Process and outcome indicators for infection control and prevention in European acute care hospitals in 2011 to 2012 – Results of the PROHIBIT study

Introduction Hospitals from 24 European countries were asked for information on infection prevention and control (IPC) indicators as part of the Prevention of Hospital Infections by Intervention and Training (PROHIBIT) survey. Methods: Leading IPC personnel of 297 hospitals with established healthcare-associated infection (HCAI) surveillance provided information on local surveillance and feedback by using a questionnaire. Results: Most hospitals focused on bloodstream infection (BSI) (n = 251) and surgical site infection (SSI) (n = 254), with a SSI post-discharge surveillance in 148 hospitals. As part of the HCAI surveillance, meticillin-resistant Staphylococcus aureus (MRSA) was the leading multidrug-resistant organism (MDRO) under surveillance. Seventy-nine per cent of hospitals (n = 236) monitored alcohol-based hand rub (ABHR) consumption. Feedback to the local IPC committees mainly included outcome data on HCAI (n = 259; 87%) and MDRO among HCAI (n = 245; 83%); whereupon a feedback of MDRO data depended on hospital size (p = 0.012). Discussion/conclusion: Objectives and methods of surveillance vary across Europe, with BSI, SSI and MRSA receiving considerably more attention than indicators such as pneumonia and urinary tract infection, which may be equally important. In order to maximise prevention and control of HCAI and MDRO in Europe, surveillance should be further improved by targeting relevant HCAI. The role of feedback should be explored in more detail.


Introduction
Based on the results of the first European point prevalence survey (PPS) in 2011-12 an estimated 3.2 million patients acquire a healthcare-associated infection (HCAI) in acute care hospitals in Europe every year [1]. The most common types of HCAI are surgical site infections (SSI), urinary tract infections (UTI), pneumonia (PN), bloodstream infections (BSI), and gastrointestinal infections, with Clostridium difficile infection (CDI) accounting for a high proportion in the latter. HCAIs result in increased morbidity and mortality, and emerging antibiotic resistance complicates their treatment. The cumulative burden of HCAIs is higher than the total burden of other communicable diseases in Europe [2].
Surveillance as the 'ongoing systematic collection and analysis of health data for the planning, implementation, and evaluation, of public health practice' [3] is a key measure in HCAI prevention and control. Even in the absence of specific prevention actions, surveillance and feedback of outcome indicators decrease HCAI by raising awareness for the issue among healthcare professionals [4][5][6][7]. Surveillance, preferably as part of a network, was identified as one of the key components in effective HCAI prevention and an important tool for monitoring the effectiveness of prevention and control measures by the 'Systematic Review and Evidencebased Guidance on Organization of Hospital Infection Control Programmes' (SIGHT) project [8].
Since the 1990s, many European countries have been developing national surveillance networks, either by applying the United States' (US) Centers for Disease Control and Prevention (CDC) National Nosocomial Infection Surveillance/National Healthcare Safety Network (NNIS/NHSN) protocol, or by using adapted methods to better take into account local diagnostic practices [9]. In 2010, the European Centre for Disease Prevention and Control (ECDC) established the Healthcare-Associated Infections surveillance Network (HAI-Net), integrating the Hospitals in Europe Link for Infection Control through Surveillance (HELICS) project  and the Improving Patient Safety in Europe (IPSE) network   [10].
Surveillance of alcohol-based hand rub (ABHR) consumption has become a mandatory quality indicator with public reporting in France since 2006 [11], and was integrated into the national Krankenhaus-Infektions-Surveillance-System (KISS) in 2008 in Germany [12]. Furthermore, national strategies on measuring hand hygiene compliance by direct observation have been organised in a number of European countries [13].
Aspects of specific surveillance activities in European hospitals were obtained as part of the Prevention of Hospital Infections by Intervention and Training (PROHIBIT) project. The PROHIBIT survey was conducted as the first pan-European survey on infection prevention and control (IPC) in order to describe which IPC recommendations are actually being used across Europe and to provide information on gaps in hospitals' IPC policies and practices for policymakers, hospital managers and healthcare workers for further improvement of HCAI prevention. This article summarises data on findings from 24 European countries.

Participating countries and hospitals
ECDC national contact points (NCPs) and IPC experts of European countries outside of the European Union (EU) were invited to organise national polls. The NCPs invited national hospitals for participation between September 2011 and March 2012. Participation in the PROHIBIT survey was based mainly on hospital interest rather than on a systematic sampling process.

Survey description
The survey was developed by an interdisciplinary group and discussed with the HAI-Net representatives. It included four questionnaires in order to assess IPC structure and process indicators (i) at the hospital level, (ii) in intensive care units (ICU), (iii) in non-ICU medical wards and (iv) in non-ICU surgical wards. Questionnaires addressed organisation and activities of IPC at those various levels.
The complete method of the survey and the characteristics of the participating hospitals are described in more detail elsewhere [14].
Local IPC professionals were also asked to provide data on hospital characteristics such as status (public/private) and size of the hospital (number of beds), and the full time equivalent (FTE) of infection control personnel.
Furthermore, country characteristics such as the United Nations (UN) European geographical region, and healthcare expenditure (HCE) as share on the national gross domestic product (GDP) were collected [15,16].

