A systematic review on the causes of the transmission and control measures of outbreaks in long-term care facilities: Back to basics of infection control

Background The unique characteristics of long-term care facilities (LTCFs) including host factors and living conditions contribute to the spread of contagious pathogens. Control measures are essential to interrupt the transmission and to manage outbreaks effectively. Aim The aim of this systematic review was to verify the causes and problems contributing to transmission and to identify control measures during outbreaks in LTCFs. Methods Four electronic databases were searched for articles published from 2007 to 2018. Articles written in English reporting outbreaks in LTCFs were included. The quality of the studies was assessed using the risk-of-bias assessment tool for nonrandomized studies. Findings A total of 37 studies were included in the qualitative synthesis. The most commonly reported single pathogen was influenza virus, followed by group A streptococcus (GAS). Of the studies that identified the cause, about half of them noted outbreaks transmitted via person-to-person. Suboptimal infection control practice including inadequate decontamination and poor hand hygiene was the most frequently raised issue propagating transmission. Especially, lapses in specific care procedures were linked with outbreaks of GAS and hepatitis B and C viruses. About 60% of the included studies reported affected cases among staff, but only a few studies implemented work restriction during outbreaks. Conclusions This review indicates that the violation of basic infection control practice could be a major role in introducing and facilitating the spread of contagious diseases in LTCFs. It shows the need to promote compliance with basic practices of infection control to prevent outbreaks in LTCFs.


Methods
Four electronic databases were searched for articles published from 2007 to 2018. Articles written in English reporting outbreaks in LTCFs were included. The quality of the studies was assessed using the risk-of-bias assessment tool for nonrandomized studies.

Findings
A total of 37 studies were included in the qualitative synthesis. The most commonly reported single pathogen was influenza virus, followed by group A streptococcus (GAS). Of the studies that identified the cause, about half of them noted outbreaks transmitted via person-toperson. Suboptimal infection control practice including inadequate decontamination and poor hand hygiene was the most frequently raised issue propagating transmission. Especially, lapses in specific care procedures were linked with outbreaks of GAS and hepatitis B and C viruses. About 60% of the included studies reported affected cases among staff, but only a few studies implemented work restriction during outbreaks.

Introduction
Outbreak of an infectious disease is defined as the occurrence of a disease above the expected level [1]. Over the past several years, many countries have experienced serious economic and health consequences due to outbreaks of infectious diseases such as the Middle East Respiratory Syndrome in 2015 and Severe Acute Respiratory Syndrome in 2003. Long-term care facilities (LTCFs) are facing a great need for preparation for infection outbreaks because of an increase in the number of residents with global aging. LTCFs are exposed to the risk of outbreaks owing to several factors. First, older residents in LTCFs are susceptible to infectious diseases because of aging and health conditions [2] and are known to be dependent with regard to activities of daily living. Thus, among residents, self-hygiene is observed to be poor. Loss of independence in residents creates unique and frequent contact opportunities between healthcare workers (HCWs) and residents [3]. Second, HCWs in LTCFs tend to be poorly informed about infection prevention and control (IPC), and compliance with IPC is generally low [2,4]. Third, the environment in LTCFs offers challenges for IPC, like the sharing of rooms, group living, and difficulty with the isolation of infected persons [5,6]. Finally, LTCFs have limited resources and capacities for diagnosis of infection [7]. This leads to a delay in the detection of hidden carriers and infection. All these factors contribute to the onset and spread of outbreaks in LTCFs. Outbreaks in LTCFs threaten the life and health of both residents and HCWs, and thus, eliminating the risk of outbreaks is a matter of concern in such facilities. However, LTCFs vary in their individual capacities to respond to outbreaks [8]. The keys to outbreak control are as follows: identification of the transmission causes and minimization of the spread through early initiation of control measures. Therefore, it is essential to understand the causes of transmission and the applied measures to control outbreaks in LTCFs.
