Understanding adverse events following immunisation: evidence of vaccine safety and coincidental events from 3 years of surveillance

Abstract

Background

Periodic vaccine safety checks are a crucial component of any country’s pharmacovigilance system. Unwanted reactions following childhood vaccination are unacceptable to the public, making rapid safety checks and alerts essential for maintaining public confidence.

Objectives

The study evaluates the safety of childhood vaccines in the paediatric population through active surveillance and explores various predictors for the development of adverse event following immunisation (AEFI) based on population characteristics.

Methodology

A cohort event monitoring study was designed to identify AEFIs in children receiving at least one childhood vaccine under the Universal Immunisation Programme (UIP) at the study site. Both active surveillance (via telephone follow-up calls) and passive surveillance (from healthcare workers) were included. Potential predictors of AEFIs were evaluated using the χ² test, and relative risk was calculated.

Results

The study estimated an AEFI incidence rate of 10.37%, with 3.47% (1.27 events/1000 doses) being serious reactions requiring hospitalisation. The highest incidence was with the Pentavac vaccine (diphtheria, tetanus, pertussis, hepatitis B, Hib) at 19.21% (218 events/1000 doses), while the lowest incidence was observed with the oral polio vaccine at 0.13% (1.32 events/1000 doses). Most reported Adevrse Events were solicited (34.83 events/1000 doses), while unsolicited events accounted for 1.69/1000 doses. The most common reaction was fever (23.75 events/1000 doses), followed by pain at injection site (4.98 events/1000 doses). One death following Pentavac vaccination was reported but later deemed coincidental. Overall, 90.62% of the reactions were vaccine product-related, 4.76% coincidental, 2.70% immunisation error-related and 0.39% were indeterminate. Predictors of AEFIs included paediatric population, delayed immunisation, nutritional deficiencies and comorbid status of the child and children from affluent families.

Conclusion

Childhood vaccines included in the UIP have shown a favourable safety profile. Serious reactions are infrequent, emphasising the need for thorough investigation and clear communication to differentiate them from coincidental reactions, which can occur more frequently in this age group.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Vaccine safety monitoring mainly relies on passive national pharmacovigilance systems, where under-reporting of adverse event following immunisation (AEFI) remains a major challenge in low- and middle-income countries. Public concerns, coincidental adverse events and inadequate communication can contribute to misinformation and reduced vaccine acceptance.

WHAT THIS STUDY ADDS

• This study demonstrates the value of active surveillance in improving the completeness and timeliness of AEFI detection in a tertiary care setting. Delayed immunisation timelines and selected comorbidities were identified as key predictors of reported AEFIs, underscoring the need for risk-informed surveillance approaches. The contrast between active and passive surveillance demonstrates the importance of integrated monitoring strategies in minimising underreporting.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• This study highlights the value of integrating active surveillance into routine immunisation programmes to strengthen vaccine safety assessment. The findings encourage future research on risk-based monitoring and may inform policy decisions aimed at strengthening surveillance systems and optimising immunisation practices within healthcare settings.

Introduction

Many infections that were once devastating to humanity are now unheard of or rare, thanks largely to the incredible impact of effective vaccination coverage. However, since the invention of vaccines, concerns and misconceptions about their safety have been the greatest challenge for this preventive measure to overcome.¹ The persistent threads exerted by various vaccine-preventable diseases still underscore the importance of vaccines in the twenty-first century. Several unpredictable factors such as microbial resistance, climate change, emerging new strains and poor living conditions potentially exacerbate the challenges faced in combating these diseases and emphasise the need for continued vaccination efforts globally.²

The eradication of smallpox through effective vaccination stands as a timeless success story.³ However, today’s generation is more concentrated on the safety of the vaccines rather than the disease complications due to their unfamiliarity with those diseases.⁴ The Latin phrase ‘primum non nocere’ or ‘first do no harm’ reminds the healthcare professionals (HCPs) ‘to consider all possible harms before an intervention’, and this principle applies to the safe usage of vaccines.⁵ All the licensed vaccines available on the market undergo rigorous testing to ensure their safety, quality and efficacy. However, these clinical trials often involve smaller groups, which can limit the detection of rare safety issues.⁶ The lack of adequate sample size and insufficient time to identify rare and delayed adverse events following immunisation (AEFI) can result in poor safety outcomes during clinical trials. This discrepancy between clinical experiments and real-world scenarios can be mitigated through an effective monitoring system.

