 1990 was used. In the history of the EPINE project, 1990 was the only year in which data regarding retrospective exposure to extrinsic risk factors were collected. The main known risk factors for the most frequent HAIs to which patients had been exposed during their hospital stay were considered to be extrinsic. Specifically, information regarding exposure to central venous catheters, peripherally inserted central venous catheters,
Results
In 1990, the first year that the EPINE survey was conducted, 38 489 patients in 123 Spanish hospitals were studied. On the survey day, 18 899 exposures to the specific extrinsic risk factors occurred. In addition, 13137 exposures occurred during the week prior to the survey but were not present on the survey day (41%). The average prevalence of exposure to the factors studied ranged between 1.5% for tracheostomy and 48.2% for peripheral vascular catheters (Table I). Except for mechanical
Discussion
The diversity of methodological approaches that exists for the study of risk factors in prevalence surveys4, 5, 6, 7, 8, 9, 10, 11 is evidence of the lack of consensus. The considerable effort required to obtain retrospective clinical information, and the risk of error, may also explain the presence of a variety of approaches. Retrospective searching for adverse events documented in medical records is recognized as inexact,13, 14 particularly in determining the start- and endpoints of exposure
Acknowledgements
Supported by Grant PI20765 FIS (Fondo de Investigación Sanitaria, Spanish Ministry of Health).
References (14)
- et al.
Clinical management of catheter-related infections
Clin Microbiol Infect
(2002) - et al.
Management and outcome of bloodstream infections due to Candida species in England and Wales
J Hosp Infect
(2003) - et al.
Urinary tract infection in surgical patients
Am J Surg
(2003) - et al.
Pediatric prevention network. Prevalence of nosocomial infections in neonatal intensive care unit patients: results from the first national point-prevalence survey
J Pediatr
(2001) - et al.
A one-day prevalence survey of hospital-acquired infections in Lebanon
J Hosp Infect
(2001) - et al.
Repeated prevalence investigations on nosocomial infections for continuous surveillance
J Hosp Infect
(2000) - et al.
Prevalence of and risk factors for hospital-acquired infections in Slovenia—results of the first national survey, 2001
J Hosp Infect
(2003)
Cited by (6)
The accuracy and validity of a weekly point-prevalence survey for evaluating the trend of hospital-acquired infections in a university hospital in Turkey
2011, International Journal of Infectious DiseasesCitation Excerpt :However, incidence studies are expensive because data have to be collected over a long period and require more experienced investigators.6,8,11 However, point-prevalence studies are less expensive and time-consuming, and can be performed more easily than incidence studies.6–8,13–16 In addition, these studies increase awareness of the problem at hospitals and are widely accepted and recommended by many investigators, particularly when they can be repeated at regular intervals.7,10,17
Prevalence of healthcare device-associated infection using point prevalence surveys of antimicrobial prescribing and existing electronic data
2011, Journal of Hospital InfectionCitation Excerpt :Finally, in any prevalence survey of device-associated infection it may be difficult to determine whether the device preceded the infection, was in place simultaneously with the infection, or had been removed as a consequence of the associated infection. Rosello-Urgell et al. found that collecting data on the presence of devices on the date of the survey was as valid as collecting information within the seven days before the survey.19 In our study, we tried to overcome this limitation by reviewing patients receiving antimicrobials for BSI, UTI and pneumonia but without the specific-associated device in situ on the day of the survey, and we determined whether they still met the criteria for device-associated infection.
Epidemiology and impact of nosocomial infections
2010, Medicina IntensivaEvaluation of add-on devices for the prevention of phlebitis and other complications associated with the use of peripheral catheters in hospitalised adults: a randomised controlled study
2009, Journal of Hospital InfectionCitation Excerpt :Peripheral venous catheters are the most frequently used medical devices for hospitalised patients. As reported by Spanish nationwide serial prevalence surveys, use of peripheral cannulae steadily increased from 28% in 1990 up to 44% in 1999, which agrees with data from other settings.1,2 The most frequent complication associated with the use of these devices is infusion phlebitis, whose prevalence ranges from 2.5% to 70% with a mean of 25–35%.3,4
Four Country Healthcare Associated Infection Prevalence Survey 2006: risk factor analysis
2008, Journal of Hospital InfectionCitation Excerpt :This is in contrast to the study reported here where we confined the recording of exogenous risk factors (e.g. presence of urinary catheter) to the previous seven days for many but not all of the risk factors, as well as the day on which the survey was carried out. Those extrinsic risk factors which had high odds ratios in the Spanish survey were: a central intravascular catheter (15.4), parenteral nutrition (17.3), tracheostomy (19.7) and mechanical ventilation (38.8).15 The authors concluded that limiting data collection to recording the presence of a risk factor on the day of the survey provided an adequate estimate of the role that that risk factor played in the prevalence of infection.
- †
The complete list of participants in the EPINE Working Group can be found in Rosselló-Urgell et al. Infect Control Hosp Epidemiol 2004;25:41–46.