There are several gaps and limitations in previous studies to comprehensively understand outbreaks in LTCFs. First, in many previous researches that addressed outbreaks, the focus was on pathogens, burdens, and adverse outcomes such as mortality [9,10]. Based on the perceived importance of control measures, one of the purposes of this review was to explore and analyze in detail the control measures reported in studies. Secondly, pharmaceutical measures may have some limitations on effectiveness against newly emerging infectious diseases or resistant strains [3] and some pathogens may not have pharmaceutical interventions to be considered during outbreaks. Non-pharmaceutical interventions (NPIs) such as hand hygiene and precautions should be utilized to prevent the transmission of outbreak pathogens, regardless of the evolution of infectious diseases. However, little attention has been paid to NPIs in studies concerning outbreaks in LTCFs [3]. This review focused on NPIs and ascertains the control measures based on the guideline for prevention and control of influenza outbreaks in LTCFs of the World Health Organization (WHO) [11]. Finally, to our knowledge, there is no published systematic review addressing overall outbreaks in LTCFs in the past 5 years.

Data extraction
After pilot data extraction, two reviewers independently extracted data such as information on participants, pathogens, case definitions, number of cases and non-affected persons, overall attack rate, causes and problems that led to transmission, and control measures.
Control measures. General control measures considered were the formation of outbreak control team, active surveillance, standard precautions, transmission-based precautions, training and education, employee work restriction, environmental control, containment measures, and prophylaxis based on the WHO guideline [11].

Quality assessment
The quality of studies was assessed using Risk of Bias Assessment tool for Nonrandomized Study (RoBANS) by two reviewers [13]. RoBANS is an evaluation tool for the risk of bias of non-randomized studies, with moderate reliability and acceptable validity and compatible with domains of the Cochrane risk-of-bias tool [13]. Six domains were evaluated including the selection of participants, confounding variables, exposure measurement, blinding for outcome assessment, incomplete outcome, and selective outcome reporting. According to the instruction for evaluation [13], the risk of bias for each domain was determined as low risk, unclear risk and high risk. Studies with full-text including case-control analysis, cohort study was involved in the quality assessment. Studies that simply described the results of investigation without comparative analysis were not able to evaluate the domains of the tool. Thus, the quality evaluation was not conducted with this type of study. Any difference was discussed between the two reviewers, and, if necessary, an agreement was reached with the corresponding author. CINAHL TI (infection OR infections OR outbreak � OR transmission) AND TI ("nursing home � " OR "skilled nursing � " OR "long-term care") AND TI (control � OR outcome � OR factor � ) NOT TI (surgery OR cancer OR "neoplasm" or "intensive care unit" OR child OR children OR "operative") The result of quality assessment was displayed using Review Manager (RevMan) version 5.3 software (The Cochrane Collaboration, Oxford, UK).

Search results
A total of 2,789 studies were retrieved from 4 databases and hand searched. The duplicate records were removed (n = 1180), and the eligibility criteria were applied for the selection process. After reviewing the full text, 76 articles were excluded for the following reasons: irrelevant for the research topic (n = 38) and population (n = 4), unavailable full-text (n = 32), review article (n = 1), and duplicated report (n = 1). Finally, 37 articles were included in this review (Fig 1).

Characteristics of the included studies
Characteristics of the eligible studies are presented in Table 2. Over half of the included studies (n = 22) were published since 2013. The majority of the included studies were reported in the United States (n = 15) and Europe (n = 13) followed by Asia (n = 5).

Risk of bias in the included studies
The quality of 20 studies was assessed and the results are summarized in Fig 2. One study was at high risk for five criteria [14]. Six studies were at low risk for all criteria [15][16][17][18][19][20]. Problems related to recall bias and standardization of self-reported measurement created a high risk of bias for the measurement of the exposure domain in seven studies [14,[21][22][23][24][25][26]. Lack of consideration for confounders led to a high risk of bias in four studies [24,[27][28][29]. The problem related to missing data resulted in a high risk of bias for the incomplete outcome domain in five studies [14,26,28,30,31].