Most vaccines have faced numerous challenges from anti-vaccination movements throughout their successful development. These challenges must be addressed promptly by the appropriate authorities through detailed investigations to strengthen public confidence in vaccines. The importance of vaccine safety monitoring was evident from the ‘Cutter incident’ during the American polio vaccine programme in 1955.⁷ Consequently, various events such as Guillain-Barré Syndrome following swine influenza vaccination (1976),⁸ intussusceptions observed in infants following rotavirus vaccine (1998–1999)⁹ and narcolepsy following H1N1 vaccination (Pandemrix) (2009–2010) were¹⁰ reported in different time periods. Recently, the rare event of vaccine-induced thrombotic thrombocytopenia following COVID-19 vaccines underscores the effectiveness of these monitoring systems.¹¹

Although global initiatives for vaccine safety surveillance are commendable, variations exist from country to country based on infrastructure and facilities. In particular, suboptimal reporting of AEFIs is highly observed in low- and middle-income countries due to the lack of sophisticated surveillance systems.^{12 13} In India, the surveillance system relies heavily on voluntary spontaneous reporting by HCPs, with limited participation from the private sector.¹⁴ Additionally, factors such as diverse population, numerous private healthcare settings and a higher focus on non-communicable diseases (which are more prevalent in the Indian population) contribute to suboptimal AEFI reporting in the country.12,14

The active vaccine safety surveillance conducted in this study confirms the safety of childhood vaccines in the paediatric population and significantly contributes to post-marketing vaccine safety research. Timely surveillance aids in the differential diagnosis of coincidental events, particularly in vulnerable populations such as children, and helps identify any delayed onset of AEFIs. The study also explores various predictors for the development of AEFIs based on population characteristics, emphasising its epidemiological importance. Hence, the significance of this study extends beyond the present, offering valuable insights for future research.

Methodology

Active vaccine safety surveillance was conducted among children under 16 years who received any of the recommended childhood vaccines at a tertiary care hospital’s immunisation centre. This cohort event monitoring study aimed to understand the trends and patterns of AEFIs in the paediatric population over 3 years from May 2021 to December 2023, with an additional 5 months for follow-up. It also included all passively reported AEFIs by HCPs and parents, as well as suspected cases admitted to the paediatric ward of the study site through routine clinical reporting. The study community was identified from a single immunisation centre of tertiary care hospital located in a south Indian city, with a population density of 476 persons per square kilometre.¹⁵ The district’s total child population (0–6 years) was estimated at 305 561,¹⁵ with the study enrolling 2.15% (n=6571) of children from this age group within the district.

Each vaccine beneficiary was assigned a unique registration number in the Digital Immunisation Registry, where all demographic, immunisation and AEFI data were recorded, preventing double reporting. Consent was obtained from parents or the legal guardians of each child before enrolling into the study. The informed consent form was designed according to the Indian Council of Medical Research guidelines for biomedical research on human subjects. The consent form was developed in English and back-translated to the local language, Kannada. The study excluded those unwilling to participate or unreachable during the study period. The contact numbers of the parents or the legal guardians, along with their availability to respond to the calls, were collected and communicated while obtaining consent.

Active surveillance was conducted through telephone contact with parents or guardians, conducted in two phases, to gather information on immediate and delayed adverse reactions. The phase I surveillance is 8±2 days of post-immunisation, and phase II surveillance is 28±2days of post-immunisation. The selection of these time windows was informed by prior on-site active surveillance experience and supported by evidence from the literature,^{14 16} which indicates that most immediate and local adverse reactions typically occur within the first week following immunisation, whereas delayed reactions are commonly reported within 1 month. Given that individual follow-up was performed via telephone for each vaccine beneficiary, a grace period of ±2 days was incorporated into both surveillance phases to accommodate logistical constraints and ensure complete follow-up without compromising data quality.

Follow-up calls were conducted by the study team, with additional support from interns posted at the immunisation centres. All calls were made from the immunisation centre in the presence of the immunisation nurse for verification of vaccine details only and without involvement in the surveillance process. Each call lasted approximately 3 to 5 min and followed a standardised set of surveillance questions, beginning with the screening question, ‘Has your child experienced any health problems or unusual symptoms after vaccination?’ This was followed by confirmation of vaccination details (date and vaccine administered), and where applicable, systematic documentation of the nature, onset and duration of symptoms, healthcare consultation or treatment and current status or resolution of any reported event. When additional medical care was sought elsewhere, parents were requested to share relevant medical records (soft or hard copies) for further assessment. Both surveillance calls follow a similar pattern, and in the absence of any AE reported during these calls, parents were advised to contact the study team if any significant symptoms were observed subsequently. All details of AEFIs include the onset of the reaction, description of the event, management (symptomatic or specific), time to recovery and severity of the reaction. These details were documented in the ‘AEFI case report form’ which was specially designed according to WHO core variables for AEFI collection and reporting. Additionally, information on the management of the reaction, diagnosis and details of hospitalisation (if any) was obtained from the patient medical records provided by the parents during their visit to the study site (if the patient was admitted outside). For in-patients, details related to AEFIs were directly obtained from the treating physician, medical records and the Hospital Information System. The severity of each reported event was determined according to the revised national AEFI guidelines issued by the Ministry of Health & Family Welfare, GoI.¹⁷

Subjects who did not respond to the first phase of the surveillance (telephone follow-ups) were labelled ‘surveillance I absentees’. Those who did not respond to both the surveillances and did not visit the study site at any time during the study period were considered ‘dropouts’. If participants who missed both surveillances visited the study site for subsequent vaccine doses, AEFI details for the previous dose were obtained through face-to-face interviews to recall any adverse events. Causality assessment of reported AEFI was performed in accordance with the updated WHO causality assessment manual, wherein eligibility assessment is defined as the initial step of the evaluation.¹⁷ Accordingly, each reported AEFI was first screened for eligibility. Then, the ‘Causality assessment checklist’ was used to gather evidence for case review, and the ‘Causality assessment Algorithm’ was determined whether the AEFI was ‘consistent’ or ‘inconsistent’ with immunisation, ‘indeterminate’ or ‘unclassifiable’. Final classification was based on Council for International Organizations of Medical Sciences (CIOMS)/WHO definitions for vaccine pharmacovigilance.

