Nosocomial Infections at Three Regional Tertiary Hospitals in Trinidad and Tobago

Objective: The main objectives of this research were to conduct and provide accurate and original findings related to the epidemiological study of nosocomial infections at three regional tertiary hospitals in Trinidad and Tobago. Specifically, the researcher determined; the frequency of nosocomial infections (NI), the frequency of multiple drug resistance among bacterial organisms associated with NI, infection control measures practiced at the research hospitals and the cost of such NI in terms of morbidity and mortality. Methods: A one-year prospective cross-sectional study was carried out. The nosocomial pathogens were retrieved from the microbiology laboratory. Antimicrobial susceptibility test by the disk diffusion method were done on all bacterial isolates. Data was analysed using SPSS version 20.0. Results: This research revealed that 450 inpatients suffered nosocomial infections, with thirty (30) mortalities during the twelve (12) months that the study lasted (June 2013 to May 2014) at three regional hospitals of Trinidad and Tobago. The incidence of nosocomial infections was 5.8% and the nosocomial infection rate was 3.6 per 1000 (450/126,668). The highest rate (30.1%) was observed in the Intensive Care Unit (82/272 admissions). The most frequent type of nosocomial infection was Skin and Soft Tissue Infections 168 (37.3%). Staphylococcus sp. (22.5%), Pseudomonas aeruginosa sp. (12.7%), Acinetobacter (11.8%) and Klebsiella sp. (11.6%) were the most frequently occurring nosocomial pathogens. Conclusion: Consistency in performing good hygiene practices is vital for reducing the high nosocomial rate found at the research sites. Prediction of these infections is very important as a part of clinical surveillance programs to take preventive measures in advance. The antimicrobial susceptibility pattern rate (ASPR) showed that only 8.3 % (5/60) of the isolates were antibiotic-susceptible strains.


Introduction
Limited studies have been done on nosocomial infections in Trinidad and Tobago. Nosocomial infection is important to study so as to reduce the incidence, mortality and morbidity associated with nosocomial infections. Neonatal and elderly admitted patients are mostly at risk due to their weak immune. Orrett, 2002 indicated in his research in Trinidad and Tobago that 139 NI were identified from 629 admissions to ICU. The main NI was from respiratory tract, 41 (29.5%) followed by surgical wounds, 35 (25.2%), urinary tract, 28 (20.1%) then bloodstream, 24 (17.3%). From the 165 bacterial organisms, 80% of these organisms were gram negative bacilli, with Pseudomonas aeruginosa, 48 (36.6%), being the most common isolate followed by Klebsiella pneumoniae, 27 (20.6%) then Enterobacter sp. 22 (16.8%). The major gram positive isolates were Staphylococcus aureus, 23 (41.8%), coagulase-negative Staphylococci, 17 (30.9%) and Enterococci, 11 (20.0%) [9]. In Europe, incidences varied from 1% for all types of nosocomial infections and up to 23.6% in paediatric ICUs [16]. In the United States of America (USA), the Centre for Disease Control and Prevention (CDC) calculated approximately 1.7 million nosocomial infections from all types of microorganisms resulting in 99,000 deaths annually [17]. The American Thoracic Society/Infectious Disease Society of America (ATS/IDSA) introduced the concept of healthcare-associated pneumonia (HCAP) in 2005 and among its guidelines includes the recommended broadspectrum antibiotics therapy as the treatment of hospital-acquired pneumonia [18]. These nosocomial pathogens affect other organs and tissues including the urinary tract [3,5,10,19], bloodstream [20], brain [15,21], and digestive tract [22] and the skin and soft tissues [23].
The standard criteria that were used along with WHO's definition to confirm nosocomial cases were: the patient had no growth of organism from laboratory culture present at day one of their admission and no signs and symptoms present the same day including temperature spikes, worsening coughing or dyspnea, tachypnea, bronchial rates, breath sounds, vomiting, leukopenia <4000, low white blood cell (WBC) counts or leukocytosis, sloppiness at wound, redness, swelling, warmth, hotness around wound area, septicemia, hypotension, dysuria, urgency and tenderness. [31].
The aims of this study were to carry out a hospital microbiological survey for detecting the most prevalent pathogens that were causing nosocomial infections and to study their antimicrobial susceptibility patterns to prevent and control existing outbreaks and to insight in control measures to prevent nosocomial infection in Trinidad and Tobago. The research was a Cross Sectional Study conducted at three major hospitals in Trinidad and Tobago namely; Eric Williams Medical Science Complex, San Fernando General Hospital and Port of Spain General Hospital.

