Serial serum C-Reactive Protein in diagnosis and estimate duration of antibiotic therapy among neonates with bacterial septicemia at Mulago Hospital
Abstract
Description
A thesis submitted in partial fulfillment of the requirements for the award of the degree of Master of Medicine in Paediatrics and Child Health of Makerere University.
Bacterial septicemia is one of the most common diagnostic challenges in newborn medicine. A definitive diagnosis is usually reached only after a delay of a day or two, yet rapid progression of untreated infection may greatly increase morbidity or mortality. So empiric therapy based on signs and symptoms of septicemia is recommended which may result in treatment of as many as 30 uninfected infants for every 1 who is ultimately determined to have been infected. Recent reports indicate that serial CRP levels during this interval may be useful for early identification of neonates with septicemia and also for early identification of neonates for whom antibiotic therapy can be safely discontinued. Objective: To determine whether serial CRP can be used as a parameter to identify neonatal septicemia and the time point when antibiotic treatment can be safely discontinued in neonates treated for suspected bacterial infection in Mulago Hospital. Study design and settings: A cross sectional study with a cohort follow up which was carried out in ACU and general Pediatric wards in Mulago Hospital. Participants: Infants 1-28days old with suspected neonatal septicemia. Methods: Prospectively 221 neonates with suspected bacterial septicemia were enrolled consecutively over a period of 4 months and 203 neonates were followed until clinical improvement or for 10 days. All enrolled children had baseline blood cultures, CBC, CSF and CRP levels performed on admission. Thereafter, serial CRP levels for each study participant were performed once daily till two consecutive CRP values were normal and neonates were also clinically monitored daily. Neonates with positive blood culture received antibiotics for 10 days. While for culture negative neonates, antibiotics were stopped once two consecutive CRP values were normal and neonate were clinically stable. Clinical improvement was evaluated for all study participants at the time of CRP reversion to normal values. The primary outcome variable of the study was the proportion of neonates with culture proven bacterial septicemia and proportion of children with positive CRP. Proportions of neonates with clinical improvement or failure of clinical improvement at time of CRP reversion were computed among culture positive and culture negative study arms. Relative risk of failure to have clinical improvement at the time of CRP reversion were estimated with 95% CI. Results: Of the 221 enrolled, 213 were analyzed of which 7 died and 3 ran-away. CRP was positive in 87(40.8%) and negative in 126(59.2%) neonates. Blood culture proven septicemia was found in 30(14.1%) neonates of which 20 were CRP positive and 10 were CRP negative. Fever, refusal to breast feed and tachypnoea were the most common clinical presentation. The most common organism isolated was Staph.aureus(50%). Sensitivity, Specificity, Positive Predictive value and Negative Predictive value of serial CRP were 66.7%, 63.4%, 22.9%, and 92.7% respectively. The sensitivity, specificity and predictive value of hematological indices were not statistically significant when compared to blood culture. Factors associated with elevated CRP were blood culture proven septicemia, presence of jaundice and leucocytosis. The mean duration for CRP to revert to normal in the culture positive arm was 4.5days and for culture negative arm was 3.5days. The percentage of neonates who improved at CRP reversion was 80% and 63% in culture positive and negative arm respectively. While percentage of neonates who failed to improve were 20% and 36% in culture positive and negative arm. The relative risk for failure to improve of 0.55(0.216-0.139) was insignificant with p=0.276. Conclusions: A very high negative predictive value of serial CRP makes it a valuable parameter to exclude infections but a low sensitivity, specificity and positive predictive value makes it a poor diagnostic marker for infection. The mean time for CRP normalization was 4days in culture negative group and 5days for culture positive group which would imply the mean duration for treatment in respective groups. The percentage of neonates with clinical improvement at CRP reversion to normal was significantly high in both culture positive and culture negative groups. Recommendation: In neonates with negative CRP on day 1 and day 2, antibiotics can be safely discontinued if 48-72hour culture results are negative and have clinically improved. In neonates with a positive CRP on day1 or day 2 with negative blood culture, antibiotic treatment can be given for 5days. In neonates with a positive blood culture, antibiotics treatment can be given for 7days. It is important to measure the CRP at discharge on all neonates to ensure a truly negative neonate at discharge.
