Pertussis (whooping cough) is a commonly overlooked diagnosis in adults with respiratory symptoms, accounting for infection in 20% of adults presenting with prolonged cough in settings with high immunisation coverage [1]. The bacterial infection caused by Bordetella pertussis is characterised by a prolonged cough (lasting more than two weeks) with paroxysms, inspiratory whoop and post-tussive vomiting. The incubation period varies between six and 20 days and cases are infectious from six days after exposure and may last up to three weeks after the onset of typical paroxysms [2]. Complications of pertussis such as pneumonia and otitis following bacterial superinfection occur usually in childhood; with the highest number occurring in the first six months of life [3]. In adults and adolescents, it often presents atypically, making diagnosis difficult. Adults and school-age children are a known source of infection for younger family members who are too young to be immunised [2].

Pertussis is a vaccine-preventable disease with routine immunisation in the UK given to children at two, three and four months and a subsequent pre-school booster. The vaccine currently in use in the UK is an acellular vaccine that was first introduced in 2001 as a pre-school booster. The Health Protection Agency figures for England and Wales show statutory notifications of pertussis averaging 665 per year over the period 2000-2005, while during 2002-2004 on average 292 of these cases were laboratory confirmed diagnoses (serology, PCR or culture) per year . Increased notification as well as ascertainment has occurred among adults during this time [4].

The following outbreak report describes a cluster of infections affecting mostly adult members of a family and the subsequent infection of an infant. The Oxfordshire Health Protection Team was notified of a case of pertussis in a 20-year-old student in West Yorkshire on 27 October 2006. Contact tracing revealed that she had been in contact with several members of her family over the August Bank Holiday weekend. This article describes the identification and management of the cluster and seeks to highlight the importance of clinical awareness of pertussis in adolescents and adults and the performing of thorough contact tracing.

Methods and results

Outbreak investigation

Index case
The index case was in a 20-year-old university student in West Yorkshire. She developed a cough with post-tussive vomiting on 21 August 2006, which persisted for eight weeks. She had not been immunised against pertussis as a child and was never known to have suffered from pertussis. She had returned from a two-month trip to East Africa two weeks prior to the onset of her symptoms and had no recollection of any contact persons with similar symptoms whilst on the trip. During the three weeks following onset of symptoms, her ‘close/ household contacts’ (Figure 1) were her boyfriend, who lived in the same flat with her, and seven members of her family, who shared the same accommodation with her for three days over the Bank Holiday weekend at the end of August.

Secondary cases
Sister 1 developed a chronic cough with a whoop two weeks after contact with the index case (Figure 1). She was eight months pregnant at the time of contact and symptoms persisted until the week after delivery. However, she did not seek medical attention for her symptoms at any point in time. The investigation revealed that she was not immunised against pertussis during childhood.

Sister 2 also developed symptoms that fit the case definition for pertussis [3], two weeks after the contact with the index case (Figure 1). She presented to her General Practitioner (GP) with these symptoms at the time and was treated with a five-day course of Erythromycin for a suspected chest infection. She was not tested for pertussis and she had also not been immunised against pertussis as a child.

Tertiary case
The baby born to Sister 1, who was symptomatic at the time, developed a cough with apnoeic attacks three weeks after birth (Figure 1) and was admitted to hospital where pertussis was suspected and diagnosed on a nasal swab culture. She was treated with intravenous erythromycin and made an uneventful recovery.

The remaining five contacts in the family (four adults and a three-year-old niece) were fully immunised and did not meet the case definition for pertussis (Figure 1). The boyfriend of the index case was also fully immunised and showed no symptoms.

Measures taken

Following the confirmed diagnosis of pertussis in the three-week-old baby above, the unimmunised mother was treated as a ‘vulnerable contact’ and provided with oral Erythromycin prophylaxis for 10 days. Her husband, who was fully immunised, had symptoms of a non-specific cough at the time and tested negative for B. pertussis. The index case presented to her GP with information that her three-week-old niece had been diagnosed with whooping cough. A blood sample was sent for pertussis PCR and she was commenced on a 10-day course of oral Erythromycin. However, the result of the blood test did not confirm the diagnosis. The negative result could have been due to the delay of over two months between the onset of symptoms and testing. Paired serology would have been a better choice of test given the duration of symptoms [3].

All the other family contacts and the boyfriend remained asymptomatic and did not meet the requirements for prophylaxis as they had all received childhood immunisation.

Discussion

Discussion Professional and public loss of confidence in the pertussis vaccine due to a belief that there was a link to a group of children with brain damage published in a paper in 1974 [2] led to a drop in immunisation coverage of the whole cell pertussis containing DPT vaccine to 30% in 1975 in the UK (Figure 2). Many of those unvaccinated during the late 1970s and 1980s would have been infected with pertussis at this time and the drop in vaccination cover corresponds to the rise in pertussis notifications as shown in Figure 2.

This outbreak report seeks to highlight the importance of suspecting and recognising symptoms of pertussis in the adult population, especially in those who should but may not have been unimmunised during the pertussis vaccine ‘scare’ in the late 1970s. A causal relationship has already been noted with the decrease in vaccination coverage and a following rise in pertussis notifications [5]. We could be seeing a second impact now, 30 years on, with those born in the 1970s and not having been immunised presenting now with pertussis as adults but being diagnosed wrongly with other respiratory illnesses like asthma.

Pertussis is a statutorily notifiable disease, so immunisation history and contact tracing for close contacts in the preceding three weeks of the onset of symptoms in the patient are vital. In the case reported above, this would have led to antibiotic prophylaxis being prescribed to the pregnant sister and possibly the prevention of transmission to the newborn child. Close contacts of pertussis cases who are either unvaccinated, partially vaccinated or under five years of age should be given erythromycin treatment or prophylaxis [6]. The low rates of laboratory confirmation seen nationally compared to the notification rates is probably due to reluctance of physicians to subject patients to a nasopharyngeal swab [8] or the fact that culture diagnosis loses sensitivity with prolonged duration of symptoms [5].

Children and adults might be infected with pertussis even if they were vaccinated in the past because both immunity following vaccination and natural infection wane over time [2]. In the UK, where vaccination coverage by two years of age is currently around 94% [9], pertussis is often overlooked as a differential diagnosis. This is particularly so in adults, as noted above. It is possible that current notification rates that are based on clinical suspicion do not reflect the true incidence of pertussis, especially in the adult population.

Conclusion

This report highlights two major issues with regards to pertussis. Firstly, that a diagnosis of pertussis should form part of the differential diagnosis in any adult with prolonged cough. The sensitivity of this case definition is as high as 84-92% [10].

Secondly, due to the high secondary attack rate associated with unimmunised household contacts, a good family history and prompt notification and testing while providing prophylaxis is crucial in avoiding infection in the most vulnerable population – unimmunised infants.