Introduction

This concept will discuss the methods used by the Manitoba Centre for Health Policy (MCHP) to evaluate adverse events potentially associated with vaccines provided in Manitoba to infants, children and adults.


As described by the Public Health Agency of Canada (PHAC) (2006), there are multiple steps taken to ensure the safety of vaccines prior to and during use in Canada. First, the vaccine undergoes pre-clinical testing to ensure that the desired immune response is produced by the vaccine recipient without experiencing toxicity (PHAC, 2006). Human clinical studies are subsequently completed to assess the safety of the vaccine, its immunogenicity, optimal dose/schedule and local/systemic reactions to the vaccine. Multiple phases are conducted in which the number of human subjects increases from ten to up to 30,000 (PHAC, 2006). Post-marketing evaluation of the vaccine is also completed in the general public. More information on vaccine evaluation and regulation is available in Part 2 of the Canadian Immunization Guide (7th edition) .

In Hilderman et al. (2011), Adverse Events Following Immunization (AEFI) are defined as "any undesirable or unexpected event that occurs following immunization, which may or may not be caused by the administration of the vaccine" (p.195). AEFI may range from very rare to very common, depending on its frequency (less than one in 10, 000 and more than one in ten, respectively) (PHAC, 2006). Separating evidence based vaccine safety issues from perceived public concerns is important to foster public confidence in immunization programs and ensure success of these programs (PHAC, 2006).

Methods

In Hilderman et al. (2011) , the following methodology was used to assess the rate of AEFI in Manitobans.

Background Information

Vaccine safety is assessed by various bodies in Manitoba and Canada. The Biologics and Genetic Therapies Directorate (BGTD) of Health Canada regulates the safety, efficacy, and quality of vaccines (Hilderman et al., 2011). Passive and active surveillance of adverse effects of vaccinations are completed by the Canadian Adverse Event Following Immunization Surveillance System (CAEFISS) and the Immunization Monitoring Program ACTive (IMPACT), respectively (PHAC, 2006). In Manitoba, adverse events from these bodies are reported to the Communicable Disease Control Branch of Manitoba Health, as well as the Public Health Agency of Canada.

Using the PHAC's list of Adverse Events of Special Interest (AESI), Hilderman et al. (2011) evaluated AEFIs from administrative data; due to the unproven validity of single visit specific ambulatory care diagnostic codes, only AESIs recorded in hospital discharge abstracts were included in the analyses. The adverse events of interest are generally severe enough to warrant hospitalization should they occur. The diagnostic codes for the following conditions were collected from hospital discharge abstracts to determine whether an individual experienced an event that could have been associated with immunization.

ICD Codes for Adverse Events

Condition	ICD-9-CM Codes	ICD-10-CA Codes
Encephalitis/Myelitis/Acute disseminated encephalomyelitis (EMA)	323.5, 320.9, 323. 8, 323. 9, 323.4, 323.0, 049.9	G04.0, G04.2, G04.8, G04.9 G05.0, G05.1, G05.2, G05.8, A86
Guillain-Barré Syndrome (GBS)	357.0	G61.0
Idiopathic thrombocytopenic purpura (ITP)	287.3	D69.3

Vaccine Tariff Codes

Data from the Manitoba Immunization Monitoring System (MIMS) was used to collect vaccine tariff codes to determine whether an individual received a vaccine in a specified 12 week time period. The vaccine tariffs used for influenza were 8791, 8792, and 8799. Appendix 1: Tariff Codes for Vaccinations in the Manitoba Immunization Study by Hilderman et al. (2011) provides the tariff codes for all other vaccines.

Description of Steps

All cases of adverse events were included for Manitobans who were 6 months and older at the diagnosis date. Infants younger than 6 months could not be immunized for influenza, therefore they were excluded from the analysis. Individuals for which the number of Aggregated Diagnosis Groups (ADGs) was missing, were also excluded.

Only the first diagnosis, of any of the conditions mentioned above, was counted per person during a specified time period. This ensured that an individual was only counted once per condition. Immunization records from the 12 weeks prior to hospital admission were used to specify what, if any, vaccine(s) was received prior to diagnosis of the above conditions. In the case of a person receiving more than one immunization within these 12 weeks, the immunization closest to the diagnosis date was used in the analysis.

"Onset of Guillain-Barré Syndrome/Encephalitis or ITP within 42 days of immunization, it could be considered to be plausibly associated with immunization (Glanz et al., 2006)." (Hilderman et al., 2011, p.184).

"A case-control analysis method was used to look at the relationship between vaccination and development of each of the three aforementioned rare conditions...Cases were matched to controls (four controls for each case) by birth year, sex, region or residence, and number of ADGs (0, 1-3, 4 or more) in the year prior to diagnosis date. The date of admission to hospital where condition was diagnosed was used as the diagnosis date. The following ADGs were excluded: 15, 19, 27, 29, 30, 31, and 34. A conditional logistic regression analysis was then done, looking at immunization for any antigen for cases and controls. The conditional logistic regression analysis was then repeated, looking at influenza immunizations for cases and controls" (Hilderman et al., 2011, p.184).

Limitations/Cautions

Not all vaccinations, such as those given in private settings (work places), are entered into MIMS. This can lead to underreporting of vaccines. Hilderman et al. (2011) indicate that the rates of the influenza immunization are most commonly affected by this limitation. As well, tariff codes are not assigned if an individual receives an immunization when he/she does not fall under Manitoba's eligibility criteria for publicly funded vaccines, thus this vaccine would not be entered into MIMS.