Infection control in settings with limited resources
What is the impact of healthcare-acquired infections in settings with limited resources?
One of the central premises of healthcare-acquired infection (HAI) control is that thorough surveillance knowledge of the occurrence of infections is essential to effectively address this public health burden. In settings with limited resources, such accurate knowledge is many times underestimated, and the actual, critical impact that HAI have on the population of limited-resource settings is difficult to assess.
To determine which countries are referred to as “limited-resource countries”, the World Bank categorizes countries worldwide into four economic strata based on 2015 gross national income (GNI) per capita: (1) low-income economies, $1,025 or less; (2) lower middle-income economies, between $1,026 and $4,035; (3) upper middle-income economies, $4,036 and $12,475; and (4) high-income economies, $12,476 or more. Within this categorization, 144 out of 209 (68%) are low-income and lower middle-income economies, which can also be referred to as lower-income countries, low resources countries, developing economies, or developing or emerging countries. Developing economies represent more than 75% of the world population, and it is in these settings where the issue concerning HAI remains many times unresolved and needs to be highlighted and dealt with as a public health priority.
In this chapter the author analyzed the studies on HAI in settings with limited resources, with emphasis on countries of all the six WHO regions, such as Latin America, Africa, Eastern Europe, South East Asia, Eastern Mediterranean, and Western Pacific, published since 2002 up to mid 2016. From the available literature, it is highly visible that the adverse consequences of HAI in settings with limited resources, that is, attributable mortality, prolonged length of stay, extra hospital costs, and increased bacterial resistance, are more far-reaching in terms of severity than in the high-income economies. The prevalence of HAI in settings with limited resources was found to at least double the rates published by the European Centre for Disease Prevention and Control, and triple those found in the USA (see Table I).
In the case of device-associated healthcare-associated infections (DA-HAIs), the rate of device use was found to be analogous to or even lower than the one reported of U.S. ICUs by the Centers for Disease Control and Prevention’s National Healthcare Safety Network (CDC/NHSN) System; however, pooled mean rates identified in intensive care units (ICUs) internationally by the International Nosocomial Infection Control Consortium (INICC) were found to be exceedingly higher than those reported from U.S.’s ICUs by the National Healthcare Safety Network (NHSN) (see Table I).
Among the most serious consequences attributable to HAI in settings with limited resources, it has been shown in the mainstream literature that mortality can range from 3 to 75.1%.
Rosenthal et al., have shown mortality due to central line-associated bloodstream infections (CLABs) has rates that ranged from 4 to 75.1% (see Table II). Cost, length of stay (LOS) and mortality attributable to DA-HAI have been determined by INICC internationally, through prospective, matched analyses. For measuring LOS and mortality attributable to DA-HAI, the INICC applied a new multi-state model, including specific censoring to ensure the estimation of the independent effect of each DA-HAI, and not the combined effects of multiple DA-HAIs.
In a review to analyze the incidence of CLAB in limited-resource countries performed by Rosenthal in 2009, it was demonstrated that the CLAB rate was associated with significant extra mortality, with an odds ratio ranging from 2.8 to 9.5 (see Table II).
Similarly, mortality attributable to ventilator-associated pneumonia (VAP) has been found to be as high as 56.7% (see Table III).
With respect to mortality due to catheter-associated urinary tract infection (CAUTI), reports are scarce and there has been diversity in the interpretation of findings. In some publications, it was stated that CAUTI was not associated to mortality, but other findings specified rates up to 21.3% (see Table IV).
Within the adverse effects of HAI, prolonged LOS and the correlated extra hospital costs have been shown to cause a high impact at hospital and national levels. In this respect, in Argentina, Rosenthal et al., found that the extra LOS of patients with CLAB was 11.9 days, and that the mean extra cost was US$4,888.
Also in Argentina, Rosenthal et al., found that the mean extra LOS of patients with VAP was 8.95 days and the mean extra total cost was $2,255, and the extra mortality was 30.3%.
Higuera et al., have shown in a study conducted in Mexico that the mean extra LOS of patients with CLAB was 6.1 days, and that the mean extra hospital cost of patients with CLAB was US$ 11,591. In the case of surgical site infections (SSI), a systematic review and meta-analysis by the World Health Organization (WHO) showed that the incidence of SSI in settings with limited resources was markedly higher than in developed countries, with rates ranging from 1.2 to 23.6 per 100 surgical procedures (Table X).
In 2013, the INICC reported the results of a cohort, prospective surveillance study on SSIs conducted on patients undergoing surgical procedures (SPs) from January 2005 to December 2010 in 82 hospitals of 66 cities in 30 countries (Argentina, Brazil, Colombia, Cuba, Dominican Republic, Egypt, Greece, India, Kosovo, Lebanon, Lithuania, Macedonia, Malaysia, Mexico, Morocco, Pakistan, Panama, Peru, Philippines, Poland, Salvador, Saudi Arabia, Serbia, Singapore, Slovakia, Sudan, Thailand, Turkey, Uruguay, and Vietnam) from 4 continents (America, Asia, Africa, and Europe). Data from 7,523 SSIs were associated with 260,973 surgical procedures. SSI rates were significantly higher for most types of surgical procedures in INICC hospitals compared with US CDC/NHSN data, including the rates of SSI after hip prosthesis (2.6% vs. 1.3%; relative risk [RR], 2.06 [95% confidence interval (CI), 1.8-2.4]; P < .001), coronary bypass with chest and donor incision (4.5% vs. 2.9%; RR, 1.52 [95% CI, 1.4-1.6]; P < .001); abdominal hysterectomy (2.7% vs. 1.6%; RR, 1.66 [95% CI, 1.4-2.0]; P < .001); exploratory abdominal surgery (4.1% vs. 2.0%; RR, 2.05 [95% CI, 1.6-2.6]; P < .001); ventricular shunt, 12.9% vs. 5.6% (RR, 2.3 [95% CI, 1.9-2.6]; P < .001), and others.
