Breakthrough COVID-19 infections among healthcare workers at a tertiary care hospital in Beirut: Prevalence and determinants

Highlights

• •A higher number of vaccine doses was associated with fewer breakthrough infections.
• •Nausea and vomiting were less frequent among healthcare workers with breakthrough infections.
• •The findings support booster vaccination and increased vaccine coverage among healthcare workers.

Abstract

Objectives

Despite mass vaccination campaigns to limit the spread of COVID-19, breakthrough infections have been reported in many countries. This raises global concern regarding the longevity and scope of vaccine-induced immunity, especially among high-risk populations such as healthcare workers (HCWs). This study aimed to examine the prevalence and predictors of breakthrough infections among HCWs at a tertiary care hospital in Beirut.

Methods

A cross-sectional study was conducted among 300 HCWs from clinical and non-clinical departments. Data were collected using a structured questionnaire assessing demographics, vaccination, COVID-19 exposure, compliance with preventive measures, symptoms, and clinical outcomes with bivariate and multivariate analysis. Data were collected from participants between November and December 2023.

Results

Of the 300 HCWs, 36.5% (n = 46) had breakthrough infections. In multivariate analysis, receiving more vaccine doses was associated with reduced odds of breakthrough infection (odds ratio = 0.315, 95% confidence interval: 0.116-0.855, P = 0.023). No significant associations were found with age, comorbidities, work area, number of exposures, or compliance with preventive regulations.

Conclusion

Receiving more COVID-19 vaccine doses was significantly associated with a reduced likelihood of breakthrough infection among HCWs. These findings emphasize the importance of booster vaccination and support continued efforts to promote full vaccine coverage within healthcare settings.

Introduction

Since its emergence in late 2019, the COVID-19 pandemic has caused substantial morbidity and mortality worldwide [1,2]. In response, the rapid development and deployment of COVID-19 vaccines played a crucial role in limiting transmission, reducing hospitalization, and mitigating death [3,4]. Several studies have demonstrated that vaccination significantly lowers the likelihood of severe illness and shortens the duration of infectivity among those who become infected [4,5].

However, it has become increasingly evident that immune protection following the primary vaccination series against COVID-19 may wane over time [[6], [7], [8]], leading to the emergence of breakthrough infections. In response, booster vaccines were introduced to prolong immune protection [[9], [10], [11]]. While studies have demonstrated that booster vaccines reduce the risk of infection and severe outcomes, their effectiveness in preventing breakthrough infections remains a subject of ongoing research [12,13].

This especially becomes a concern for healthcare workers (HCWs), as they represent one of the most high-risk populations [14]. Given their significant role at the forefront of the response, HCWs are placed at a heightened risk of infection and transmission, as they directly deal with patients with suspected or confirmed COVID-19 [15,16]. However, concerns remain regarding the durability of vaccine-induced immunity, especially in the context of emerging new variants [17].

To our knowledge, there is no published literature documenting the prevalence of COVID-19 breakthrough infections among HCWs in a tertiary care setting in Lebanon. Given this gap, this study aims to assess prevalence and predictors of COVID-19 breakthrough infections among HCWs at a tertiary care hospital in Beirut.

Methodology

The study adopted a cross-sectional design to assess the prevalence of COVID-19 breakthrough infections among HCWs at a tertiary care hospital in Beirut, Lebanon. Breakthrough infections were defined as confirmed SARS-COV-2 infections occurring between 14 days and 6 months after administration of the second vaccine dose or any subsequent dose within the same period [18].

A total of 300 HCWs were recruited from both clinical and non-clinical departments using a convenience sampling approach. Eligibility included all HCWs irrespective of the number of COVID-19 vaccine doses received.

Data were collected using a structured questionnaire. The data was collected from participants between November to December 2023. Informed consent was obtained from all participants before data collection. The questionnaire captured information across five key domains:

• 1)Demographic information: this section collected basic demographic information, including age, sex, occupation, work area category, and comorbidities.
• 2)Vaccination history: participants were asked to provide information regarding COVID-19 vaccination status, including the number of doses received, vaccine type, consistency of vaccine, and the dates of each corresponding dose.
• 3)Exposure history: to assess the likelihood of COVID-19 exposure, participants were asked to report on whether they had been in contact with confirmed or suspected COVID-19 cases, the number of suspected exposures, and whether these occurred either in or outside the hospital setting. Questions were also used to capture the level of compliance of HCWs with the use of personal protective equipment and distancing protocols both inside and outside of the hospital setting.
• 4)Infection history and clinical outcomes: to gain insight into the clinical course of both breakthrough and non-breakthrough cases, participants were asked to indicate whether they had tested positive via positive polymerase chain reaction (PCR) result, the number of times they tested positive, and the corresponding dates for each infection. Additional questions addressed the nature and duration of symptoms, time to recovery, and time to the first negative PCR test following each infection.

