Surveillance of adverse events following the introduction of 13-valent pneumococcal conjugate vaccine in infants, and comparison with adverse events following 7-valent pneumococcal conjugate vaccine, in Victoria, Australia

Abstract

The 13-valent pneumococcal vaccine (PCV13) replaced the 7-valent vaccine (PCV7) on the Australian National Immunization Program (NIP) in 2011. Post-marketing surveillance of adverse events following immunization (AEFI) is crucial for detecting potential safety signals and maintaining confidence in the NIP. This study describes all AEFI reported to Surveillance of Adverse Events following Vaccination in the Community (SAEFVIC), Melbourne, Australia, following both the primary series of PCV13 (children <7 months) and the catch-up dose (12 months–35months) in its first year of inclusion on the NIP. AEFI reporting rates per 100,000 doses of vaccine administered were compared for the PCV13 primary series and PCV7 primary series in the previous year. SAEFVIC received 229 reports describing 406 AEFI following PCV13 vaccine in the 12 months post introduction. There was no difference in the total number of AEFI cases reported between the vaccines but 7 AEFI categories were reported at a significantly higher rate following PCV13 compared with PCV7. No difference in reporting rate was observed for serious AEFI (p = 0.25). Post-hoc analysis of a further 12 months of PCV13 data revealed that all 7 AEFI categories that were initially reported at a significantly higher rate following PCV13 compared to PCV7 in the first 12 months post introduction, were no longer significantly increased in the 13–24 month period. The initial high reporting rate for several common AEFI post PCV13 compared to PCV7 may be explained by heightened awareness of the new vaccine. There were no safety signals detected for rare or serious AEFI that would require further investigation at this time.

Keywords:

• adverse events
• immunisation
• pneumococcal vaccines
• post-marketing surveillance
• vaccine safety

Introduction

The bacterium Streptococcus pneumoniae causes a wide range of disease including pneumonia, septicaemia, meningitis and otitis media with the burden of disease greatest in infants and the elderly.¹ Pneumococcal disease remains a major cause of morbidity and mortality in children worldwide, particularly in countries where non-vaccine serotypes are common, and is a leading cause of vaccine preventable deaths in children under 5.¹ Significant reductions in invasive pneumococcal disease have occurred in countries where pneumococcal conjugate vaccines are included in routine childhood immunization schedules.¹ The World Health Organization recommends that pneumococcal vaccines should be included in immunizations programs worldwide.²

In Australia, the 7-valent pneumococcal vaccine (PCV7) was introduced onto the National Immunization Program (NIP) for Indigenous children in 2001 and all children in 2005.³ The impact of routine immunization with PCV7 on rates of pneumococcal disease has been considerable. In non-Indigenous children aged less than 5 years, the notification rate for invasive pneumococcal disease decreased by 68%, from 52.5 to 16.8 cases per 100,000 population from 2002–2008,⁴ and was also associated with a significant reduction in hospitalization rates for pneumonia.⁵ Introduction of PCV7 was associated with an increased burden of disease caused by non-vaccine serotypes, with a 4-fold increase in disease caused by serotype 19A.⁴

In March 2010, the Australian Therapeutic Goods Administration licensed a 13-valent pneumococcal vaccine (PCV13) for the prevention of pneumococcal disease for use in children aged between 6 weeks and 6 y.⁶ PCV13 contains the 7 polysaccharide serotypes included in PCV7 (4, 6B, 9V, 14, 18C, 19F and 23F) with additional coverage of 6 serotypes (1, 3, 5, 6A, 7F, 19A), the selection of which was based on serotype replacement studies post PCV7 introduction.⁶ In July 2011, PCV13 replaced PCV7 on the Australian NIP. In addition, children aged from 12 to 35 months who had completed a primary pneumococcal vaccination course with PCV7 were eligible to receive a free single supplementary dose of PCV13.⁷ In Victoria, PCV13 is given as a 3-dose primary series at 2, 4 and 6 months of age concomitantly with other routine vaccines listed on the NIP (Table 1).⁸

