A retrospective cohort study on the protective effect of influenza vaccination during pregnancy in Gansu Province Authors: Xue-Feng Liang, Juan-Xia Peng, Di Sun, Bing-Ling Zhang, Jing An, Xiao-Shu Zhang Published: 2024 ABSTRACT ABSTRACTTo evaluate the Vaccine Effectiveness (VE) of the Influenza Vaccine (InfV) during pregnancy in Gansu province. A retrospective cohort of pregnant women vaccinated with InfV was constructed using the electronic case system and immunization planning system of Gansu province. Outcome information of Influenza-Like Illness (ILI) in pregnant women after vaccination with InfV during the influenza epidemic seasons from 2019 to 2021 was collected. The log-binomial model was used to estimate the adjusted Risk Ratio (RR), VE, and their Confidence Intervals (CI). The incidence rates of the InfV vaccination group and the non-vaccination group during pregnancy in Gansu province were 4.70% (54/1,150) and 6.91% (16,990/245,754), respectively. The adjusted RR was 0.78 (95%CI 0.43–0.75), and the adjusted VE was 22.20 (95%CI 0.79–40.30). Vaccination with InfV during pregnancy in Gansu province from 2019 to 2021 has a protective effect against ILI in pregnant women, and there is no difference in vaccination at different stages of pregnancy.KEYWORDS: Pregnancy, influenza vaccine, protective effect, vaccination Introduction Acute infections caused by influenza viruses during pregnancy often lead to more severe health risks for both the mother and the infant, resulting in higher rates of hospitalization and mortality compared to non-pregnant women of childbearing age.1 Studies have found that vaccinating pregnant women against influenza not only reduces the risk of hospitalization due to severe influenza2 but also prevents the occurrence of in newborns.3 The World Health Organization (WHO) has recommended InfV during pregnancy.4 However, the vaccination rate among pregnant women remains low due to the fact that some China InfV instructions list pregnant women as a contraindicated population and concerns about the safety of vaccination during pregnancy.5,6 Conducting randomized cohort studies during pregnancy also faces challenges. There is an urgent need for relevant research on InfV vaccination during pregnancy in China to comprehensively understand the risks and benefits of InfV vaccination during pregnancy and to improve public awareness and understanding of InfV vaccination during pregnancy. Since 2013, Gansu Province has been continuously improving its province-wide electronic medical record (EMR) system and immunization information system. By 2018, the EMR system covered all public healthcare institutions at all levels and 90% of private healthcare institutions within the province, including all outpatient, emergency, and inpatient electronic medical records.7 The immunization information system covered 2,831 vaccination units, with all vaccination records being completed through the scanning of vaccine traceability codes. Linking the established multiple database resources has made it possible to evaluate the protective effect of InfV vaccination during pregnancy. Retrospective cohort studies by linking multiple databases can avoid some common problems in prospective cohort studies, particularly the inaccurate or missing data caused by study subjects’ refusal to participate or recall bias. This study aims to construct a retrospective cohort of InfV vaccination during pregnancy by obtaining data from the immunization program information system and electronic case system of Gansu Province from 2019 to 2021, and to analyze the impact of InfV vaccination during pregnancy on ILI in pregnant women, providing a scientific basis for evaluating the effectiveness of InfV vaccination implementation in pregnant women. Materials and methods Study subjects Data from the Gansu Province Electronic Case System and Immunization Program System from January 1, 2019, to December 31, 2021, were obtained. A retrospective cohort of pregnant women who received InfV during pregnancy was constructed by linking databases through identification numbers. The inclusion criteria for study subjects were as follows: pregnant women aged between 18 and 45 years with valid identity information in the immunization program system database, influenza vaccine vaccination records between September 1 and