In total, 37 studies reported 1,332 outbreaks (affecting 1,122 residents and 385 staff members) in 1,182 facilities. There were three prolonged GAS outbreaks of multiple consecutive clusters for over 6 months. The overall attack rates ranged widely from 0.84% to 73.17% in 29 studies. Among the 29 studies, the median of the overall attack rate was 15.73%: 8.27% for bacterial outbreaks and 19.25% for viral outbreaks (Table 3). In 8 studies, it was not possible to calculate the rate due to lack of information. The highest attack rate of 73.17% was reported in an outbreak of respiratory syncytial virus (RSV) and human metapneumovirus (HMPV) [50], followed by Clostridium difficile (51.97%) [27] and viral gastroenteritis caused by norovirus and rotavirus (48.6%) [26]. Influenza-like illness had the median overall attack rate of 24.50%. The median attack rate among staff was highest for the acute gastroenteritis outbreaks.

PLOS ONE
Outbreaks in long-term care facilities: SR  If only one study of the outbreaks was reported, the attack rate of the study was displayed. If the number of reports was 2, only the range of attack rate was displayed.

PLOS ONE
Outbreaks in long-term care facilities: SR  Causes and critical problems contributing to transmission. Causes of transmission in the eligible studies were reported as person-to-person transmission, contaminated water and food, and problems in practice (Tables 2 and 4). The following studies (n = 12) did not report or could not identify the cause of the outbreaks: 5 studies on influenza viruses, 2 on non-typeable Haemophilus influenzae, 1 on hepatitis C virus, 1 on C. difficile, 1 on adenovirus, 1 on norovirus, and 1 on RSV and HMPV [14,27,29,30,32,36,40,42,44,45,48,50]. The most commonly reported route of transmission was person-to-person [17, 22-24, 26, 31, 33, 35, 38, 41, 43, 46, 47, 49]. Of these studies, the human source of transmission was identified as the HCWs in six outbreaks [22,24,26,38,41,49] and the residents in two outbreaks [17,33]. The skin of a HCW was a reservoir for Methicillin-resistant Staphylococcus aureus (MRSA) outbreaks, leading to cross-infection, in the study by Maltezou et al. [22]. The large gastroenteritis outbreak affecting 394 people in 8 facilities was attributed to staff who worked at multiple facilities [49]. One study showed that three outbreaks of GAS recurred for three years because of continued person-to-person transmission from colonized residents [17]. In the study by Šubelj and Učakar [23], person-to-person transmission resulted from delayed implementation of control measures. Contaminated water and food were sources of infection in five studies [16,25,26,33,46]. Hepatitis E outbreak was caused by contaminated tap water after heavy rain [25], while the consumption of tap water was the suspected cause of one viral gastroenteritis outbreak [26]. Foodborne causes such as contaminated cake or meals were noted in three studies regarding Clostridium perfringens, Salmonella enteritidis, and gastroenteritis [16,33,46].
Most of the reviewed studies pointed out several issues in practice that might have facilitated the occurrence and spread of outbreaks. The most frequently observed problem was suboptimal hand hygiene, followed by personal protective equipment (PPE), and cleaning and disinfection. Investigation for the GAS outbreak in the study by Nanduri et al. [37] revealed that hand hygiene compliance among employees was 14-25%. Additionally, poor hand hygiene became a more critical factor that facilitated the transmission of acute gastroenteritis, particularly in LTCFs having close living conditions with frequent close contact between the staff and dependent residents [31]. Issues related to PPE had been addressed including inappropriate use of glove and improper storage of PPE [20,24,30,31,34,37]. There were reports indicating the potential to cross-contamination by not-changing gloves between residents or by storage of PPE in the room of the index case [31,34]. Breaches in disinfection and cleaning of the environment and equipment were associated with many outbreaks, most of them were GAS [20,37,38] or hepatitis B and C outbreaks [28,30]. Three reports of GAS outbreaks found lapses in wound care practice such as inconsistent cleaning and disinfection [20,37,38]. The outbreaks of MDROs and hepatitis B reported device related issues including sharing of a device and inappropriate use of reusable devices [15,18,21,34]. Hepatitis B and C outbreaks commonly reported that lapses during podiatry care and point-of-care testing procedures (blood glucose test and international normalized ratio monitoring) caused the transmission of bloodborne pathogens among residents [18,19,21,28,30]. The lapses included the sharing of contaminated equipment, improper disinfection, and poor hand hygiene adherence.