Table 1. Concomitant vaccines given with the primary series or catch up pneumococcal vaccines in the Victorian Immunization schedule,⁸ over the study period

Age	Routine immunization	Vaccine brand
Birth	Hepatitis B (hepB)	H-B-Vax II Paediatric®
2 months	Hepatitis B, diphtheria, tetanus, acellular pertussis, Haemophilus influenza type b, inactivated poliomyelitis (hepB-DTPa-Hib-IPV)	Infanrix hexa®
	Pneumococal conjugate (PCV13)	Prevenar 13®
	Rotavirus	Rotateq®
4 months	Hepatitis B, diphtheria, tetanus, acellular pertussis, Haemophilus influenza type b, inactivated poliomyelitis (hepB-DTPa-Hib-IPV)	Infanrix hexa®
	Pneumococcal conjugate (PCV13)	Prevenar 13®
	Rotavirus	Rotateq®
6 months	Hepatitis B, diphtheria, tetanus, acellular pertussis, Haemophilus influenza type b, inactivated poliomyelitis (hepB-DTPa-Hib-IPV)	Infanrix hexa®
	Pneumococcal conjugate (PCV13)	Prevenar 13®
	Rotavirus	Rotateq®
12 months	Haemophilus influenza type b	Hiberix®
	Meningococcal C	Neisvac C®
	Measles, mumps and rubella (MMR)	Priorix®
18 months	Varicella	Varilrix®

Pre-licensure, the safety of PCV13 was assessed in 13 clinical trials where the vaccine was administered with concomitant vaccines according to country-specific vaccination schedules. In the 3 primary US safety studies, the percentage incidence and severity of both local and systemic reactions were similar between PCV7 and PCV13 recipients.^9,10,11 Across all studies, in general there were no major differences demonstrated in serious or non-serious adverse events following immunization (AEFI), suggesting a comparable safety profile for the 2 vaccines.^6,12

The safety of vaccines included on the Australian NIP is of significant public health interest and is critical in maintaining confidence in the vaccine program. In Victoria, an enhanced passive surveillance system called Surveillance of Adverse Events Following Vaccination in the Community (SAEFVIC) has been in operation since 2007.¹³ This passive surveillance system has improved AEFI reporting rates in Victoria from 2.6 per 100,000 in 2003 to 14.6 per 100,000 in 2012.¹⁴ Passive or post-marketing surveillance systems, such as SAEFVIC, play a key role in signal detection of potential AEFI to guide further investigation if warranted. Rates of AEFI detected through passive surveillance systems are likely to be lower than the true rate, due to under-reporting, but may still flag important AEFI for follow up. AEFI detected following the introduction of a new vaccine into a population may be different to those detected in pre-implementation controlled clinical trials, as these are rarely powered sufficiently to detect rare AEFI.

This study aimed to describe all AEFI and AEFI reporting rates from SAEFVIC surveillance data following the PCV13 primary series in its first year of inclusion on the NIP, compared to those reported following the PCV7 primary series in the previous year. As SAEFVIC was only established in 2007, AEFI data from the first year post introduction of PCV7 was unavailable to be used as a comparator.

Results

PCV7 primary series (Cohort 1)

Following PCV7, there were 221 reports to SAEFVIC, describing 309 adverse events. PCV7 was co-administered with RotaTeq® and Infanrix hexa® in 92.8% (n = 205) of reports, administered with other vaccines in various combinations in 6.3% of reports (n = 14) and administered alone in 0.9% (n = 2) of reports. The most frequently reported AEFI were rash (21.1 per 100,000 doses), diarrhea (16.9 per 100,000 doses), generalized allergic reaction (15.3 per 100,000 doses) and fever (14.8 per 100,000 doses) (Table 2).

