Pertussis is a vaccine-preventable infectious disease caused by Bordetella pertussis. In recent years, there has been evidence of an increased incidence in many countries with a high vaccination coverage.1,2 Variations in published incidence rates may be due to differences in case reporting systems and access to diagnostic tests,2–4 but the highest rates are always reported in infants aged less than 1 year, who also experience the most complications and highest mortality. The Canary Islands is one of the autonomous communities in Spain with the highest rates of infection.5 In the 2003–2007 period, we performed a study where we found a high incidence of complications and mortality in this age group.6 For this reason, since 2015, vaccination of pregnant women was introduced in our autonomous community. In order to determine the subsequent trends of infection, we have conducted a clinical-epidemiological study of all the cases diagnosed in infants aged less than 1 year that visited the emergency department of the only public paediatric hospital in the island of Gran Canaria between 2008 and 2016.

All patients underwent testing for detection of B. pertussis by means of polymerase chain reaction (PCR) (LightMix Kit Bordetella pertussis/parapertussis or Diagenode Bordetella pertussis/parapertussis) in samples of nasopharyngeal exudate obtained with a flexible swab. Viral detection tests were also performed 581patients (55.9%) in nasopharyngeal aspirate samples.

We performed a retrospective review of the health records of patients who tested positive for B. pertussis, collecting data on demographic variables, vaccination history, probably source of infection (contacts with clinical features compatible with pertussis), symptoms, laboratory data and treatment.

In the statistical analysis, we used the Mann–Whitney U test to compare quantitative variables and the chi square test to compare proportions. We defined statistical significance as a p-value of less than .05. The study was approved by the local bioethics and research committee.

Despite the high vaccination coverage in the paediatric population, pertussis continues to be a significant health problem that recently led to the introduction of vaccination of pregnant women in the autonomous community of the Canary Islands with the intent to protect infants, the most vulnerable subset.7–9 Pertussis is a disease that causes epidemics on a cyclical pattern, although since 2010 in Spain there has been a sustained epidemic,5 which was reflected in our study. This phenomenon coincides with the re-emergence of pertussis observed both in the United States and in other European countries due to reasons that are yet unknown; some of the reasons that have been proposed are loss of immunity over time, antigenic drift in circulating strains relative to those used to make the vaccine and lack of vaccination in certain subsets of the population.9–13 In our region, between 2003 and 2010, the incidence in infants aged less than 1 year who required hospital admission only exceeded 100 cases in years 2003 (103 cases) and 2006 (204.5 cases).6 Since 2010, the annual incidence of cases with hospital admission exceeded 130 per 100000 inhabitants, with peaks of 365.0 and 425.0 cases in 2011 and 2015, respectively. However, in 2016 there was a decrease in incidence that coincided with the introduction of vaccination of pregnant women. This measure should be analysed in the long term to determine whether this trend is sustained, as it could also have been due to the data being collected in an interepidemic period.

Comparing incidence data between countries poses significant difficulties due to differences in the case definition, the diagnostic techniques used and reporting systems.2,14 According to data from the European Centre for Disease Prevention and Control, in 2014 the highest reported age-specific rate corresponded to infants aged less than 1 year, with an incidence of 51.6 cases per 100000 inhabitants in Europe overall. Spain, Lithuania, the Netherlands, Slovenia and Sweden were the countries with the highest incidence rates, with more than 100 cases per 100000 inhabitants in this age group.15 According to several studies, pertussis may be underdiagnosed, so its actual incidence could be higher. Basing the case definition on clinical variables could contribute to underdiagnosis, as such a definition is limited because signs and symptoms vary with age and vaccination status and also because physicians still maintain a low level of suspicion, especially in adolescents and adults. Recently, a research group proposed a diagnostic algorithm based on clinical signs and symptoms by age group.16 According to this algorithm, pertussis should be suspected in infants aged less than 3 months if they present with coughing and coryza, especially in cases where the cough increases in frequency and severity, independently of its duration or whether or not it is paroxysmal, or infants presenting with leucocytosis and lymphocytosis. In our study, most patients presented with cough, accompanied by cold symptoms in 2/3 of cases, and most also had lymphocytosis. We ought to note that pneumonia and apnoea developed mainly in infants aged less than 3 months, as described in previous studies,17 but we found no other differences in other signs and symptoms between the 2 age groups.

The population aged less than 1 year is at higher risk of complications. We observed complications in 15.4% of patients, mostly pneumonia. Younger age and unvaccinated status were the most significant factors associated to the development of complications, although none of the patients died. Compared to the preceding period, we found a lower rate of complications (23.4% in the 2003–2007 period) and a lower mortality (6.5% in the 2003–2007 period). Other studies have found a higher rate of complications in this age group.17 Several studies have analysed different variables and have found a higher risk of death or more severe disease in infants aged less than 4 months with low birth weight, hyperleukocytosis (≥100000 leukocytes/μL), who develop pulmonary hypertension or seizures or with unvaccinated status, among others,18–20 and a high degree of suspicion in these patients would probably improve clinical outcomes.

Viruses were detected in a high proportion of patients in whom B. pertussis infection was ruled out. The virus detected most frequently was respiratory syncytial virus, which demonstrates the overlap of signs and symptoms in this subset of the population, a fact that has been described by other authors.21,22 Furthermore, a viral coinfection was detected in a small proportion of the sample (most commonly by rhinoviruses), which was consistent with previous studies in which rhinoviruses and coronaviruses were most frequently detected in cases of coinfection.22

In conclusion, in recent years there has been an increase in the incidence of pertussis in infants aged less than 1 year in our region associated with decreases in the incidence of complications and in mortality compared to the previous period. Younger age and unvaccinated status were the main risk factors associated with the development of complications. The incidence of pertussis and its complications would probably decrease with vaccination of pregnant women, so it is important that long-term followup studies be conducted to assess the effectiveness of this preventive measure.

The authors have no conflicts of interest to declare.