Some studies noted failure of environmental infection control [32,35,39,43]. Two of those studies were TB epidemics, and the investigation revealed that the case residents were exposed to insufficient room ventilation. An influenza outbreak in the summer was facilitated by a heating preventive measure that placed all the residents in one limited area [43]. The response to outbreaks also could influence the progress of outbreaks. Nine reports underlined early notification of outbreaks to public health authorities and implementation of control measures within 3 days of onset of the first case, which affected the attack rates and duration of the outbreaks [23,29,35,41,44,46,[48][49][50].
Some studies on influenza outbreaks discussed issues related to vaccines. Of the three influenza outbreaks in a well-vaccinated population, two studies pointed out that a mismatch between the circulating strains and the vaccine strains affected this population [41,45], and the other study noted an insufficient vaccine effectiveness [42]. Especially, the study by Burette et al. [41] identified that in addition to the mismatch, several defects including a vaccination rate of 0% among staff and untimely vaccination among residents led to the outbreaks. In addition, they raised the issue of the knowledge and proficiency of general practitioners in influenza diagnosis, treatment, and prevention.
Moreover, three studies suspecting transmission from staff to residents placed emphasis on work restriction of the ill staff [38,40,49]. There were other problems including poor personal hygiene of staff members [22], lack of communication between institutions [17], and understaffing [24].
Although not the result of the case-control analysis, the study by Spires et al. [50] reported that all cases of RSV and HMPV were dependent dementia patients, implicating that dependence was an important factor.
Control measures. Strategies to control outbreaks were reported in 30 of the 37 reviewed papers, as summarized in Table 5. All 30 studies reported that one or more NPIs were applied to control the outbreaks. From a rigorous perspective, only one study on a multidrug-resistant Pseudomonas aeruginosa (MRPA) outbreak implemented all the measures recommended to control an outbreak by pathogens [15]. Work restriction of ill workers was less frequently reported compared to other measures. Only five studies reported the creation of outbreak control teams for effective management of the outbreaks [15,17,23,27,43]. Most facilities notified public health authorities or institutions about the outbreaks and received advice and assistance to manage the outbreaks. All four studies applying limitation or cessation of group activities were recently published since 2014 [31,42,44,50].
Gastroenteritis outbreaks (n = 5). Three studies on gastrointestinal infection, in which adherence to hand hygiene among HCWs was crucial to prevent its spread, reported control measures including stringent hand hygiene practice and reinforcement of standard precautions [23,27,31]. Only two studies implemented barrier precautions by use of PPE [23,31]. All five studies that reported control measures used various types of social distancing measures including isolation, restriction of new admission and visitors, or cessation of group activities [23,26,27,31,49]. Active surveillance by symptom reporting for early detection of new cases was reported in two studies [27,49]. Although four studies reported that staff members were affected by the outbreaks, only one study implemented work exclusion for ill employees and showed the lowest attack rate among staff [23]. Four studies reported intensive cleaning and disinfection of the environment [23,26,27,31]. The implementation of more stringent procedures for cleaning and disinfection with diluted bleach was reported for outbreaks of C. difficile [27]. The study by Luque et al. [26] on viral gastroenteritis reported a relatively small number of interventions, showing a high attack rate of 48.63%. On the other hand, the study by Šubelj and Učakar [23] with the largest number of control measures among the five studies had a lower attack rate of 11.21% compared to the other outbreaks.