The relationship of antibiotic use and the emergence of antibiotic-resistant HAI is an issue that epidemiologists and hospital authorities in settings with limited resources must be aware of. In the last INICC Report, which contained a data summary of the device-associated module of 50 countries for 2010 – 2015, antimicrobial resistance rates found for Pseudomonas isolates to amikacin (29.87% vs. 10%) and to imipenem (44.3% vs. 26.1%), and of Klebsiella pneumoniae isolates to ceftazidime (73.2% vs. 28.8%) and to imipenem (43.27% vs. 12.8%), were far higher than US CDC-NHSN ICUs’ rates.
In several studies, researchers have highlighted the extreme vulnerability of neonates hospitalized in neonatal intensive care units (NICUs) to mortality attributable to DA-HAI, with rates ranging from 24% in the pre-surfactant era to 11% in the post-surfactant era in the developed countries. The burden of CLAB in the NICU is not limited to mortality, and newborn sepsis was associated with adverse consequences in the central nervous system, longer duration of mechanical ventilation, and hepatic fibrosis and chronic lung disease higher incidence. However, within the context of settings with limited resources, access to knowledge regarding DA-HAI is scarce, and there is an insufficient recognition of the importance of surveillance for measuring the infection risks, outcomes and processes concerning the neonatal patient hospitalized in the NICU.
In this respect, a study was performed to evaluate the impact of country socioeconomic status and hospital type on DA-HAIs in 30 NICUs, from hospitals members of INICC in 15 settings with limited resources. Its findings revealed that DA-HAIs were significantly lower in private than academic hospitals (10.8 versus 14.3 CLAB per 1,000 catheter-days [p<0.03]), but not different in public and academic hospitals (14.6 versus 14.3 CLAB per 1,000 catheter-days [p.0.86]). Furthermore, CLAB rates found in NICUs enrolled from low-income countries were significantly higher than in lower middle-income countries or upper middle-income countries, and VAP rates in patients hospitalized in NICUs from academic hospitals were significantly higher than rates found in private or public hospitals.
These findings are a clear indication of the influence that economics, as a surrogate of available supplies, outdated technology, and scarce human resources availability, have on settings with limited resources, and of the close relation between hospital type and limited access to health care resources. In public and academic hospitals, the limitation to sufficient resources in terms of adequate number of trained and specialized staff, budget, medical supplies, and hospital administrative support is markedly more serious than in private hospitals, as they are more dependent on the socio-economic category of the country concerning the budget allocation.
Limited-resource countries are confronted with aspects that transcend clinical findings and good delivery of healthcare practices; the harsher reality suffered by patients hospitalized in the ICUs of settings with limited resources lies outside the scope of the hospital itself, and reflects the country’s social and political situation, poor living conditions, difficult or differentiated access to labor market and precarious labor conditions, diversity of cultural values, unequal allocation of assets among population resulting in unsatisfied basic needs, including sanitary infrastructure and limited access to the education and health system. As long as these conditions prevail, healthcare workers from settings with limited resources are urged to focus their best efforts on improving healthcare and clinical practices, and disseminating their successful achievements, so as to be able to counteract the many social factors that cannot be directly controlled by clinical practices alone.
What elements need to be adhered to for prevention and control of healthcare-acquired infection in settings with limited resources?
There are several measures to be considered as basic recommendations for the implementation of an infection control program, which should be consistent with the actual capabilities of the healthcare facility and personnel.
In this respect, the recommendations described in the guidelines published by the Institute for Healthcare Improvement (IHI), Centers for Disease Control and Prevention (CDC), Society of Health Care Epidemiology of America (SHEA), Association for Professionals of Infection Control, and Joint Commission International (JCI) provide cost-effective preventative measures, feasibly applicable to infection control programs in settings with limited resources.
There are available comprehensive and detailed evidence-based recommendations for preventing catheter-related bloodstream infection (CRBSI) previously edited and published by organizations from high-income industrialized countries, such as CDC in 2011, JCI in 2012, SHEA and the Infectious Diseases Society of America (IDSA) in 2014, National Evidence-Based Guidelines for Preventing Healthcare-Associated Infections in National Health Service (NHS) Hospitals in England (EPIC 3) in 2014, Infusion Nurses Society (INS) in 2016, from Asia Pacific Society of Infection Control in 2016 and from the International Nosocomial Infection Control Consortium, specifically adapted to settings with limited-resources.
However, as identified in assessments undertaken by the International Nosocomial Infection Control Consortium (INICC) team at healthcare facilities worldwide during the last 25 years, there is a significant gap between the contents of the above-mentioned available recommendations and the feasibility of their implementation at hospitals in resource-limited countries, due to the actual structure, supplies, technology, knowledge, skills and practices at their disposal. It was commonly observed that hospitals in resource-limited countries apply the five classic components of the Institute for Health Improvement (IHI) bundle along with its check list ―which are, undoubtedly, significantly effective in industrialized countries.
However, their effectiveness is hindered by the existence of unfavorable situations, including overcrowded intensive care units (ICUs), insufficient rooms for isolation, lack of sinks, lack of medical supplies, including but not limited to, alcohol hand rub products, antiseptic soap, paper towels, chlorhexidine solutions, single-use flush syringes, prefilled syringes (leading to manual admixture of all saline solutions and drugs), closed intravenous systems (replaced with vented intravenous containers) and needleless connectors devices (which are replaced with three-way stop cocks). It was also observed that limited supplies serve as an obstacle to applying maximal barriers precautions during catheter insertion (See Figure 1). Moreover, adopting unsafe practices, such as using cotton balls already impregnated with antiseptic and placed in a contaminated container, not covering an insertion site with sterile dressing, using drugs in already-open single-dose vials, reusing single-dose vials, leaving needles inserted in multiple-dose vials, and taking fluids from a 500-ml container for dilution of parenteral solutions, were also common practices.
As a result of this situation, although healthcare workers (HCWs) in limited-resource settings may comply with a comprehensive bundle perfectly well, CLABSI rates at their health care facilities continue to be above 10 CLABSIs per 1000 central line (CL) days, instead of 1 CLABSI per 1000 CL days ― that is, 10 times higher CLABSI rates than US CDC’s National Healthcare Safety Network (NHSN) rates ― causing frustration and disappointment among HCWs, and leading to higher morbidity and mortality rates.
