Baby Ryan is delivered vaginally, 11 hours after premature rupture of the membranes. Although 1 week premature, he appears healthy and is breathing without distress. His 1-minute and 5-minute Apgar scores are both 9.
Approximately 3 hours later, his respiratory rate climbs to 66 breaths/minute (bpm); however, his skin is still pink as he breathes room air. One hour later, he starts to grunt and exhibit chest retractions; his skin is mottled and his capillary refill has decreased slightly. His respiratory rate drops to 55, and he seems less active than before, with decreased muscle tone. When the nurse performs venipuncture to obtain samples for blood cultures and a complete blood count (CBC), she notes that Ryan doesn’t cry.
Based on these assessment findings, the nurse wonders if Ryan is developing an infection or a respiratory problem. On the other hand, she reasons, his signs aren’t uncommon among newborns—even some healthy ones. By the time she and the rest of the healthcare team realize Ryan has sepsis, his condition may be irreversible.
Subtle signs masking a sinister threat
Newborn sepsis is a severe bacterial infection that can cause disability and death. Despite recent technological advances in the care of sick newborns, sepsis remains a dire peril. Deaths among infants who don’t receive treatment may be as high as 50%.
If sepsis is detected early enough, antibiotics can cure it. But in many cases, initial signs are gradual and subtle, and may not raise suspicion of sepsis—or may go undetected. To identify early signs and help prevent septic shock and death, healthcare providers must gain expert assessment skills.
Categorizing newborn sepsis
Early-onset sepsis typically emerges rapidly during the first week after delivery. Late-onset sepsis arises after the first week.
Group B streptococcus (GBS) infections are the most common cause of newborn sepsis. In about 10% to 30% of pregnant women, GBS bacteria colonize the vaginal canal; the fetus may contract sepsis when these bacteria ascend into the uterus during pregnancy or when the fetus aspirates infected amniotic fluid or passes through the birth canal.
Non-GBS infections, such as Escherichia coli, Haemophilus influenzae, and other gram-negative bacteria, also can cause newborn sepsis.
Clinical signs of sepsis
Every nurse who works in a newborn nursery or couplet care program must become skilled in quickly identifying the signs of infection. Ideally, newborns should be assessed for signs of early-onset sepsis during the first 24 to 48 hours after delivery. Nurses working in neonatal intensive care units (NICUs) must stay alert for sepsis signs during the infant’s entire hospital stay, which could last up to several months.
Many newborns with early-onset sepsis look healthy at birth, but within 2 or 3 hours they deteriorate and must be placed on a ventilator, unable to oxygenate and in septic shock. About half of newborns with sepsis are symptomatic at birth.
Newborns with gradual, subtle signs initially may seem to be just “not doing well.” Or they may seem healthy at first, but then exhibit a slow progression of troubling respiratory signs, including mild tachypnea, soft grunting, mild retractions, and progressive cyanosis. However, these signs also occur in many newborn respiratory disorders, such as transient tachypnea or pneumonia.
Other subtle signs of sepsis include decreased temperature, reduced responsiveness, lethargy, poor feeding, pale or grayish skin, and poor sucking. But these findings occur in many healthy newborns as well as those with such problems as hypothermia, hypoglycemia, and respiratory distress.
Laboratory markers of sepsis
For a newborn with signs of sepsis or a high risk for it, the initial workup includes a CBC with white cell differential and platelet count, plus blood cultures. Although a blood culture is the only diagnostic indicator of sepsis, CBC results aid early assessment. Yet interpreting a newborn’s CBC can be challenging because of wide variations in normal values and variations based on gestational age, weight, and postconceptional age. Nonetheless, a low white blood cell (WBC) count (leukopenia), low platelet count (thrombocytopenia), or low neutrophil count (neutropenia) indicates bacteremia. Neutropenia and an abnormal immature-to-total (I:T) neutrophil ratio are strong laboratory markers of infection.
Many newborns with sepsis also have metabolic acidosis, as indicated by a pH below 7.35. In this condition, poor tissue oxygenation leads to anaerobic metabolism and increased lactic acid production. Signs of metabolic acidosis include pallor, poor perfusion, hypotension, tachypnea, and tachycardia.
Understanding the Newborn Scale of Sepsis
I developed the Newborn Scale of Sepsis (SOS) to help new nurses objectify clinical observational data and thus more accurately assess at-risk newborns. This tool helps bridge the gap between the new nurse’s lack of experience and the need for early diagnosis of sepsis. I believe it helps all nurses—new or experienced—detect sepsis using both clinical signs and laboratory data.