Each AEFI reported at the study site was screened by the AEFI investigation team comprising two senior professors of paediatrics, two clinical pharmacists and a clinical pharmacologist who have experience in the area of vaccine safety. Based on availability AEFI details, events were primarily classified into two groups: (i) adequate information available and (ii) adequate information not available. AEFIs with adequate information were further classified as follows: category A: consistent with causal association to immunisation (subdivided to A1, vaccine product-related reactions; A2, vaccine quality defect-related reaction; A3, immunisation error-related reactions and A4, immunisation anxiety-related reactions (ISRR)), category B: indeterminate (B1, temporal relationship is consistent, but there is insufficient definitive evidence for vaccine causing event and B2, qualifying factors result in conflicting trends of consistency and inconsistency with causal association to immunisation), category C: inconsistent with causal association to immunisation (coincidental underlying or emerging condition(s), or condition(s) caused by exposure to something other than vaccine). However, events with no adequate information were classified as category D: unclassifiable (also specify the additional information required for the classification).¹⁷

Apart from causality assessment, the study also categorised the reported AEs as solicited and unsolicited reactions. To examine the patterns and trends of AEFIs, this classification was performed in accordance with the guidelines published by the CIOMS/WHO Working Group on Vaccine Pharmacovigilance.¹⁸ Adverse events with a pre-defined frequency or those considered expected reactions, as documented in the fact sheets of the respective vaccines, were categorised as solicited events. In contrast, unsolicited events were defined as unexpected adverse reactions that are infrequently observed, have limited supporting evidence in the literature and are not described in the vaccine fact sheets of the respective vaccines.

Statistical analysis

Statistical analysis was performed using the SPSS V.22 (IBM Corp: licensed to JSS University). Quantitative variables were summarised using mean and SD, while categorical variables were expressed as frequencies and percentages. Pre-defined predictor variables for the occurrence of AEFIs were selected a priori based on biological plausibility, prior active surveillance experience at the study site and evidence from published literature. These included demographic factors (age group, sex), vaccine-related factors (type of vaccine/antigens, concomitant vaccine administration and timelines of immunisation) and beneficiary-related factors (presence of comorbidities, low birth weight ≤2 kg and family economic status). Associations between these predictors and AEFI occurrence were assessed using χ² tests and bi-variant analysis. Statistical significance was set at a ‘p’ value<0.05 with a 95% CI.

Patient and public involvement

Patients and the public were not involved in the design, conduct, reporting or dissemination plans of this research. The study was based on routine vaccine safety surveillance data collected as part of institutional monitoring activities.

Results

During the study period, 79 520 children visited the paediatric Out Patient Department ward at the study site. Among these, 6699 children attended the facility for the purpose of immunisation and received at least one childhood vaccine, resulting in an immunisation rate of 8.42%. A total of 21 303 vaccine doses were used by the study population, accounting for an average visit of 3.2±1.6 times per child. Consequently, the study team conducted an average of 4.1±2.2 telephone calls per child as part of the active surveillance. The study population was categorised by age according to the categories defined by National Institute of Child Health and Human Development for the ease of evaluation.¹⁹ The highest vaccine utilisation was observed in infants (28 days to 12 months), while the lowest utilisation was seen in the early adolescent group (12 to 16 years). A decreasing trend in vaccine utilisation was noted as the age of the children increased. A detailed description of the study population and the childhood-vaccine utilisation is presented in table 1.

Table 1. The distribution of vaccine beneficiaries and the utilisation percentage of childhood vaccines across different age groups.

Sl no.	Study population	Age category	No. of vaccine beneficiaries (n=6699)	Percentage of vaccine beneficiaries	No. of vaccine doses used (n=21 303)	Percentage of vaccine utilisation
1	Newborns	Birth to 27 days	2276	33.98%	6828	32.05%
2	Infants	28 days to 12 months	2996	44.72%	10 842	50.89%
3	Toddler	13 months to 2 years	701	10.46%	2122	9.96%
4	Early childhood	2 to 5 years	448	6.69%	1039	4.88%
5	Middle childhood	6 to 11 years	246	3.67%	430	2.02%
6	Early adolescent	12 to 16 years	32	0.48%	42	0.20%

Overall, the study population is predominantly composed of boys, who account for 53.65% (n=3594), compared with girls, who make up 46.35% (n=3105). During the study period, a total of 695 children reported 778 adverse events. This accounted for an AEFI incidence rate of 10.37% (total number of children reported with AEFIs divided by total study population multiplied by 100). The distribution of AEFIs among different age groups among the study population is depicted in table 2. The majority of the reported AEFIs were identified in the study through active surveillance, that is, 97.68% (n=760), while only 2.31% (n=18) were reported passively during the study period. However, only 2.70% (n=21) AEs were serious and required hospitalisation for their management, of which 18 events were passively reported from the study site hospital. All the reported AEFIs were classified as ‘adequate information available’ for assessing its causality. The different steps involved in the identification of AEFIs among the study population are presented in figure 1.