Target Population:
The study population was all patients hospitalized on the following wards: Medical, Surgical, Paediatrics, Intensive Care Unit (ICU) and Obstetrics & Gynaecology in three public hospitals in Trinidad during the period June 1, 2013 to May 31, 2014.

Inclusion and Exclusion Criteria:
All patients with features suggestive of nosocomial infections, who were willing to participate and gave their written or verbal consents on ward, were included in the study. Out-patients from accident & emergency wards and nephrology wards were excluded from this study, if suspected infections were considered community-acquired infections.

Study Design
This research was a cross-sectional study conducted at three major hospitals in Trinidad and Tobago. These medical facilities were: Eric Williams Medical Sciences Complex (EWMSC), San Fernando General Hospital (SFGH) and Port of Spain General Hospital (POSGH). Ethics approvals were granted by the University Campus Ethics Committee and the three Regional Health Authorities. Clinical symptoms and laboratory diagnosis of nosocomial infections were noted and reviewed by the attending physician. Each week, prospective cases of nosocomial infections were reviewed from Doctor's notes in patient's dockets for any clinical signs of nosocomial infections three days (72hrs) following patient admission for all types of nosocomial infections, except for nosocomial bloodstream infections (BSI) of which was observed two days (48hrs) after admission. A total of 2600 patient dockets for the duration of the research period were obtained at each of the research hospitals and reviewed for prospective cases of nosocomial infections. Codes were assigned for the nosocomial patient's names on the data collection sheets. In addition, the patient's names, their laboratory registration numbers, dates of samples collection and dates of sending samples to the laboratory by doctor or nurses were extremely useful to trace in the log book for their sample number, which helped to distinguish their respective nosocomial isolates from the others. The type of pathogen was confirmed by standard laboratory and biochemical tests. Samples were sent day one of patient's admission to confirm that the patients had no growth of organism on admission.

Data Analysis
Descriptive statistics was used to calculate the sample size, incidence and rate of nosocomial infections and the percentage of nosocomial infections by age, months, research wards and nosocomial pathogens involved. Data related to patients and nosocomial infections were entered in a Microsoft® Access 2010 data bank and statistically processed using IBM SPSS® Statistics (version 20). The statistical review of the study was done by a biomedical statistician.

Sample Size Calculation:
The sample size was calculated via a cluster sampling method by wards and patient's files [33]. The most practical sampling approach, and via the WHO and standard criteria for prospective nosocomial cases were selected from the defined population. Surveillance was conducted on a weekly basis for fifty-two (52) weeks, for prospective nosocomial cases and pathogen isolates that were obtained from the laboratory during the research period. The average confirmed number of nosocomial cases by WHO's definition was two per week. Therefore, via cluster sample size calculation equals number of wards multiplied by number of surveillance days and multiplied again by average number of patients with nosocomial infection confirmed during each week's surveillance. That is 5 wards x 52 surveillance days x 2 = 520 patients were expected to acquire nosocomial infection during the data collection period. Hence, 520 patients from clustered sample size calculation were expected to have been associated with nosocomial infections with confidence interval of 95% and relative precision of 10% within the research period.