In light of these findings, there still exists a compelling need to bridge the gap between infection prevention theory and practice in healthcare facilities with limited resources. Thus, the specific objective of this publication is to select, highlight and comment on a few practical and essential components of a bundle to counteract those adverse differences in practice and use of outdated supplies and technologies that are evident when comparing healthcare facilities in high-income and resource-limited countries. The publication of a new abridged set of recommendations as part of a Bundle edited by INICC in collaboration with a group of 27 international experts from 25 countries of the 6 World Health Organization (WHO) regions is, therefore, justified as an attempt to assist hospitals worldwide ― but with special emphasis on limited resources settings ― in implementing and prioritizing effective efforts to prevent CRBSI associated with CLs and peripheral lines (PLs).
The logical initial step is the organization of a surveillance system, as it permits the identification of local problems, distinctively specific to a particular institution, and will thus serve as guide for subsequent changes. Targeted surveillance and calculation of device-associated infection rates per 1000 device-days also allows benchmarking with other similar institutions. In this respect, “Outcome Surveillance” developed by INICC includes the systematic standardized measurement of DA-HAI rates and their associated effects: mortality, morbidity, extra length of stay, extra hospital costs, and bacterial resistance.
Surveillance data are essential to have an accurate knowledge of the burden of HAI and focus efforts on the areas that need more attention. Hospitals with limited resources need to start surveillance of critical areas, such as intensive care units, where DA-HAI pose the most threatening risks for patient safety. This first approach needs to be followed by the surveillance and monitoring of processes. Process surveillance is necessary to monitor compliance with infection control prevention guidelines and basic measures, such as hand hygiene, vascular catheter care, urinary catheter care, and measures to prevent VAP. Thirdly, a continuing education program on HAI control and prevention must be addressed to healthcare-workers, particularly nurses, who have the greatest risk of transmission of organisms, and are essential to interrupt the transmission of HAI.
To reduce the incidence of these higher rates internationally, and particularly, in limited-resource countries, INICC adopted the “INICC Multidimensional Approach” (IMA). The IMA is a system to measure and reduce HAI rates, mortality, LOS, costs, bacterial resistance and antibiotic consumption that comprises the simultaneous implementation of 6 components: (1) bundles, (2) education and training, (3) online outcome surveillance of HAI rates and their adverse consequences; (4) online process surveillance to evaluate compliance with bundles; (5) online feedback of HAI rates and their adverse consequences; and (6) online performance feedback. As part of the IMA, the INICC uses an online platform called INICC Surveillance Online System (ISOS), which includes components (3), (4), (5) and (6) of the IMA.
On the one hand, the ISOS enables ICPs to conduct online prospective, active, surveillance cohort studies designed to collect specific data per patient from all patients, both those with and those without HAI; this allows the identification of risk factors of HAIs ― such age, gender, severity illness score, invasive devices utilization, and several surrogates of HAIs ― and the validation of HAIs, thereby ensuring that the latest published CDC-NHSN criteria are met in each DA-HAI diagnosis, so as to avoid under-reporting and inconsistent selections due to oversight. On the other hand, the ISOS has tools to assess compliance with each bundle element through the online process surveillance.
The successful application of the IMA and ISOS resulted in significant reductions in the rates of CLAB, VAP and CAUTI in pooled multinational studies in ICUs of many countries, and also at national level.
Finally, it is to be noted that a reduction in DA-HAI rates cannot be expected to derive from surveillance by itself, and such educational efforts may be short-lived if regular reinforcement is absent. For this reason, in a context where there is lack of financial resources, it is compelling to find and show the information on the incidence and magnitude of the burden of HAI at the hospital level. The collection of this data must be used for improvement of patient care practices, higher adherence to published infection control guidelines, and performance feedback.
As reported in different studies from limited-resource countries, disseminating data on morbidity and mortality due to HAI, and avoidable patient suffering and economic impact, is a necessary approach to move the hospital administration and healthcare workers into supporting the infection control program.
What are the key conclusions from available meta-analyses or systematic reviews related to healthcare-acquired infections in settings with limited resources that guide infection control practice and policies?
Meta-analyses and systematic reviews on HAI have been scant in settings with limited resources. Furthermore, such analyzes could not retrieve enough data from some regions and many countries were not even represented.
The systematic review and meta-analysis on the burden of endemic HAI in settings with limited resources by Allegranzi et al., concluded that HAI prevalence was significantly higher in low and middle low-income countries compared to high-income countries. The density of DAI in critically ill patients was found to be from two- to 19-fold higher than those reported from developed countries, and SSI was found to be the leading infection, with rates eight-fold higher than those reported from high-income countries.
In a review on incidence of CLABs in limited-resource countries by Rosenthal et al., in 2009, it was reported that the CLAB rate ranged from 1.6 to 44.6 cases per 1,000 central line days in adult and pediatric ICUs and from 2.6 to 60.0 cases per 1000 central line days in NICUs, and was associated with significant extra mortality.
Similarly, in a systematic review by Arabi et al., on VAP in adults in settings with limited resources, from 1966 to 2007, the rates of VAP were higher overall than NHSN benchmark rates, and ranged from 10 to 41.7 per 1,000 ventilator-days. The review found that the crude mortality attributable to VAP ranged from 16% to 94%.
According to this review, a small number of VAP intervention studies were performed, which found that staff education programs, implementation of hand hygiene, VAP prevention guidelines, and implementation of sedation protocol were related to a significant reduction in VAP rates.
A meta-analysis of time-sequence cohort studies performed in four countries with limited resources analyzed the specific impact of switching from an open to a closed infusion system on rates of central line-associated bloodstream infection. The results of this analysis indicated that switching from an open to a closed infusion container resulted in a 67% reduction in the overall CLAB incidence (from 10.1 to 3.3 CLABs per 1,000 central line-days) and all-cause ICU mortality. The data reported suggested an association between open infusion containers and an increased risk of infusion-related bloodstream infection and increased ICU mortality.
The findings of these reviews and meta-analyses evidence the urgent need to improve surveillance, infection control practices, update outdated technology, and to increase the number of intervention studies to reduce these high DA-HAI rates in settings with limited resources. Therefore, additional epidemiological studies are to be performed to develop more-definitive approaches for DA-HAI prevention in the form of practical, low-cost, cost-effective technology measures that are feasible to implement in limited-resource countries.