Table 2. The distribution of reported adverse event following immunisation (AEFI) across different age groups and vaccine types within the study population.

Sl no.	Vaccine	No. of AEFIs reported (n=695)	No. of individuals developed AEFIs (n=778)	No. of vaccines administered	AEFI incidence rate (%)	Type of reaction with number
Neonates (n=115 events reported from 111 vaccine beneficiaries)
1	Bacillus Calmette-Guerin vaccine	27	23	2276	1.01	Fever (10), injection site abscess (1), pus at injection site (1), swelling at the injection site (7), redness turned to white pimple (1), pain at injection site (5), runny nose (1), fatigue (1)
2	Oral polio vaccine (OPV)	3	3	2276	0.13	Loose stools (1), cough (2)
3	Hepatitis B vaccine	85	85	2276	3.73	Fever (79), rashes (1), jaundice (1), excessive crying (4)
Infants (n=518 events from 459 vaccine beneficiaries)
4	Pentavac Dose 1	244	226	1895	19.21	Fever (196), adenopathy (1), febrile convulsion (3), persistent crying (2), pain at injection site (29), swelling at injection site (9), pus at injection site (1), headache (1), chills (1), breathlessness (1)
	Dose 2	72	55			Fever (44), pain at injection site (18), swelling at injection site (6), convulsion (1), cough (1), exaggeration of pre-existing rashes (1), brought dead (1)
	Dose 3	97	83			Fever (64), convulsion (1), swelling at injection site (3), pain at injection site (24), lethargy (1), fever and body pain (2), bronchopneumonia (1), persistent crying (1)
5	Rota vaccine	4	4	1895	0.21	Diarrhoea (3), constipation (1)
6	Pneumococcal conjugate vaccine (PCV) Dose 1	12	12	1652	1.51	Fever (10), extensive crying (2)
	Dose 2	4	4			Fever (3), pain at injection site (1)
	Booster 1	9	9			Fever (5), rashes (1), swelling at injection site (1), febrile convulsion (1), cough (1)
7	Pentavac and PCV	1	1	3320	0.03	Fever with multiple seizure episode later diagnosed as meningitis (with pneumococcal bacterial culture) (1)
8	Influenza vaccine	21	20	571	4.58	Fever (16), chills (2), anaphylactic reaction (1), diarrhoea (2)
9	Typhoid conjugate vaccine (TCV)	18	16	332	4.82	Fever (9), pain at injection site (6), swelling at injection site (3)
10	MMR vaccine Dose 1	23	18	686	2.62	Fever (13), pain at injection site (3), swelling at injection site (3), diarrhoea (1), headache (1), chills (1), cough (1)
11	Meningitis conjugate vaccine (MCV)	3	2	22	9.09	Fever (2), pain at injection site (1)
12	Hepatitis A vaccine (inactivated)	5	4	343	2.62	Fever (8), pain at injection site (1), swelling at injection site (1)
	Hepatitis A (live vaccine)	5	5
Toddler (n=80 events reported from 75 vaccine beneficiaries)
13	Hepatitis A vaccine (live vaccine)	3	3	123	2.44	Fever (3)
14	Influenza vaccine	8	8	41	19.51	Fever (8)
15	MMR vaccine Dose 2	11	9	189	4.76	Fever (9), rashes on thigh and arms (1), pain at injection site (1)
16	MMR vaccine and TCV	1	1	1245	0.08	Increased urination (1)
17	Varicella vaccine	5	5	420	1.20	Fever (3), body pain (1), constipation (1)
18	Typhoid conjugate vaccine (TCV)	7	6	38	0.16	Fever (6), headache (1)
19	DTP vaccine Booster 1	45	43	288	14.93	Fever (40), pain at injection site (2), swelling at injection site (2), cough (1)
Early childhood (n=52 events from 41 vaccine beneficiaries)
20	DTP vaccine Booster 2	52	41	212	19.34	Fever (25), body pain (1), convulsion (4), dry cough (1), adenopathy (1), pain at injection site (15), swelling at injection site (4), cold (1)
Middle childhood (n=8 events from five vaccine beneficiaries)
21	Tdap vaccine (10 years)	8	5	77	6.50	Fever (7), pain at injection site (2), swelling at injection site (2)
Early adolescents (n=5 events reported from four vaccine beneficiaries)
22	Tdap vaccine (16 years)	4	3	16	18.75	Fever (3), swelling at injection site (1)
23	HPV vaccine	1	1	12	8.33	Pain at injection site (1)

Figure 1. Systematic representation of study flow. AEFI, adverse event following immunisation.