Influenza virus outbreaks (n = 6). Five of the six reports implemented prophylactic oseltamivir for the residents and/or HCWs [40][41][42][43][44]. Both droplet precaution and active surveillance were reported in 3 of the six studies. Five of the studies on influenza outbreaks reported a total of 172 cases among staff, but only three of the studies implemented the measure of work restriction [40,42,44]. The study by Burette et al. [41] reported the lowest number of control measures including prophylaxis and isolation and had the highest attack rate of 42.05% among the five reports on influenza A.
Tuberculosis outbreaks (n = 2). Following the detection of the index case, two reports on tuberculosis outbreaks conducted case finding among residents and staff by contact tracing [35,39]. Responding to the outbreaks, measures for the cases included isolation and transfer to a hospital in one study [39], but the other study only restricted new admissions [35]. Neither of them mentioned airborne precautions taken such as N95 respirators. Investigations in both reports found that the air exchange rates of the rooms were inadequate. The study by Lai et al. [39] corrected the failure of the environmental infection control by increasing the ventilation rates in the building. Both outbreaks involved cases among workers, but there was no description about the work status of the affected staff after the occurrence.
MDROs outbreaks (n = 3). Three outbreaks of MDROs were caused by MRSA, MRPA, or Klebsiella pneumoniae carbapenemase-producing Klebsiella pneumonia (KPC-KP) [15,22,34]. The MRSA outbreak study applied mupirocin eradication for the residents and staff [22]. All three outbreaks used transmission-based precautions and quarantine measures to prevent the spread of MDROs. In addition, they all provided re-education for the staff to improve infection control practice. Furthermore, two of the reports on MRPA and KPC-KP intensified the cleaning of the environment to interrupt the contamination of the environment. The study by Kanayama et al. [15] demonstrated that the sharing of a device such as a suction device was linked with the MRPA cases; thus, the control measures included stopping the sharing of devices.
The unexpected occurrence of the KPC-KP cases led to contact surveillance for additional exposure cases [34]. The investigation for the KPC-KP outbreak revealed poor hand hygiene compliance among staff; thus, interventions including frequent audit and feedback were implemented.
GAS outbreaks (n = 5). Three of the five GAS outbreaks provided antibiotic prophylaxis to the residents and staff [17,24,37]. All five studies conducted surveillance culture for active case finding. None of the five outbreaks reported droplet precautions, but the study by Thigpen et al. [24] mentioned an enhanced respiratory hygiene practice. Although three of the GAS outbreaks lasted for a long period due to an unsolved person-to-person transmission [17,20,37], none of the studies implemented social distancing measures. Two studies improved the availability of hand dispensers to address the suboptimal hand hygiene practice that was revealed during their observation [17,24]. None of the studies on the three outbreaks involving sick employees reported encouragement of work exclusion for ill staff [20,37,38], but some studies reported that there were voluntary sick leaves of employees before the recognition of the outbreaks.
Hepatitis virus outbreaks (n = 5). Prophylaxis of hepatitis B vaccine and immuno-globulin were implemented for hepatitis B virus outbreaks in two studies [21,28]. There is not much generally recommended NPIs for the hepatitis B and C outbreaks; thus, the studies on these outbreaks reported fewer NPIs than those on the other outbreaks. Three of the studies on the hepatitis virus outbreaks tried to find additional cases by serologic screening [21,25,30]. All the studies on the hepatitis B and C virus outbreaks employed the principle of single-use device or individual equipment to break the chain of infection [18,21,28,30]. Improvement of the care room was done in two studies that found lapses in the environment of the procedure room [28,30]. Interventions for drinking water standards and toilets were reported in the hepatitis E virus outbreak caused by contaminated water [25].
Heamophilus influenzae outbreaks (n = 2). One of the two H. influenzae outbreaks reported droplet precaution during the outbreak [32], and the other study restricted new admissions to prevent additional transmission [36].
Other outbreaks (n = 2). The study on the RSV and HMPV outbreak reported various measures including active surveillance, isolation, contact precaution, antiviral prophylaxis for residents and work restriction for ill staff to control respiratory pathogen transmission [50].
In the epidemic keratoconjunctivitis outbreak, control measures involved universal precaution with enhanced hand hygiene, isolation and restriction of visitors, and work restriction for affected workers [29].

