325
Journal of Neonatal-Perinatal Medicine 6 (2013) 325–331 DOI 10.3233/NPM-1372113 IOS Press
Original Research
Antimicrobial prescription and usage in the neonatal intensive care units of a Caribbean country: A prospective observational study S. Hariharan∗ , D. Chen, C. Harry, R. Ragobar, R. Boodoosingh, C. Gangoo, A. Hector, J. Reid and S. Raju Anesthesia and Intensive Care Unit, Department of Clinical Surgical Sciences, Faculty of Medical Sciences, The University of the West Indies, St. Augustine, Trinidad & Tobago
Received 11 April 2013 Revised 19 July 2013 Accepted 25 July 2013
Abstract. OBJECTIVE: To study the prescribing patterns and usage of antimicrobials in the Neonatal Intensive Care Units (NICU) of Trinidad & Tobago. METHODS: A 3-month prospective observational study was conducted at the three NICUs at the major public hospitals. Data included antimicrobials prescribed, route of administration, culture and sensitivity reports, leukocyte count, length of stay and outcome of patients. RESULTS: 353 patients were studied, 57.5% of the patients were males. Mean birth weight was 2.96 ± 0.94 (Standard Deviation) kg. Admission diagnoses included meconium stained liquor, preterm, respiratory distress, sepsis, etc. Length of stay ranged between 1 to 76 days, (median 4, Interquartile Range 1–8). The mean leukocyte count was 15.7 ± 8.5 × 103 per L. Overall, 645 culture specimens were sent; umbilical swab (27.6%), throat swab (27.0%) and blood (16.4%) being the most common specimens. 310 (48.1%) showed no bacterial growth. Overall, 16 different antimicrobials were used. First line antibiotic of choice was a combination of ampicillin and gentamicin (85.8 %). Second line antibiotic of choice was cefotaxime. The overall mortality was 7.6%. CONCLUSION: The choice of antimicrobials in the NICUs of major public hospitals is mostly empirical and not primarily dictated by the culture and sensitivity reports, emphasizing the need for antibiotic stewardship programme in Trinidad & Tobago. Keywords: Antimicrobials, prescription patterns, usage, neonatal intensive care
1. Introduction Antibiotics are one of the most frequently prescribed drugs worldwide. There has been a wider concern ∗ Corresponding author: Dr. Seetharaman Hariharan, Anesthesia & Critical Care Medicine, Faculty of Medical Sciences, The University of the West Indies, Eric Williams Medical Sciences Complex, Mount Hope, Trinidad (West Indies). Tel./Fax: +1 868 6624030; E-mail: [email protected].
voiced regarding the inappropriate use of antibiotics with consequent wastage of resources [1]. Failure to adhere to hospital guidelines and inappropriate use of antibiotics have accounted for the emergence of resistant organisms. Drug use evaluations are essential in every unit of the hospitals to complement the efforts taken to conserve the antibiotics. The assessment of utilization of antimicrobials is particularly important in developing countries such as Trinidad and Tobago,
1934-5798/13/$27.50 © 2013 – IOS Press and the authors. All rights reserved
326
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean
where they account for a vast majority of the healthcare budget. Antimicrobials are one of the critical drugs in saving the lives of the most fragile subpopulation of the humans – the newborn and are widely used in Neonatal Intensive Care Units (NICU). Neonates admitted to NICUs present with a wide range of symptoms, many of which are typical of bacterial infections. Symptoms may include fever, jaundice, persistent crying, irritability, sleeping more than usual, lethargy, breathing problems, rashes, hypoglycemia or hyperglycemia, abnormally fast or slow heart beat, bruising or bleeding and seizures [2, 3]. Because the mortality rate of untreated sepsis can be quite high, and sepsis may manifest with non-specific clinical signs with serious adverse effects, it is recommended that rapid empiric therapy is initiated until a more specific drug may be used according to culture test results, sensitivities and site of infection [4]. Patients in NICUs pose unique characteristics such as low birth weight, under developed humoral and cellular immune