Overall, a higher AEFI incidence rate was observed among boys, at 10.91% (n=392/3594) compared with girls at 9.76% (n=303/3105) in the study population. Infants reported the highest AEFI incidence rate at 15.32% (n=459/2996), while the lowest occurrence was seen in the middle-childhood group (2.03%, n=5/246). The detailed AEFI reporting rates and their distribution across different age groups are depicted in figure 2a. The study found that 13.49% (n=904/6699) of vaccine beneficiaries followed a delayed immunisation timeline, and 19.91% (n=180/904) of these children reported an AEFI during the study period. The highest number of delayed immunisation was observed in the middle-childhood age group (n=223), predominantly involving the Diphtheria,Pertussis,and Tetanus booster vaccines. Although delayed immunisation was most frequent in this age group, the highest AEFI incidence among children with delayed immunisation was observed in early adolescents (100%), where all children in this category reported an AEFI. In contrast, the middle-childhood group demonstrated the lowest AEFI incidence among delayed immunisers. Overall, except for the middle-childhood group, all other age groups showed a higher incidence of AEFIs among children with a delayed immunisation compared with those immunised on schedule. A detailed comparison of AEFI incidence between delayed and non-delayed immunisation timelines across age groups is presented in figure 2b.

Figure 2. The population characteristics of the study population. AEFI, adverse event following immunisation; DM, diabetic mellitus. RT, Respiratory Tract; GIT, Gastro Intestinal Tract.

The study population included 0.16% (n=11) of children with congenital disorders (congenital hypothyroidism and HIV) and 1.63% (n=109) with chronic disorders such as thalassaemia, heart disorders, asthma and type 1 diabetes mellitus. Additionally, 2.00% (n=134/6699) had a history of recent hospitalisation, and 8.88% (n=595/6699) had nutritional deficiencies (stunting, wasting, underweight and low birth weight). The incidence of AEFI among children with any of the aforementioned health conditions was estimated to be 40.28% (n=342/849), whereas among children without these conditions, the AEFI incidence was substantially lower at 6.03% (n=353/5850). A detailed distribution of comorbid conditions among the vaccine beneficiaries and the incidence of AEFIs is shown in figure 2c. The majority of vaccine beneficiaries were from the lower middle class of society (49.08%, n=3288), with the least enrolment observed in the lower (4.91%) and upper (5.24%) classes. An increased incidence of AEFI was observed in the lower middle class (4.72%), while fewer incidences were seen in both the lower (0.34%) and upper (0.85%) classes. The incidence of AEFIs across different socioeconomic categories is detailed in figure 2d.

Overall, the most commonly reported event among the study population was fever (72%, n=560), followed by pain at injection site (n=106, 13.62%) and swelling at injection site (5.27%, n=41). According to the system organ classification, 89.72% (n=698) of the reported events were categorised under ‘general disorders and administration site conditions’, followed by 5.40% (n=42) who were skin and subcutaneous tissue disorders. Among these events, 56.64% (n=464) were reported in the infant age group, and 49.36% (n=384) occurred following Pentavac (diphtheria, tetanus, pertussis, hepatitis B, Hib) vaccination among infants. The categorisation of reported AEFIs based on the system organ classification is depicted in figure 3a.

Figure 3. The trends and types of reported AEFIs among the study population. (a) Categorising of AEFIs based on system organ classification. (b) Types of AEFIs based on WHO causality assessment algorithm. AEFI, adverse event following immunisation; MCV, meningitis conjugate vaccine; MMR, Measles, Mumps, and Rubella; OPV, oral polio vaccine; PCV, pneumococcal conjugate vaccine; TCV, typhoid conjugate vaccine; HPV, Human Papillomavirus Vaccine.

Among the total number of AEFIs reported, 95.37% (n=742) were identified as solicited AEs while 4.65% (n=36) of the events were unsolicited events as per the respective vaccine fact sheet and case definitions. The majority of solicited events (51.93%) and unsolicited events (1.30%) were reported among Pentavac vaccine beneficiaries. However, HPV vaccine beneficiaries reported the fewest AEFIs (0.13%) among the study population. Unsolicited AEFIs were infrequently reported across all vaccines, and no vaccine demonstrated a disproportionately higher occurrence of unexpected events. The causality assessment of each reported AEFI was determined based on the WHO’s causality assessment algorithm. The majority of the reported events (94.86%, n=738) were classified under category A: consistent with causal association to immunisation. Additionally, 4.76% (n=37) were categorised as category C: inconsistent with causal association to immunisation, and 0.39% (n=3) of them were category B: indeterminate events. The events reported under category A were further subdivided according to the type of reaction. Of these, 90.62% (n=705) were identified as ‘vaccine product-related reactions’ (category A1), 2.70% (n=21) of the reactions were ‘immunisation error-related reactions’ (category A3) and 1.54% (n=12) of the events were ‘immunisation anxiety-related reactions’ (category A4). The causality assessment of each reported event following respective vaccination was depicted in figure 3b. The majority of unsolicited events were reported under the System Organ Classification ‘respiratory thoracic and mediastinal disorders’ (n=13), while solicited events were reported under ‘general disorders and administration site conditions’ (n=698). The causality assessment of 90.62% (n=704) of the solicited events reported in the study was determined to be vaccine product-related reactions (88.95%) and immunisation error-related reactions (1.67%). However, 4.68% (n=36) of unsolicited events were assessed as 0.39% (n=3) indeterminate and 3.98% (n=31) as coincidental events.

