The authors have declared that no competing interests exist.
Infections are common complications in critically ill patients with associated significant morbidity and mortality.
This study determined the prevalence, risk factors, clinical outcome and microbiological profile of hospital-acquired infections in the intensive care unit of a Nigerian tertiary hospital.
This was a prospective cohort study, patients were recruited and followed up between September 2011 and July 2012 until they were either discharged from the ICU or died. Antimicrobial susceptibility testing of isolates was done using CLSI guidelines.
Seventy-one patients were recruited with a 45% healthcare associated infection rate representing an incidence rate of 79/1000 patient-days in the intensive care unit. Bloodstream infections (BSI) 49.0% (22/71) and urinary tract infections (UTI) 35.6% (16/71) were the most common infections with incidence rates of 162.9/1000 patient-days and 161.6/1000 patient-days respectively.
Our findings demonstrate that healthcare associated infections is a significant risk factor for ICU-mortality and morbidity even after adjusting for APACHE II score.
The prevalence of ICU-acquired infections is significantly higher in developing countries than in industrialised countries, varying between 4.4% and 88.9% [
A recent European multicentre study posted that the proportion of infected patients in intensive care units can be as high as 51%; most of these are health care associated [
Critically ill patients with severe sepsis in intensive care units (ICUs) require lengthy and expensive management, with an associated high mortality, with rates ranging from 30% to 50% [
To initiate necessary policies that are critical to effective treatment of ICU-acquired infections and prevent antibiotic resistance development, there should be surveillance of bacterial aetiologies and infection patterns. This study determined the prevalence, risk factors, clinical outcome and microbiological profile of hospital-acquired infections in the intensive care unit of a Nigerian tertiary hospital.
This was a prospective and observational study. It was conducted in the ICU of the Lagos University Teaching Hospital, Nigeria (LUTH), from September, 2011 to July, 2012. LUTH is a 761 bed tertiary hospital with a six-bed ICU that admits critically ill patients from all specialties. The ICU admits approximately 220 patients annually. This study was approved by the Health Research and Ethics Committee of the Lagos University Teaching Hospital, Lagos (
All patients that were fifteen years of age and above whose surrogates gave informed written consent (
Health-care associated infections (HAIs) are defined as an infection developing >48hours after hospital admission or within 30 days after discharge from a hospital [
Patient-days is the total number of days that patients were in the ICU during the period of study.
Multi-drug resistance (MDR) was defined only for gram-negative bacteria, as resistance to three or more groups of antibiotics [
Urine, blood, endotracheal aspirate, cerebrospinal fluid, wound aspirate and stool samples were aseptically collected from 71 recruited patients on the first day of admission into the ICU (no bronchoalveolar lavage specimen was collected). Repeat samples were taken after 48 hours of admission. Further samples were collected whenever there was clinical suspicion of infection; otherwise, they were collected weekly. All samples were transported to the clinical microbiology laboratory for immediate processing. Blood was cultured in the BACTEC culture system 9050 (Becton Dickinson, New Jersey, US). The other samples were processed according to established standardized protocol. Anaerobic cultures were not done. Isolates were identified to the species level using MicroBact® (Oxoid, UK). Quality control was done using
Antimicrobial susceptibility testing was determined by Kirby-Bauer disc diffusion method [
The presence of ESBLs was suspected if an isolate of K. pneumoniae or E. coli demonstrated resistance to one or more of the indicator beta-lactam antibiotics–ceftriaxone, cefotaxime or cefepime [
All Staphylococcus spp. isolated were subjected to testing which was performed according to the CLSI guidelines [
All Enterococcus spp. isolated were tested for vancomycin resistance using E-test (AB, Biodisk, Solna, Sweden), manufacturer’s instructions were strictly adhered to. The plate was incubated at 37°C for 24 hours. Minimum Inhibitory Concentration (MIC) result was interpreted according to CLSI guideline [
Data was entered into Epi Info software version 3.4(CDC, Atlanta GA, USA). The analysis was done with Statistical Package for Social Sciences (SPSS) software version 19.0(SPSS Inc., Chicago, IL., USA). Categorical variables were compared using Pearson’s Chi-square test or Fisher’s exact test. P-values < 0.05 were considered significant for all tests. Mutivariate logistic regression analysis was employed to determine the independent contribution of clinical variables to the prediction of acquisition of ICU infections in the hospital as dependent variables. The same statistical test was used to determine independent predictors of ICU mortality except that APACHE II-adjusted model was used. Variables that had a value of P≤ 0.2 on univariate analysis were entered into a forward stepwise logistic regression model. Goodness-of-fit was evaluated by Hosmer-Lemeshow test. Two tailed p values were reported. The association between the independent determinants of ICU-acquired infections and hospital mortality were estimated using odds ratios and 95% confidence intervals. For outcome analysis, patients were distributed into two subgroups according to survival status (died or discharged).