defense mechanisms, instrumentation with indwelling devices such as endotracheal tubes for mechanical ventilation and intravascular catheters, which increase their susceptibility to acquire bacterial infections [5]. NICU patients sometimes require a longer hospital stay and are therefore at a higher risk of acquiring nosocomial infections due to poor infection control practices. Neonatal sepsis has been defined as a syndrome within the first forty days of life manifested by symptoms and signs of bacterial infection accompanied by bacteremia. A previously conducted study at one of the NICUs in Trinidad during 1996-97 revealed that the overall incidence of neonatal sepsis was 10/1000 live-births [6]. In the same study, the incidence of sepsis due to group B Streptococcus was found to be 3/1000 live-births [6]. Also, Gram-negative organisms accounted for 63% of the positive cultures. There were also three outbreaks of nosocomial infection, two caused by Enterobacter spp. and one caused by Pseudomonas aeruginosa [6]. In another NICU during the period 1998, there was an outbreak of Pseudomonas Aeruginosa, where six infants died following sepsis with multi-resistant Pseudomonas Aeruginosa [7]. Another study from Trinidad investigating the etiology and risk factors of neonatal sepsis and mortality at an NICU showed that the major drugs used empirically in suspected cases of sepsis were ampicillin or
ceftazidime plus gentamicin [8]. The study also found that early-onset neonatal sepsis was 86.4% and may be nosocomial in origin. Antibiotic prescription practices and usage is one of the major factors which influence the spectrum of microbes prevalent in a unit as well as their resistance patterns. There has been little quantification of the antimicrobial usage in NICUs in the Caribbean. With this background, we conducted an observational study of the pattern of antimicrobial use in all the three NICUs at the public hospitals of Trinidad.
2. Methods This is a prospective observational study that evaluated the prescribing patterns of parenteral and oral antimicrobials in the Neonatal Intensive Care units at the three major public hospitals in Trinidad & Tobago. Approval was obtained from the Ethics Committee of the University of the West Indies and the Medical Chiefs of Staff of the hospitals for collection of data. Waiver for informed consent was requested since all relevant information was collected only from the charts and the study was purely observational. All patients receiving antimicrobials in the NICUs at the three hospitals during a time period of three months in the year 2008 were included. The study was conducted simultaneously in all the three hospitals during September through November months. Patients not receiving any antimicrobials were excluded. 2.1. Data collection The required data were obtained from the patient records and was transcribed with a coded number to maintain confidentiality. Demographic data collected included age, gender, height and weight. Clinical data included diagnoses on admission, antimicrobials used, dosage and duration, culture and sensitivity reports, leukocyte counts, co-morbid illnesses, diagnosis on discharge/death and patient outcome. Descriptive analyses of the data were done. Central tendencies were expressed as mean and standard deviation (SD) or median and interquartile ranges (IQR). Comparison of variables between survivors and non-survivors was done using the MannWhitney-U test. Statistical significance was fixed at p < 0.05. Statistical analyses were done using SPSS version-12 (Chicago IL, USA) software.
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean
327