December 31 each year from 2019 to 2021, which overlapped with the pregnancy period and influenza season, and medical records of outpatient (emergency) or inpatient visits in the electronic case system. The exclusion criteria were as follows: individuals who received vaccines other than the influenza vaccine during pregnancy and those who received immunoglobulin or blood product treatment during pregnancy. Data collection and definitions Data on education level, household registration, gestational week, vaccine name, vaccination time, delivery mode, gestational hypertension, gestational diabetes, history of thyroid disease, chronic respiratory system disease, and nervous system disease were collected for the study subjects. The study outcome was the occurrence of ILI within 6 months after InfV vaccination in pregnant women. Fever (axillary temperature ≥ 38°C) accompanied by cough or sore throat was used as the diagnostic criteria for ILI. The InfV collected in this study were split trivalent influenza vaccine and split quadrivalent influenza vaccine, both administered through intramuscular injection. The follow-up endpoint was the occurrence of ILI symptoms within 6 months after delivery or reaching the study termination time, which was 6 months after delivery. InfV vaccination within 12 weeks of gestation was considered early pregnancy vaccination, vaccination between 12 and 27 weeks of gestation was considered mid-pregnancy vaccination, and vaccination after 28 weeks of gestation was considered late pregnancy vaccination. Gestational hypertension and gestational diabetes were determined according to the diagnostic criteria of the International Classification of Diseases, 10th Revision (ICD-10). Histories of thyroid disease, chronic respiratory system disease, and nervous system disease were determined by text matching in the Gansu Province Electronic Case System, and if a match was found, the subject was considered to have the corresponding disease history. Statistical analysis Data processing and statistical analysis were performed using R software version 4.3.1, with the main packages being EpiStats, Stats, Robust, and Gtsummary. The basic characteristics of the vaccinated and unvaccinated groups were compared using the χ2 test for qualitative data and the tt-test for quantitative data. A log-binomial model was used to estimate the RR and its 95% CI of InfV vaccination after adjusting for demographic factors, disease history, and other factors.8 The test level α was set at 0.05 (two-sided), and the VE was calculated as (1-RR)*100. Result Basic characteristics analysis of study subjects The study included 246,904 pregnant women, with 1,150 in the InfV vaccination group and 245,754 in the unvaccinated group. The mean age was 29.95 ± 4.29 years for the vaccinated group and 28.26 ± 4.88 years for the unvaccinated group. Significant differences were observed between the two groups in terms of maternal age, household registration, and education level. However, no significant differences were found in gestational age, delivery mode, gestational hypertension, gestational diabetes, history of thyroid disease, chronic respiratory system disease, or nervous system disease (Table 1). Table 1.Basic characteristics analysis of the InfV vaccinated and unvaccinated groups during pregnancy.CharacteristicVaccination group, N = 1,150Unvaccinated group, N = 245,754StatisticP-valueGestational week38.98 (0.99)38.96 (1.01)0.66.508Mode of delivery 0.77.380Cesarean section957 (83.22%)206,942(84.21%) Vaginal delivery193 (16.78%)38,812(15.79%) ILI 8.4.004No1,096(95.30%)228,764(93.09%) Yes54 (4.70%)16,990 (6.91%) Maternal age29.95 (4.29)28.26 (4.88)13<.001Household registration 1,172<.001City1,144(99.48%)120,066(48.86%) Countryside6 (0.52%)125,688(51.14%) Educational level 1,492<.001Junior High School64(5.57%)110,784(45.08%) University and above1,066(92.70%)91,869(37.38%) High School19 (1.65%)39,621(16.12%) Elementary School1 (0.09%)3,480 (1.42%) Hypertension during pregnancy 0.00>.999No1,099(95.57%)234,853(95.56%) Yes51(4.43%)10,901(4.44%) Gestational diabetes during pregnancy 0.07.796No1,125(97.83%)240,783(97.98%) Yes25 (2.17%)4,971(2.02%) Thyroid disease history 1.5.224No1,133(98.52%)240,770(97.97%) Yes17 (1.48%)4,984(2.03%) History of chronic respiratory diseases 1.4.244No1,144(99.48%)243,572(99.11%) Yes6 (0.52%)2,182 (0.89%) History of neurological diseases 0.00.982No1,142(99.30%)243,923(99.25%) Yes8 (0.70%)1,831 (0.75%) Basic characteristics analysis of the InfV vaccinated and unvaccinated groups during pregnancy. ILI Incidence and VE The incidence rates of ILI were 4.71% in the vaccinated group and 6.13% in the unvaccinated group. After adjusting for maternal age, gestational age, delivery mode, household registration, education level, gestational hypertension, gestational diabetes, history of thyroid disease, chronic respiratory system disease, and nervous system disease using a log-binomial model, the adjusted RR was 0.78 (95% CI 0.43–0.75), and the adjusted VE was 22.20 (95% CI 0.79–40.30). Significant differences in ILI incidence were noted among populations defined by maternal age, gestational age, delivery mode, and gestational hypertension (Table 2). Table 2.ILI incidence rates and VE in different populations [N (%)].VariableAdj.RR (95%CI)Z-valueP-valueAdj.VE (95%CI)Group0.688 (0.523-0.884)−2.797.00531.158 (11.615-47.683)Maternal age0.993 (0.990-0.996)−4.65.0000.708 (0.412-1.005)Gestational week0.969 (0.956-0.983)−4.439.0003.097 (1.722-4.435)Mode of delivery1.030 (0.990-1.073)1.459.145−3.047 (−7.312-1.005)Household registration0.999 (0.970-1.028)−0.068.9460.100 (−2.841-2.958)Educational level0.999 (0.984-1.015)−0.091.9270.072 (−1.493-1.613)Hypertension during pregnancy1.069 (0.998-1.144)1.924.054−6.911 (−14.351-0.207)Gestational diabetes during pregnancy0.964 (0.867-1.069)−0.677.4983.556 (−6.885-13.328)Thyroid disease history1.067 (0.964-1.176)1.277.202−6.690 (−17.630-3.586)History of chronic respiratory diseases0.933 (0.791-1.090)−0.847.3976.689 (−9.023-20.900)History of neurological diseases1.000 (0.840-1.177)−0.003.9980.023 (−17.700-15.983) ILI incidence rates and VE in different populations [N (%)]. Results of vaccination at different stages of pregnancy During the early, middle, and late stages of pregnancy, 325, 323, and 502 women were vaccinated, respectively. No significant difference in ILI occurrence was found among women vaccinated at different stages of pregnancy (Table 3). Table 3.Results of vaccination at different stages of pregnancy.Vaccination during pregnancyILITotalYesNoearly16309325mid12311323late26476502 Results of vaccination at different stages of pregnancy. Annual comparison of ILI incidence rates and vaccine effectiveness in vaccinated vs. unvaccinated groups Each year, the ILI incidence rate was lower in the vaccinated group compared to the unvaccinated group. However, the 95% confidence intervals for VE each year included zero, suggesting that the vaccine effectiveness might not be statistically significant (Table 4). Table 4.ILI incidence rates and VE in different years.YearVaccination groupUnvaccinated groupAdj VE(95%CI)Enrollment CountILI CaseIncidence %Enrollment CountILI CaseIncidence %(%)2019381194.995160735996.9729.82 (−11.16 -55.70)2020369164.347370250846.9038.70 (−1.08 - 62.82)2021400194.7512044583076.9032.60 (−6.78 - 57.46)Total1150544.70245754169906.9133.55 (12.71 – 49.41) ILI incidence rates and VE in different years. Results of across different characteristics Pregnant women with urban household registration (VE 33.03%, 95% CI: 12.07%-49.00%), having a college degree or higher (VE 31.87%, 95% CI: 9.85%-48.52%), undergoing cesarean section (VE 32.31%, 95% CI: 9.02%-49.64%), without gestational hypertension (VE 32.79%, 95% CI: 11.25%-49.10%), without gestational diabetes (VE 33.37%, 95% CI: 12.25%-49.41%), without a history of thyroid disease (VE 33.72%, 95% CI: 12.71%-49.67%), with a history of chronic respiratory disease (VE 33.22%, 95% CI: 12.28%-49.17%), and without a history of neurological disease (VE 33.06%, 95% CI: 12.06%-49.17%) achieved better vaccine effectiveness after receiving the InfV. However, in the characteristics of junior high school and high school educational levels, vaginal delivery, hypertension during pregnancy, gestational diabetes, and thyroid disease history, the 95% CI for vaccine effectiveness (VE) included zero, indicating that the vaccine effectiveness in these characteristics might not be statistically significant. Therefore, further research is needed to verify the effectiveness of the vaccine in these specific populations (Table 5). Table 5.ILI incidence rates and VE in different