Bacterial septicemia is one of the most common diagnostic challenges in newborn medicine. A definitive diagnosis is usually reached only after a delay of a day or two, yet rapid progression of untreated infection may greatly increase morbidity or mortality. So empiric therapy based on signs and symptoms of septicemia is recommended which may result in treatment of as many as 30 uninfected infants for every 1 who is ultimately determined to have been infected. Recent reports indicate that serial CRP levels during this interval may be useful for early identification of neonates with septicemia and also for early identification of neonates for whom antibiotic therapy can be safely discontinued. Objective: To determine whether serial CRP can be used as a parameter to identify neonatal septicemia and the time point when antibiotic treatment can be safely discontinued in neonates treated for suspected bacterial infection in Mulago Hospital. Study design and settings: A cross sectional study with a cohort follow up which was carried out in ACU and general Pediatric wards in Mulago Hospital. Participants: Infants 1-28days old with suspected neonatal septicemia. Methods: Prospectively 221 neonates with suspected bacterial septicemia were enrolled consecutively over a period of 4 months and 203 neonates were followed until clinical improvement or for 10 days. All enrolled children had baseline blood cultures, CBC, CSF and CRP levels performed on admission. Thereafter, serial CRP levels for each study participant were performed once daily till two consecutive CRP values were normal and neonates were also clinically monitored daily. Neonates with positive blood culture received antibiotics for 10 days. While for culture negative neonates, antibiotics were stopped once two consecutive CRP values were normal and neonate were clinically stable. Clinical improvement was evaluated for all study participants at the time of CRP reversion to normal values. The primary outcome variable of the study was the proportion of neonates with culture proven bacterial septicemia and proportion of children with positive CRP. Proportions of neonates with clinical improvement or failure of clinical improvement at time of CRP reversion were computed among culture positive and culture negative study arms. Relative risk of failure to have clinical improvement at the time of CRP reversion were estimated with 95% CI. Results: Of the 221 enrolled, 213 were analyzed of which 7 died and 3 ran-away. CRP was positive in 87(40.8%) and negative in 126(59.2%) neonates. Blood culture proven septicemia was found in 30(14.1%) neonates of which 20 were CRP positive and 10 were CRP negative. Fever, refusal to breast feed and tachypnoea were the most common clinical presentation. The most common organism isolated was Staph.aureus(50%). Sensitivity, Specificity, Positive Predictive value and Negative Predictive value of serial CRP were 66.7%, 63.4%, 22.9%, and 92.7% respectively. The sensitivity, specificity and predictive value of hematological indices were not statistically significant when compared to blood culture. Factors associated with elevated CRP were blood culture proven septicemia, presence of jaundice and leucocytosis. The mean duration for CRP to revert to normal in the culture positive arm was 4.5days and for culture negative arm was 3.5days. The percentage of neonates who improved at CRP reversion was 80% and 63% in culture positive and negative arm respectively. While percentage of neonates who failed to improve were 20% and 36% in culture positive and negative arm. The relative risk for failure to improve of 0.55(0.216-0.139) was insignificant with p=0.276. Conclusions: A very high negative predictive value of serial CRP makes it a valuable parameter to exclude infections but a low sensitivity, specificity and positive predictive value makes it a poor diagnostic marker for infection. The mean time for CRP normalization was 4days in culture negative group and 5days for culture positive group which would imply the mean duration for treatment in respective groups. The percentage of neonates with clinical improvement at CRP reversion to normal was significantly high in both culture positive and culture negative groups. Recommendation: In neonates with negative CRP on day 1 and day 2, antibiotics can be safely discontinued if 48-72hour culture results are negative and have clinically improved. In neonates with a positive CRP on day1 or day 2 with negative blood culture, antibiotic treatment can be given for 5days. In neonates with a positive blood culture, antibiotics treatment can be given for 7days. It is important to measure the CRP at discharge on all neonates to ensure a truly negative neonate at discharge.
Keywords
Infant mortality, Neonatal septicemia, Antibiotic therapy