3. Results Overall, 353 patients were studied during the three months; 99, 190 and 64 patients from the hospital I, II and III respectively. Of these 203 (57.5%) were male patients. All patients were similar in age (less than forty days) and birth weight ranged between 0.59–5.2 kg, mean 2.96 ± 0.94 (SD). The average birth weights in all the three units were similar. Length of stay for all three hospitals ranged between 1 to 76 days. The overall median length of stay was 4 days. There was a mild variation in the median length of stay between the three units. The overall mean leukocyte count was 15.7 ± 8.5 (SD) × 103 per L and this was similar in all the three units. Table 1 shows the distribution of the different variables in the three hospitals. Patients were admitted with a wide range of diagnoses which included meconium stained liquor, preterm, respiratory distress, sepsis, maternal pyrexia, neonatal seizures, neonatal jaundice, infant of diabetic mother, low APGAR score and other miscellaneous. The overall distribution of the patients according to the diagnostic categories is shown in Fig. 1. Overall, 645 culture specimens were sent for microbial analyses, of which 310 (48.1%) showed no bacterial growth. Percentage of cultures which showed no bacterial growths were 34.9%, 66.7% and 77.8% for hospital I, II and III respectively. The nature and number of culture specimens in all the three hospitals are shown in Table 2. Umbilical swab (27.6%), throat swab (27.0%) and blood (16.4%) were the most common specimens sent for culture and sensitivity tests. The most common organism grown in the umbilical swab, throat swab and blood was Coagulase Negative Staphylococcus aureus. The most common organism grown Table 1 Comparison of variables in the three hospitals Variable
Hospital I Hospital II Hospital III (n = 99) (n = 190) (n = 64)
Male (%) 62.6 55.8 Birth weight (kg) 2.3 (0.9) 3.1 (0.8) Mean (SD) Length of stay (d) 7 (4, 15) 2 (1, 6) Median (IQR) Leukocyte count 15.4 (8.3) 16.0 (8.9) (×103 per L) Mean (SD) Ampicillin + Gentamicin 78.7 85.3 as the first line antibiotic (%) Mortality (%) 11.1 7.4
53.1 2.5 (0.8) 5 (3, 9) 14.8 (6.5)
95.3 3.1
Fig. 1. Diagnostic categories. Table 2 Nature of specimens in all the three hospitals Specimen Umbilical swab Throat swab Blood Urine Cerebrospinal fluid Ear swab Gastric contents Umbilical venous catheter tip Umbilical arterial catheter tip Eye swab Central venous catheter tip Endotracheal tube tip Suction catheter tip Skin swab Abdominal drain Chest drain tip Fistula contents Nasopharyngeal swab Oropharyngeal swab Wound swab Total
Number of specimens Percentage 178 174 106 44 27 26 22 17 14 10 8 7 4 2 1 1 1 1 1 1 645
27.6 27.0 16.4 6.8 4.2 4.0 3.4 2.6 2.2 1.6 1.2 1.1 0.6 0.4 0.2 0.2 0.2 0.2 0.2 0.2 100.0
in the urine culture was Candida albicans (4.5%). The positive cultures showing the various organisms are shown in Table 3. Irrespective of the cultures, antimicrobials were used empirically on admission in the majority of the neonates admitted to the NICUs. Overall, 16 different antimicrobials were used. The duration of use of the first antibiotic on average was 4.0 ± 2.4 (SD) days. The frequencies for the routes of administration were intravenous (71.1%), intramuscular (26.1%), oral (1.6%) and topical administration (1.2%). First line antibiotic of choice was ampicillin + gentamicin in all the three units with a frequency of use at 85.8%. The frequency of use of other antibi-
328
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean Table 3 Positive cultures from specimens Organisms
Positive cultures from specimens (%) Umbilical Swab
Throat Swab
Blood cultures
Urine
47.2 12.4 7.3 6.7 3.9 1.1 NIL NIL
15.5 9.2 6.9 3.4 3.4 NIL NIL NIL
17.9 1.9 NIL NIL NIL NIL 5.7 NIL
NIL NIL NIL 2.3 4.5 NIL NIL 2.3
Coagulase negative Staphylococcus aureus Escherichia coli Group B Streptococcus Enterococcus spp. Klebsiella spp. Pseudomonas aeruginosa Staphylococcus epidermidis Enterobacter spp.
Table 4 Antimicrobial regimens in the three hospitals Hospital
First-line
Second-line
Third-line
Hospital I
Ampicillin (25–50 mg/kg q6h) + Gentamicin (2.5 mg / kg q12h) Ampicillin (25–50 mg/kg q6h) + Gentamicin (2.5 mg / kg q12h) Ampicillin (25–50 mg/kg q6h) + Gentamicin (2.5 mg / kg q12h)
Piperacillin/Tazobactam (100 mg/kg/day)* + Cefatoxime (25–50 mg/kg q8h) Amoxicillin/clavulanate (22.5–45 mg/kg q12h)¶ + Ceftazidime (25–50 mg/kg q8h) Meropenem (20 mg/kg q12h)
Meropenem (20 mg/kg q12h)
Hospital II
Hospital III
∗ Piperacillin
Meropenem (20 mg/kg q12h) + Vancomycin (10 mg/kg q6h) Cotrimoxazole (2–4 mg /kg q12h)§
component. ¶ Amoxicillin component. § Trimethoprim component (PO).
