14 Pertussis (whooping cough)

14.3 Epidemiology

The epidemiology of B. pertussis infection and pertussis disease differ. Infection occurs across the age spectrum and repeated infection without disease is common.12 The endemic circulation of B. pertussis in older children and adults provides a reservoir for spread of the infection and the development of severe disease in incompletely vaccinated infants. Young infants have always been particularly vulnerable to pertussis disease. For example, in the US during the 1940s pertussis resulted in more infant deaths than measles, diphtheria, poliomyelitis and scarlet fever combined.13

14.3.1 Global burden of disease

Pertussis mortality

Pertussis mortality rates have always been highest in the first year of life.13, 14 Beyond age 3 years mortality rates have always been relatively low. In immunised populations virtually all deaths occur in the first two months of life, and deaths in toddlers and preschool-aged children have largely disappeared. Among infants, younger age, lack of immunisation, low socioeconomic status, premature gestation, low birthweight and female gender are associated with an increased risk of fatal pertussis.14

Pertussis deaths are under-reported. It is estimated that in the developed world three times more deaths are due to pertussis than are reported.15–18 Infants continue to die from pertussis despite state-of-the-art intensive care.11, 19–22

Pertussis morbidity

The majority of national epidemiological data on pertussis is collected via passive notification systems. The proportion of pertussis cases that are notified is estimated to vary between 6 and 25 percent. As well as underestimating disease incidence, passive notification systems are biased: a larger proportion of more clinically severe cases are notified and the proportion of cases that are notified may decrease with increasing age.17

Since the introduction of mass immunisation, countries with consistently low pertussis incidence rates have had consistently high immunisation coverage rates.23, 24 Higher pertussis incidence rates are due primarily to lower immunisation coverage, but also in some instances to lower vaccine efficacy or less-than-optimal immunisation schedules.25–29

The decrease in incidence following the introduction of mass immunisation has been most pronounced in those aged under 10 years. Despite this, the reported pertussis disease rates have remained highest in infants and young children.30–32 Infants aged under 3 months have the highest rate of notification and hospitalisation.33, 34

Pertussis is an epidemic disease with two- to five-yearly epidemic cycles. Epidemics are frequently sustained over 18 months or more, during which there are dramatic increases in hospital admission rates. Pertussis does not show the seasonal variability that is typical of most respiratory infections.

The epidemic periodicity of pertussis has not lengthened with the introduction of mass immunisation. This contrasts with the increase in time between epidemics that has occurred with other epidemic diseases for which mass immunisation is used, such as measles. This lack of lengthening of the pertussis epidemic cycle implies minimal impact of mass immunisation on the circulation of B. pertussis in the human population.12, 35, 36

14.3.2 New Zealand epidemiologyTop

Pertussis mortality in New Zealand

On average, there are zero to one deaths from pertussis each year in New Zealand. During the current pertussis epidemic (see below), there have been three deaths in children: two in infants aged under 6 weeks and one in an unimmunised preschooler.

Pertussis morbidity in New Zealand

Pertussis morbidity in New Zealand has usually been described using hospital discharge data. National passive surveillance data has been available since 1996, when pertussis became a notifiable disease.

Pertussis morbidity in New Zealand as described by notification data

Three epidemics have occurred since pertussis became a notifiable disease, with an epidemic peak annual number of notified cases of 4140 in 2000, 3485 in 2004, and 5902 in 2012 (see Figure 14.1).37 Although pertussis notifications fell in 2013, they still remained well above those seen in 2010 and 2011.38

Figure 14.1: Pertussis notifications and hospitalisations, 1998–2013

Figure 14.1: Pertussis notifications and hospitalisations, 1998–2013

Note: Includes confirmed, probable and suspect cases, and notifications still under investigation.

Source: Institute of Environmental Science and Research

Since pertussis became notifiable, the annual proportion of notified cases aged 30 years or older has increased from 23 percent (in 1997) to 48 percent in 2013. However, the highest proportion of hospitalised cases continues to be in infants aged under one year. Of the 898 notified cases in infants from 2010 to 2013, 673 (75 percent) were hospitalised (Figure 14.2).

Figure 14.2: Age distribution of notified and hospitalised pertussis cases, 2010–2013 cumulative data

Figure 14.2: Age distribution of notified and hospitalised pertussis cases, 2010–2013 cumulative data

Source: Institute of Environmental Science and Research (notifications) and the Ministry of Health (hospitalisations)

Pertussis morbidity in New Zealand, as described by hospital discharge data

Hospitalisation rates for pertussis, as measured by ICD discharge diagnostic codes, provide a measure of severe pertussis disease. The discharge rate in the 2000s was lower than in the 1990s (2000s versus 1990s, relative risk 0.79 [95% CI: 0.74–0.84]). Despite this decrease, the infant hospitalisation rate for pertussis in New Zealand in the 2000s (196 per 100,000) remained three times higher than contemporary rates in Australia (2001 infant rate: 56 per 100,000) and the US (2003 infant rate: 65 per 100,000).39–41

Pertussis hospital admission rates vary with ethnicity and household deprivation. From 2006 to 2010 the infant (aged under 12 months) pertussis hospital discharge rate (per 1000) was higher for Māori (1.49; relative risk 2.29 [95% CI: 1.77–2.96]) and Pacific people (2.03; relative risk 3.11 [95% CI: 2.30–4.22]) and lower for Asian/Indian (0.31; relative risk 0.47 [95% CI: 0.25–0.90]) compared with European/Other people (0.65 per 1000).42

From 2006 to 2010 an infant living in a household in the most deprived quintile was at a four-fold increased risk of being hospitalised with pertussis compared with an infant in the least deprived quintile (1.89 versus 0.39 per 1000; relative risk 4.81 [95% CI: 2.99–7.75]).42