Background
Infection prevention and control (IPC) programs have been repeatedly shown to be effective at decreasing the incidence of healthcare-associated infections (HAIs). A landmark paper on this topic in 1985 showed a 32% decrease in the hospital infection rate after 5 years of an ongoing IPC program [
1]. In 1999 the CDC identified seven key evidence-based elements of an effective IPC strategy including voluntary participation of all hospitals, standardized case definitions and protocols, targeted interventions for high risk patient populations, risk adjusted comparisons of infection rates across hospitals, education and adequacy of resources, and feedback to healthcare providers [
2]. The elements of an IPC program have since been significantly updated, forming the concept of “multimodal strategy” [
3].
To prevent HAIs, the WHO recommends implementing an IPC program in every acute healthcare facility [
4]. However, according to the most-recent survey, only 29% of 133 countries surveyed have IPC programs in all tertiary hospitals [
3]. In Russia, IPC programs are also not widely used. The rate of HAIs in Russia has been heavily underestimated for decades. In 2016 it was reported to be approximately 0.08% (24,771 [
5] cases per 31.3 million hospitalized patients [
6]) yet a concurrent meta-analysis which included Russia reported the prevalence of HAIs at 15.5% [
7]. According to the latest World Bank report, Russia has a gross national income per capita of US $9720, corresponding to a middle-income country [
8].
Besides significant underreporting of HAIs, Russia faces other challenges in establishing IPC programs, such as lack of commitment, punishment-based HAI reporting systems, lack of expertise, and inadequate allocation of resources [
9]. Since the dissolution of the Soviet Union, Russia has made some progress in adopting the IPC programs [
10]. A pioneering Russian hospital where an evidence-based IPC program was implemented in 2010 is Burdenko National Medical Research Center of Neurosurgery (NSI) in Moscow. Herein we report the results of our study which aimed to evaluate the impact of this program on HAI prevention in the ICU.
Discussion
A comprehensive IPC program with a focus on hand hygiene and patient isolation was started in NSI’s ICU in 2010 (Fig.
1a). By that time, the use of our IPC program to prevent HAIs in the ICU became a paradigm-shifting solution across Russia as HAI prevention strategies had previously remained unchanged for years and had become outdated [
20].
The importance of HAI prevention programs is clearly indicated by the observation that HAIs directly deteriorate patient survival. It was found that HAIs increased the probability of death by 1.4–1.5 [
21] and odds of mortality increased 1.5 to 1.9-fold [
22]. In our study, we found that HAVM decreased the probability of survival by 1.43, while other HAIs did not significantly influence survival. It has been previously reported that HAVM increased mortality rate approximately three times [
23]. Although the exact mechanism is not yet understood, prospective studies have found that in ICU patients, gastrointestinal dysfunction is also an independent risk factor for increased mortality [
15]. We can postulate that the intestinal microbiome serves an important role in immune function and consequently, is a well described reservoir for antibiotic resistance [
24]. Additionally, in critically ill patients intestinal dysbiosis could be postulated as a potential contributor to gut translocation of pathogens and may play a role in enteric absorption. In our study, ICU-acquired intestinal dysfunction decreased the probability of survival by 1.46, which is consistent with earlier studies. The implementation of IPC initiatives and the accompanying reduction in the incidence of infections, thereby reducing the requirement for antibiotics, can be assumed to at least in part account for reduction in gastrointestinal dysbiosis. This finding further highlights the potential unseen morbidity impact of IPC beyond simple measures of antibiotic utilization and resistance rates.
The implementation of the IPC program was followed by significant reduction of HAIs in the ICU. In fact, the impact of this program may actually be under-estimated. Our IPC program was implemented in 9/2010 whereas study data collection began 1/2011. Therefore, although adherence to IPC protocols would be expected to improve with greater time and familiarity, the totality of impact of this program may be under-estimated. Key initiatives, such as early removal of indwelling catheters, would be expected to have an immediate impact in the reduction of nosocomial infections. Even discounting the IPC impact in the initial months after implementation, the fact that a sustained and continued reduction in HAI rate occurred is both meaningful and serves as a reinforcement of overall utility. In high-risk ICU patients we observed a substantial decrease in HAI incidence: cumulative incidence of respiratory HAIs declined by 1.47 (from 36.1 to 24.5%), urinary tract HAIs by 1.4-fold (from 29.1 to 21.3%), HAVM by twofold (from 16 to 7.8%), CAUTI by 1.93 (from 35.4 to 18.3%) (Fig.