Table 5 Antimicrobials usage as Prescribed Daily Dosage (PDD) Antimicrobial
Hospital I Hospital II Hospital III Total
Ampicillin 138.1 Gentamicin 2.8 Amoxicillin-Clavulanate NU Cefotaxime 42.5 Ceftriaxone NU Ceftazidime NU Cefuroxime NU Piperacillin-Tazobactam 33.5 Cotrimoxazole 0 Meropenem 0.9 Metronidazole 0.5 Vancomycin NU Benzyl Penicillin 0.3
80.1 4.2 14.1 1.7 10.5 2.6 1.3 NU 0.07 2.2 0.9 0.4 NU
49.8 2.0 1.8 NU NU 0.3 0.5 4.0 0.6 4.7 0.7 NU 0.03
268.0 9.0 15.9 1.7 10.5 2.9 1.8 37.5 0.67 7.8 2.1 0.4 0.33
otics was amoxiclav 4%, gentamicin 3.4%, cefotaxime 1.2%, penicillin 1.2%, piperacillin-tazobactam 0.9%, cotrimoxazole 0.9% and others 2.8%. Second line antibiotic of choice was ampicillin and cefotaxime (4.6%). Table 4 shows the antimicrobial regimens in the three hospitals. Table 5 shows the overall usage of antimicrobials in the three NICUs as Prescribed Daily Dosage (PDD). Of the 353 patients studied, 326 patients survived and 27 patients died. The overall mortality was 7.6%.
Table 6 shows a comparison of survivors versus nonsurvivors. There was no evidence of any difference in the number of cultures performed and the antimicrobial usage between survivors and non-survivors. Of the non-survivors, prematurity was the most common diagnosis recorded (66.6%) and sepsis was recorded as the cause of mortality in only 18.5% of the neonates.
4. Discussion This study aimed to assess the pattern of antimicrobial use in three NICUs in Trinidad to be compared with institutional and international protocols. There are significant regional differences in incidence of prevailing microbes, thus region-specific selection of antimicrobials is warranted. Group B streptococci (GBS) is of particular importance in neonatal sepsis in some African countries (but absent in others), Middleeastern countries, and the Caribbean Islands, but is less important in South Asia [9]. Selection of empiric antibiotic therapy is dependent upon target organisms and their antibiotic susceptibility, spectrum of antibiotic activity, association with
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean
329
Table 6 Comparison of survivors and non-survivors Variable (Mean ± SD)
Overall (n = 353)
Survivors (n = 326)
Non-survivors (n = 27)
Significance*
Birth weight Gender (n) Male Female Length of stay (days) Leucocyte count (×103 per L) Duration of the first antibiotic (days)
2.96 ± 0.9
2.9 ± 0.9
1.7 ± 0.9
p
Journal of Neonatal-Perinatal Medicine 6 (2013) 325–331 DOI 10.3233/NPM-1372113 IOS Press
Original Research
Antimicrobial prescription and usage in the neonatal intensive care units of a Caribbean country: A prospective observational study S. Hariharan∗ , D. Chen, C. Harry, R. Ragobar, R. Boodoosingh, C. Gangoo, A. Hector, J. Reid and S. Raju Anesthesia and Intensive Care Unit, Department of Clinical Surgical Sciences, Faculty of Medical Sciences, The University of the West Indies, St. Augustine, Trinidad & Tobago
Received 11 April 2013 Revised 19 July 2013 Accepted 25 July 2013
Abstract. OBJECTIVE: To study the prescribing patterns and usage of antimicrobials in the Neonatal Intensive Care Units (NICU) of Trinidad & Tobago. METHODS: A 3-month prospective observational study was conducted at the three NICUs at the major public hospitals. Data included antimicrobials prescribed, route of administration, culture and sensitivity reports, leukocyte count, length of stay and outcome of patients. RESULTS: 353 patients were studied, 57.5% of the patients were males. Mean birth weight was 2.96 ± 0.94 (Standard Deviation) kg. Admission diagnoses included meconium stained liquor, preterm, respiratory distress, sepsis, etc. Length of stay ranged between 1 to 76 days, (median 4, Interquartile Range 1–8). The mean leukocyte count was 15.7 ± 8.5 × 103 per L. Overall, 645 culture specimens were sent; umbilical swab (27.6%), throat swab (27.0%) and blood (16.4%) being the most common specimens. 310 (48.1%) showed no bacterial growth. Overall, 16 different antimicrobials were used. First line antibiotic of choice was a combination of ampicillin and gentamicin (85.8 %). Second line antibiotic of choice was cefotaxime. The overall mortality was 7.6%. CONCLUSION: The choice of antimicrobials in the NICUs of major public hospitals is mostly empirical and not primarily dictated by the culture and sensitivity reports, emphasizing the need for antibiotic stewardship programme in Trinidad & Tobago. Keywords: Antimicrobials, prescription patterns, usage, neonatal intensive care
1. Introduction Antibiotics are one of the most frequently prescribed drugs worldwide. There has been a wider concern ∗ Corresponding author: Dr. Seetharaman Hariharan, Anesthesia & Critical Care Medicine, Faculty of Medical Sciences, The University of the West Indies, Eric Williams Medical Sciences Complex, Mount Hope, Trinidad (West Indies). Tel./Fax: +1 868 6624030; E-mail: [email protected].