people.CharacteristicVaccination groupUnvaccinated groupAdj VE(95%CI) (%)Enrollment CountILI CaseIncidence %Enrollment CountILI CaseIncidence %Household registration City1144544.7212006682966.9133.03 (12.07 – 49.00)Countryside60012568886946.92–Educational level Elementary School10034802447.01 Junior High School6423.1311078476256.8854.59 (−14.04 - 100)High School1915.263962127887.0425.2 (−128.15 - 100)University and above1066514.789186963336.8930.59 (11.20 - 48.31)Mode of delivery Cesarean section957464.81206942143876.9532.31 (9.02 – 49.64)Vaginal delivery19384.153881226036.7139.72 (−22.19 – 70.26)Hypertension during pregnancy No1099524.73234853161866.8932.79 (11.25 – 49.10)Yes5123.92109018047.3848.62 (−110.88-87.48)Gestational diabetes during pregnancy No1125534.71240783166586.9233.37 (12.25 – 49.41)Yes251449713326.6841.65 (−329.95 – 92.08)Thyroid disease history No1133534.68240770166236.933.72 (12.71 – 49.67)Yes1715.8849843677.3621.31 (−491.67 – 89.53)History of chronic respiratory diseases No1144544.72243572168496.9233.22 (12.28 – 49.17) Yes60021821416.46− History of neurological diseases No1142544.73243923168636.9133.06 (12.06 – 49.04)Yes80018311276.94− Total1150544.7245754169906.9133.55 (12.71 – 49.41) ILI incidence rates and VE in different people. Discussion Vaccination against influenza during pregnancy can reduce the risk of influenza infection and its complications.9 The VE of InfV in pregnant women reflects the actual performance of InfV in preventing the disease. Evaluating the effectiveness of InfV vaccination during pregnancy can provide scientific evidence for immunization strategies during pregnancy. The results of this study show that the incidence rates of ILI in the InfV-vaccinated and unvaccinated groups during pregnancy in Gansu Province were 4.71% and 6.13%, respectively, with an adjusted RR of 0.78 (95% CI: 0.43–0.75) and an adjusted VE of 22.20% (95% CI: 0.79–40.30). In a study by D Mølgaard-Nielsen the protective effect of InfV vaccination during pregnancy against laboratory-confirmed influenza was 63.9% (95% CI: 29.1–81.6),10,11 which is higher than the VE obtained in this study. This discrepancy may be attributed to the non-specificity of ILI, and the estimation of the protective effect using ILI is still subject to a certain degree of interference. This study also shows that vaccination with InfV during early, middle, and late pregnancy can reduce the occurrence of ILI, and there is no difference in the incidence of ILI among these periods. This finding is similar to the results of Helen Ding et al., who found that vaccination with seasonal influenza vaccine during any stage of pregnancy can reduce the risk of respiratory diseases.12 This study effectively controlled for recall bias among the study subjects by utilizing electronic medical records,13 which improved the accuracy and reliability of the data. Connecting multiple databases increases the complexity of data management and analysis, requiring more resources and technology to handle the integration, cleaning, and analysis of large-scale data. The log-binomial model was used to address the issue of estimating RR in the presence of multi-categorical and continuous covariates, avoiding the biased estimation of RR by the logistic regression model.14 Although the “Guidelines for Influenza Vaccination in China” recommend InfV vaccination during pregnancy, the vaccination rate during pregnancy in Gansu Province is extremely low, at only 0.04%, due to factors such as InfV being a self-paid vaccine in Gansu Province and being listed as a contraindication in some instructions.1 The public’s doubts about the safety and VE of InfV vaccination during pregnancy, along with ethical considerations and implementation difficulties, have limited the conduct of prospective cohort studies. Although InfV vaccination during pregnancy plays a significant role, there are currently no similar domestic studies evaluating the protective effect of InfV vaccination during pregnancy on maternal and child health. This study is the first in China to use a retrospective cohort analysis to investigate the relationship between influenza vaccination during pregnancy and the risk of ILI in pregnant women in Gansu Province, providing evaluation data for the protective effect of influenza vaccination during pregnancy. Due to the limited patient