Overall, the influenza vaccine reported highest incidence of AEFIs at the study site, accounting for 19.51% followed closely by the DTP booster 2 with 19.34% and Pentavac with 19.21%. In contrast, the lowest incidence was observed with the oral polio vaccine at 0.13%, followed by the varicella vaccine at 1.20%. The incidence rates of AEFIs across different age groups were as follows: 15.32% among infants (n=459/2996), 12.5% in early adolescents (n=4/32), 10.70% among toddlers (n=75/701), 9.15% in early childhood (n=41/448), 4.88% in neonates (n=111/2276) and 2.03% in middle childhood (n=5/246). The distribution of reported AEFIs and the incidence rate of each childhood vaccine are represented in table 2.

Predictors associated with the development of adverse event following immunisation

Various parameters were analysed to identify potential predictors of AEFIs in the study population. Gender was not a significant predictor for developing AEFI (p value 0.63), whereas different age groups were significantly associated with the reporting of AEFIs. The study group with the lowest incidence rate of AEFIs served as the reference for comparison (refer to table 3). Although all the paediatric age groups showed a significant association with AEFIs at the study site, infants (χ² value: 32.73), toddlers (17.68) and early childhood (13.05) were identified as strong age predictors for reported AEFIs.

Table 3. Various parameters analysed for identifying AEFI predictors in the study population.

Sl no.	Predictors	No. of vaccine beneficiaries with AEFI	No. of vaccine beneficiaries without any AEFIs	Incidence rate (%)	χ2 value	P value	Relative Risk
1	Gender
	Boys	391	3211	10.86	1.93	0.16	1.11
	Girls	304	2793	9.82			1 (ref)
2	Age group
	Newborns	111	2165	4.88	4.09	0.004	0.42
	Infants	459	2537	18.09	32.73	<0.0001	0.13
	Toddler	75	626	11.98	17.68	0.0002	0.19
	Early childhood	41	407	9.15	13.05	0.0003	0.22
	Middle childhood	5	241	2.03	*Ref	*Ref	1 ref
	Early adolescents	4	28	14.28	9.90	0.001	0.16
3	Type of vaccine
	Pentavac	364	1531	19.21	406.82	<0.0001	0.05
	DTP boosters	84	416	16.8	275.78	<0.0001	0.06
	Hepatitis B (birth dose)	83	2193	3.65	34.77	<0.0001	0.28
	Influenza vaccine	32	669	4.56	37.34	0.00001	0.22
	PCV	27	1678	1.58	2.58	0.108	0.64
	MMR vaccine	27	876	2.99	16.36	0.00005	0.34
	BCG	23	2253	1.01	*Ref	*Ref	1 ref
	TCV vaccine	22	370	5.95	42.70	<0.0001	0.18
	Hepatitis A	12	453	2.56	7.55	0.005	0.39
	Tdap vaccine	8	93	8.60	35.87	0.0001	0.13
4	Concomitant vaccine administration
	Single vaccine	144	1352	9.63	1.16	0.28	1 (ref)
	More than one vaccine	551	4652	10.59			0.91
5	Immunisation timeline
	Delayed	180	1038	14.78	31.05	<0.0001	1.57
	On time	515	4966	9.40			1(ref)
6	Comorbid status
	Underweight	133	106	55.65	52.79	<0.0001	0.43
	Stunted	12	34	26.08	0.08	0.76	0.92
	Wasted	29	15	95.48	31.48	<0.0001	0.37
	Lower birth weight (≤2 kg newborns)	64	202	24.06	*Ref	*Ref	1 (ref)
	Congenital disorders	4	7	57.14	0.86	0.35	0.66
	Chronic disorders	44	65	67.69	10.03	0.001	0.6
	Recent hospitalisation	56	78	71.79	13.34	0.0002	0.58
7	Socioeconomic class
	Upper class	57	294	16.24	13.99	0.0001	0.43
	Upper middle class	191	1039	15.53	15.97	0.0006	0.45
	Lower middle class	316	2972	9.61	2.41	0.12	0.73
	Upper lower class	108	1393	7.20	0.017	0.89	0.97
	Lower class	23	306	6.99	*Ref	*Ref	1 (ref)

^*Reference: group with least incidence rate. Result significance if p value is less than 0.05.

^†P-values less than 0.05 were considered statistically significant and are presented in bold, indicating significant predictors for the development of AEFI within the study population.

AEFI, adverse event following immunisation.

Among the top 10 vaccines reported highest number of AEFIs, children immunised with Pentavac (χ square value: 406.82) and DTP booster dose (275.78) were identified as having a higher likelihood of reporting an AEFI compared with other childhood vaccinations. However, several other childhood vaccines were also significantly associated (p value less than 0.05) with the reporting of an AEFI (detailed in table 3). No significant association was observed among children receiving single or multiple vaccines at a single visit. However, the development of AEFIs among children with delayed immunisation timelines showed significant (p value: 0.0001) association among the study population.