voiced regarding the inappropriate use of antibiotics with consequent wastage of resources [1]. Failure to adhere to hospital guidelines and inappropriate use of antibiotics have accounted for the emergence of resistant organisms. Drug use evaluations are essential in every unit of the hospitals to complement the efforts taken to conserve the antibiotics. The assessment of utilization of antimicrobials is particularly important in developing countries such as Trinidad and Tobago,
1934-5798/13/$27.50 © 2013 – IOS Press and the authors. All rights reserved
326
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean
where they account for a vast majority of the healthcare budget. Antimicrobials are one of the critical drugs in saving the lives of the most fragile subpopulation of the humans – the newborn and are widely used in Neonatal Intensive Care Units (NICU). Neonates admitted to NICUs present with a wide range of symptoms, many of which are typical of bacterial infections. Symptoms may include fever, jaundice, persistent crying, irritability, sleeping more than usual, lethargy, breathing problems, rashes, hypoglycemia or hyperglycemia, abnormally fast or slow heart beat, bruising or bleeding and seizures [2, 3]. Because the mortality rate of untreated sepsis can be quite high, and sepsis may manifest with non-specific clinical signs with serious adverse effects, it is recommended that rapid empiric therapy is initiated until a more specific drug may be used according to culture test results, sensitivities and site of infection [4]. Patients in NICUs pose unique characteristics such as low birth weight, under developed humoral and cellular immune defense mechanisms, instrumentation with indwelling devices such as endotracheal tubes for mechanical ventilation and intravascular catheters, which increase their susceptibility to acquire bacterial infections [5]. NICU patients sometimes require a longer hospital stay and are therefore at a higher risk of acquiring nosocomial infections due to poor infection control practices. Neonatal sepsis has been defined as a syndrome within the first forty days of life manifested by symptoms and signs of bacterial infection accompanied by bacteremia. A previously conducted study at one of the NICUs in Trinidad during 1996-97 revealed that the overall incidence of neonatal sepsis was 10/1000 live-births [6]. In the same study, the incidence of sepsis due to group B Streptococcus was found to be 3/1000 live-births [6]. Also, Gram-negative organisms accounted for 63% of the positive cultures. There were also three outbreaks of nosocomial infection, two caused by Enterobacter spp. and one caused by Pseudomonas aeruginosa [6]. In another NICU during the period 1998, there was an outbreak of Pseudomonas Aeruginosa, where six infants died following sepsis with multi-resistant Pseudomonas Aeruginosa [7]. Another study from Trinidad investigating the etiology and risk factors of neonatal sepsis and mortality at an NICU showed that the major drugs used empirically in suspected cases of sepsis were ampicillin or
ceftazidime plus gentamicin [8]. The study also found that early-onset neonatal sepsis was 86.4% and may be nosocomial in origin. Antibiotic prescription practices and usage is one of the major factors which influence the spectrum of microbes prevalent in a unit as well as their resistance patterns. There has been little quantification of the antimicrobial usage in NICUs in the Caribbean. With this background, we conducted an observational study of the pattern of antimicrobial use in all the three NICUs at the public hospitals of Trinidad.
2. Methods This is a prospective observational study that evaluated the prescribing patterns of parenteral and oral antimicrobials in the Neonatal Intensive Care units at the three major public hospitals in Trinidad & Tobago. Approval was obtained from the Ethics Committee of the University of the West Indies and the Medical Chiefs of Staff of the hospitals for collection of data. Waiver for informed consent was requested since all relevant information was collected only from the charts and the study was purely observational. All patients receiving antimicrobials in the NICUs at the three hospitals during a time period of three months in the year 2008 were included. The study was conducted simultaneously in all the three hospitals during September through November months. Patients not receiving any antimicrobials were excluded. 2.1. Data collection The required data were obtained from the patient records and was transcribed with a coded number to maintain confidentiality. Demographic data collected included age, gender, height and weight. Clinical data included diagnoses on admission, antimicrobials used, dosage and duration, culture and sensitivity reports, leukocyte counts, co-morbid illnesses, diagnosis on discharge/death and patient outcome. Descriptive analyses of the data were done. Central tendencies were expressed as mean and standard deviation (SD) or median and interquartile ranges (IQR). Comparison of variables between survivors and non-survivors was done using the MannWhitney-U test. Statistical significance was fixed at p < 0.05. Statistical analyses were done using SPSS version-12 (Chicago IL, USA) software.