Table 2. AEFI reported frequencies and rates per 100,000 doses administered for each cohort and comparison data

	Cohort 1		Cohort 2		Cohort 3		Cohort 4
	PCV7		PCV13 0–12 months		PCV13 13–24 months		PCV13 catch-up		Comparison cohort	Comparison cohort
	n = 221		n = 229		n = 176		n = 52		1 to cohort 2	1 to cohort 3
AEFI type	No (%)	Reporting rate per 100,000 doses	No (%)	Reporting rate per 100,000 doses	No (%)	Reporting rate per 100,000 doses	No (%)	Reporting rate per 100,000 doses	IRR (CI)	IRR (CI)
		n = 189,381		n = 177,984		n = 195,646		n= 76,708
Total number of case reports	221	116.7	229	128.7	176	90.0	52	67.8	0.91 (0.75–1.10)	1.30 (1.06–1.59)
Total number of AEFI	309	163.2	406	228.1	292	149.2	92	119.9	1.40 (1.20–1.63)	0.91 (0.77–1.07)
Fever (all)	28 (9.1)	14.8	55 (13.5)	30.9	39 (13.4)	19.9	13(14.1)	16.9	2.09 (1.30–3.42)	1.35 (0.81–2.27)
Not recorded	12 (42.9)		19 (34.5)		22 (56.4)		4 (30.8)
≥38°C to <40°C	15 (53.6)		32 (58.2)		15 (38.5)		8 (61.5)
≥ 40°C	1 (3.6)		4 (7.3)		2 (5.1)		1 (7.7)
Diarrhea	32 (10.4)	16.9	53 (13.1)	29.8	35 (12.0)	17.9	3 (3.3)	3.9	1.76 (1.12–2.82)	1.06 (0.64–1.77)
Irritability	19 (6.1)	10.0	50 (12.3)	28.1	30 (10.3)	15.3	3 (3.3)	3.9	2.80 (1.62–5.03)	1.53 (0.83–2.87)
Vomiting	25 (8.1)	13.2	41 (10.1)	23.0	23 (7.9)	11.8	7 (7.6)	9.1	1.75 (1.04–2.99)	0.89 (0.48–1.63)
Rash	40 (12.9)	21.1	37 (9.1)	20.8	30 (10.3)	-	9 (9.8)	11.7	0.98 (0.61–1.58)	—
Abdominal pain	6 (1.9)	3.2	20 (4.9)	11.2	2 (0.7)	1.0	0 (0.0)	0.0	3.55 (1.37–10.79)	0.32 (0.03–1.80)
Lethargy	10 (3.2)	5.3	20 (4.9)	11.2	12 (4.1)	6.1	6 (6.5)	7.8	2.13 (0.95–5.09)	1.16 (0.46–3.00)
Skin	26 (8.4)	13.7	195 (4.7)	10.7	18 (6.2)	—	11 (12.0)	14.3	0.78 (0.41–1.46)	—
Allergic reaction (generalized)	29 (9.4)	15.3	17 (4.2)	9.6	17 (5.8)	—	6 (6.5)	7.8	0.62 (0.32–1.17)	—
Hypotonia	11 (3.6)	5.8	11 (2.7)	6.2	6 (2.1)	—	1 (1.1)	1.3	1.06 (0.42–2.71)	—
Crying (persistent >3hrs)	9 (2.9)	4.8	10 (2.5)	5.6	6 (2.1)	—	0 (0.0)	0.0	1.18 (0.43–3.29)	—
Gastrointestinal	3 (1.0)	1.6	10 (2.5)	5.6	5 (1.7)	2.6	0 (0.0)	0.0	3.55 (0.91–20.06)	1.61 (0.31–10.38)
Injection site reaction (minor)	6 (1.9)	3.2	8 (2.0)	4.5	18 (6.2)	—	18 (19.6)	23.5	1.42 (0.43–4.96)	—
Respiratory	9 (2.9)	4.8	8 (2.0)	4.5	10 (3.4)	—	0 (0.0)	0.0	0.95 (0.32–2.76)	—
HHE	21 (6.8)	11.1	8 (2.0)	4.5	1 (0.3)	0.5	0 (0.0)	0.0	0.41 (0.16–0.95)	0.05 (0.00–0.29)
Intussusception	6 (1.9)	3.2	5 (1.2)	2.8	4 (1.4)	—	0 (0.0)	0.0	0.89 (0.21–3.49)	—
Seizure(afebrile)	4 (1.3)	2.1	5 (1.2)	2.8	8 (2.7)	—	4 (4.3)	5.2	1.33 (0.29–6.70)	—
Seizure (febrile)	1 (0.3)	0.5	5 (1.2)	2.8	5 (1.7)	—	2 (2.2)	2.6	5.32 (0.59–251.63)	—
Apnoea	12 (3.9)	6.3	4 (1.0)	2.2	5 (1.7)	—	1 (1.1)	1.3	0.36 (0.08–1.17)	—
Cardiovascular	1 (0.3)	0.5	4 (1.0)	2.2	5 (1.7)	—	1 (1.1)	1.3	4.26 (0.42–209.60)	—
Neurological	6 (1.9)	3.2	3 (0.7)	1.7	4 (1.4)	—	1 (1.1)	1.3	0.53 (0.09–2.49)	—
Injection site reaction (severe)	1 (0.3)	0.5	3 (0.7)	1.7	2 (0.7)	—	2 (2.2)	2.6	3.19 (0.26–167.58)	—
Presumed infection	3 (1.0)	1.6	3 (0.7)	1.7	5 (1.7)	—	1 (1.1)	1.3	1.06 (0.14–7.94)	—
Nodule at injection site	1 (0.3)	0.5	2 (0.5)	1.1	0 (0.0)	—	0 (0.0)	0.0	2.13 (0.11–125.55)	—
Metabolic	0 (0.0)	0.0	2 (0.5)	1.1	1 (0.3)	—	1 (1.1)	1.3	—	—
Anaphylaxis	0 (0.0)	0.0	1 (0.2)	0.6	0 (0.0)	—	0 (0.0)	0.0	—	—
Meningitis	0 (0.0)	0.0	1 (0.2)	0.6	0 (0.0)	—	0 (0.0)	0.0	—	—
Haematological	0 (0.0)	0.0	1 (0.2)	0.6	1 (0.3)	—	0 (0.0)	0.0	—	—
Death	0 (0.0)	0.0	0 (0.0)	0.0	0 (0.0)	—	1 (1.1)	1.3

