Pediatricians from Baylor College of Medicine reviewed the medical records of all neonates ≤ 30 days of age hospitalized at Texas Children’s Hospital from July 1, 2000 through December 31, 2007 who had a pertussis polymerase chain reaction (PCR) test performed. Infants with a positive PCR were compared with infants who had a negative PCR but had a confirmed alternative viral microbiologic diagnosis. Severity of illness was determined by assessing outcomes such as length of hospitalization, supplemental oxygen requirement, need for mechanical ventilation, and death.
During the study interval, of 216 neonates tested by PCR for pertussis, 33 (15.2%) were positive and 35 (16.2%) had a negative pertussis PCR and an alternative microbiologic diagnosis. A total of 33 medical records from patients in each group were available for review. In the pertussis PCR-negative group infections were caused by respiratory syncytial virus (22), rhinovirus (6), parainfluenza virus (3), and influenza A virus (2). Children in the pertussis PCR-positive group had a significantly greater mean number of days of illness prior to presentation to the hospital (6.8 vs 4 days; P<.001), longer median length of hospital stay (10 vs 4 days; P<.001), longer median number of days on oxygen (7 vs 4 days; P=.018), and were more likely to have paroxysmal cough, apnea, and a cyanotic spell than those in the PCR-negative group. Tachypnea was significantly more likely (P=.02) in the non-pertussis comparative group. Five infants with pertussis compared to none of the infants in the non-pertussis group were discharged home with a respiratory support device (non-significant, P=.053). No deaths occurred in either group.
Approximately 77% of infants with a recorded exposure history had a sick household contact. Compared to PCR-negative infants, pertussis PCR-positive patients were more likely to have a sick household member and the household member was more likely to be a parent than a sibling.
The authors conclude that neonates who are PCR-positive for pertussis are likely to have a more severe, prolonged course of disease than neonates with a similar presentation who had an established alternative microbiologic diagnosis.
PCR, now in widespread use, is a rapid and highly sensitive diagnostic test for the detection of pertussis (see AAP Grand Rounds, December 2009;22:621). However, there is no standardization of methods, nor is the test currently licensed by the FDA. Comparing neonates with pertussis based on a positive PCR with other neonates with respiratory symptoms who had a negative pertussis PCR and an alternative microbiologic (viral) diagnosis, the authors of the current study have shown that the pertussis PCR-positive neonates have a more serious disease. A major limitation of this study is that the practitioners knew the diagnosis, pertussis or rhinovirus for instance, and this knowledge may have strongly influenced the variables assessed such as length...