Fever Among Pertussis Cases Reported Through Enhanced Pertussis Surveillance, 2015–2022

Abstract

Pertussis, also known as whooping cough, is a highly contagious respiratory disease caused by Bordetella pertussis. During 2015–2022, approximately 13 000 people were reported with pertussis annually in the United States (US) [1]. Pertussis is cyclic, with peaks in incidence every 3–5 years. Pertussis poses a significant health threat, particularly to infants and young children who are not fully vaccinated [2]. Vaccination is the best way to prevent pertussis and reduce the severity of infection [3–5].

Pertussis typically starts with early symptoms similar to a common cold, but generally, instead of resolving, it progresses to severe coughing fits. Pertussis is likely underdiagnosed, and atypical clinical presentation, especially in adolescents and adults, may make it difficult to identify [6]. Fever is often not thought of as being part of the classical pertussis presentation nor included in the US national case definition for reporting [7], but understanding the prevalence of fever among pertussis cases and associated factors will help clinicians better identify and treat pertussis cases.

We aimed to determine the prevalence of fever in pertussis cases reported in the US, describe characteristics associated with fever, and describe temperatures in those with fever.

METHODS

Cases of pertussis with cough onset from 1 January 2015 through 31 December 2022 were identified through Enhanced Pertussis Surveillance in 7 sites as part of the Centers for Disease Control and Prevention's Emerging Infections Program Network (Supplementary Figure 1) [8, 9]. According to the pertussis case definition, pertussis cases were classified as confirmed or probable [7]. From 2015 through 2019, a probable case was defined as a cough illness lasting ≥2 weeks and at least 1 clinical symptom (paroxysms, inspiratory whoop, or posttussive vomiting) in the absence of a more likely diagnosis; a confirmed case was defined as either (1) a probable case with a positive polymerase chain reaction (PCR) result or epidemiologic linkage to a laboratory-confirmed case; or (2) a cough of any duration with the isolation of B pertussis; specific differences in symptoms and cough duration for infants aged <1 year were also included. In 2020, Council of State and Territorial Epidemiologists updated the pertussis case definition by classifying PCR-positive cases as confirmed, limiting confirmed cases to those with laboratory confirmation, and eliminating age-specific classifications [10].

For all identified pertussis cases, public health surveillance personnel completed a standardized case report form that included demographics, clinical presentation, and pertussis vaccination history. Vaccination history was determined by reviewing medical records, state immunization registries, patient vaccination cards, school vaccination records, or self-report. People were considered ever vaccinated if they had ever received at least 1 dose of a pertussis-containing vaccine.

A fever was considered “documented” if either the individual with pertussis or their doctor reported a specific temperature ≥38°C (≥100.4°F) during the 2 weeks before cough onset through the date of the interview. Fevers were considered subjective if a fever was reported but no temperature ≥38°C was recorded. Among people with documented fever, we characterized the timing of fever onset (before, at, or after cough onset) and the temperature distribution.

Among cases with known fever status (documented fever or no fever; excludes subjective fever), we used bivariate log-binomial regression models to estimate the prevalence ratio of fever by pertussis case characteristics, adjusted for age category.

A P value <.05 was used to determine statistically significant differences. Analyses were conducted in R version 4.3.1 software. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy.

RESULTS

Between 1 January 2015 and 31 December 2022, there were 8992 reported cases of pertussis; 802 (9%) had documented fever, 661 (7%) had subjective fever, 7038 (78%) had no fever reported, and 491 (5%) had unknown fever status (Supplementary Figure 1, Supplementary Table 1). The analyses included individuals with documented fever status, totaling 7840 cases: 802 (10%) with documented fever and 7038 (90%) without. Among people with known timing of fever (751/802), most fevers (75%) began on the same date or after cough onset. Documented temperatures ranged from 38°C to 41.7°C (median, 38.7°C) (Figure 1).

Figure 1.

Distribution of temperature among pertussis cases with documented fever (n = 802).

Open in new tabDownload slide

People aged <2 months with pertussis were less likely to have documented fever (6%) compared to people aged 2 months to <6 years (16%) (Table 1). Adjusting for age, Hispanic individuals were 1.3 times more likely to have fever compared to non-Hispanic individuals. This association remained even when stratifying the proportion with fever by site, surveillance year, or age group (data not shown). People with cough onset in fall, winter, and spring were more likely to have fever than cases with cough onset in summer. People classified as probable cases were 1.6 times more likely to have fever than people with a confirmed pertussis case status. There were differences in the prevalence of reported fever among people with pertussis between surveillance sites, ranging from 6% in Connecticut to 13% in New Mexico. There were no differences in fever by surveillance year, except in 2022, when the prevalence of fever among those with pertussis was 20% compared to <12% in previous years. Clinically, symptoms such as whooping, posttussive vomiting, apnea, and cyanosis, and being hospitalized for pertussis infection were all associated with an increased likelihood of fever. People with pertussis who were ever vaccinated against pertussis were 2 times more likely to have fever than those who were never vaccinated with a pertussis-containing vaccine.