What are the key conclusions from available clinical trials related to healthcare-associated infections in settings with limited resources that guide infection control practices and policies?
Regarding clinical trials performed in INICC’s ICUs, conclusions were consistent in that there is a paramount need to implement a multi-faceted infection control program including education, outcome and process surveillance, and performance feedback (in particular, in a multidimensional approach that includes a bundle of targeted infection control measures) since they were associated with significant reductions in rates of DA-HAI and their adverse effects. After reviewing the studies published from 2002 to 2016, the author summarized some findings below, whose details are shown in Table XI.
There are different multi-national, multi-centric publications from settings with limited resources.
Different clinical trials: Impact of intervention on rates of Device-Associated Healthcare-Acquired Infections
In Argentina, at a hospital member of the INICC, a program was implemented which consisted of focused education, process surveillance, and frequent performance feedback which produced a sustained improvement in compliance with hand hygiene, coinciding with a reduction in HAI rates in the ICUs. Hand hygiene adherence improved progressively from 23.1% (268/1160) to 64.5% (2056/3187) (RR: 2.79, 95% CI: 2.46-3.17, P value: < .0001). During the same period, overall HAI rate in both ICUs decreased by 41% from 47.55 to 27.93 HAIs per 1,000 patient-days (RR: 0.59, 95% CI: 0.46-0.74, P-value: < .0001).
Impact of Intervention on Rates of Central line-associated bloodstream infection
With regard to the reduction of central line-associated bloodstream infection (CLAB), in a time-sequence analysis of the effectiveness of this multi-faceted approach in reducing rates of CLAB in 15 countries with limited resources from INICC, it was concluded that after implementing the infection control program, adherence to infection control compliance significantly improved, the CLAB incidence was reduced by 54% (16.0 to 7.4 CLABs per 1,000 CL-days; RR 0.46, 95% CI 0.33 – 0.63, P< 0.001) and the number of CLAB-associated deaths decreased by 58%.
A study was performed by INICC on pediatric intensive care units (PICUs) of five countries with limited resources to analyze the impact of a multidimensional infection control approach on CLAB rates. The approach included (1) a bundle of infection control interventions, (2) education, (3) outcome surveillance, (4) process surveillance, (5) feedback of CLAB rates, and (6) performance feedback of infection control practices. After intervention, the CLAB was reduced from baseline by 52% (10.7 to 5.2 CLABs per 1000 CL-days; RR 0.48, 95% CI 0.29 – 0.94, P 0.02).
A similar multidimensional approach for CLAB reduction was adopted in another study conducted by INICC in NICUs of 4 countries with limited resources. During baseline, the CLAB rate was 21.4 per 1,000 CL days, and after intervention, the CLAB rate decreased to 9.7 per 1000 CL days [RR 0.45 (95% CI 0.33 – 0.63)], showing a 55% CLAB rate reduction.
In India, a study was conducted to evaluate the impact of the INICC multidimensional infection control approach on CLABSI rates in 16 adult ICUs of 11 hospitals, members of INICC. During the baseline period, outcome surveillance of CLABSI was performed, applying the definitions of the CDC/NHSN (US Centers for Disease Control and Prevention/National Healthcare Safety Network). During the intervention, the INICC approach was implemented, which included a bundle of interventions, education, outcome surveillance, process surveillance, feedback on CLABSI rates and consequences, and performance feedback. Random effects Poisson regression was used for clustering of CLABSI rates across time periods. The baseline rate was 6.4 CLABSIs per 1000 CL-days, which was reduced to 3.9 CLABSIs per 1000 CL-days in the second year and maintained for 36 months of follow-up, accounting for a 53% CLABSI rate reduction (incidence rate ratio 0.47, 95% confidence interval 0.31-0.70; p=0.0001).
In Turkey, a study was conducted to analyze the effect of education on the rate of CLAB. During the pre-education period, the CLAB rate was 8.3 infections per 1,000 CL-days, and during the post-education period, the CLAB rate was 4.7 infections per 1,000 CL-days. In another study conducted in Turkey, 133 patients requiring CL were chosen at random to receive either an antiseptic-impregnated triple-lumen line (N=64) or a standard triple-lumen line (N=69). The CLAB rates were 5.3/1,000 CL-days for the antiseptic line group and 1.6/1,000 CL-days for the standard line group (P=0.452). The results of this study indicated that the use of antiseptic-impregnated central lines had no effect on the incidence of either line colonization or CLAB in critically ill patients.
In Saudi Arabia, a prospective before-after surveillance study was conducted on 3,769 patients hospitalized in 4 adult ICUs and 1 paediatric ICU of 5 hospitals of 5 cities. During baseline, performed outcome and process surveillance of CLABSI was conducted applying CDC/NHSN definitions. During intervention, the INICC Surveillance Online System (ISOS) International and INICC Multidimensional Approach (IMA) were applied: 1- a bundle of infection prevention practice interventions, 2- education, 3- outcome surveillance, 4- process surveillance, 5- feedback on CLABSI rates and consequences and 6- performance feedback of process surveillance. During baseline, 4,468 CL days and 31 CLABSIs were recorded, accounting for 6.9 CLABSIs per 1000 central line (CL) days. During intervention, 12,027 CL days and 37 CLABSIs were recorded, accounting for 3.1 CLABSIs per 1000 CL days. The CLABSI rate was reduced by 56% (incidence-density rate: 0.44; 95% CI 0.28–0.72; P 0.001).
According to the first randomized controlled trial (RCT) conducted to compare rates of CLAB between patients using a closed system with a pre-pierced septum (Split Septum) and single-use prefilled flushing devices (SUF) and those using an open system (three-way stopcocks) and manual admixture (MA), which was conducted by the INICC in India, a significantly lower incidence of CLABs and higher cost-effectiveness were observed in the Split Septum +SUF group compared with the three-way stopcocks +MA group. Coincidentally, the use of the Split Septum +SUF significantly improved the cumulative infection-free catheter survival compared with the three-way stopcocks +MA (hazard ration, 0.33; 95% CI, 0.15-0.73; P = .006). Using a Split Septum + SUF represented savings of $402.88 and an increase in quality-adjusted life years of 0.0008 per patient. For each extra dollar invested in a Split Septum + SUF, $124 was saved. In conclusion, the use of Split Septum + SUF is cost-effective and associated with a significantly lower CLAB rate compared with the use of three-way stopcocks. Nevertheless, the extended suffering of patients and their relatives cannot be estimated in terms of economic costs only (see Table V, Table VI, Table VII, Table VIII, Table IX).