^a there was 1 or more AEFI associated with each case report;

^b significantly increased reporting rate in PCV7 cohort;

* significant p value (p < 0.05).

PCV13 primary series (Cohort 2)

Following PCV13, there were 229 reports to SAEFVIC, describing 406 adverse events. In most reports (93.4%, n = 214), PCV13 had been co-administered with RotaTeq® and Infanrix hexa® as per the NIP schedule (Table 1). PCV13 had been administered with other vaccines in various combinations in 5.7% (n = 13) of reports and administered alone in 0.9% (n = 2) of reports. The most frequently reported AEFI were fever (30.9 per 100,000 doses), diarrhea (29.8 per 100,000 doses), irritability (28.1 per 100,000 doses) and vomiting (23 per 100,000 doses) (Table 2).

Comparison of PCV7 vs PCV13 (Cohort 1 vs Cohort 2) and post-hoc analysis comparison

There was no difference in the total case reporting rate for Cohort 2 (PCV13)(128.7 per 100,000 doses) compared with Cohort 1(PCV7) (116.7 per 100,000 doses)(IRR 0.91, p = 0.30). However, the total AEFI reporting rate was higher in Cohort 2 (228.1 per 100,000 doses) compared with Cohort 1 (163.2 per 100,000 doses) (IRR 1.4, p < 0.001). Seven AEFI categories were reported at a significantly higher rate in Cohort 2 compared to Cohort 1 (Table 2). Conversely, HHE was reported at a significantly higher rate in Cohort 1 compared to Cohort 2 (IRR 0.41, p = 0.025).

The total case reporting rate in the post-hoc analysis, Cohort 3,(90.0 per 100,000 doses) was much lower than in Cohort 1 (PCV7)(IRR 1.30, p = 0.01). Similarly, no significant difference was seen between the total AEFI reporting rate for Cohort 3 (149.2 per 100,000 doses) compared to Cohort 1(163.2 per 100,000 doses) (IRR 0.91, p = 0.276) or for the 7 AEFI categories that were reported more frequently in the initial 12 months (Table 2). The reporting rate for HHE remained consistently lower in Cohort 3 compared with Cohort 1 (IRR 0.05, p < 0.001).