All information provided by participants covered the time period since the start of the vaccination campaigns in Lebanon, reflecting the time interval from 2020 to 2023.

Statistical analysis

Descriptive statistics were used to summarize the demographic characteristics of the participants included in the study. In addition, variables related to comorbidities, vaccination, COVID-19 exposure, and compliance with measures, symptoms, and recovery information were analyzed.

For inferential statistics, nonparametric tests were primarily used given the nature of our dataset, which included variables that were categorical or continuous that did not follow a normal distribution.

Bivariate analyses were conducted to explore associations between participant characteristics and the occurrence of COVID-19 breakthrough infections. Variables showing statistical significance at the bivariate level were subsequently included in a binary logistic regression model to assess their independent contribution as risk factors to breakthrough infections. A 95% confidence interval was used for all estimates, and P-values <0.05 were considered statistically significant. The analysis was conducted using IBM SPSS Statistics Version 27.

The data collection tool was reviewed for clarity and relevance by the study team before administration. Responses were anonymized to maintain confidentiality. The institutional review board of the Lebanese American University approved the study protocol.

Results

Descriptive statistics

The demographic and occupational characteristics of the study population are summarized in Table 1. A total of 300 HCWs from the hospital were included in the analysis; three participants had incomplete questionnaires. Analyses were conducted using available case data. The median age of participants was 28 years (interquartile range: 24-40). In terms of occupation and work area categories, 297 participants provided information. Of these, 30.7% (n = 92) were nurses, followed by 21.0% (n = 63) medical students, and 12.3% (n = 37) residents/fellows. Regarding work area categories, most participants were working on non-COVID floors, whereas 16.2% (n = 48) passed through COVID floors and the remaining 11.1% (n = 33) were on rotation between both. Overall, of the 283 participants who provided information on comorbidities, only 90 (31.8%) reported at least one comorbidity related to either the gastrointestinal, respiratory, or cardiovascular systems. The most reported condition was hypertension, affecting 7.4% (n = 21) of participants. This was followed by 6.4% (n = 18) participants who reported asthma or chronic obstructive pulmonary disease and 4.9% (n = 14) with high cholesterol.

Table 1.

Demographic and occupational characteristics of participants.

Variable	Frequency (n)	Percentage (%)
Age
Mean ± SD	32.40 ± 10.89
Median	28
Mode	23
Occupation
Administrative staff	22	7.3%
Attending physician	5	1.7%
Call center	2	0.7%
Housekeeping	3	1.0%
Lab technician	4	1.3%
Laboratory staff	10	3.3%
Laundry	1	0.3%
Medical student	63	21.0%
Midwife	2	0.7%
Nurse	92	30.7%
Nursing student	4	1.3%
Nursing supervisor	2	0.7%
Nutrition student	2	0.7%
Nutritionist	9	3.0%
Patient transporter	3	1.0%
Pharmacy	10	3.3%
Pharmacy student	7	2.3%
Practical nurse	4	1.3%
Radiology technician	9	3.0%
Resident/fellow	37	12.3%
Secretary	2	0.7%
Security	3	1.0%
Work area category (n = 297)
COVID floor	48	16.2%
Non-COVID floor	216	72.7%
Rotational	33	11.1%

The vaccination characteristics of the HCWs showed that of the 299 participants who provided information, 97.2% (n = 291) received at least one dose of a COVID-19 vaccine at the time of data collection. Primary vaccination series (two doses) were completed by 19.7% (n = 59) of participants, whereas 63.5% (n = 190) completed three doses and 10.7% (n = 32) received a fourth dose. Only 2.7% (n = 8) did not receive any vaccination. In terms of the type of vaccine product received, 95.3% (n = 285) were vaccinated with Pfizer-BioNTech vaccine. Vaccine consistency was high among participants, with 97.9% (n = 281) receiving the same vaccination product across all doses, most commonly Pfizer-BioNTech. The median interval between the first and second dose was 22 days. For participants who received a third dose, the median interval from the second dose was 204.5 days (approximately 6 months). Among those who received a fourth dose, the interval between the third and fourth doses was slightly longer, with a median of 9 months.

Overall, of the 297 participants 89% (n = 267) reported being exposed to a confirmed or suspected case of COVID-19. Among the 262 exposed participants, 43.9% (n = 116) reported multiple exposure events exceeding five times. Among those exposed, over half of the participants reported adhering to regulations both inside and outside the hospital, reflecting overall strong compliance to infection prevention measures. An additional 30.6% (n = 90) of participants complied only within the hospitals, whereas 5.1% (n = 15) did so outside the hospital. The remaining 9.5% (n = 28) reported that they did not comply in either setting.