Children with comorbid conditions were analysed as potential predictors for reported AEFIs. Among these, children with deficient nutritional status, particularly those who were underweight and wasted, were found to be significant predictors of AEFIs. Children with chronic disorders and a history of recent hospitalisation were also found to have a higher likelihood of reporting an AEFI. Additionally, the socioeconomic status of the vaccine beneficiaries was significantly related to the development of AEFIs, with children from upper and upper middle-class families more likely to report an AEFI compared to those from other socioeconomic classes. A detailed representation of various predictors and their significance is shown in table 3.

Discussion

Overall, the study reported an AEFI incidence rate of 10.37%, with 2.70% being serious reactions. Similar single-centred studies conducted in different regions across the country reported an AEFI incidence ranging from 33% to 40%, with no or very few serious reactions.^{20 21} This variation is unlikely to be explained solely by differences in enrolment size; rather, it may reflect surveillance-related selection bias, whereby passive reporting systems preferentially capture symptomatic cases and thereby inflating AEFI incidence estimates. No AEFIs reported at the study site were categorised as unclassifiable, likely due to the active vaccine safety surveillance implemented. All the relevant information required for the causality assessment of each reported event was collected through follow-up calls, reducing misunderstandings and under-reporting. However, a 9.97% dropout rate was observed in the study, potentially due to reasons such as migration and altered preferences for immunisation centres based on convenience. Previous studies have reported that although active surveillance enhances the quality and completeness of AEFI reporting, its large-scale implementation is often limited by resource and logistical constraints.²² In India, AEFI reporting largely relies on passive surveillance through pharmacovigilance channels and is not routinely integrated into centralised digital immunisation databases.²³ However, platforms such as UWIN²⁴ function primarily as vaccination registries; they can be used as a foundation for integrating active surveillance components. Scalable alternatives to telephone follow-up, including SMS-based reporting and automated mobile or interactive voice response systems, could enable nationwide active AEFI monitoring with reduced time burden and maintained data quality.

All passively reported (2.31%, n=18) events at the study site involved children who were admitted to the hospital after being immunised at vaccination centres outside the study site and were therefore not included in the active surveillance cohort. The active and passive surveillance populations were distinct, with no overlap, minimising the likelihood that active follow-up discouraged passive reporting among parents or healthcare providers, thereby reducing reporting bias. Although most serious AEFIs requiring hospitalisation were identified through passive surveillance, this does not diminish the value of active surveillance, as no serious events occurred among vaccine beneficiaries immunised at the study site during the study period. Active surveillance primarily facilitated the identification of mild to moderate events that were otherwise under-reported. Experience from COVID-19 vaccine surveillance at the same study site further demonstrated that active monitoring enabled detection of serious and rare events, with higher AEFI reporting compared with national passive systems.¹⁶ Given its resource-intensive nature, nation-wide implementation of active surveillance may not be feasible; however, targeted use at sentinel hospitals or district-level referral centres may provide an effective and practical approach for strengthening vaccine safety monitoring.

The majority of the reactions reported at the study site were solicited events, with fever being the most common (26.3 events per 1000 doses), followed by pain at injection site (4.96 events per 1000 doses). However, 2.31% (n=18/778) of the reactions were serious reactions and required hospitalisation. Febrile convulsions following DTP vaccination (Pentavac or DTP booster doses) were considered solicited serious reactions, with an incidence rate of 4.17 events per 1000 doses. Similar studies have reported an increased risk of vaccine-proximate convulsions in children following DTP and MMR vaccination.^{25 26} In contrast, few studies suggest that children under 2 years of age are inherently prone to seizure due to developmental factors, febrile illnesses, intercurrent infections and underlying neurological conditions.^{27 28} Seizures observed during the surveillance period were therefore interpreted in the context of the expected background frequency of both febrile and non-febrile seizures in early childhood, and temporal association alone was not considered sufficient to imply a vaccine-associated risk. However, such reactions following vaccination have been shown to create high parental unacceptability and unwarranted concerns.²² All of the children who reported febrile convulsion at the study site had received DTP vaccines containing whole-cell pertussis (wP) as a common ingredient. Previous studies have observed increased reactogenicity associated with (wP in DTP vaccination compared with the acellular pertussis.^{28 29} On the other hand, population-based studies have shown that wP vaccines provided better disease protection and longer duration of immunity.^{30 31} Therefore, wP vaccines were introduced to UIP considering that the benefits outweigh the risks of higher AEs.³²

The study reported an incidence of 1.67 unsolicited events per 1000 doses (n=36) among the study population. Of these, 63.89% (n=23) were identified as coincidental to the vaccination. One serious unsolicited reaction involved the death of a two-and-a-half-month-old child 3 days after receiving the Pentavac vaccination, which was later determined to be coincidental to vaccination. Gold et al explain that consistent causal reactions resulting in death are rare, and these coincidental deaths may provide insights into the country’s infant mortality rates.³³ Although an increased risk of AEFI was found among boys (RR: 1.11) than girls, gender was not found to be a significant predictor for reported AEFIs among the study population (p value: 0.16). This may be because boys were the highest vaccine users, and therefore, an increased incidence of AEFI was found in boys. Numerous studies worldwide have concluded a gender risk for females in developing AEFIs, regardless of the vaccines.^{34 35} Additionally, these studies determined that AEFIs were reported throughout the paediatric age spectrum. A registry-based observational study reported age- and vaccine-specific variations in AEFIs, rather than a uniform increase across paediatric age groups.³⁶ In the present study, a higher number of AEFIs were reported following Pentavac and DTP booster vaccines. Furthermore, a significant association with AEFI occurrence was observed for most childhood vaccines (apart from pneumococcal conjugate vaccine) used in the first 1 year of the child.