Data analysis
Descriptive data analysis was performed and results summarised as totals and strata of the following four parameters: hospital size (small: ≤ 300 beds; medium: 301-600 beds; large: > 600 beds), UN European geographical regions, full-time-equivalent (FTE) infection control nurses (ICN) (internal staff and external staff) per 1,000 acute care hospital beds ≤ / > the median of all participating hospitals (3.72 FTE ICN/1,000 beds) and HCE as share on GDP [15,16]. HCE was modelled as a dichotomous variable and considered low or high if below or above the European mean HCE of 9%. Differences in the process and outcome indicators between the strata of the four parameters described above were tested by logistic regressions models. In the regression analysis with indicator parameters as outcome, only the independent variables were included in the generalised estimating equation (GEE) models, adjusting for cluster effects by country. A two-sided p value < 0.05 in the type III test was considered significant. All analyses were performed with SPSS (IBM SPSS statistics, Somer, NY, US) and SAS (SAS Institute, Cary, NC, US).

Participating countries and hospitals
Of 32 invited countries, 24 participated (Table 1)

Medical conditions/disease outcome where HCAI surveillance applies
Surveillance in the hospitals mainly focused on SSI and BSI, and less often on PN, CDI and UTI ( Table 2). Surveillance of UTI depended on countries' HCE with significantly higher proportions in countries with low HCE. For PN, significant differences were observed in accordance to the hospital size, with more medium to large hospitals having PN surveillance in place compared to smaller hospitals. Significantly more hospitals from countries with low HCE performed hospital-wide surveillance of PN and UTI. Hospital-wide surveillance of BSI varied significantly with the UN regions.
Surveillance of CDI was reported most often by hospitals in Northern Europe and more often by hospitals in countries with high HCE; but these differences were not statistically significant.
Hip prosthesis implantation (HPRO) was the most common indicator operation of SSI surveillance with higher percentages in countries with high HCE; but these differences were not statistically significant. In 148 of 254 hospitals with SSI surveillance (58%), post-discharge surveillance (PDS) was in place.

Monitoring hand hygiene compliance
Consumption of ABHR and hand hygiene compliance was observed in 79% and 78% of the hospitals, respectively (Table 2). Hospitals in countries in Northern Europe preferred monitoring hand hygiene compliance to monitoring ABHR consumption, while hospitals in countries in Western Europe preferred monitoring ABHR consumption to monitoring hand hygiene compliance (Table 2). Table 3 summarises the operators involved in HCAI surveillance. Most surveillance activities are performed by IPC personnel. Although the overall numbers are low, it appeared that hospitals in countries of Northern Europe had a higher percentage of specific staff dedicated to surveillance of HCAI compared to other regions. In almost half of the hospitals, data on HCAI were collected by IPC personnel, whereas surveillance exclusively performed by ward personnel was by reported by 9% of all hospitals, and 18% of hospitals in Eastern Europe.

Feedback on the HCAI situation and hand hygiene compliance
In almost all hospitals, healthcare workers (HCW) received feedback on HCAI (n = 106 more than twice a year, n = 61 twice a year, n = 115 once a year, n = 15 less than once a year). Of the 236 hospitals that performed ABHR consumption surveillance, 200 hospitals provided feedback at least once a year (n = 41 more than twice a year, n = 32 twice a year, n = 127 once a year), while in 35 hospitals feedback was given less than once a year and for one hospital information on the frequency of feedback was unavailable. Concerning direct hand hygiene compliance observations, 152 hospitals provided immediate feedback to the observed personnel and 131 hospitals provided a later summary feedback.
As shown in the Figure, IPC committees mainly received data on HCAI (n = 259; 87%) and the proportion of MDRO among HCAIs (n = 245; 83%) but less often on hand hygiene performance indicators. Feedback on MDRO among HCAIs was most often provided in larger hospitals (p = 0.012). IPC committees in hospitals with ICN rates above the European median received significantly more often feedback on hand hygiene compliance data compared to IPC committees in hospitals with ICN rates below the median (p = 0.039). Feedback on hand hygiene performance (ABHR consumption and/ or hand hygiene compliance) was significantly more often provided in countries with high HCE (p = 0.042). No feedback was given to IPC committees in 23 (8%) hospitals.