Table 2 summarizes the infection-related characteristics of the study population. Among the 299 HCWs, 42% (n = 126) reported being infected with COVID-19. Of these, 18.3% (n = 23) experienced at least more than one infection. The median time interval between the first and second infection was approximately 7 months. Among those infected, 36.5% (n = 46) were classified as breakthrough infections, whereas the remaining 63.5% (n = 80) were categorized as non-breakthrough infections.

Table 2.

Frequency and timing of infections.

Variable	Frequency (n)	Percentage (%)
Breakthrough vs non-breakthrough infection (n = 126)
Breakthrough infection	46	36.5%
Non-breakthrough infection	80	63.5%
Group categories (n = 126)
Breakthrough after dose 2 within 6 months	4	1.2%
Breakthrough after dose 2 within 6 months and before dose 3	7	2.0%
Breakthrough after dose 3 within 6 months	28	8.1%
Breakthrough after dose 3 within 6 months and before dose 4	5	1.4%
Breakthrough after dose 4 within 6 months	2	0.6%
Infection after dose 2 post 6 months	22	6.4%
Infection after dose 2 post 6 months and before dose 3	13	3.8%
Infection after dose 3 post 6 months	38	11.0%
Infection after dose 3 post 6 months and before dose 4	1	0.3%
Infection after dose 3 post 6 months and before dose 4	2	0.6%
Infection after dose 4 post 6 months	1	0.3%
Infection between dose 1 and dose 2	3	0.9%
Time between first and second infection (n = 23)
Mean ± SD	233.26 ± 118.5
Median	216
Mode	216
Min	58
Max	530

The most reported symptoms among infected HCWs were fatigue 65.4% (n = 166), fever 61.4% (n = 156), headache 53.5% (n = 136), and cough 46.1% (n = 117). Gastrointestinal symptoms were less commonly reported, only affecting 10% (n = 27) of respondents. With regard to recovery times, 52.2% (n = 107) of participants reported that symptom recovery took from 2-7 days, whereas 33.2% (n = 68) of participants reported that the time to the first negative PCR ranged from 7-14 days. Additionally, 7.8% (n = 16) of participants reported a longer time for symptom resolution. The same was observed in 17.8% (n = 36) of participants, where it took more than 14 days to get their first negative PCR since their onset symptoms.

Bivariate analysis

No significant association was found between infection type and demographic characteristics. Both age (Z = –1.149, P = 0.251) and work area (Z = 4.699, P = 0.95) did not demonstrate a significant association with infection type. A significant association was found between the number of vaccine doses and infection type, with fewer doses being associated with breakthrough infections. No correlation between any of the common comorbidities and infection type was found. These included conditions such as obesity (P = 0.299), hypertension (P = 0.485), asthma or chronic obstructive pulmonary disease (P = 1.000), and diabetes (P = 0.414). Similarly, both exposure status (Z = 0.000, P = 1.000) and number of exposures (Z = –1.344, P = 0.179) were not associated with the likelihood of breakthrough infections.

In addition, most symptoms were not significantly associated with the type of infection (breakthrough vs non-breakthrough). Symptoms such as chills (P = 0.440), cough (P = 1.000), fatigue (P = 0.314), and fever (P = 1.000) were not significantly different between groups. The only symptom significantly associated with infection type was nausea and vomiting, which was reported exclusively among participants with non-breakthrough infections (P = 0.003). No statistically significant differences were found between recovery time variables and infection type. Recovery time until symptoms subsided did not differ significantly between the two groups (Z = –1.303, P = 0.193). Similarly, the recovery time until achieving the first negative PCR test was not significantly different (Z = –0.632, P = 0.527).

Multivariate analysis

Multivariate logistic regression analysis was conducted to examine the association between various factors and the likelihood of breakthrough infections (Table 3). The model included age, work area, number of vaccine doses, comorbidities, exposure status, and compliance with COVID-19 regulations. The number of vaccine doses was the only significant predictor of breakthrough infections (odds ratio = 0.315, 95% confidence interval: 0.116-0.855, P = 0.023). Other factors, including age, work area, comorbidities, and number of exposures, were not significantly associated with breakthrough infections.

Table 3.

Multivariate analysis between various factors and breakthrough infection.

Variable	Exp[B]	95% confidence interval	P-value
Age	0.978	[0.936-1.022]	0.321
Work area category(Reference: COVID floor)
COVID floor			0.171
Non-COVID floor	2.803	[0.478-16.435]	0.253
Rotational	0.851	[0.184-3.947]	0.837
Number of vaccine doses	0.315	[0.116-0.855]	0.023a
Comorbidities	0.573	[0.166-1.979]	0.379
Number of exposure(Reference: Once)
Once			0.481
Two-five times	1.100	[0.156-7.768]	0.924
More than five times	0.583	[0.231-1.475]	0.255

^aStatistically significant at P <0.05.