The scale incorporates five laboratory markers and eight clinical indicators of sepsis. The clinical indicators can be used alone or in combination with laboratory markers. (See Newborn Scale of Sepsis in the pdf version of this article.)
Be aware that repeat assessment is important. You can repeat the SOS hourly—or more often if the infant’s condition is deteriorating quickly, as commonly occurs in sepsis. If the infant’s condition worsens, as shown by a rising SOS score, notify the physician or neonatal nurse practitioner immediately.
Is the SOS reliable and valid?
Reliability and validity testing found that the Newborn SOS has an internal consistency of 0.65 (slightly lower than the preferred level of 0.70 for a new scale) and an inter-rater reliability of 96.3%. Although its face and content validity were acceptable, the scale was found to be nonspecific for indicating sepsis, with a cut-off point of 10. (The scale was specific for indicating sepsis when 14 was used as the cutoff point. However, I decided to use a lower cutoff point so as not to miss any sepsis cases, as it’s better to overtreat than fail to treat sepsis.)
The same testing showed that infants with respiratory distress had high SOS scores (10 or above); those with scores below 10 did not have sepsis (with a negative predictive value of 97%). Thus, the SOS is better at determining absence of newborn sepsis than presence. A score that stays below 10 indicates the newborn does not have sepsis. Although the Newborn SOS doesn’t predict sepsis, it can help nurses recognize the clinical pattern of sepsis and interpret a newborn’s CBC.
Using the Newborn SOS with Baby Ryan
Let’s take another look at Baby Ryan in light of his serial SOS scores. Three hours after delivery, when his respiratory rate increases to 66 bpm and his skin appears pink, the nurse would give him a score of 3 points on the SOS.
When he develops chest retractions with grunting (5 points), mottled skin (2 points), slightly decreased capillary refill (1 point), a respiratory rate of 55 bpm (0 points), and decreased muscle tone (3 points), his score rises to 11 points.
When the nurse draws blood for CBC and cultures and notes Ryan’s lack of response to the pain of venipuncture, she increases his score. His rising SOS score during the first 4 hours after delivery strongly suggests sepsis.
Now let’s examine Ryan’s laboratory findings: The CBC shows a WBC count of 42,000/mm3 (2 points), with 33% segmented neutrophils, 10% bands, 2% metamyelocytes, 45% lymphocytes, 5% monocytes, 3% eosinophils, and 2% basophils. Ryan’s platelet count is 180,000/mm3 (0 points)—well above the normal level of 150,000/mm3.
To determine the I:T ratio, the nurse divides the number of Ryan’s immature neutrophils (bands and metamyelocytes) by his total number of neutrophils. Absolute neutrophil count (ANC, also called total neutrophil count) is the total number of WBCs multiplied by the percentage of neutrophils. Ryan’s immature neutrophils are 12% of his total WBCs; his total neutrophils (bands, metamyelocytes, and segmented neutrophils) are 45%. Thus, Ryan’s I:T is 0.26 (12 ÷ 45), giving him an SOS score of 3. To calculate ANC, the nurse multiplies 45% (0.45) by 42,000/mm3; the result is 18,900/mm3 (0 points)—showing that Ryan has plenty of neutrophils to fight a bacterial infection.
Combined with his clinical indicators, these laboratory findings give Ryan a total SOS score of 21 (16 points for clinical indicators and 5 points for laboratory findings). Based on his high—and climbing—SOS score, the healthcare team should transfer him to the NICU and start antibiotic therapy at once.
Rising to the challenge
Newborn sepsis must be identified and treated as early as possible to avert a catastrophic outcome. Yet its clinical signs resemble those of other neonatal health problems, and many of the laboratory indicators are inadequate positive predictors. To help overcome these challenges, I recommend that nurses and other healthcare providers use the Newborn SOS.
Anderson-Berry A, Bellig L, Ohning B. Neonatal sepsis, eMedicine. August 18, 2006. www.emedicine.com/ped/topic2630.htm. Accessed February 21, 2008.
Centers for Disease Control and Prevention. Prevention of perinatal group B streptococcal disease: revised guidelines from CDC. MMWR. 2002;51(RR11):1-22. www.cdc.gov/mmwr/preview/mmwrhtml/rr5111a1.htm. Accessed February 21, 2008.
Lori Baas Rubarth is a Neonatal Nurse Practitioner and NNP Program Coordinator at Creighton University Medical Center and Alegent Health System in Omaha, Neb.