Measurements
In calculating the incidence of nosocomial infections in Trinidad and Tobago (2013 -2014), the number of surveillance days (52) was multiplied by fifty (50) that was the average number of dockets reviewed each surveillance day from all the research wards: surgical, medical, paediatrics, ICU and Obstetrics & Gynaecology. Therefore, fifty-two (52) multiplied by 50 equals 2600 cases reviewed at each research site. If this product is multiplied by three (3) equals 7800 files reviewed during the research period. Four hundred and fifty (450) nosocomial cases were recovered during the research period. The incidence of all three (3) regional hospitals was calculated by dividing number of confirmed cases over total number of reviewed files and multiplied by one hundred: (450/7800) X 100 = 5.8% and this was the incidence for inpatients associated with nosocomial infections at all major regional hospitals of Trinidad and Tobago during the specified period. The incidence at site A was calculated by dividing the number of cases with nosocomial infection (265) divided by the total files reviewed (2600) and multiplied by 100, which equaled to 10.2%. Similarly, the incidence at site B was calculated by dividing the number of cases with nosocomial infection (63) divided by the total files reviewed (2600) and multiplied by 100, which equaled to 2.4%. The incidence at site C was calculated by dividing the number of hospital associated cases (122) divided by the total files reviewed (2600) and multiplied by 100, which equaled to 4.7%. The nosocomial rate was calculated by dividing the number of nosocomial cases recovered during the research period, divided by the total number of patients admitted on the research wards and multiplied by 100. The number of patients admitted on research wards for all three regional hospitals was 126,668 and the total number of nosocomial cases recovered for the research period was 450. Hence, nosocomial rate for research period was; 450/126,668 x 100 = 0.36%; 3.6 per 1000 patients. The nosocomial rate for research site A was 265/48057 = 0.55% (5.5 per 1000), whereas the nosocomial rate for research site B was 63/39950 = 0.16%; 1.6 per 1000 and the nosocomial rate for research site C was 122/38,661= 0.32%; 3.2 per 1000 patients. Calculation of the antimicrobial susceptibility pattern rate (ASPR) for each pathogen strain was done as follows: Number of antibiotics with susceptibility for a specific strain ASPR= ______________________________________________ x 100 Total number of antibiotic tested for that specific strain

Result
From out of 126668 admissions 450 nosocomial infections were recorded.                                Table 13 represents seven (7) distinct susceptibility patterns of Serratia marcescens, indicating that seven (7) Table 18 features two (2) susceptibility patterns of Alcaligenes sp. indicating that two strains of the same organism were present. This pathogen was sensitive to CAZ, CIP and TZP. ____________________________________         Fever or rigors and at least one positive blood culture   [79] and Burkholderia cepacia [80] have been reported. Antibiotic resistance can be acquired by horizontal transfer of a resistance gene or mutation and generally an acquired mechanism results in a predictable increase in phenotypic resistance [78]. A surveillance programme is in place at the 3 regional hospitals to monitor susceptibility patterns of these microorganisms, to delay antibiotic resistance in the future. The fact that antibiotic resistance has not been developed for these four (4) [9]. The researcher observed from the research that the incidence of nosocomial infections was 5.8% which was lesser when compared to Benin, another developing country, whose incidence was 19.1% [6,26]. There were other nations (developing countries) with higher incidences when compared to the researcher's findings [27]. The age of patients who had nosocomial infections ranged from eight (8) days old to ninety-six (96) years old. Neonates were more susceptible of acquiring nosocomial infections (108/450; 24%) and this was due to their weak immune system. It was reported in the literature that children, who developed nosocomial infections had a three-fold increase in hospital stay (27 days versus 9 days, p<0.001) compared to those who did not [14]. An increased expression of cytokine genes (IL1B and IL10) was observed in patients, who developed nosocomial infections, in addition to a pro-apoptosis pattern due to a lower expression of BCL2. CD3D, a key TCR co-factor that was found significantly downmodulated in children, which developed nosocomial infections [34]. Frequency of nosocomial cases amongst elderly inpatients were higher than other age group, and this was supported by several studies [35][36][37][38][39][40][41][42], which reported that it may be due to risk factors such as impaired immunity, chronic diseases, medications, malnutrition, and functional impairments, among others.