PCV13 catch-up cohort (Cohort 4)

In Cohort 4 there were 52 reports to SAEFVIC (total case reporting rate 67.8 per 100,000 doses), describing 92 adverse events. In most reports, PCV13 had been co-administered with other vaccines in various combinations (55.8%, n = 29) and administered as a single vaccine in 44.2% (n = 23) of reports. The most frequently reported AEFI were injection site reaction (minor) (23.5 per 100,000 doses), fever (16.9 per 100,000 doses), skin reactions (14.3 per 100,000 doses) and rash (11.7 per 100,000 doses) (Table 2).

Serious AEFI

There was no difference in total case reporting rates for serious AEFI between Cohort 1 (n = 37, 19.54 per 100,000 doses) and Cohort 2 (n = 45, 25.28 per 100,000 doses) (IRR 0.77, p = 0.25). Most serious reports in both groups were due to hospitalizations. In Cohort 2 these were due to fever, either on its own or with a combination of rash or infective symptoms (n = 8); intussusception (n = 2); HHE (n = 2) and; apnoea (n = 2). In comparison, apnoea was the most common reason for hospitalization in Cohort 1 (n = 9) followed by allergic reaction (generalized) (n = 2).

There was one report of anaphylaxis in Cohort 2 (PCV13). In both Cohorts, seizures were the most frequently reported medically significant AEFI where the recipient was not hospitalized (Cohort 2 n = 7, Cohort 1 n = 4). Other medically significant AEFI were one case of apnoea and one case of supra-ventricular tachycardia that occurred in a child with known Wolf-Parkinson-White Syndrome in Cohort 2. In Cohort 1 there was one case of Horner's syndrome and 3 cases of apnoea.

In Cohort 4 the total cases reporting rate for serious AEFI was 11.7 per 100,000 doses (n = 9). Most were classified as medically important events (n = 6) all of which were afebrile seizures. There was one death reported in a child who had received PCV13, Priorix®, Hiberix® and NeisVac-C® 2 weeks previously. The cause of death determined by the coroner was cardiac arrest from myocarditis. A causality assessment of this case was made by SAEFVIC staff using the WHO causality approach.¹⁵ The death was determined to be unrelated to vaccine administration.

Discussion

This study presents an evaluation of AEFI reported to a passive surveillance system in Victoria, Australia, following the first year of inclusion of PCV13 vaccine on the NIP and provides a comparison between total case reporting rates and AEFI reporting rates for PCV13 compared to the previously administered PCV7 vaccine.

Overall, the PCV13 primary series AEFI profile was similar to the PCV7 profile. Most reactions reported following PCV13 were either consistent with well-recognized vaccine side effects or with common symptoms of illness expected in infants that may be unrelated to vaccination. In the first year of PCV13 on the NIP, there was no significant difference in the rate of total case reports when compared with PCV7. However, the PCV13 cohort cases did report more AEFI reactions than were reported by the PCV7 cases. In addition, an increased reporting rate was initially detected for 7 specific AEFI categories. This raised the question of whether PCV13 was more reactogenic than PCV7 prompting a post-hoc analysis of a subsequent 12 months of PCV13 data (Cohort 3).

In PCV13s second year on the NIP (Cohort 3) there were significantly fewer AEFI case reports following PCV13 compared with PCV7 (Cohort 1) and no difference was demonstrated in total individual AEFI rates. Furthermore, the increased reporting of the 7 individual AEFI categories was no longer evident. This analysis shows that in the first year, reporters were just as likely to report AEFI as with the PCV7 vaccine, however, more individual AEFI reactions were reported for each case. This difference was not sustained in the second year suggesting that infants who experienced AEFI following PCV13 were not likely to experience more AEFI than with the PCV7 vaccine. One possible explanation for the initial difference is heightened reporting sensitivity post introduction of a new vaccine. In the first year reporters may have been more likely to report more symptoms for their patient/child due to awareness of PCV13 as a new vaccine. This is an important finding for future vaccine introduction in order to avoid false signal generation and a caution not to focus only on individual numbers of AEFI but to examine the total number of cases as well. Overall, these findings indicate that PCV13 is not more reactogenic than PCV7.