In a prospective before/after trial performed in Argentina, at hospitals members of INICC, the rates of CLAB determined during a period of active surveillance without education or performance feedback (phase 1) were compared to rates of CLAB after sequential implementation of an infection control program that included education (phase 2) and performance feedback (phase 3). Overall rates of CLAB were reduced by 75%, from 46.63 to 11.10 BSIs per 1,000 IVD-days (RR = 0.25, 95% CI = 0.17-0.36, P-value = <0.0001).
A prospective, controlled, time-series, cohort trial was undertaken in adult patients admitted to four level-III adult ICUs in Argentina to study the impact of open and closed infusion systems on rates of CLAB. Rates of CLAB during a period of active surveillance with an open infusion system (baseline; externally vented, semi-rigid, non-collapsible, 1-port plastic bottles) were compared with rates after switching to a closed system (intervention; non-vented, collapsible, 2-port plastic bags). The incidence of CLAB during use of the closed system was significantly lower than during use of the open system (2.36 versus 6.52/1,000 central line days, relative risk=0.36, 95% confidence interval=0.14-0.94, P=.02).
An open-label, prospective cohort, active healthcare-associated infection surveillance, sequential study was conducted in three intensive care units in Brazil at hospitals members of INICC, to determine the rate and time to develop first CLAB when comparing open and closed infusion containers. The probability of acquiring CLAB was assessed over time and compared between open and closed infusion container periods; 3-day intervals were examined. CLAB rate was significantly higher during the open compared with the closed infusion container period (6.5 versus 3.2 CLAB/1000 CL days; RR=0.49, 95% CI=0.26-0.95, p=0.031). During the closed infusion container period, the probability of acquiring a CLAB remained relatively constant along the time of central line use (0.8% Days 2-4 to 0.7% Days 11-13) but increased in the open infusion container period (1.5% days 2-4 to 2.3% days 11-13). Combined across all time intervals, the chance of a patient acquiring a CLAB was significantly lower (55%) in the closed infusion container period (Cox proportional hazard ratio 0.45, p=0.019).
In Brazil, an educational program was developed by a multidisciplinary task force to highlight correct practices for central line (CL) care. Before intervention, the CLAB rate was 20 per 1000 CL days, and after the educational intervention and policy change, such as standardized use of povidone-iodine during dressing care, the number of CLAB dropped to 11 per 1,000 CL-days.
In Mexico, a prospective before/after trial was performed at level III adults ICUs in one public university hospital member of INICC. During a period of active surveillance without process control, rates of CLAB were determined (phase 1) and were then compared to rates of CLAB after implementing an infection control program that applied process surveillance and performance feedback (phase 2). Compliance with CL site care and hand hygiene improved significantly from baseline during the study period: placing a gauze dressing over the catheter insertion site improved from 86.69% to 99.24% (RR:1.14, 95% CI:1.07-1.22, P-value: 0.0000.), proper use of gauze central line insertion site improved from 84.21% to 97.87% (RR: 1.16, 95% CI:1.09-1.24, P-value: 0.0000.), documentation of date of placement of administration set of vascular catheter improved from 40.69% to 93.85% (RR: 2.34, 95% CI:2.14-2.56, P-value: 0.0000), hand hygiene prior to contact with the patient improved from 62% to 84.9% (RR:1.37, 95% CI:1.21-1.51, P-value: 0.0000). Overall rates of CLAB were significantly reduced by 58% after implementing a process control program, from 46.3 to 19.5 CLABs per 1000 CL-days (RR = 0.42, 95% CI = 0.27-0.66, P-value = 0.0001). Finally, overall rates of crude unadjusted mortality were lowered significantly from baseline rates, from 48.5% per 100 discharges to 32.8% (RR: 0.68, 95% CI: 0.50-0.31, P-value: 0.01).
An open label, prospective cohort, active DA-HAI surveillance, sequential study was performed in four ICUs in Mexico at hospitals members of INICC to determine the effect of switching from open (glass, burettes, and semi-rigid) infusion containers to closed, fully collapsible, plastic infusion containers on the rate and time to onset of CLABs. The CLAB rate was significantly higher during the open versus the closed container period (16.1 versus 3.2 CLAB/1000 central line days; RR = 0.20, 95% CI = 0.11-0.36, P < 0.0001). The probability of developing CLAB remained relatively constant in the closed container period (1.4% days 2-4 to 0.5% Days 8-10), but increased in the open container period (4.9% days 2-4 to 5.4% days 8-10). The chance of acquiring a CLAB was significantly decreased (81%) in the closed container period (Cox proportional hazard ratio 0.19, P < 0.0001), and mortality was statistically significantly lower during the closed compared with the open container period (23.4% versus 16.1%; RR = 0.69, 95% CI = 0.54-0.88, P < 0.01).
In Tunisia, a randomized, controlled trial was conducted in which 246 patients with non-tunneled CL were randomly assigned to receive a heparin-coated line with 50 mL/d of normal saline solution as a continuous infusion (heparin-coated group) or a non-coated catheter with a continuous infusion of low-dose unfractionated heparin (control group: continuous infusion of 100 U/kg/d). CLAB occurred in 0.9 events per 1,000 days in the heparin-coated group and in 3.5 events per 1,000 days in the control group (3.5 events per 1,000 days; P = 0.027). The conclusion of this study stated that the use of heparin-coated lines could be a safe and effective approach to the prevention of CLAB in patients with hemato-oncologic disease.