Discussion
To our knowledge, the data of the PROHIBIT survey offer the first broad analysis of HCAI and MDRO surveillance activities in European acute care hospitals. The findings show that content and methods of surveillance and the role of feedback vary widely across Europe. Hospitals focused more frequently on the surveillance of outcome indicators as BSI and SSI than on PN, CDI or UTI. This may be due to numerous success stories of BSI and SSI preventability, which raised hospitals' awareness towards these two infection types [17][18][19].
Fifty-eight per cent of hospitals with SSI surveillance reported to have PDS in place. Such additional surveillance as described by Woelber et al. in 2016, partly prevents under-reporting of SSI in Europe [20]. The finding that HPRO is the most common indicator procedure for SSI surveillance corresponds to the results of ECDC's HAI-Net surveillance of SSI with HPRO being the most frequently reported type of surgery, representing 33% of all operations in 2010-11 [21].
Nevertheless, successful preventability of HCAIs such as PN and UTI has also been described [22,23] and data of the PPS from 2011-12 indicate that respiratory tract infections and UTI are common HCAIs all over Europe. Interestingly, UTI surveillance was significantly more frequently reported in countries with low HCE. Data of the PPS for countries with high HCE however, showed frequencies of UTIs up to 31%, indicating that patients hospitalised in such countries are also at risk for this type of HCAI [1].
Resources are limited; and thus, priorities must be made in HCAI surveillance, even if a broad surveillance strategy including process and outcome indicators is considered helpful to tailor intervention activities for HCAI prevention. Prospective hospital-wide HCAI surveillance is resource-intensive, and in this sense, it was surprising to find that the proportion of hospitals   to perform detailed risk factor collection is indispensable [27].
In addition to the surveillance of outcomes, many hospitals assess data on hand hygiene performance indicators. Monitoring process indicators and assessment of adherence to IPC measures such as hand hygiene observation, enables hospitals to identify gaps and improve adherence to IPC measures more promptly than by focusing on outcome data alone. Interestingly, our study identified a discrepancy between a relatively high number of hospitals monitoring ABHR consumption and a relatively low number of hospitals giving feedback on ABHR consumption data to their IPC    committees. This may be due the fact that many hospitals start surveillance activities focusing on outcome indicators and still work on the feedback of these indicators rather on reporting ABHR. In addition, ABHR data are often collected yearly, and thus, may be reported less frequently to IPC committees. On the other hand, process indicators are better candidates to be used for a realistic target-setting both at ward and hospital level. Reference data on these indicators facilitate inter-hospital comparison to support improving their processes [28]. Europe-wide surveys as the ECDC-PPS or PROHIBIT [1,14] already offer reference data on factors such as IPC personnel or isolation capacities and future projects may generate more, possibly stratified reference data for relevant structural and process IPC parameters.
In order to alter their behaviour in HCAI prevention HCWs have to be aware of the problem of HCAI in their setting. Data of our survey indicate that HCWs do receive feedback on HCAI rates in order to raise awareness. However, more research is needed to explore how surveillance data are communicated and perceived, and how this process can be further optimised.
Feedback of data may be combined with behaviourally informed approaches such as the setting of long-term goals and encouraging involvement/participation of HCWs for creating local ownership and reflection on achievements and further activities.
Since successful implementation of IPC measures requires the participation of HCWs and other stakeholders, feedback to members of the IPC committee is essential. Especially in smaller hospitals, feedback is not always established yet. In which way the size of a hospital influences feedback of MDRO data to hospitals' stakeholders cannot be fully answered. It can be speculated that larger hospitals see more MDROs, and thus, data are perceived more relevant, particularly because they care more frequently for patients with severe and/or chronic diseases.
In the future, all hospitals' IPC committees should be encouraged to work with MDRO data in order to address supporting organisational factors such as leadership support and communication in MDRO transmission prevention and antibiotic stewardship programmes [29,30].
The current survey gives insight into established surveillance activities of European hospitals. However, there are some limitations: Participation in the survey was voluntary, and thus, based mainly on hospitals' interest rather than on a randomised sampling process. Therefore, the data may have overestimated surveillance activities in European hospitals. A randomly selected sample would have improved representation of hospitals in Europe. However, the questionnaire could not have been imposed on hospitals, and thus, data quality and the number of participating hospitals might have been lower.
The UN geographical regions are not homogeneous in terms of GDP, healthcare organisation and culture. However, by also reporting data in reference to countries' HCE, we tried to take into account such heterogeneity.
Our findings show that objectives and methods of surveillance vary across Europe. Some outcome indicators, such as BSI, SSI and MRSA, seem to receive considerably more attention than others that are equally important, such as PN, UTI or CDI.
Hospitals' IPC committees mainly receive data on outcome indicators as HCAI and MDRO, but less often on process indicators as hand hygiene performance indicators.
In order to better address prevention of HCAI and antimicrobial resistance in Europe surveillance should be further improved by targeting all relevant HCAI and MDRO and providing active surveillance by trained personnel. To what extent surveillance of process indicators prevent HCAI must be further analysed. In addition, the role of feedback and behaviourally informed approaches should be explored in more detail.