Discussion
We updated the understanding of outbreaks in LTCFs with more recently published reports. This review also explored and summarized critical issues facilitating the spread of the outbreaks and the control measures, which have not been addressed in detail in the previous review [9]. Lessons learned from the results of this review would enable better prevention and control of outbreaks in LTCFs in the future. Implications and suggestions for achieving the best response to epidemics by LTCFs, and for future research concerning outbreaks, have been described in this review.
The most common outbreaks in LTCFs in this review were respiratory infections followed by GI infections, showing consistency with the findings of a previous study on nursing homes [51]. Interestingly, there is a difference in the outbreak reports for MDROs compared to the previous review [9]. This review identified three reports including MRSA, MRPA, and KPC-KP, suggesting the increase of multidrug-resistant organisms, given that the prior review found only two reports of MRSA [9]. As the prevalence of MDROs is increasing in LTCFs [2], they become a particular concern in these facilities. With drug resistance on the rise, MDRO related outbreaks may occur in LTCFs with growing frequency. It shows that staff and managers of LTCFs need to be aware of the significance of this trend and to prepare a plan.
Influenza viruses and GAS accounted for a large number of single pathogens. This is similar to the results of the previous review showing that the largest number of aetiologic agents affecting outbreaks was influenza viruses in LTCFs from 1966 to 2008 [9]. First, in this review, five of the six influenza virus outbreaks occurred by the influenza A virus and the other by influenza B virus. Influenza-like illness included in the studies showed a median attack rate of 24.50%, similar to that of seasonal influenza, usually 20-30% [40]. Vaccine-related issues have been raised in influenza outbreaks that have occurred in LTCFs of highly immunized residents. This finding suggests some implications to prevent influenza outbreaks in LTCFs. Regarding vaccination coverage among staff, the study by Thomas [52] found that influenza episodes were reduced if an employee was vaccinated, and the Centers for Disease Control and Prevention (CDC) recommend that all healthcare worker get vaccinated annually [53]. Therefore, influenza vaccination among healthcare personnel should be considered to mitigate the risk of influenza outbreaks in LTCFs. Additionally, because vaccination does not provide complete protection, active daily surveillance for influenza-like illness is still recommended for all persons in LTCFs during influenza season [53], which is evident by the outbreaks occurring in highly immunized LTCF population. Secondly, the GAS infection rate among older adults in LTCFs is from 3 to 8-fold higher than that of community-dwelling older adults, due to risk factors such as grouped living conditions, and underlying diseases [54,55]. Five of the studies on the GAS outbreaks in this review reported a median attack rate of 2.43% in LTCFs, which is within the range of 1-30% reported in the previous study [56]. Three of them were long-lasting outbreaks with multiple clusters for more than 6 months, which suggested that an accurate identification of how pathogens spread was a fundamental step in outbreak control.
This review also explored critical issues on practices that propagated the occurrence and spread of outbreaks. Consequently, failure to adhere to basic infection control practice, including hand hygiene, disinfection, and cleaning, was found to be a practical issue of great importance on the transmission of the outbreaks in LTCFs. Some reports even mentioned that this issue ultimately caused their outbreaks [19,21,28]. Most studies showed that this problem contributed to their outbreaks by causing cross-contamination between hands, environments, and equipment. First, the hands of HCWs may be the sources of the outbreaks. Frequent close contacts between residents and HCWs in LTCFs increase the risk of widespread outbreaks. Incorrect hand hygiene among HCWs can result in hands remaining contaminated, and this may lead to the transfer of organisms to the environment and to other residents [57]. Like previous studies that already confirmed poor compliance with hand hygiene among HCWs [4], one of the included studies reported that hand hygiene compliance was less than 30% [37]. Semmelweis demonstrated the role of hand hygiene in preventing infections transmitted by person-to-person [58]. Hand hygiene has a significant effect on reducing GI and respiratory infections [59]. The WHO recommends that hand hygiene should be performed at the following 5 key moments: before and after touching a patient, before clean/aseptic procedures, after body fluid exposure risk, and after touching a patient's surroundings [57]. Promotion of hand hygiene compliance through multimodal strategies has been proven to reduce healthcare-associated infection [60]. Multifaceted interventions such as WHO-5 strategies (including system change, training and education, monitoring and feedback, reminder and communication, and culture of safety) are generally effective in increasing and sustaining hand hygiene compliance at various settings [61][62][63][64][65]. The same evidence has been reported from studies on LTCFs, suggesting improved hand hygiene reduces the infection rate or respiratory outbreaks [66,67].