DISCUSSION

This provides the first population-based estimate of the occurrence of fever with pertussis in the US. In our analysis of 8 years of pertussis surveillance, fever was an infrequent clinical presentation of reported pertussis infections, reported in approximately 1 in 10 pertussis cases, with a relatively stable prevalence across surveillance years and the change to the case definition in 2020. An exception was in 2022 when there were few pertussis cases and a high prevalence of fever among cases, which may be attributable to an increase in other pathogens causing fever or health-seeking behavior that would make fever documentation more likely as severe acute respiratory syndrome coronavirus 2 pandemic restrictions lifted. Previous studies report varied proportions of fever in pertussis cases. Fever occurred in about 25% of international travelers (median age, 44 years) [11], 29% of cases in a population mostly <1 year old [12], and only 2% of hospitalized infants ≤90 days old [13]. A systematic review found that, in adult patients, the presence of fever should rule out pertussis [14]. The between-study variation could be due to various factors, including studied age groups, study setting, method of ascertaining pertussis case status, and prevalence of circulating coinfections.

Among pertussis cases with fever, temperatures were generally mild, with most fevers reported <38.9°C. The low fevers are consistent with our understanding of pertussis’ optimal temperature growth range of 35–37°C [15]. In a pertussis baboon model, B pertussis avoided inducing fever (only observed minor [1°C increase] and transient temperature elevations). One immunological reason may be the loss of the key virulence factor, adenylate cyclase, at temperatures >38.9°C [16].

Exploration of expanding pertussis testing to all cases presenting with influenza-like illness may increase sensitivity and identify a broader spectrum of cases with potentially a different prevalence of pertussis and associated symptoms. For example, in a non-US population-based study in a whole cell pertussis vaccine setting, using active surveillance with weekly follow-up and with no requirement for cough duration, fever was reported in 50% of infants aged ≤6 months with pertussis; this higher detection of fever could be due to the increased sensitivity to detect atypical pertussis infection [17]. Expanding the clinical symptoms considered potentially part of pertussis symptomology may increase detection in clinical settings. Furthermore, the feasibility of maintaining pertussis surveillance may be increased if pertussis testing can be added to existing surveillance systems that detect a broader spectrum of respiratory illnesses.

A limitation of this study is that the catchment area was restricted to 7 sites participating in Enhanced Pertussis Surveillance and is not representative of the US. Mild cases would not have been captured if they had not sought care or had laboratory testing. The population included only those who met the pertussis case definition, which may exclude cases with an atypical presentation. There may be biases in which people with pertussis and true fever were able to obtain a documented fever versus subjective or unknown fever; those with a home thermometer or better access to care potentially may be more likely to have documented fever, although the percentage with unknown fever status was low (∼5%). There were also likely biases in those with documented versus suspected fever. Since coinfection data were collected but highly missing (77%), we could not look directly at coinfections’ association with fever. Higher testing and collection of coinfection data for all pertussis cases would aid in our understanding of the contribution of pertussis versus a coinfection with other pathogens in fever symptomology.

These findings can help clinicians better identify pertussis cases by highlighting the possibility of pertussis in patients with cough and a history of or current low-grade fever. We hope these findings will also inform future integration efforts by aiding public health practitioners in better understanding how pertussis may be integrated with other respiratory disease surveillance systems.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Acknowledgments. The authors thank the Enhanced Pertussis Surveillance staff for collecting the data on pertussis cases used for this analysis.

Author contributions. All authors contributed significantly to this work. M. M. H. and T. H. S. conceived the study design and research methodology. M. M. H. performed the data analysis, drafted the manuscript, and ensured its integrity and accuracy. T. H. S., A. R., A. M. A., V. A., P. F., K. H., E. S., R. W., E. Y., and V. C. performed the data collection, contributed to the interpretation of the data, and provided critical revisions for intellectual content. All authors reviewed and approved the final version of the manuscript for submission.

Data availability. The data are not publicly available.

Disclaimer. The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the CDC.

Financial support. This work was part of the Enhanced Pertussis Surveillance through the Emerging Infections Program Network, which is supported through a Centers for Disease Control and Prevention cooperative agreement.

Patient consent. This activity was reviewed by the Centers for Disease Control and Prevention (CDC) and was conducted consistent with applicable federal law and CDC policy: see, eg, 45 Code of Federal Regulations (C.F.R.) part 46.102(l)(2), 21 C.F.R. part 56; 42 United States Code (U.S.C.) §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq. Our study used data collected for public health surveillance purposes and thus individual patient consent was not required.

References

Author notes