3), and ICU-acquired intestinal dysfunction by 2.3 fold. These results are consistent with previously reported evidence, demonstrating a reduction of HAI prevalence by approximately 1.7 fold (from 11.7 to 6.8%) [
25].
We also found that the risk-adjusted incidence of EVD-related HAVM reduced 1.64 fold (from 22.2 to 13.5 cases per 1000 EVD-days) over the six-year study period. The impact of an IPC program on decreasing DA-HAI incidence has been previously reported. For example, one publication reported a 2.7-fold decrease in CAUTI episodes per 100 patients within a year after IPC implementation [
26]. However, for some HAIs, like HAVM, such statistics are absent. In addition, the changes in the incidence of intestinal dysfunction could be confounded by the implementation of an advanced nutritional protocol in 2012 at the ICU.
We did find that in 2012 the rates of several infection subcategories did increase in comparison to 2011. The rate of respiratory and urinary HAIs had increases ranging from 4 to 14% compared to 2011. The reason for this increase is unclear, but we postulate that this may be related to several factors. One contributor may be that staff were educated on the appropriate identification of HAIs and utilized clear standardized case definitions. As staff became more familiar with these definitions, they may have been able to better identify cases leading to an apparent increase in infection rates. Additionally, during initial implementation of IPC protocols, staff underwent in-service training and consequently there was a specific focus on the strict adherence to protocols. However, adherence to infection control practices may wane with time, and that probably what happened in 2012. Therefore, continued reinforcement of best practices along with feedback to healthcare teams is necessary for sustained adherence to IPC initiatives. Following the re-education of staff, a renewed attention to IPC may have contributed to reductions seen in 2013 HAI rate.
Additionally, both the length of patients’ stays in the ICU and the incidence of patient mortality did decrease over the study period. Although a direct causality cannot be determined, it would be fair to postulate that the associated decrease in HAI incidence may at least have been a partial contributor for this reduction. Thus, a reduction in the rate of HAIs may result in a meaningful reduction in healthcare cost, and potential benefit in overall mortality. However, we did not monitor all other parameters that could have influenced the mortality and the length of stay, thus other explanations should be investigated. Additionally, we admit that the overall approach in patient treatment did not change much, and the DUR did not change for any of the devices we monitored.
The prevention of the spread of carbapenem-resistant, Gram-negative bacteria was named the first priority of IPC efforts by the latest WHO guidelines because these strains pose significant threat to global health [
27]. We found firstly that the proportion of such Gram-negatives as
K. pneumoniae and
A. baumannii in the spectrum of bloodstream HAIs decreased and secondly that the resistance of both pathogens to carbapenems was significantly reduced. In our study the initial percentage of isolates resistant to imipenem was 34.5% for
K. pneumonia and 77.7% for
A. baumannii. By the end of the study, the percentage decreased 1.7- and 2-fold, respectively (Fig.
4b). The initial prevalence of carbapenem-resistant isolates in the NSI neuro-ICU was shown to be higher than the mean prevalence in Europe (8.1% for
K. pneumoniae and 50% for
A. baumannii), and in the U.S. (7.9% for
K. pneumoniae and 49.5% for
A. baumannii) [
27]. This finding could partly be explained by the study population because we analyzed only intensive care unit patients which may be a higher risk population. However, we postulate these initial rates of carbapenem resistance were at least in part due to nosocomial cross-infection of patients.
Our hypothesis is that the implementation of IPC protocols acted in a two-fold manner with an initial reduction in nosocomial patient-to-patient transmission which consequently lead to a reduction in nosocomial infection rate. Our most critical interventions involved implementation of contact precautions utilizing gloves, gown, and mask, isolation of patients identified with carbapenemase resistance genes, and cohorting of patients with
Acinetobacter or
Klebsiella (Fig.