In this study we are not comparing passive surveillance data to active surveillance data. However, active surveillance data is usually more complete and less susceptible to case under ascertainment and thus more stable than passive surveillance data. As such, initial passive data collected after the introduction of a new vaccine may more closely mirror actively collected data as there is usually heightened awareness of a new vaccine and thus higher case ascertainment. However, this study did not test this hypothesis. If possible, future studies comparing AEFI rates obtained from passive and active surveillance systems after the introduction of a new vaccine would address this question.

Australia-wide data also showed an increase in AEFI reporting rates in the first 6 months following PCV13 implementation (48 per 100,000 doses) compared with the rate seen in the previous year for PCV7 (23 per 100,000 doses) in children under 7 y¹⁶ This elevated rate showed a decline in the first 6 months of 2012 to 35 per 100,000 doses.¹⁷ Our data showed a similar trend in reporting rates, however the reporting rates were higher, which was most likely due to state-based differences in passive reporting systems.

Similar rises in reporting rates following vaccine introduction have been demonstrated in the literature, a phenomenon known as the Weber effect.¹⁸ This describes an increased rate of adverse event reporting that occurs in the first 2–3 y of drug marketing followed by a decline and has been demonstrated with a variety of drugs, including vaccines.^18,19 This effect was demonstrated following introduction of PCV7 in the US where an increase in passively reported events was seen from 2000 to 2001 followed by a steady decrease and stabilization of reported events from 2002–2007.²⁰ In Australia, a quarterly increase followed by stabilization in reporting rates was demonstrated following introduction of both meningococcal C vaccine (2003) and PCV7 (2005) on the Australian NIP.²¹ More recently, AEFI reporting rates following quadrivalent human papillomavirus vaccine (HPV) introduction showed an impressive increase both from the US Vaccine Adverse Event Reporting System²² and in Australian surveillance data,²³ and showed significant decline the following year.²¹ HPV vaccine received high media coverage and included an episode of mass psychogenic illness occurring in Melbourne within weeks of program commencement.²⁴ This highlights the impact that heightened public awareness can have on reporting rates collected by passive reporting systems.²⁵

Frequently reported AEFI following primary series PCV13 in our study were similar to the most frequently solicited reactions reported in pre-licensure studies.^9,10,11 Published post-licensure experience with PCV13 is limited, consisting of one study that analyzed AEFI following PCV13 administration in children aged 1 to 24 months in the 2 y following inclusion on the US NIP.²⁶ The study focused only on pre-defined AEFI categories based on those rare and serious AEFI detected during the PCV7 post-licensure surveillance phase. A comparison was made to a historical population of children who received PCV7 in the 2 y prior to PCV13 introduction. No signal of excess risk for febrile seizures, urticaria, encephalopathy, asthma, anaphylaxis or thrombocytopenia was found, compared with PCV7.²⁶ A non-significant fold2- increased relative risk for Kawasaki disease was found.²⁶

HHE were reported less frequently following PCV13 than following PCV7 and this effect was sustained throughout the 24-month study period. HHE reactions are traditionally associated with pertussis containing vaccines but have also been reported uncommonly following other vaccines.²⁷ There is no reported association of HHE with PCV7. The difference in HHE rates between PCV vaccines noted in our study was unexpected. Due to our small sample size we were unable to draw any definite conclusions regarding a true difference between the vaccines with respect to HHE occurrence and further research would be necessary to explore this difference. In a majority of reports (>90 % for each primary series cohort), PCV vaccines were administered concomitantly with vaccines listed on the Australian NIP. This was consistent across the cohorts and over time and therefore presumed to not be a factor in accounting for potential differences. In a small number of reports, PCV vaccines were administered with other unscheduled vaccines but this would be unlikely to account for any differences in AEFI reporting between the PCV7 and PCV13 cohorts.