Impact of intervention on rates of Ventilator-Associated Pneumonia
As regards the reduction of VAPs, in another multi-center study conducted by INICC in adult ICUs of 14 countries with limited resources, a multi-dimensional approach was applied with the aim of reducing the rates of VAP. The VAP rate at baseline was 22.0, and after intervention, it decreased to 17.2 per 1,000 MV days (RR; 0.78; 95% CI 0.68-0.90; P 0.0004), showing a 55.83% VAP rate reduction.
With the same approach, in a study conducted in PICUs of five countries with limited resources it was shown that the rate of VAP at baseline was 11.7, and after intervention, it had decreased to 8.1 per 1,000 MV days (RR; 0.69; 95% CI 0.5-0.96; P 0.02), showing a 31% VAP rate reduction.
Another similar study to assess the effectiveness of a multidimensional approach on VAP rates was recently performed by INICC in NICUs of 10 countries with limited resources. The VAP rate during Phase 1 period was 17.8, and during Phase 2 period was 12.0 per 1000 MV days (RR; 0.67; 95% CI 0.50-0.91; P 0.001), showing a reduction in the VAP rate of 33%.
In Turkey, a prospective before-after study evaluated the impact of the INICC multidimensional approach on the reduction of VAP in adult patients hospitalized in 11 ICUs, from 10 hospitals, members of the INICC, in 10 cities of Turkey. The baseline rate of VAP was 31.14 per 1,000 MV-days, and was reduced to 16.82 per 1,000 MV-days during intervention, amounting to a 46 % VAP rate reduction (RR, 0.54; 95 % CI, 0.42-0.7; P value, 0.0001.).
In India, the INICC multidimensional approach for the reduction of VAP was assessed in adult patients hospitalized in 21 ICUs from 14 INICC member hospitals in 10 Indian cities. The VAP rate was 17.43/1000 mechanical ventilator days during baseline, and 10.81 for intervention, showing a 38% VAP rate reduction (relative risk 0.62, 95% confidence interval 0.5-0.78, P = 0.0001).
In a before-after study performed in four level III adult ICUs in two Argentinian hospitals, members of INICC, it was reported that after the implementation of a multi-faceted infection control program, the rate of VAP was successfully reduced by 31%, from 51.28 to 35.50 episodes of VAP per 1,000 MV-days (RR = 0.69, 95% CI: 0.49-0.98, P
In China, a before-after study was conducted by INICC members from January 2005 to July 2009, to evaluate the implementation of a multidimensional approach for VAP reduction. The VAP baseline rate was 24.1 per 1000 ventilator-days, which was significantly decreased to 5.7 per 1000 ventilator-days in 2009 (2009 vs 2005: relative risk, 0.31; 95% confidence interval, 0.16-0.36; P = .0001), amounting to a 79% cumulative VAP rate reduction. In Cuba, a pre-post study in AICU patients in an INICC member hospital assessed the effect of the multidimensional approach on the reduction of VAP rates. The baseline rate of VAP was 52.63 per 1000MV days and 15.32 per 1000MV days during the intervention, showing a 70% VAP rate reduction at the end of the study period.
In Pakistan, an observational pre and post-intervention study was conducted to assess whether an educational program focusing on preventive practices for VAP could reduce its incidence. An evidence-based guideline for preventive practices at the bedside was developed and disseminated to the intensive care unit staff. VAP infection rates were reduced by 51%, from a mean of 13.2 VAP in the pre-intervention period to 6.5 VAP per 1,000 device days in the post-intervention period (mean difference 6.7; 95% CI: 2.9-10.4, P =0.02).
In Thailand, a study was performed to determine the long-term effect of an educational program to prevent VAP in a medical ICU (MICU). The educational program involved respiratory therapists and nurses, and included a self-study module with pre-intervention and post-intervention assessments, lectures, fact sheets, and posters. Before the intervention, there were 20.6 cases per 1,000 ventilator-days in the MICU, and after intervention the rate of VAP decreased by 59% to 8.5 cases per 1000 ventilator-days; P=. 001.
Impact of Intervention on Rates of Catheter-Associated Urinary Tract Infection
In relation to CAUTI, a before-after study conducted in 15 countries with limited resources, at INICC member hospitals, evaluated the impact of a multidimensional infection control strategy for the reduction of the incidence of CAUTI in patients hospitalized in adult ICUs. Before the intervention, the CAUTI rate was 7.86 per 1,000 UC-days, and after intervention, the rate of CAUTI decreased to 4.95 per 1,000 UC-days [relative risk (RR) 0.63 (95% confidence interval [CI] 0.55-0.72)], showing a 37% rate reduction.
Likewise, a study was conducted by INICC in PICUs from six countries with limited resources; the study analyzed the impact of a multi dimensional approach developed by INICC to reduce CAUTI rates. In Phase 1, the CAUTI rate was 5.9 per 1,000 UC days, and in Phase 2, after implementing the multidimensional infection control approach for CAUTI prevention, their rate of CAUTI decreased to 2.6 per 1,000 UC days [RR 0.43 (95% CI 0.21–1.0)], showing a rate reduction of 57%.
In Turkey, a before-after prospective active surveillance study evaluated the effectiveness of the INICC multidimensional infection control approach for the reduction of CAUTI in 13 ICUs in 10 hospital members of the INICC. During phase 1, the rate of CAUTI was 10.63 per 1,000 UC-days and was significantly decreased by 47% in phase 2 to 5.65 per 1,000 UC-days (relative risk, 0.53; 95% confidence interval: 0.4-0.7; P value = .0001).
In an open trial in an Argentinian hospital, members of INICC, performed by Rosenthal et al., rates of CAUTI were determined during a baseline period of active surveillance without education and performance feedback, and were then compared with rates of CAUTI after implementing education, process surveillance, and performance feedback regarding catheter care measures and hand hygiene compliance. The findings showed that the CAUTI rate decreased significantly by 42%, from 21.3 to 12.39 CAUTIs per 1,000 catheter-days (RR, 0.58; CI%, 0.39 to 0.86; P = .006). With regard to hand hygiene compliance, three Argentinian hospitals members of INICC were studied for adherence to a hand hygiene protocol, and 15,531 patient contacts were observed. The baseline rate of hand hygiene before contact with patients was 17%. The implementation of a program consisting in education, hand hygiene before contact with patients increased to 44% (RR 2.65; 95% CI 2.33-3.02, P-value: <0.001), and with education and performance feedback, hand hygiene further increased to 58% (RR 1.86; 95% CI 1.38-2.51; P value: <0.001).
