Increasing Incidence of Life-threatening Pertussis: A Retrospective Cohort Study in New Zealand.

BACKGROUND: Pertussis immunization programs aim to prevent severe infant disease. We investigated temporal trends in infant pertussis deaths and pediatric intensive care unit (PICU) admissions and associations of changes in disease detection and vaccines used with death and PICU admission rates. METHODS: Using national data from New Zealand (NZ), we described infant pertussis deaths and PICU admissions from 1991 to 2013, over which time national immunization coverage at 2 years of age increased from <80% to 92%. In NZ, pertussis became a notifiable disease with polymerase chain reaction (PCR) diagnosis available in 1997 and acellular replaced whole-cell vaccine in 2000. We used Poisson regression to model temporal trends and compared rates in time intervals using rate ratios (RRs) with 95% confidence intervals (CIs). RESULTS: There were 10 pertussis deaths and 159 infant PICU admissions with pertussis from 1991 to 2013. The annual number of infant pertussis PICU admissions increased from 1991 to 2013 (P = 0.02) but the number of pertussis deaths did not (P = 0.09). The risk of PICU admission during infancy with pertussis was increased in the notification/PCR versus the non-notification/PCR era (RR: 1.12; 95% CI: 1.02-1.19) and when acellular replaced whole-cell vaccine (RR: 1.19; 95% CI: 1.06-1.31). Median Pediatric Index of Mortality scores during 2001-2013 were lower than during 1991-1999 (P < 0.001). CONCLUSIONS: Infant PICU pertussis admission rates have increased in NZ despite improvements in immunization coverage. Higher rates have occurred since pertussis notification/PCR became available and since acellular replaced whole-cell vaccine. The severity of disease in infants admitted to PICU with pertussis has decreased in recent years.

Pertussis requiring intensive care.

OBJECTIVES: To describe children with pertussis who require intensive care. DESIGN, SETTING AND PATIENTS: An audit in Auckland, New Zealand, of pertussis admissions to the national paediatric intensive care unit (PICU) from 1991 to 2003. RESULTS: 72 children, 97% of whom were <12 months old. The annual number of cases increased with time (p = 0.04). Forty patients (56%) were coughing for less than 8 days before admission. Apnoea or paroxysmal cough was present in 33 (83%) of these children. Thirty five (49%) received assisted ventilation. Four died. 19% were readmitted to PICU. Those readmitted presented with more atypical disease and had a shorter first admission but longer total PICU admission (9 vs 5 days, p = 0.009). Of the 58 children from Auckland, nine either died (three) or had subsequent respiratory or neurodevelopmental problems (six). There was an increased risk (relative risk, 95% CI) of death or disability associated with having a co-morbidity (RR = 5.56, 1.50 to 8.15), an elevated lymphocyte count (RR = 5.75, 1.54 to 13.65), presenting with seizures/encephalopathy (4.87, 1.18 to 8.34) or shock (6.50, 1.89 to 8.94), having a PIM score of 1% or more (RR = 6.20, 1.22 to 21.72), any abnormal neurological signs (RR = 9.65, 3.32 to 15.23) or being readmitted to PICU (RR = 4.63, 1.44 to 8.82). CONCLUSIONS: Apnoea and paroxysmal cough are key symptoms of pertussis in those with shorter cough duration. Death or disability are frequent. Clinical factors define children at increased risk of these poor outcomes. Early discharge from PICU is associated with an increased risk of readmission and poor outcome.