Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

In the unwell neonate is a raised heart rate a useful sign in identifying the likelihood of serious illness?

Three Part Question

In a [unwell neonate] does a [raised heart rate] [increase the likelihood of serious illness]?

Clinical Scenario

A previously well, term 20 day year old male is brought into the emergency department one evening. His mother reports that he is hot, irritable and not feeding. Examination does not find a source of infection but his observations show that he is febrile with a temperature of 38.1 degrees and his heart rate is raised at 190 bpm (normal range 120-160 bpm). You wonder if his raised heart rate increases the likelihood that this neonate has a serious illness?

Search Strategy

Using the Medline database: 1950- week 1 June 2010 via OVID.
Embase database: 1980- week 23 2010 via OVID
CinAHL: CINAHL plus: 1937- June 2010
[exp Heart Rate OR exp Tachycardia OR fast heart rate.mp OR raised heart rate.mp] AND [exp Critical Illness OR severe illness.mp. OR serious illness.mp. OR serious bacterial infection.mp. OR exp Sepsis] LIMIT to [english language and humans and "newborn infant (birth to 1 month)"]

Search Outcome

Medline: 47 papers, 1 was relevant. Embase: 29 papers, 1 additional relevant paper. CINAHL: 29 papers, no new papers.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Singh SA et al
2003
India
80 neonates (all between 3 and 28 days old) in a hospital neonatal unit with suspected sepsis, 30 of which were found to have sepsis. prospective cohort study sensitivity 30%91% of the participants were pre-term (mean gestational age 31.4 weeks +/- 3.4 weeks, range 25-38 weeks). 93% of participants had a low birth weight (mean 1378g +/- 507g). Also the study was not blinded.
specificity81%
positive predictive value39%
negative predictive value74%
Graves GR et al
1984
USA
82 neonates (all less than 72 hours old) with suspected sepsis in a large university hospital obstetric unit, 13 neonates had proven sepsis. prospective cohort studysensitivity 92%There was a small number of neonates which were actually septic. Premature and low birth weight neonates were included in the study: In the septic infants, the mean gestational age was 35.5 weeks (range 29-40 weeks) and the mean weight was 2430g (range 1300-3800g). In the non-septic infants, the mean gestational age was 34.8 weeks (range 28-41 weeks) and the mean weight was 2385g (range 970-4000g). The study was not blinded.
specificity91%

Comment(s)

There is a large gap in the literature for data regarding predictive values and likelihood ratios for tachycardia as a clinical sign in the diagnosis of sepsis in the neonate. Although laboratory tests have been suggested to aid diagnosis there has been a lack of research into the use of the clinical signs. The studies in the field of neonatal tachycardia are mostly done in neonatal intensive care units with extremely sick neonates with complex cardiac anomalies. The two above studies were done in relatively well neonates but were still not ideal for answering the proposed question. This is because, Singh et al's study involved an extremely high percentage of neonates which were preterm and/or of low birth weight which may have affected the results, and in Graves et al's study the actual number of preterm or low birth weight neonates involved was unknown but the mean and ranges were given which again showed the involvement of preterm and low birth weight participants. It is also important to note that in the Graves paper, all 13 septic neonates were diagnosed with Group B Streptococci (GBS), this may affect the implication of the results as not all neonates presenting in the ED will have GBS. The Graves study only involved neonates less than 3 days old meaning that only the relationship of early onset septicaemia and tachycardia is studied. Although both studies have flaws, which need to be taken into consideration, generally the studies were well executed, reliable and involved 80 or more neonates. Both papers came to similar conclusions and found tachycardia to be a useful clinical sign in helping to diagnose sepsis in a neonate. Graves et al found tachycardia inisolation to be helpful in differentiating between those infants who were subsequently septic from those who had less life-threatening problems. However, Singh et al found tachycardia to be one of the most common signs of neonatal sepsis and when used with other clinical signs very helpful in diagnosing sepsis but in isolation tachycardia is not solely enough to diagnose sepsis. Both studies acknowledged that heart rate is simply obtained and universally available and therefore very useful in limited resource settings.

Clinical Bottom Line

Tachycardia is a useful indicator in neonatal sepsis. Tachycardia alone is not specific for sepsis, however this finding in an unwell neonate presenting to the ED warrants further investigation. A raised heart rate can be used, along with other clinical signs, to help to differentiate between the seriously ill septic neonate and one with a less serious condition.

References

  1. Singh SA, Dutta S and Narang A Predictive clinical scores for diagnosis of late onset neonatal septicemia Journal of Tropical Pediatrics 2003;235-239
  2. Graves GR and Rhodes PG Tachycardia as a sign of early onset neonatal sepsis Pediatric Infectious Disease 1984; 404-406