In Lebanon, a study assessed the impact of a multidimensional infection control approach for the reduction of CAUTI adult ICU patients of a hospital member of the INICC. The baseline rate of CAUTI was 13.07 per 1000 urinary catheter-days, and was decreased by 83% to 2.21 per 1000 urinary catheter-days (risk ratio 0.17; 95% confidence interval 0.06-0.5; p=0.0002).
In the Philippines, a before-after prospective active surveillance study was conducted to assess the impact of the INICC multidimensional infection control approach on the reduction of CAUTI rates in adult ICUs in two hospitals in the Philippines, members of the INICC. The rate of CAUTI was 11.0 per 1000 UC-days at baseline and was decreased by 76% to 2.66 per 1000 UC-days during intervention (rate ratio [RR], 0.24; 95% confidence interval [CI], 0.11-0.53; P-value, 0.0001).
The extracted findings from the available clinical trials are representative and consistent evidence of the effectiveness that multi-faceted infection control strategies can have in settings with limited resources. Within the broad spectrum of infection control, to successfully address the burden of HAI in limited-resource healthcare facilities, it has been key to implement surveillance of DA-HAI rates and of processes related to appropriate use and care of devices, educate healthcare workers, assesses their practices, and provide them with feedback of observed processes, and ensure adequate observations of the recommendations set forth in published guidelines. These findings reveal that the reduction of DA-HAIs is feasible and cost-effective in settings with limited resources; therefore, this valid evidence should lead to the mandatory organization of multi-dimensional infection control programs at every hospital.
What are the consequences of ignoring key concepts related to infection control in settings with limited resources?
As described in this chapter, the consequences of ignoring infection control are high HAI rates, higher mortality, high extra cost, high bacterial resistance, and excessive use of antibiotics and apparition of super bugs.
What other information supports the key conclusions of studies of or advice from infection control in settings with limited resources?
The increasing recognition of the preventive value of implementing an organized infection control program has allowed the publication of data reflecting economic concerns. In different published scientific articles, the economic impact that HAI have on the hospital administration, and at regional and national level in the form of “extra cost” has been analyzed as another preventable consequence in the developing nations.
Case control studies
The following case control studies were published showing extra cost, extra length of stay and extra mortality attributable to HAI.
In order to calculate the cost of CLAB in intensive care units, a 5-year prospective nested case-control study was undertaken in six adult ICUs from three hospitals of Argentina, members of INICC. One hundred and forty-two patients with CLAB (cases) and 142 patients without CLAB (controls) were matched for hospital, type of ICU, year of admission, length of stay, gender, age, and average severity of illness score. The mean extra LOS for cases (compared to the controls) was 11.90 days, the mean extra antibiotic defined daily doses was 22.6, the mean extra antibiotic cost was $1,913, the mean extra cost was $4,888.42, and the excess mortality was 24.6%.
With a view to calculating the cost of CLABs in intensive care units, an 18- month prospective nested case-control study was undertaken at three hospitals in Mexico City, members of INICC, in four ICUs. Fifty-five patients with CLAB (cases) and 55 patients without CLAB (controls) were compared by analyzing hospital, type of ICU, year of admission, length of stay, gender, age, and average severity of illness score. The results indicated that extra LOS of patients with CLAB was 6.05 days. The mean extra cost of antibiotics amounted to $598, the mean extra cost of other drugs was $25.77, and the mean extra cost of hospitalization was $8,326. The mean extra cost for cases (compared to the controls) amounted to $11,591. Finally, the extra mortality attributable to BSI was 20%.
A study to estimate the excess LOS in an intensive care unit (ICU) due to CLAB was performed in hospitals members of INICC in three Latin American countries (Argentina, Brazil, and Mexico). An analysis was made by means of a statistical model that accounted for the timing of infection. A cohort of 3,560 patients hospitalized in 11 ICUs was followed for 36,806 days. The average excess LOS due to a CLAB increased and varied between –1.23 days to 4.69 days.
In order to calculate the cost of nosocomial pneumonia in intensive care units, a 5-year matched cohort study was undertaken at six ICUs of three hospitals in Argentina members of INICC. Three hundred and seven patients with VAP (exposed) and 307 patients without VAP (unexposed) were matched for hospital, ICU, period, LOS more than 7 days, gender, age, and average severity of illness score (ASIS). The mean extra LOS for 307 cases (compared to the controls) was 8.95 days, the mean extra antibiotic defined daily doses (DDD) was 15, the mean extra antibiotic cost was $996, the mean extra total cost was $2,255, and the extra mortality was 30.3%.
In another study from northern India, patients with VAP experienced significantly longer hospital stay [21 (IQ=14-33) days versus 11 (IQ=6-18) days, P<0.0001)] and incurred greater hospital costs [USD $6250.92 (IQ=3525.39-9667.57) versus $2598.84 (IQ=1644.33-4477.65), P<0.0001]. Multiple regression analysis revealed that the cost-driving factors in this study population were the occurrence of VAP infections (P<0.0001) and the duration of hospital stay (P<0.0001). The attributable cost of VAP infection was calculated to be USD $5200 (95% CI=3245-7152).
In a study performed in hospitals member of INICC in 10 countries with limited resources to estimate extra LOS and mortality in an intensive care unit (ICU) due to a VAP, a cohort of 69,248 admissions were followed for 283,069 days in ICUs. Data were arranged according to a multi-state format. Extra LOS and increased risk of death were estimated independently in each country, and their results were combined using a random effects meta-analysis. The findings of the analysis showed that a VAP prolonged LOS by an average of 2.03 days (95% CI: 1.52, 2.54 days), and increased the risk of death by 14% (95% CI: 2, 27%).