information collected by the electronic medical record system in Gansu Province, the lack of influencing factors, including lifestyle, may affect the accuracy and comprehensiveness of the study results. Most of the medical information exists in text form, and these data have not been fully organized into structured data, increasing the difficulty of data processing and analysis. Therefore, it is impossible to rule out the possibility that the study results may be influenced by unknown or unconsidered confounding factors. Since the number of people in the vaccinated group was significantly smaller than that in the unvaccinated group, there may be significant differences in other key characteristics, making the comparison results between the two groups inaccurate or subject to selection bias. As this study is a retrospective observational study, the study subjects cannot be randomly grouped. The log-binomial model technique was used to deal with the influence of confounding factors; however, when using this model, the error distribution must be correctly specified; otherwise, it may lead to biased study results.15 Due to the characteristics of retrospective studies and the current influenza surveillance sentinel sites in Gansu Province, it is impossible to uniformly use laboratory-confirmed influenza or hospitalized influenza incidence as outcome indicators. Therefore, the estimation of VE may be biased. The ILI cases collected in this study included non-influenza sentinel surveillance hospitals, and there may be a certain degree of underreporting of ILI due to inconsistent diagnostic criteria in non-influenza sentinel surveillance hospitals. Subgroup analysis indicated that vaccination with InfV could increase the VE against ILI among pregnant women who have urban residency, a university education or higher, and no gestational diabetes or hypertension. However, there are currently no other studies confirming the vaccine’s efficacy in this specific subgroup of pregnant women. Further research is needed to validate these findings. According to relevant studies, InfV vaccination during pregnancy can reduce the risk of pregnant women and infants under 6 months of age suffering from severe influenza-related diseases or hospitalization.16 It not only has a protective effect on pregnant women but also saves each woman approximately $50, with a net gain of about 45 quality-adjusted hours, showing significant cost-effectiveness.17,18 In summary, incorporating InfV vaccination during pregnancy into a free program and removing contraindications from the vaccine instructions can provide broader protection for pregnant women and reduce the risk of influenza-related diseases in maternal and child health. Conclusion This study shows that administering InfV during pregnancy has a protective effect in preventing ILI in pregnant women. The administration of InfV during pregnancy provides a statistically significant protective effect against ILI in pregnant women. It is recommended to include InfV vaccination during pregnancy in free vaccination programs and to remove contraindications from the vaccine’s instructions. This can provide broader protection for pregnant women and reduce the risk of influenza-related diseases among pregnant women. Biography Xue-Feng Liang Since 2002, I have been working at the Gansu Provincial Center for Disease Control and Prevention, focusing on vaccine research and immunization program development. Over the past decade, I have been deeply involved in the formulation and implementation of immunization policies in Gansu Province, with a specific focus on vaccine-preventable disease research. Additionally, I have participated in numerous vaccine-related research projects, accumulating substantial practical experience. Funding Statement This study was funded by the Gansu Provincial Health and Health Industry Scientific Research Plan [GSWSKY2023-21] Disclosure statement No potential conflict of interest was reported by the author(s). Ethical approval The study protocol and informed consent form were approved by the Medical Ethics Committee of the Gansu Provincial Center for Disease Control and Prevention (Approval notice: 2023007). Abbreviations Vaccine Effectiveness Influenza Vaccine Influenza-Like Illness Relative Risk Confidence Intervals World Health Organization References