In the PCV13 catch-up cohort (Cohort 4), AEFI reporting rates were generally low apart from injection site reactions (ISR). This was consistent with data from PCV13 clinical studies, which demonstrated an increase in ISR and systemic reactions following the fourth dose of either PCV7 or PCV13 in the second year of life compared with infant cohorts.^10,11 As the Victorian catch-up program was unique to the introduction of the new PCV13 vaccine there was no comparison PCV7 cohort available for this older age group.

There were no specific rare or serious AEFI signals detected in this study that require further investigation. When introduced into the community, PCV13 appears to be a safe, well-tolerated vaccination.

Our data came from an enhanced passive reporting system, which has several inherent limitations. Passive reporting systems rely on self-initiated notifications and will under-report AEFI.²⁸ as well as being subject to inconsistency in the quality and completeness of data. Therefore, an estimation of the absolute or true incidence of AEFI cannot be made. However, there is currently no other way to gain surveillance data to evaluate vaccines on the NIP. Causality assessment was not undertaken for reported AEFI in this study so no inference can be made regarding potential causal associations. Passive reporting systems are susceptible to influential factors such as media attention or heightened awareness of a new vaccine program that can result in increased reporting rates following commencement of a new vaccine. However, this reporting bias also allows detection of potential safety signals of rare or previously unknown AEFI, which can then be further investigated. The SAEFVIC surveillance system targets healthcare provider reports from doctors and nurses, as well as parents, but is subject to changes in reporter trends.²⁹

A further limitation of this study relates to the comparison PCV7 data. SAEFVIC was established in 2007, several years after the commencement of the PCV7 program and therefore no data from the first year of this program was available to use as a comparator. ACIR data were used to calculate AEFI reporting rates per 100,000 doses administered in Victoria. While the recording of childhood vaccines on ACIR is very high, there is recognized under-reporting,³⁰ which may have resulted in over estimation of reporting rates. However this would be assumed to be consistent across both vaccines.

Conclusion

This study presents a description of AEFI reported to SAEFVIC, Victoria, Australia following vaccination with the newly introduced PCV13 vaccine in 2011. Comparison data revealed that several AEFI were reported at a significantly higher rate following PCV13 compared with PCV7 in the previous year, although the overall case reporting rate was not elevated. This difference may be explained by heightened awareness of the new vaccine as supported by the decline in reporting rates seen in the second 12 months. There were no specific safety signals detected for rare or serious AEFI that would require further investigation at this time. Our findings support the favorable safety profile of PCV13, however ongoing post-marketing surveillance remains essential to detect any rare but potentially important AEFI.

Methods

Ethics approval for the study was obtained from the Royal Children's Hospital Human Ethics Committee (DA 017-2013-01).

Saefvic

SAEFVIC maintains a database of all reported AEFI for both children and adults and is integrated with a clinical support service whereby all persons reporting a serious AEFI or who have concerns about future vaccination are offered a medical review.¹³ Reports made to SAEFVIC are recorded using a standardized proforma that includes a set of pre-defined criteria as well as information regarding the circumstances and description of the AEFI that occurred. AEFI are categorized post medical review by SAEFVIC nurses and clinicians according to standardized definitions.¹³

Data collection

De-identified data were extracted from the SAEFVIC database by date of vaccination and PCV vaccine type for the 4 study cohorts; PCV7 primary series (Cohort 1), PCV13 primary series 0–12 months post introduction (Cohort 2), PCV13 primary series 13–24 months post introduction (cohort 3)and PCV13 catch-up cohort (Cohort 4) (Table 3). Information extracted included vaccinee demographic details, reporter professional type, vaccines received and adverse event details. Most reports came from nurses followed by health professionals, parents and general practitioners. Data for the PCV7 and PCV13 primary series cohorts were limited to all AEFI reports made to SAEFVIC in the 12 months prior to and following introduction of PCV13 respectively (Table 3). For both primary series cohorts, the upper age limit of vaccine administration was defined as <7 months of age. For the PCV13 catch up cohort, data were limited to age at vaccination between 12 and 35 months (Table 3).