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean
327
3. Results Overall, 353 patients were studied during the three months; 99, 190 and 64 patients from the hospital I, II and III respectively. Of these 203 (57.5%) were male patients. All patients were similar in age (less than forty days) and birth weight ranged between 0.59–5.2 kg, mean 2.96 ± 0.94 (SD). The average birth weights in all the three units were similar. Length of stay for all three hospitals ranged between 1 to 76 days. The overall median length of stay was 4 days. There was a mild variation in the median length of stay between the three units. The overall mean leukocyte count was 15.7 ± 8.5 (SD) × 103 per L and this was similar in all the three units. Table 1 shows the distribution of the different variables in the three hospitals. Patients were admitted with a wide range of diagnoses which included meconium stained liquor, preterm, respiratory distress, sepsis, maternal pyrexia, neonatal seizures, neonatal jaundice, infant of diabetic mother, low APGAR score and other miscellaneous. The overall distribution of the patients according to the diagnostic categories is shown in Fig. 1. Overall, 645 culture specimens were sent for microbial analyses, of which 310 (48.1%) showed no bacterial growth. Percentage of cultures which showed no bacterial growths were 34.9%, 66.7% and 77.8% for hospital I, II and III respectively. The nature and number of culture specimens in all the three hospitals are shown in Table 2. Umbilical swab (27.6%), throat swab (27.0%) and blood (16.4%) were the most common specimens sent for culture and sensitivity tests. The most common organism grown in the umbilical swab, throat swab and blood was Coagulase Negative Staphylococcus aureus. The most common organism grown Table 1 Comparison of variables in the three hospitals Variable
Hospital I Hospital II Hospital III (n = 99) (n = 190) (n = 64)
Male (%) 62.6 55.8 Birth weight (kg) 2.3 (0.9) 3.1 (0.8) Mean (SD) Length of stay (d) 7 (4, 15) 2 (1, 6) Median (IQR) Leukocyte count 15.4 (8.3) 16.0 (8.9) (×103 per L) Mean (SD) Ampicillin + Gentamicin 78.7 85.3 as the first line antibiotic (%) Mortality (%) 11.1 7.4
53.1 2.5 (0.8) 5 (3, 9) 14.8 (6.5)
95.3 3.1
Fig. 1. Diagnostic categories. Table 2 Nature of specimens in all the three hospitals Specimen Umbilical swab Throat swab Blood Urine Cerebrospinal fluid Ear swab Gastric contents Umbilical venous catheter tip Umbilical arterial catheter tip Eye swab Central venous catheter tip Endotracheal tube tip Suction catheter tip Skin swab Abdominal drain Chest drain tip Fistula contents Nasopharyngeal swab Oropharyngeal swab Wound swab Total
Number of specimens Percentage 178 174 106 44 27 26 22 17 14 10 8 7 4 2 1 1 1 1 1 1 645
27.6 27.0 16.4 6.8 4.2 4.0 3.4 2.6 2.2 1.6 1.2 1.1 0.6 0.4 0.2 0.2 0.2 0.2 0.2 0.2 100.0
in the urine culture was Candida albicans (4.5%). The positive cultures showing the various organisms are shown in Table 3. Irrespective of the cultures, antimicrobials were used empirically on admission in the majority of the neonates admitted to the NICUs. Overall, 16 different antimicrobials were used. The duration of use of the first antibiotic on average was 4.0 ± 2.4 (SD) days. The frequencies for the routes of administration were intravenous (71.1%), intramuscular (26.1%), oral (1.6%) and topical administration (1.2%). First line antibiotic of choice was ampicillin + gentamicin in all the three units with a frequency of use at 85.8%. The frequency of use of other antibi-
328
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean Table 3 Positive cultures from specimens Organisms
Positive cultures from specimens (%) Umbilical Swab
Throat Swab
Blood cultures
Urine
47.2 12.4 7.3 6.7 3.9 1.1 NIL NIL
15.5 9.2 6.9 3.4 3.4 NIL NIL NIL
17.9 1.9 NIL NIL NIL NIL 5.7 NIL
NIL NIL NIL 2.3 4.5 NIL NIL 2.3
Coagulase negative Staphylococcus aureus Escherichia coli Group B Streptococcus Enterococcus spp. Klebsiella spp. Pseudomonas aeruginosa Staphylococcus epidermidis Enterobacter spp.