Our study did not find any association between the risk of AEFIs and children receiving more than one vaccine at a time. Most of the UIP childhood vaccines are recommended for concomitant administration in a single visit, while a few optional vaccines, such as typhoid conjugate vaccine and hepatitis A, are given as single doses. Evidence from a large real-world cohort study by Bauwens et al supports the overall safety of routine childhood vaccine coadministration, demonstrating no significant increase in AEs for most outcomes, although a modest rise in expected and generally mild reactions was observed with certain vaccine combinations.³⁷ In contrast, the findings reported by Goldbaum et al were specific influenza vaccination programmes, where the observed reduction in AEFIs was attributed to the use of less reactogenic vaccine brands and variations in concomitant vaccination and therefore may not be directly applicable to routine childhood immunisation schedules.³⁵ This study identified delayed immunisation as a potential predictor of reported AEFIs, with children receiving vaccines later having a 1.57 times higher risk of reporting an AEFI compared with those immunised on time. Evidence from a population-based study using national survey data from India indicates that delayed immunisation affects a substantial proportion of children, with low birth weight being the second most common reason.³⁸ In the present study, low birth weight, recent hospitalisation and chronic diseases were identified as reasons for the delayed timelines. Further studies are required to ascertain the various factors associated with these conditions, in combination with delayed immunisation, that lead to an increased AEFI reporting rate. The proportion of delayed immunisation observed in this hospital-based population appears broadly comparable with national data, suggesting that the study population may not differ markedly from the general population in terms of AEFI reporting. A parallel study conducted at the same study site provides a more detailed description of immunisation timelines and the factors contributing to delayed vaccination, offering further insight into this context.³⁹

In the present study population, nutritional deficiencies (underweight (p value: <0.0001) and stunted (<0.0001)) and other comorbid conditions (chronic disorders (0.001) and recent hospitalisation (0.0002)) were associated with an increased likelihood of reporting an AEFI. Lawan et al compared the pattern of AEFI development among well-nourished and malnourished children in North Western Nigeria and found that there was a 15.7% higher prevalence of AEFIs in malnourished infants, with more incidence of fever and local site reactions.⁴⁰ However, this study observation contradicts a similar study that found pre-existing diseases in children were not a significant factor in reported AEFIs.⁴¹ National survey data further indicate that nutritional deficiencies among Indian children remain widespread, particularly among socioeconomically disadvantaged populations, potentially affecting immune responses and increasing vulnerability to adverse health outcomes.⁴²

Additionally, this study identified a significant association between socioeconomic class and the development of AEFIs. Although the majority of the children were from lower middle families (44.36%), the significant association with AEFI occurrence was found among children from upper (p value: 0.0001) and upper middle families (0.0006). This may be attributed not only to the lower enrolment of children from upper segments but also to the higher response rates to the surveillance observed among these groups compared with the lower segments of the society. Higher dropout rates and lower AEFI reporting were observed among middle and lower sections of society. A cross-sectional study conducted in the country explains that awareness regarding AEFIs was significantly associated with mother’s age, educational status and employment status of the parents.³⁸,⁴³

Conclusion

Active vaccine safety surveillance plays a critical role in strengthening immunisation programmes, particularly in settings where passive reporting remains suboptimal. Through systematic follow-up and detailed event documentation, this study demonstrated that routine childhood vaccines under UIP exhibit a favourable safety profile, with the majority of reported AEs being expected, solicited reactions (34.83 events per 1000 doses) and infrequent serious events at a rate of 1.26 events per 1000 doses. The identification of key predictors of AEFIs underscores the importance of active surveillance in supporting causality assessment and in differentiating consistent vaccine-related reactions from inconsistent (coincidental) events. Importantly, these findings provide robust evidence to support transparent and effective communication of vaccine safety to caregivers and healthcare providers, particularly in addressing concerns arising from coincidental post-vaccination events. Integrating such surveillance approaches within immunisation services can enhance early signal detection, strengthen public trust and thereby reinforce confidence in childhood vaccination programmes and reaffirm that the benefits of immunisation far outweigh the risks posed by adverse events.

Limitation of the study

This study was limited to a single tertiary care setting and therefore cannot be generalised to large-scale or population-level contexts. The study did not assess the feasibility, cost or resource requirements needed for broader implementation of active surveillance systems. The findings primarily highlight the functional aspects and potential advantages of continuous vaccine safety monitoring within the study site. Further multicentre and operational research is required to evaluate the scalability, sustainability and practical integration of active surveillance approaches within national immunisation programmes.

Data availability statement

Data are available upon reasonable request.