Secondly, lapses in cleaning and disinfection could make equipment and the environment become a reservoir for transferring pathogens [68]. Most of the studies regarding this issue were on outbreaks of GAS, gastroenteritis, and hepatitis B and C virus, and they found a failure to adhere to proper disinfection and cleaning principles. First of all, the outbreaks of GAS and hepatitis B and C were linked with breaches in specific procedures. GAS outbreaks were usually relevant to wound care and hepatitis B and C to point-of-care testing. With the aforementioned hand hygiene, disinfection and cleaning were basic infection control practices that are included in standard precautions. The standard precautions consist of hand hygiene, environmental cleaning, reprocessing and disinfection of care equipment, waste and linen management, the prevention of needle stick injuries, and the use of personal protective equipment (PPE), if necessary [11]. The practice of standard precautions is the imperative basic approach for IPC that was applied to all residents assuming they had the potential for pathogen transmission [11]. Standard precautions are necessary practice, especially in LTCFs where the systems for diagnostic tests are poor and active surveillance is not generally done. Tailored ongoing education with multimodal strategies for HCWs would ensure that basic infection control principles and standard precautions are integrated into daily practice such as point-ofcare testing [69]. As a result, a reduction in threats of outbreaks can be guaranteed, as well as the safety and health of persons residing or working in LTCFs. Meanwhile, the studies on the outbreaks of gastroenteritis reported that there were lapses in decontaminating environment. Environmental contamination may have a critical role in the spread of these outbreaks. Importantly, norovirus and C. difficile that are capable of surviving in the environment for long periods of time require more consideration in environmental disinfection [70,71]. For norovirus, the CDC has recommended more frequent cleaning and disinfection of rooms and high-touch surfaces with a hypo-chlorite (1000-5000 ppm) or other proper disinfectant [72]. The most effective control method for C. difficile was reported as disinfection and cleaning of rooms and high-touch surfaces with a chlorine-based solution (5000 ppm) [70,71].
This systematic review identified control measures taken during the outbreaks, especially non-pharmaceutical interventions. The results showed that the actual application of control measures may be far from what is recommended and implied that there are several challenges to overcome in outbreak management at LTCFs. First, acute care facilities like hospitals can successfully manage outbreaks through collaborative efforts with multiple experts [73]. However, most of the LTCFs in this review requested advice from public health authorities and organizations for unexpected outbreaks instead of organizing a multidisciplinary team. This may imply that LTCFs do not have sufficient capacity and expertise to individually plan, implement, and evaluate the management of outbreaks. Forming a local support network between acute hospitals and LTCFs at a regional level would be a potential way to close the gaps and to enhance outbreak control practices in LTCFs without adequate capacity [74]. Furthermore, training infection control professionals in facilities could facilitate early detection of outbreaks and timely interventions.