During the study period, 139 patients were admitted into the ICU, but only 71 patients were eligible for the study. Among the 68 patients that were excluded, 39 were admitted for less than 48 hours in ICU, 23 were children under the age of 15 years and 6 declined. The male female ratio was 1:0.8 with an age range of 15 years to 69 years, with a mean of 38.7 (±14.9). Length of stay (LOS) in the ICU ranged from 2 to 38 days with a median of 5 and Interquartile range (IQR) of 5–12 days. All patients had urinary catheters inserted for urine output monitoring and 35 (49.3%) had Central venous catheter (CVC) in-situ. Thirty-seven (52.1%) of the patients were either transferred from the ward or another hospital while 34 (47.9%) came via the Accident and Emergency unit (
Variables | n | frequency | Mean (SD) | Median (IQR) |
---|---|---|---|---|
Gender | ||||
Female | 32 | 45.1 | ||
Male | 39 | 54.9 | ||
Age | 71 | 38.7(±14.9) | ||
15–40 | 43 | 60.6 | ||
41–69 | 28 | 39.4 | ||
Apache II Score | 71 | 21.2 (± 8.07) | ||
<20 | 35 | 49.3 | ||
≥20 | 36 | 50.7 | ||
Length of ICU stay | 71 | 9.1 (± 7.3) | 7 (5–12) | |
2–7 | 41 | 57.7 | ||
8–38 | 30 | 42.3 | ||
Duration of antibiotic administration in ICU (days) | 69 | 8.0 (±6.4) | ||
Locations before ICU admissions | ||||
Hospital | 37 | 52.1 | ||
Home | 34 | 47.9 | ||
Wards admitted in hospital before transfer to ICU | ||||
Outside hospital | 20 | 28.2 | ||
Accident and Emergency | 34 | 47.9 | ||
Surgical ward | 8 | 11.3 | ||
Medical ward | 3 | 4.2 | ||
Obstetrics and Gynaecology | 3 | 4.2 | ||
Neuro-ward | 3 | 4.2 |
Thirty-five (49.3%) of the patients had an APACHE II score of less than 20 while 36 (50.7%) had a score of 20 and above, the mean APACHE II score was 21.2 (±8.07). Admission diagnoses were respiratory failure 19 (26.8%); severe sepsis 11(15.5%); eclampsia 5(7.0%) and others 21 (29.6%) and these included post-surgical patients, patients with chronic renal failure and malignancies) (
“Others” as a group of admitting diagnosis included: post-surgical patients, patients with chronic renal failure and malignancies.
A total of 45 laboratory-confirmed infections were identified in 32 patients, representing a prevalence rate of 45.1%, and an incidence rate of 79/1000 patient-days (Incidence rate was derived by dividing the number of new nosocomial infections acquired in a period by Total number of patient-days for the same period x 1000). The total patient-days in the study was 405. The most common infection were bloodstream infections accounting for 49.0% (22/45) of all infections. Two patients had 3 episodes and 3patients had 2 episodes of BSI which were catheter associated. Eleven patients had a total of 16 (35.6%) urinary tract infections (UTI); 1 patient had 3 episodes and 3 patients had 2 episodes each of UTI. A total of 4 (8.9%) skin-soft tissue infections (SSTI) were reported amongst 3 patients, one had 2 episodes of SSTI. Three (6.7%) RTIs were observed in three patients (
Bloodstream infection (48.9%); Urinary tract infection (35.6%); Skin-soft tissue infection (8.9%); Respiratory tract infection (6.7%).
Twenty different species of pathogenic microorganisms were identified in the 45 infections recorded.
KEY: RTI = Respiratory tract infections; SSTs = Skin-soft tissue infections; UTI = Urinary tract infections; BSI = Blood stream infections; CoNS = Coagulase negative staphylococcus.