CAZ CIP TZP SXT TOB CN TE ____________________________________ a) S S S S R R R b) S S S R S S S ____________________________________
This study showed that surgical wards had the highest nosocomial infection rate (33.8%). But in a surgical emergency department was verified or suspected aspiration as the most important risk factor for the development of nosocomial infection [43] and we agreed with that. Another research showed that the highest surgical site infection (SSI) incidence was after dirty surgery and was 17.8% (102 out of 574) in patients living in high human development index (HDI) countries. Patients in middle-HDI countries had an incidence of 31.4% (74 out of 236), and the poorest countries had an incidence of approximately 40% (72 out of 181) [44]. The use of prophylactic antibiotics for the prevention of SSI has been cited but it may not be available for third world countries [45]. This data indicate that poverty and hospital environment are also risk factors for nosocomial infections and we stressed the need of measures to control and prevent infections from pathogens that live in the hospital microbiota. In addition, a high incidence of nosocomial infection (28.15%) was seen in intensive care units in a study carried out in Ethiopia, where the risk of nosocomial infection was found to be higher in patients with chest tube, on mechanical ventilation or underlying diseases [2,46].
Skin and soft tissue infection was the most frequent type of nosocomial infection shown in this study, accounting for 37.3% (168 out of 450). It was reported that the most frequent clinical syndromes in 174 infection episodes in patients ≥65 years old were osteoarticular (40%) and skin and soft-tissue infections (30%) in a research aiming to study the frequency distribution of syndromes in the elderly, where bacteraemia was found in 46% of the skin and soft tissue nosocomial infections [47]. Despite the considerable effort devoted to observing each ward, it is of vital importance that key focus be placed on the surgical, medical and paediatric facilities, where the nosocomial cases were highest. Another aspect of great interest is to put special attention in sterilizing surgical tools, frequent hand washing and changing of gloves as often as possible. Reichmann DE & Greenberg JA, 2009 reported that patient skin preparation in the operating room, usage of chlorhexidine-based preparations, pre-operative hand/forearm antisepsis, hair removal, etc. should be considered for safe procedures, when carried out on the surgical wards to reduce the nosocomial rate [48].
In this investigation Staphylococcus sp. (22.4%; 193/859) and Pseudomonas aeruginosa (12.6%; 109/859) organisms were one of the principal gram-positive and gram-negative nosocomial pathogens respectively connected with hospital acquired infections during the research period. The cumulative occurrence of multi-drug-resistant Staphylococcus sp. and Pseudomonas aeruginosa strains were alarming as effective antibiotic choices were severely limited. Other frequently associated multidrug resistant nosocomial organisms included Acinetobacter sp. (11.8%; 101/859) and Klebsiella sp. (11.6%; 100/859). In addition, Klebsiella and Staphylococcus sp. were the most frequently observed causative organisms in nosocomial bloodstream infections in our investigation. Several scientists studied nosocomial bloodstream infections caused by Pseudomonas sp. in new-borns and concluded that they were very frequent in neonates and a cause of bacteraemia and mortality, which may be due to low birth weight, underlying disorders and invasive procedures as predisposing factors [49][50]. Other authors published the most commonly isolated microorganisms in new-born ICU in another research. It was Klebsiella sp. (39.6%) followed by Pseudomonas aeruginosa (11.3%) and coagulase-negative staphylococci (9.4%) [7][8]51]. Our results along with what has been previously reported in the literature suggest that gram-negative bacteria (especially P. aeruginosa and Klebsiella sp.) play an important role in the pathogenesis of nosocomial infections in neonates admitted in ICU. The neonate's immune system is not mature enough to fight off these infections caused by these organisms and thus linked to their virulence factors create a perfect scenario for bacteraemia, septic shock and other complications [4,15].
Nosocomial skin and soft tissue infections (SSTI) and bloodstream infections (BSI) were the most common types of nosocomial infections with occurrences of 37.3% and 28.4% respectively, but their incidence can be dropped by good hygiene practices, and this has been observed by other authors [72][73][74][75]. These infections were commonly transmitted via direct contact. Respiratory tract infections (RTI) and bloodstream infections (BSI) were observed to be most frequently associated on the medical wards. Also, BSI was mostly observed on the paediatric medical wards, whereas skin and soft tissue infection (SSTI) was most frequently observed on the surgical wards. Nosocomial central nervous system infections and nosocomial urinary tract infections were more commonly observed on the medical