Table 3. Characteristics of the four study cohorts

Cohort number	Cohort description	Age	Male n (%)	Time period	Doses administered in Victoria over 12 month period
Cohort 1	PCV7 primary series	<7 months	109/221 (49%)	01 July 2010 – 30 June 2011	189,381
Cohort 2	PCV13 primary series	<7 months	125/229 (55%)	01 July 2011 – 30 June 2012	177,984
	0–12 months
	post PCV13 introduction
Cohort 3	PCV13 primary series	< 7 months	85/176 (48%)	01 July 2012 – 30 June 2013	195,646
	13–24 months post PCV13 introduction
Cohort 4	PCV13 catch up	12–35 months	29/52 (56%)	01 Oct 2011 – 30 Sept 2012	76,708

Study definitions

Each notification to SAEFVIC is recorded as a ‘case report’, however within each ‘case report’ up to 6 individual AEFI reactions may be recorded. In this study, both the total number of case reports and total individual AEFI reactions were used for analysis. AEFI are recorded as per standard case definitions based on those from the Brighton Collaboration, the Australian Immunization Handbook and definitions derived by SAEFVIC from published literature, or recorded verbatim as has been previously described.¹³

For this study, AEFI were categorized according to SAEFVIC case definitions. Reaction descriptions that did not meet these case definitions, due to being expressed as symptoms, were grouped into broader organ system categories based on the Medical Dictionary for Regulatory Activities (MedDRA) classes;³¹ Skin (any skin reaction other than rash), Gastrointestinal (GI)(any reaction other than diarrhea, vomiting and abdominal pain), Respiratory, Cardiovascular, Neurological (any reaction other than hypotonia and hypotonic hyporesponsive episode (HHE)), Possible infection, Metabolic and Haematological. Allergic reaction (generalized) was defined as per the SAEFVIC definition of a non-anaphylactic generalized reaction characterized by one or more skin or GI symptom and included urticaria, hives and angioedema.

Serious AEFI was defined as one that resulted in death, was life threatening, required in-patient hospitalization or prolongation of an existing hospitalization, resulted in significant or persistent disability/incapacity, caused a congenital anomaly or was medically important.³²

Data analyses

Reports relating to drug administration error were excluded from the analysis. Descriptive statistics were used to describe the epidemiology of reports for each cohort. Within each cohort, the proportion of reports where each vaccine was given alone or given with concomitant vaccines was described. The numbers and proportions for total case reports and for each individual AEFI category were described for each cohort. Crude AEFI reporting rates, per 100,000 doses of vaccine administered, were calculated for total case reports and for each AEFI category. This was achieved by dividing the total number of reports for each AEFI occurring in the cohort's 12-month period, by the total number of doses administered in Victoria (for that same time period and in the same age group). Denominator data were accessed through the Australian Childhood Immunization Register (ACIR), a national population based register that records vaccinations given on the NIP to children under 7 y of age (monthly data accessed on 27 March 2013). Reports meeting the definition of serious were described. All reports had sufficient clinical information to allow medical review, however, clinical information was only outlined for serious reports. Incidence rate ratios (IRR) were used to compare rates of AEFI following PCV7 with those following PCV13. Data were analyzed using Microsoft Excel 2007 (Microsoft, Redmond, PA) and STATA version 13.0 (Statacorp, Texas).

Post-hoc analyses

Due to the significant overall increase in individual reporting rates found in Cohort 2, a further consecutive 12 months of PCV13 primary series data were extracted and analyzed (Cohort 3- 01 July 2012 to 30 June 2013). IRR's were used to compare the AEFI reporting rate for Cohort 3 with the reporting rate for PCV7 primary series (Cohort 1). In addition, only the 7 AEFI categories that showed significant reporting rate differences between Cohort 1 and Cohort 2 were also analyzed for Cohort 3 and compared with Cohort 1 using IRR's.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors would like to thank the contributions of Mee Lee Easton for assistance with statistical analysis and Gowri Selvaraj for assistance with extraction of ACIR data.