Eleven gram-positive bacteria were isolated,
Resistance to antibiotics (R/S for single isolates; %R for n ≥ 2 isolates) | |||||||||
---|---|---|---|---|---|---|---|---|---|
Isolates | n | SAM | FEP | CRO | CIP | GEN | LVX | MEM | TZP |
5 | 60 | 0 | 60 | 100 | 60 | 20 | 0 | 60 | |
3 | 33.3 | 66.7 | 66.7 | 66.7 | 66.7 | 33.3 | 0 | 100 | |
3 | 33.3 | 0 | 100 | 66.7 | 66.7 | 33.3 | 0 | 0 | |
3 | NA | 0 | 100 | 100 | 100 | 100 | 0 | 100 | |
2 | 100 | 0 | 100 | 50 | 50 | 0 | 50 | 50 | |
2 | 100 | 0 | 50 | 0 | 100 | 0 | 0 | 100 | |
2 | 50 | 50 | 100 | 100 | 100 | 100 | 0 | 50 | |
1 | R | S | S | S | R | S | S | R | |
1 | R | R | R | R | R | R | S | R | |
1 | R | R | R | R | R | R | S | R | |
1 | S | S | S | S | S | S | S | S | |
1 | R | S | R | R | R | S | S | S | |
1 | R | S | R | R | R | R | S | S | |
1 | R | R | R | R | R | S | R | R | |
1 | R | R | R | R | R | S | R | S |
KEY: SAM = Ampicillin-sulbactam; FEM = Cefepime; CRO = Ceftriaxone; CIP = Ciprofloxacin; GEN = Gentamicin; LVX = Levofloxacin; MEM = Meropenem; TZP = Piperacillin-tazobactam; R = Resistance; S = Sensitive; NA = Not applicable
Five factors were identified as statistically significant regarding HAIs in the ICU using univariate analysis: Use of antibiotics one month before ICU admission (OR = 0.334; p = 0.03),; surgery one month before admission (OR = 0.181, p< 0.001); urethral catheterization (OR = 5.38; p<0.05), endotracheal intubation (OR = 5.78; p< 0.05), and patients’ location before ICU admission (OR = 0.11; p< 0.05) (
Factors | Infected n(%) | Not Infected n(%) | Odds ratio | p-value | |
---|---|---|---|---|---|
Use of antibiotic one month before hospital admission | 0.334 | 0.03 | |||
• Yes | 20(62.5) | 14(35.9) | |||
• No | 12(37.5) | 25(64.1) | |||
Surgery one month before admission | 0.181 | 0.001 | |||
• Yes | 21 (65.6) | (25.6) | |||
• No | 11 (34.4) | 29 (74.4) | |||
Urethral catheterization | 5.38 | 0.03 | |||
• Yes | 32(100.0) | 33(84.6) | |||
• No | 0(0.0) | 6(15.4) | |||
Endotracheal intubation | 5.78 | 0.02 | |||
• Yes | 29(90.6) | 13(33.3) | |||
• No | 3(9.4) | 26(66.7) | |||
Location before admission | 0.11 | 0.001 | |||
• Hospital wards | 13 (59.1) | 4 (13.8) | |||
• Accident/Emergency unit | 9 (40.9) | 25 (86.2) | |||
Age(years) | 3.28 | 0.17 | |||
• ≥60 | 2 (6.25) | (17.9) | |||
• <59 | 30 (93.75) | 32 (82.1) | |||
Gender | 0.44 | 0.086 | |||
• Female | 18 (56.25) | (35.9) | |||
• Male | 14 (43.75) | 25 (64.1) | |||
Malignancy | 0.46 | 0.191 | |||
• Yes | 9 (28.13) | (15.38) | |||
• No | 23 (71.87) | 33 (84.62) | |||
Length of ICU stay | 0.44 | 0.093 | |||
• ≥ 7 | 17 (53.1) | (33.3) | |||
• < 7 | 15 (46.9) | 26 (66.7) | |||
APACHE II score | 0.52 | 0.186 | |||
• ≥ 20 | 19 (59.4) | (43.6) | |||
• < 20 | 13 (40.6) | 22 (56.4) |
Septiceamic patients had higher mortality rates than non-septiceamic patients (75.0% vs 25.0%). Five (5) factors were significantly associated with this severe outcome. These factors were ICU-acquired infection (OR = 8.2; p = 0.04); endotracheal intubation (OR = 5.7; p = 0.04); urethral catheterization (OR = 7.5; p = 0.04); acquisition of infection within seven days of admission (OR = 4.9; p = 0.05) and an APACHE 11 score value greater or equal to 20 (OR = 9.04; P <0000) (
Univariate analysis | Multivariate analysis | ||||
---|---|---|---|---|---|
Risk factors | OR | p-value | OR | CI | p-value |
ICU-aquired infection | 8.2 | 0.004 | 3.83 | 1.082–13.522 | 0.04 |
LOS < 7days before infection | 4.9 | 0.05 | 0.65 | 0.190–2.208 | 0.49 |
Endotracheal intubation | 5.7 | 0.004 | 2.7 | 0.576–13.165 | 0.21 |
Urethral catheterization | 7.5 | 0.004 | 1.79 | 0.107–29.857 | 0.69 |
APACHE II score | 9.04 | 0.000 | - | - | - |
Location before admission | 1.9 | 0.17 | 2.2 | 0.708–6.501 | 0.18 |
Age | 1.7 | 0.28 | |||
Gender | 0.63 | 0.32 | |||
Use of antibiotic one month before admission | 0.77 | 0.58 | |||
Surgery one month before admission | 0.50 | 0.24 | |||
Malignancy | 1.5 | 1.00 | |||