facilities followed by surgical wards. Nosocomial urinary tract infection was least observed on the gynaecology wards. Candida albicans, Enterobacter and Escherichia coli isolates were most commonly observed in urinary tract infections. Of the susceptibility test results, Amoxicillin (AML), Amoxicillin/Clavulanic acid (AMC), Cefaclor (CEC), Cefuroxime (CXM), Ampicillin (AMP) and Trimethoprine/Sulphamethazole (SXT) were observed of being the most frequently resistant antibiotics in this study. Resistance of these antibiotics were commonly observed for the following nosocomial pathogens: Acinetobacter, Klebsiella, Staphylococcus, Enterobacter, Escherichia coli, Serratia sp. and Citrobacter koserii organisms. In the literature has been published that the highest resistance rate was against Ciprofloxacin and Imipenem in Iran [52]. A total of eight multi-resistant strains of gram-negative bacteria with ESBL-production were detected in five E. coli and three K. pneumoniae strains at the Albert Schweitzer hospital in Gabon. However, four were resistant to the whole spectrum of antibiotics available [27]. Van der Zee A et al (2014) reported that the resistance to Carbapenem antibiotics is emerging worldwide among Enterobacteriaceae and they developed a PCR technique for identification of carbapenemase genes: blaOXA-48, blaVIM, blaIMP, blaNDM and blaKPC in cultures of broth rectal swabs [53]. Lin et al (2014) reported that Tigecycline non-susceptible K. pneumoniae bacteraemia may suggest a critical problem that caused high mortality of patients at a medical centre in Taiwan over a 3-year period. It was divulged that resistance to commonly used antibiotics was observed in up to 80% of the isolates in Malawi [54]. However, Pseudomonas aeruginosa, Acinetobacter and Klebsiella were observed to be most frequently susceptible to Imipenem, Amikacin, Ciprofloxacin, Gentamicin and Cefepime. Quinupristin (QD) was observed to be most commonly susceptible for methicillinresistant coagulase negative Staphylococcus (MRCNS) organisms and Clindamycin, Rifamficin, Tigercycline, Vancomycin and Linezolid were also commonly observed to be effective for methicillin-resistant coagulase negative Staphylococcus (MRCNS) isolates. Sulphamethazole was frequently observed to be susceptible for Citrobacter koserii, Stenotrophomonas maltophilia and Alcaligenes sp. In contrast Sulphamethazole was least susceptible for Burkholderia (P) cepacia isolates. Piroth L et al, 2014 reported that E. coli were susceptible to third-generation Cephalosporins and amoxicillin-clavulanate in 89.5% and 62.5% of cases, respectively. No single antibiotic allowed antimicrobial coverage of more than 60% in 190 cirrhotic patients in France [55]. Natoli S et al, 2009 reported that bacteraemia caused by Coagulase negative Staphylococci remained susceptible to Linezolid, Daptomycin and Tigecycline in Italy [56]. In this study inpatients were more likely of being associated with nosocomial infections due to inconsistency in sanitizing work areas, lack of proper ventilation on some wards, irregularity in changing long term use of invasive devices such as urinary and central venous catheters, long term stays in hospitals and previous hospitalization exposure. Patients acquired nosocomial infections either endogenously or from external environment (exogenously). Endogenous nosocomial infections were because of opportunistic pathogens residing in or on external surfaces of patients and brought on by conditions present at or as a direct outcome of events on the wards. In contrast, exogenous hospital acquired infections were the result of pathogens being transmitted by patients as they are shed from numerous thresholds of exit while the patients were hospitalized. Zilberberg MD et al, 2014 reported several risk factors that predispose to recurrent Clostridium difficile infection including use of certain antimicrobial such as Fluoroquinolones and IV Vancomycin after completion of C. difficile treatment, community-onset healthcare associated infection, ≥2 hospitalizations within prior 60 days, age and gastric acid suppression [11,57].
Among the reported risk factors for colonization by extensively drugresistant P. aeruginosa in immunocompromised patients, the unnecessary use of antibiotics, particularly ciprofloxacin was reported; and when using medical devices, it was suggested that a high standard of infection control measures must be achieved [28,58,76]. Yamakawa K et al, 2011 reported that Healthcare-associated methicillin resistant Staphylococcus aureus infection in ICU patients causes high mortality, which is associated to risk factors such as intubation, treatment with antibiotics, open wound, and steroid administration, all occurring within 24 hours of ICU admission [12][13]59]. Vasudevan A et al, 2014 agreed with previous authors that the global increase of resistant gram-negative bacilli infections in ICU are due to various risk factors and the most important one is the widespread use of empiric broad spectrum antibiotics [1,60].