Table 4 Antimicrobial regimens in the three hospitals Hospital
First-line
Second-line
Third-line
Hospital I
Ampicillin (25–50 mg/kg q6h) + Gentamicin (2.5 mg / kg q12h) Ampicillin (25–50 mg/kg q6h) + Gentamicin (2.5 mg / kg q12h) Ampicillin (25–50 mg/kg q6h) + Gentamicin (2.5 mg / kg q12h)
Piperacillin/Tazobactam (100 mg/kg/day)* + Cefatoxime (25–50 mg/kg q8h) Amoxicillin/clavulanate (22.5–45 mg/kg q12h)¶ + Ceftazidime (25–50 mg/kg q8h) Meropenem (20 mg/kg q12h)
Meropenem (20 mg/kg q12h)
Hospital II
Hospital III
∗ Piperacillin
Meropenem (20 mg/kg q12h) + Vancomycin (10 mg/kg q6h) Cotrimoxazole (2–4 mg /kg q12h)§
component. ¶ Amoxicillin component. § Trimethoprim component (PO).
Table 5 Antimicrobials usage as Prescribed Daily Dosage (PDD) Antimicrobial
Hospital I Hospital II Hospital III Total
Ampicillin 138.1 Gentamicin 2.8 Amoxicillin-Clavulanate NU Cefotaxime 42.5 Ceftriaxone NU Ceftazidime NU Cefuroxime NU Piperacillin-Tazobactam 33.5 Cotrimoxazole 0 Meropenem 0.9 Metronidazole 0.5 Vancomycin NU Benzyl Penicillin 0.3
80.1 4.2 14.1 1.7 10.5 2.6 1.3 NU 0.07 2.2 0.9 0.4 NU
49.8 2.0 1.8 NU NU 0.3 0.5 4.0 0.6 4.7 0.7 NU 0.03
268.0 9.0 15.9 1.7 10.5 2.9 1.8 37.5 0.67 7.8 2.1 0.4 0.33
otics was amoxiclav 4%, gentamicin 3.4%, cefotaxime 1.2%, penicillin 1.2%, piperacillin-tazobactam 0.9%, cotrimoxazole 0.9% and others 2.8%. Second line antibiotic of choice was ampicillin and cefotaxime (4.6%). Table 4 shows the antimicrobial regimens in the three hospitals. Table 5 shows the overall usage of antimicrobials in the three NICUs as Prescribed Daily Dosage (PDD). Of the 353 patients studied, 326 patients survived and 27 patients died. The overall mortality was 7.6%.
Table 6 shows a comparison of survivors versus nonsurvivors. There was no evidence of any difference in the number of cultures performed and the antimicrobial usage between survivors and non-survivors. Of the non-survivors, prematurity was the most common diagnosis recorded (66.6%) and sepsis was recorded as the cause of mortality in only 18.5% of the neonates.
4. Discussion This study aimed to assess the pattern of antimicrobial use in three NICUs in Trinidad to be compared with institutional and international protocols. There are significant regional differences in incidence of prevailing microbes, thus region-specific selection of antimicrobials is warranted. Group B streptococci (GBS) is of particular importance in neonatal sepsis in some African countries (but absent in others), Middleeastern countries, and the Caribbean Islands, but is less important in South Asia [9]. Selection of empiric antibiotic therapy is dependent upon target organisms and their antibiotic susceptibility, spectrum of antibiotic activity, association with
S. Hariharan et al. / Antimicrobial usage in the neonatal intensive care units, Caribbean
329
Table 6 Comparison of survivors and non-survivors Variable (Mean ± SD)
Overall (n = 353)
Survivors (n = 326)
Non-survivors (n = 27)
Significance*
Birth weight Gender (n) Male Female Length of stay (days) Leucocyte count (×103 per L) Duration of the first antibiotic (days)
2.96 ± 0.9
2.9 ± 0.9
1.7 ± 0.9
p