Secondly, this review found that many LTCF employees were affected by the outbreaks, which is consistent with the finding of the previous review [9]. However, it also revealed that work restriction for ill staff was not implemented well during the outbreaks in the LTCFs, which was not reviewed in the prior review [9]. Gastroenteritis outbreaks in this review had a higher median attack rate among staff than the other outbreaks, but only one study among them reported the application of work exclusion. Moreover, there were some reports that implied transmission attributable to sick employees. These results pointed out the role of presenteeism in LTCFs. Presenteeism among sick employees may have a role in either introducing pathogens or facilitating the transmission of outbreaks. The CDC recommends work restrictions for health care workers infected with or exposed to diseases such as diarrheal diseases, GAS, tuberculosis, and viral respiratory infection [75]. However, it may be challenging for LTCFs with fewer available resources to implement the exclusion of ill staff during outbreaks, given the fact that one study reported difficulty from understaffing as a result of work restriction [50]. The study by Widera et al. [76] suggested that daily screening of all staff members for symptoms during outbreaks on every shift may mitigate the impact of presenteeism. Considering that presenteeism is associated with various factors such as job security and lack of paid sick leave [77], further discussion is needed for plans addressing this issue in LTCFs.
Lastly, additional challenges in managing the outbreaks in LTCFs were reported. They included understaffing, insufficient supply of products such as PPE, lack of expertise, and limited application of isolation [41,46,47,50]. Many long-term care facilities have difficulty in applying isolation of infected persons due to the limited availability of isolation rooms [5] and concern about the adverse effects of isolation and additional precautions may affect the compliance with related practices [78]. For the same reason, some studies in this review used minimal types of isolation like enteric isolation. If single rooms are not available, facilities should consider applying the cohort measure or bed curtains as another method of isolation. In regard to this challenge, a study by Dumyati et al. [73] suggested a shift towards enhanced standard precautions or risk-based application of transmission-based precautions to uphold the quality of life of residents by HCWs. Future research should identify the rationale for the safety and effectiveness of this strategy or other options. Additionally, one qualitative study found that misunderstanding of the key concepts and recommendations of IPC contributed to under-utilization of transmission-based precautions [78]. Thus, emphasis should be on training and education of HCWs on transmission-based precautions.
A majority of the reviewed studies assessed infection control practice as part of the investigation to identify the problem areas of the outbreaks. Most studies attempted on-site direct observation of infection control practices and product availability. Some studies of retrospective design used survey and interview among employees. However, it is difficult to find a study that investigated the compliance rate of control measures during the outbreaks. Only a few studies described gaps in the actual application after recommendations for control measures were made. Even several studies overlooked reporting control measures, especially NPIs [3]. Although this does not necessarily mean that they did not apply measures, for the purpose of this review some studies were excluded from the analysis of control measures. The Outbreak Reports and Intervention studies of Nosocomial Infection (ORION) statement was developed and recommended, to improve the quality of outbreak reporting [79]. According to the statement, control measures should be included in the paper. Future studies should consider following the ORION statement for reporting of outbreaks [79], which would facilitate the formation of a body of evidence for outbreak management in LTCFs.
This study has several limitations. First, we only included studies written in English. It is possible that our review missed articles of interest written in other languages. Second, we conducted only a qualitative review due to the variability of the outbreak reports. Third, the quality assessment was conducted on studies of certain design including case-control and cohort study due to applicability of the quality assessment tool. Fourth, the results of this review have limited generalizability due to publication bias, given that either successfully controlled outbreaks or outbreaks with higher attack rates or fatality rates tend to be published.

Conclusions
This update for understanding outbreaks in LTCFs by reviewing recent studies indicates that staff members and residents are still at risk for contagious disease outbreaks including influenza, gastroenteritis, and GAS infection. As for the problem aspects, rather than by new or unexpected issues, violation of basic infection control practices was found to facilitate the occurrence and onward transmission of pathogens. The results of this review suggest that LTCFs need to inspect basic infection control practice and to implement them thoroughly in daily care as priorities. Efforts should be directed to promoting consistent and optimal adherence to the basic practice of infection control among HCWs at all times in LTCFs.
When an outbreak occurs, non-pharmaceutical control measures should be utilized to interrupt transmission. However, work restriction was infrequently taken compared to other measures. Given the fact that over half of the included studies reported at least one employee ill and their possible role in the spread of pathogens, it is necessary that symptomatic staff members temporarily preclude themselves from working. In addition to work restriction, though, LTCFs with poor resources have faced various challenges in outbreak management. Further discussion and studies are needed to identify the way addressing these challenges.