1). These efforts were paired with intensive environmental disinfection measures, skin antisepsis for indwelling devices, as well as initiates focused on hand hygiene as a multi-modal strategy (Fig.
1).
Of note, hand hygiene compliance was particularly difficult to implement with a compliance rate of 27% in 2011. Compliance with hand hygiene in the subsequent years 2012 through 2016 were 40, 69, 63, 68, and 81% respectively. The reduction in infection rate over time could reasonably be postulated to result in a secondary reduction in the necessity of broad spectrum antibiotic therapy. This reduction in antibiotic utilization is underscored by the dramatic decline in the rate of antibiotic utilization over the study period. It must be noted that an antibiotic stewardship program was in existence prior to IPC implementation. Antibiotic stewardship involved institutional protocols for perioperative antibiotic prophylaxis and for empiric antibiotic therapy. However, integration of IPC protocols, including surveillance measures may have enhanced the effectiveness of antibiotic stewardship interventions. The ultimate result was that within the study period, our observed resistance rates decreased to the level of global and regional estimations.
This improvement in susceptibility rates, is in contrast to the global trend of increasing carbapenem resistance over the past decade [
27], indicating that in limited-resource settings IPC programs can be highly effective. The programs may be especially significant in healthcare settings with high levels of resistance where they can serve as a cost-effective intervention leading to a substantial clinical impact. The substantial diminution in carbapenem resistance supports the notion that implemented IPC strategies contain effective measures to prevent and control the resistance to carbapenems (Fig.
1). Moreover, this is supported by the recent WHO guidelines which affirmed that the core components of multimodal IPC strategy can help to prevent carbapenem resistance.
This paper reports a prospective study of the impact of an infection control program in a high acuity limited resource setting with regard to the reduction in HAI risk. Such studies are limited to date but have been identified by the WHO as particularly needed [
27]. Thus, this study can help to fill this research gap providing insight regarding an approach to implantation of these programs and highlighting the most essential IPC components. Our results suggest that a focus on robust surveillance paired with isolation/infection control measures can promote a sustainable and meaningful reduction in HAI incidence and antibiotic resistance.
The current study has certain limitations. It is a single-center study in a highly specialized ICU facility. Thus, one should be careful when generalizing these results to other hospitals and other wards. In addition, we only studied a cohort of high-risk patients, those staying in the dedicated neuro-ICU for > 48 h—not the entire ICU population. Thus, reported HAI incidences are higher than those calculated for the entire ICU population. However, the underlying principles of our IPC program leading to the reduction of CAUTI, CLABSI, and VAP would be expected to be generalizable to other hospitalized settings with a similar expected impact.
One aspect that was not able to be fully evaluated were Clostridium difficile infections (CDI). The prevalence of CDI, identified by a positive PCR stool assay and compatible symptoms, was measured quarterly. However, the quarterly rate included all patients in the ICU at the time of a positive diagnosis and included patients that did not meet the defined criteria for high-risk population that were studied. Additionally, the incidence rate was low throughout the six-year period with a peak rate of 1.5% in 2011 and a nadir of 0.9% in 2015. Notably, patients who were transferred out of the ICU and subsequently developed CDI would not have been identified. Therefore, we can postulate that IPC initiatives may result in a reduction in CDI as the rate did decline from 2011; however, the low overall incidence of CDI and aforementioned limitations do not allow for definitive conclusions.
By design, the study did not include a control group (i.e. a group treated in the ICU before the IPC program had been implemented), because HAI rates without surveillance are unknown. Moreover, the decrease of HAI incidence and length of stay in the ICU could be explained by modification of clinical practices and by regression to the mean. It should be mentioned that survival analysis in our study suffers from immortal time bias. Patients in the HAI group are “immortal” until they get the infection, that favors the HAI group by lowering mortality rate in this group. Thus, HAIs have a stronger influence on survival, posing a higher risk of death in patients once they get HAIs.