Nasogastric intubation | 0.97 | 0.97 | |||
LOS > 7 days | 0.81 | 0.66 |
KEY: LOS = Length of ICU stay; OR = Odds ratio; CI = Confidence interval
[-2loglikelihood 77.170;
The 45% prevalence rate of HAIs reported is very high, it demonstrates the inadequacy of the infection control processes in place in our ICU. This rate is significantly higher than the pooled prevalence of 35.2% from ICUs in LMICs [
BSIs (49.0%) were the most documented infections in our ICU and this is contrary to previous findings [
Intensive care unit (ICU)-acquired infections are a challenging health problem worldwide, especially when caused by multidrug-resistant (MDR) pathogens. Twenty different species of microorganisms were involved in 45 episodes of ICU infections in this study. While Staphylococcus aureus, coagulase-negative Staphylococcus (CoNS), and Proteus mirabilis were the most frequently isolated organisms causing ICU-acquired infections. MDR gram negative bacteria were responsible for about two-thirds of the total infections recorded. This trend contrasts with findings from studies done in developed countries where the prevalent cause of health care-associated infections is switching over to gram positive organisms [
The most worrisome finding in this study was that almost all the gram positive bacterial agents of BSIs were MDR, including the only Enterococcus spp. Previous studies have reported steady rise of resistant pathogens from patients in ICU [
The highest rates of antibiotic resistance displayed by the bacterial strains were to amoxicillin-clavulanate and ceftriaxone. During the time of the study, these two drugs were the commonest drugs given empirically to ICU patients. This practice might have built up selective pressure that had led to the evolution of resistant strains which clonally expanded over time. This is indeed a notable finding, which addresses one of the key objectives of this study that is: determining the microbiological profile of the index ICU to guide in developing an antibiotic policy for it. A recent ‘surviving sepsis’ publication by Ramsamy et al (2016) noted that ‘unnecessary administration of antimicrobial therapy not only impacts on the individual patient but also on those patients in the same ICU environment’ and that ‘.knowledge of inherent flora and their antimicrobial susceptibility patterns are crucial’ [
Our findings revealed a high rate of multi-drug resistant (MDR) gram-negative bacilli (57.1%). A similar study demonstrated comparable high rates of MDR (51%) in gram-negative bacilli among patients in Afghanistan [
Otherstudies have independently associated acquisition of MDR-ICU-acquired infections with risk factors such as use of antibiotics one month prior to ICU admission, surgery one month before admission, urethral catheterization and endotracheal intubation [
Limitations of this study includes relatively small size of study population which is a reflection of the limited ICU beds available and the fact that healthcare generally is out of pocket in the study environment. We speculate that this could have contributed to the study’s lack of power to detect some significant relationships from our data. Secondly, the convenience sampling technique we used helped us to access our study participants easily, however, it could have introduced sampling bias, distorting good representation of the entire population.
In conclusion, ICU-acquired infections remained a significant risk factor for ICU- mortality even after adjusting for APACHE II score. There is dire need to develop and entrench an antibiotic stewardship policy in our ICU and set up a national surveillance program to monitor infections in our Nigeria. A robust infection control program is also a matter of urgency in our setting.
(DOCX)
(JPG)
(DOCX)
(DOCX)
(DOCX)
We gratefully acknowledge all the healthcare workers at the Intensive care unit of the Lagos University Teaching Hospital, Lagos for their support during this study.