In this study the infection control measures practiced at the health care facilities as recorded from patient's dockets during weekly surveillance included: isolation of MRSA patients. These individuals were separated from others to avoid non-affected inpatients becoming contaminated. It was observed wearing gloves and face masks by the barrier nurses when attending to MRSA patients. Other infection control measures that were observed at the health care facilities included strict observance of contact precautions, ventilatory support, and use of antibiotics, analgesia, fluid resuscitation and frequent cleaning of infected wounds with Lysol. All were part of the management of the sick documented by health care professionals. In addition, it was advised that indwelling catheters be changed frequently to prevent persistence and reoccurrence of infections. Nosocomial Infections pose serious health problems or challenges to patient well-being. Therefore, CDC provides world-wide guidance in close watch, outbreak surveys, laboratory research and prevention of nosocomial infections. CDC uses awareness acquired through these campaigns to spot infections and develop and implement new plans to prevent and reduce nosocomial infections. Public health action by CDC and other healthcare partners demonstrated great improvements in clinical practice, medical methods and the continuing growth of infection control guidance and prevention accomplishments. Hospitals should comply with all sanitization protocols comprising uniforms, fumigating equipments, washing and other pre-emptive procedures. Proper hand washing along with usage of alcohol rubs by all health care staff prior to and after each patient contact is one of the most effective ways of fighting hospital-acquired infections. These measures have been observed by other researchers [62][63][64][65][66][67][68]. Furthermore, some health personnel have challenged the idea that the stethoscope, may essentially be a path for transmission of nosocomial infections. In a research of a hundred and fifty (150) health care staffs, fifty (50) paramedics, fifty (50) nurses, and fifty (50) doctors, Staphylococcus sp., mostly coagulase negative were cultured from 89% of the participant's stethoscopes with the average amount of colony forming units increased, when the stethoscopes were not sanitized [5,69]. In general, 48% of health care workers cleaned their stethoscopes each day or week, 37% each month, 7% annually and 7% had never cleaned them. Cleaning the stethoscope caused an immediate decline in the bacterial count by 94% with alcohol swabs, 90% with a non-ionic detergent, and 75% with antibacterial soap [69]. With regards to ventilators, which may be an aid to prevent the risk of hospital-acquired pneumonia infections by controlling hospital internal air quality must be a fulfilled requirement at the most suitable areas. To an extent nosocomial infection can be lessened to strengthen this point as reported by [70]. Cautious use of antibiotics is crucial. Despite sanitation guidelines, patients can become susceptible of acquiring hospital-acquired infections and they are often given alternative antibiotics in controlling infections that may amplify the range for the appearance of resistant strains. In addition, sterilization is further than sanitizing; it destroys all pathogens on medical devices and surfaces via contact with chemicals, ionizing radiation, dry heat or steam under pressure. Safety measures must be implemented to avoid spread of pathogens by regular paths in health care facilities. Practicing of hand washing is the solitary way of minimizing the hazards of spreading skin pathogens between patients or from one spot to another on the same patient. Frequent hand washing as often between contacts with patients and after contact with blood, body fluids, secretions, excretions, and equipment or items infected by these pathogens is a vital constituent of infection management and isolation preventative measures. The transmission of hospital-acquired infections, among immunocompromised patients is linked with health care staff's hand infectivity in nearly 40% of cases and it is a difficult predicament in the modern health care facilities. The most appropriate application for staffs to conquer this challenge is performing proper hand-hygiene actions. Therefore, the World Health Organization (WHO) initiated in 2005, the Global Patient Safety Challenge. The objective of hand sanitation is to eradicate the transitory flora with proper act of hand washing, with various types of soap; customary and antibacterial and alcohol-based gels. The major challenges observed in the performance of hand sanitation are associated with the inadequacy of sinks, lengthy time and act of hand washing procedures. A simple way in resolving this issue can be using alcohol-based hand rubs, since, it is a quicker process in contrast to accurate hand washing [71].

Conclusion
Consistency in performing good hygiene practices is vital for reducing the high nosocomial rate found at the research sites. Prediction of these infections is very important as a part of clinical surveillance programs to take preventive measures in advance. ASPR showed that only 8.3 % (5/60) of the isolates were antibiotic-susceptible strains.