Discussion

The study aimed to assess the prevalence and predictors of COVID-19 breakthrough infections among HCWs at a tertiary care hospital in Beirut. The study population was made up of a diverse group of HCWs, most had received ≥2 doses of the COVID-19 vaccine. Among all variables assessed, the number of vaccine doses received was the only significant protective factor against breakthrough infection. A higher number of vaccine doses was associated with lower odds of breakthrough infections across both bivariate and multivariate analyses.

This finding aligns with prior studies showing that booster doses enhance protection by increasing antibodies [[19], [20], [21]]. More recent studies have documented declining protection against infection 3 months after the second dose, reinforcing the importance of timely boosters [6,7]. Similarly, a systematic review and meta-analysis demonstrated that a second booster dose significantly increased neutralizing antibody titers against multiple COVID-19 variants, enhanced cellular immunity, and was associated with significant reductions in COVID-19 cases, intensive care unit admissions, and deaths [17].

An additional finding of our study was the notable similarity in symptom profiles between breakthrough and non-breakthrough infections, except for nausea and vomiting, which appeared less frequently among breakthrough infections. This may be due to emerging new variants and/or persistent immunoglobulin A-related immunity. In contrast, individuals with non-breakthrough infections may have been infected earlier in the pandemic, potentially acquiring stronger immune responses either due to more robust viral exposure or shorter time intervals between infection and vaccination. These factors could contribute to subtle differences in symptom manifestation. Many studies have found that vaccinated individuals experience lower rates of infection as well as milder and fewer symptoms [22].

Both the time interval between first and second dose, as well as the interval between the second and third dose, were not significant predictors of breakthrough infections in this study population. In contrast, one study found that antibodies produced in the upper respiratory tract during the convalescent phase, along with post-vaccination duration, are important determinants in expanding the breadth of neutralization of non-Omicron variant [23].

Despite examining a range of demographic, clinical, and occupational variables, none were significantly associated with breakthrough infections. The lack of correlation between age and breakthrough infection is likely due to the age distribution of the study population, where around 55% of participants were between the ages of 18-29 years, while only a small minority were aged ≥50years. In several studies, age has been shown to be a significant predictor of increased susceptibility to infection and severe outcomes due to poorer immunity and a higher prevalence of comorbidities [24,25].

Similarly, the lack of association between comorbidities and breakthrough infections does not align with what is currently reported in the available literature. Comorbidities are often associated with increased risks of developing infections and severe outcomes. A study found that about 51% of patients with COVID-19 had chronic illnesses such as cardiovascular diseases, respiratory problems, endocrinologic, cerebrovascular diseases, and malignancies [26]. Given the relatively young age group and healthy composition of our study population, there was restricted variability in underlying health conditions. In fact, only 31.8% of participants reported at least one comorbidity, and the prevalence of individual comorbidities was low in comparison. For example, the most reported comorbidity, hypertension, accounted for only 7.4% of 283 participants. This might explain why no association was detected.

The study also found no association between work area and breakthrough infections. This may reflect two factors: first, most HCWs (72.7%) were stationed on non-COVID floors, potentially reducing the statistical power to detect any variation. Second, the presence of negative pressure rooms in the COVID-19 units may have contributed to the lack of a significant difference between breakthrough and non-breakthrough cases across different work areas. Studies in other hospitals have found that higher viral loads detected on the hospital floors were associated with hospital outbreaks [27]. In addition, one study reported that HCWs working in COVID-19 wards had a two-fold increased risk of infection compared with HCWs who were not in direct contact with infected patients [28].

The study also did not detect any association between compliance of the staff with the COVID-19 protective measures and breakthrough infection. This finding contrasts with several studies that have emphasized the protective value of distancing, mask use, and hand hygiene in reducing transmission [19,29]. The absence of association observed in our study can be explained, in part, by limitations in how compliance was measured. Compliance data were self-reported, which makes it susceptible to recall bias.

Limitations

First, some measures used in the analysis were subject to recall bias, as they relied on participants’ self-reported information. The accuracy of these responses may have been affected by participants’ ability to recall events. Second, the study was conducted in a single tertiary hospital in Beirut, which may limit the generalizability of the findings to other health care settings. Third, the sample size of 300 participants may have limited the statistical power to detect small or moderate associations with some variables, such as age, comorbidities, or department-specific risks. Additionally, there is a potential for selection bias, as sampling was based on convenience.

Conclusion

In conclusion, our findings indicate that a higher number of COVID-19 vaccine doses was significantly associated with a reduced risk of breakthrough infection among HCWs. These findings emphasize the importance of booster vaccination in protecting HCWs and support continued efforts to promote full vaccine coverage within health care settings.

Future research should incorporate a broader array of clinical, demographic, immunological, and occupational variables to allow for more comprehensive analysis on prior infection, vaccination history, and exposure intensity.

Declaration of competing interest

The authors have no competing interests to declare.