To estimate the excess LOS and mortality in the intensive care unit (ICU) attributable to catheter-associated urinary tract infections (CAUTI), a statistical model that accounted for the timing of infection was applied in 29 ICUs of hospitals members of INICC from 10 countries: Argentina, Brazil, Colombia, Greece, India, Lebanon, Mexico, Morocco, Peru, and Turkey. In a cohort of 69,248 admissions followed for 371,452 days in 29 ICUs, a multi-state model was applied to estimate the extra LOS due to infection. This model included specific censoring to ensure that estimations considered the independent effect of urinary tract infection, and not the combined effects of multiple infections. The extra LOS and increased risk of death independently for each country, and then combined the results using a random effects meta-analysis. The conclusions showed that a CAUTI prolonged length of ICU stay by an average of 1.59 days (95% CI: 0.58, 2.59 days), and increased the risk of death by 15% (95% CI: 3, 28%).
The above-referred findings of case control studies performed in settings with limited resources stress on the adverse consequences caused by HAIs in terms of increased LOS and extra hospital costs. They emphasize how important continued surveillance is to understand all the aspects, medical and social, involved within the implementation of sound infection control programs. This knowledge is essential to lead to decreased HAI rates, but it also aids the prioritization of resources and other efforts to improve patient safety.
Does knowledge about healthcare-acquired infection in settings with limited resources help to generate infection control guidelines?
Countries with limited resources
Higher HAI rates may reflect the typical ICU situation in countries with limited-resources as a whole, and several reasons have been exposed to explain this fact. Among the primary plausible causes, it can be mentioned that, in the majority of countries with limited resources, there are still no legally enforceable rules or regulations concerning the implementation of infection control programs, such as national infection control guidelines; yet, in the few cases in which there is a legal framework, adherence to the rules is most irregular and hospital accreditation is not mandatory.
In most hospitals, this lack of official regulations is strongly correlated to the considerable variability found in the compliance with hand hygiene guidelines. This situation is further emphasized by the fact that administrative and financial support in most hospitals is insufficient to fund infection control programs. Available human resources, and supplies are different in settings with limited resources compared to those of developed countries; and this explains why it is not possible to just use guidelines elaborated in developed countries and apply with no changes to the reality of settings with limited resources. Reduced numbers of nurse to patient ratio is associated with increased HAI rates. Extremely low nurse-to-patient staffing ratios , hospital over-crowding, lack of medical supplies, and in an insufficient number of experienced nurses or trained healthcare workers have proved to be highly connected to high DAI rates in ICUs.
In a review about CLAB in settings with limited resources, published by Rosenthal in 2009, a number of structural and behavioral reasons were associated with higher rates of CLAB, and among their most common observations were overcrowded ICUs, insufficient rooms for isolation, lack of sinks, lack of medical supplies in general, including but not limited to alcohol hand rub, antiseptic soap, and paper towels. In addition, a lack of supplies for the wearing of maximal barriers during catheter insertion, a lack of clorhexidine (and thus the use of povidone iodine), a lack of needle-less connectors (and the subsequent use of three ways stopcocks), the use of vented IV containers instead of closed IV systems, a lack of ready to use drugs (and the subsequent reliance on manual admixture for all drugs) were noted.
Poor performances, outdated technology, and poor hygiene
Moreover, poor performances in infection control practices, such as the case of using cotton balls already impregnated with antiseptic contained in a contaminated container, not covering insertion site with sterile dressing, storing drugs in already open single use vials, reusing single use vials, leaving needles inserted in multiple use vials, taking fluids from a 1,000 cc container for dilution of parenteral solutions, and using tacky mats were paramount.
In a study published by INICC in 2010, applying process surveillance a number of measures were found as associated with increased risk of CLAB, and they are the following: lack of hand hygiene, hand washing with non antiseptic soup, insufficient skin antisepsis with clorhexidine, lack of sterile gauze or transparent dressing for catheter care, keep the central line in place beyond the needs, use of three ways stop cock, use of open infusion containers, among others.
The use of outdated technology is a major problem, such as lack of availability of clorhexidine for skin antisepsis instead of povidone iodine, lack of availability of dressing with clorhexidine, lack of availability of closed infusion containers instead of open infusion containers, and lack of availability of needles connectors instead of three ways stopcock. On the other hand the commercialization of chlorexidine is not yet approved in several settings with limited resources.
There is only one meta-analysis with data from settings with limited resources that compared the use of open infusion containers (glass bottle, burette, or semi-rigid plastic bottle) or closed infusion containers (fully collapsible plastic containers) on CLAB rates and mortality in Argentina, Brazil, Italy, and Mexico. CLAB incidence dropped markedly in all four countries after switching from an open to a closed infusion container (pooled results, from 10.1 to 3.3 CLABs per 1,000 central line-days; relative risk [RR], 0.33 [95% confidence interval, 0.24-0.46]; P < .001), and also mortality also decreased significantly, from 22.0 to 16.9 deaths per 100 patients (RR, 0.77 [95% CI, 0.68-0.87]; P < .001). Switching from an open to a closed infusion container resulted in a striking reduction in the overall CLAB incidence and all-cause ICU mortality. Its findings suggested that open infusion containers are associated with a greatly increased risk of infusion-related bloodstream infection and increased ICU mortality that have been unrecognized.
To conclude, it is necessary to highlight that in order to reduce the hospitalized patients’ risk of infection in limited-resource countries, a multidimensional approach is primary and essential. As a first step it is necessary to include the implementation of DA-HAI surveillance, because it effectively describes and addresses the importance and characteristics of the threatening situation created by HAIs. Additionally, surveillance of DA-HAI has played a fundamental role, not only in increasing the awareness of DAI risks, but also providing an exemplary basis for the institution of infection control practices. It is key that surveillance is implemented along with the monitoring of practices of infection control (process surveillance), education, presence of practice bundles, performance feedback, and feedback of DA-HAI rates and consequences.
The high incidence of DA-HAI and mortality has been reduced by carrying out a multidimensional approach, with targeted performance feedback programs for hand hygiene and central line, ventilator, and urinary catheter care. Finally, it is of utmost importance to restrict the administration of anti-infective in order to effectively control of antibiotic resistance; however, this subject exceeds the scope of this chapter.
Copyright © 2017, 2013 Decision Support in Medicine, LLC. All rights reserved.
No sponsor or advertiser has participated in, approved or paid for the content provided by Decision Support in Medicine LLC. The Licensed Content is the property of and copyrighted by DSM.