End-of-season outbreaks of nosocomial influenza caused by waning vaccine immunity

Abstract

The waning of vaccine protection may be responsible for outbreaks toward the end of the influenza season. Three of five outbreaks occurred at the beginning of April following an interval of >100 days from the date of vaccination; the reported index case was a nurse or office worker, and >50% of those affected were healthcare workers. The results are consistent with intra-seasonal waning of vaccine immunity that resulted in outbreaks at the end of season.

Introduction

Vaccination is essential for minimising excess morbidity and mortality caused by influenza infection. Increased vaccination coverage of healthcare workers has been shown to reduce the occurrence of nosocomial influenza in patients with cancer (Frenzel et al., 2016). However, the protection offered by the vaccine may wane during the influenza season (Belongia et al., 2015; Ferdinands et al., 2017; Ray et al., 2019; Sullivan et al., 2014). The highest vaccine effectiveness (VE) has been reported to occur shortly after vaccination (Ferdinands et al., 2017). The odds ratio of influenza has been reported to increase by 12% for every 14 days after vaccination (Belongia et al., 2015), and VE was also found to decrease by 50% in the 93 days after vaccination (Sullivan et al., 2014). Intra-seasonal waning of vaccine immunity could result in outbreaks toward the end of the influenza season. The study objective was to survey the incidence of influenza outbreaks related to the waning of VE in healthcare workers in a Japanese hospital.

Methods

This study was conducted at the Kochi Medical School Hospital, Nankoku, Japan, a 605-bed tertiary-care general hospital with 13 wards. Patients with community-acquired influenza are usually treated at other hospitals in this region, which limits admissions in this hospital. As all influenza infections are reported to the infection control department, outbreaks involving both healthcare workers and patients are well documented. Hospital policy requires all healthcare workers to receive annual influenza vaccinations. Even though vaccination was encouraged at the time of admission, almost all patients were not likely to be vaccinated. Droplet precautions are implemented in addition to standard precautions for patients with influenza. Hospital surveillance records were searched for the occurrence of influenza outbreaks that occurred within the previous 10 years. The outbreak characteristics, patient and case data were retrieved from medical records and retrospectively reviewed. Written informed consent was obtained from all patients in this study.

Results

The hospital records included data on five seasonal influenza outbreaks that occurred within the previous 10 years. The outbreaks are listed by date of onset in Table 1, and their characteristics are summarised. Of the five outbreaks, three occurred at the end of March or the beginning of April. In all three, influenza activity in the area of Japan outside of the hospital was low level (< 10 hospital visits per week to a hospital). Interval from the date of vaccination to the beginning of outbreak was > 100 days in these outbreaks. The reported index case in each of those three outbreaks was a nurse or office worker, not a patient and > 50% of the infections occurred in healthcare workers. These three outbreaks all occurred late in the influenza season, were caused by healthcare workers and involved more healthcare workers than patients.

Table 1.

Characteristics of five influenza outbreaks.

	1	2	3	4	5
Date of onset	1 February 2017	3 March 2016	29 March 2012	30 March 2011	1 April 2017
Influenza activity outside of hospital	Peak	Peak	Low level	Low level	Low level
Interval between the date of vaccination and outbreak (days)	48	90	134	140	107
Reported index case	Patient	Patient and nurse	Office worker	Nurse	Nurse
Involved number of healthcare workers	2	5	15	11	6
Involved number of patients	14	7	0	11	3
Patients likely to be vaccinated (%)	7	0	N/A	0	0
Proportion of healthcare workers (%)	13	42	100	50	67
Location	Orthopaedic surgery ward	Orthopaedic surgery ward	Regional medical cooperation room	Orthopaedic surgery ward	Haematology ward

Peak is defined as > 30 patient visits per week to a hospital.

Low level is defined as < 10 patient visits per week to a hospital.

Proportion of healthcare workers (%) was calculated as involved number of healthcare workers divided by involved number of patients plus healthcare workers.

Figure 1 shows the time course of outbreak number four. On 30 March 2011, a nurse with a sore throat and fever was diagnosed with influenza. Another nurse on the same ward began coughing on 1 April and was diagnosed with influenza. Both nurses had worked the same 8-h shift on 30 March and were in contact. On 2 April, an inpatient in the same ward who had been under the care of the second nurse for 9 h on the previous day developed influenza. Three more nurses, one doctor and three inpatients developed influenza on 3 April. This event was recognised as an outbreak by the infection control team.

Figure 1.

Nosocomial influenza infections at our institution (2011).

Discussion

The hospital records contained descriptions of five influenza outbreaks in the previous 10 years, three of which occurred toward the end of the season. All three were caused by vaccinated healthcare workers and involved more healthcare workers than patients. The results are consistent with the waning of VE in healthcare workers during the influenza season that resulted in outbreaks at the end of the season. Similar end-of-season influenza outbreaks in vaccinated patients have been reported (Tang et al., 2018). Available evidence thus suggests that improvement of vaccine immunity is necessary to reduce the risk of in-hospital influenza outbreaks.

One possible solution is to delay vaccination. However, delayed vaccination could increase influenza infection before vaccination starts. Moreover, delayed vaccination might result in decreased vaccination coverage (Rambhia and Rambhia, 2019). Therefore, delaying vaccination is not an ideal strategy. Substantial benefits have been associated with minimising the time between vaccination and influenza activity (Newall et al., 2018). Unfortunately, the variation in seasonal influenza epidemics would make it difficult to determine the ideal time to begin vaccinating hospital staff. In addition, in the northern hemisphere, the influenza season usually starts in autumn and continues until early spring. Therefore, VE should persist for approximately 22 weeks.

An additional vaccination of healthcare workers within a single influenza season may be an acceptable way to extend VE. We previously examined the effects of a booster vaccination in elderly people. The booster vaccination was administered four weeks after the primary vaccination. We found that an additional vaccination during the same influenza epidemic season promotes a long-lasting antibody response and is effective in maintaining immunity to influenza (Matsushita et al., 2018). Future clinical trials are warranted to identify the best approach to induce an antibody response to protect against influenza infection in healthcare workers.