Vaccine 33 (2015) 6902–6907
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
Strengthening of causality assessment of adverse events following immunization in the WHO South East Asia and Western Pacific regions: Lessons from the 2014 SEAR inter-country workshop Noni E. MacDonald a,∗ , Stephane Guichard b , Ananda Amarasinghe c , Madhava Ram Balakrishnan d , 2014 Inter-country SEAR Workshop Participants1 a
Department of Paediatrics, Dalhousie University, IWK Health Centre, Halifax, Canada Vaccine Supply and Quality, IVD, SEARO, Bangkok, Thailand c Epidemiology Unit, Ministry of Health, Colombo, Sri Lanka d Safety and Vigilance, Regulation of Medicines and Other Health Technologies, Department of Essential Medicines and Health Products, Health Systems and Innovation, WHO HQ, Geneva, Switzerland b
a r t i c l e
i n f o
Article history: Received 7 October 2014 Received in revised form 7 January 2015 Accepted 12 January 2015 Available online 21 January 2015 Keywords: AEFI Adverse events following immunization Causality assessment World Health Organization South East Asia Region
a b s t r a c t Background: Poorly managed AEFI undermine immunization programs. Improved surveillance in SEAR countries means more AEFIs but management varies. SEAR brought countries together to share AEFI experiences, and learn more about causality assessment. Methods: Three day 10 country workshop (9 SEAR; 1 WPR). Participants outlined county AEFI experiences, undertook causality assessment for 8 AEFIs using WHO methodology, critiqued the process by questionnaire and had a discussion. Results: All 10 valued AEFI monitoring and causality assessment, and praised the opportunity to share experiences. Participants determined a range of AEFI and causality assessment needs in SEAR such as adapting WHO Algorithm, CIOMS/Brighton definitions, WHO verbal autopsy to fit context, requesting a practical guide – AEFI definition, time interval, rates of AEFI for different vaccines and evidence for vaccine related causes of death under 24 h. Conclusions: LMIC need WHO AEFI tools adapted to better fit LMIC. Learning from each other builds capacity. Sharing AEFI experiences, case reviews help LMIC improve practices. © 2015 Published by Elsevier Ltd.
Serious or unexpected adverse events following immunization (AEFI) can raise concerns about the safety of vaccines for local health workers, communities and families leading to an undermining of public confidence in a country’s immunization program [1,2]. As vaccine adverse events differ by age group, population health status, concurrent illnesses, program delivery as well as by specific vaccine and/or combination and manufacturer, every country needs to have good AEFI surveillance systems and review causality assessment of AEFIs as part of their vaccine pharmacovigilance program. To address this need in the South East Asia Region (SEAR), the WHO Regional Office in 2003 established a Global Training Network Centre in Colombo, Sri Lanka to provide training on AEFI monitoring to SEAR Member States and to other WHO regions.
∗ Corresponding author at: Department of Paediatrics, Dalhousie University, IWK Health Centre, 5850/5980 University Avenue, Halifax, Nova Scotia, Canada B3K 6R8. Tel.: +1 902 470 8799; fax: +1 902 470 7232. E-mail address: [email protected] (N.E. MacDonald). 1 List of all participants by country, name and position is given in Appendix A. http://dx.doi.org/10.1016/j.vaccine.2015.01.033 0264-410X/© 2015 Published by Elsevier Ltd.
Initially, the course focused on AEFI due to programmatic errors and systems for early detection of serious AEFI cases and their management. By 2005, with strengthened National Regulatory Authority (NRA) capacity to regulate vaccines, the course was expanded to reach out to sub-national programme managers and immunization service providers. Since 2008, WHO SEARO has provided training support to NRA, National Immunization Programme and members of the national AEFI committee to strengthen capacity to detect, report, investigate and to carry out causality assessment for serious AEFI [3]. For some member countries, SEAR also provides support for development of infrastructure and small scale pilot projects to test and validate procedures, skills needed for quality AEFI reporting and causality assessment. These combined efforts have lead to a marked increase in AEFIs being reported in SEAR countries [4] including serious AEFIs. However, the causes of the reported AEFIs have not always been well understood leading in some instances to local health care worker, public and/or political concerns about vaccine safety that have caused disruptions in routine immunization program in many countries [5]. To help address this AEFI potential for loss of public confidence in vaccines, SEAR countries have
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
expanded the expertise on their national AEFI causality assessment committees. Recognizing that most of the countries in the regions acquire their vaccines from the same manufacturers and acknowledging that AEFI causality assessment is an important need for LMIC and that a lack of formal opportunities exist for countries to share experiences and concerns about AEFI surveillance and causality assessment, SEAR organized an inter-country workshop in February 2014. The main objective was to enhance regional capacity to evaluate investigated AEFI and carry out causality assessment of serious AEFI previously assessed by country committees. Use of harmonized terminology, standardized AEFI investigation processes, and causality assessment by countries would promote data aggregation from several countries for signal detection as many of the countries use vaccines from regional manufacturers. This is important for quality regional pharmacovigilance. We report here the lessons learned from this workshop based upon the plenary discussions, small group reports and outcomes of a participant questionnaire that reviewed each step in the WHO Causality Algorithm [6]. These findings may be of interest to immunization programs, National Regulatory Authority and vaccine pharmacovigilance groups in other LMIC, and also to WHO Regional offices, WHO Headquarters, vaccine safety educators and manufacturers. 1. Methods The WHO country offices were instrumental in mobilizing National AEFI Committee members to identify AEFI case reports that might benefit from inter-country discussion. Six of the 10 participating countries provided a total of 23 serious AEFI case reports from which 8 cases were short-listed for review at the workshop. The cases were anonymized as to patient, locale and country prior to review. A standard template was sent to the countries requesting that they provide details of the AEFI surveillance systems and the performance for the last 2 years. During the workshop each invited country shared an overview of their country’s AEFI program, specific AEFI causality assessment processes, experiences and concerns. The revised WHO causality assessment methodology (http://www.who.int/vaccine safety/ publications/aefi surveillance/en/) was then reviewed, discussed and critiqued. The participants then worked in three teams of 7–10 members with heterogeneous backgrounds and countries, and formally reviewed and assessed the 8 selected anonymized serious AEFI cases using the WHO Causality Algorithm methodology [6,7]. Following causality assessment of each case by one of the three teams, the logic and conclusions were then presented at a plenary session. This was followed by interactive plenary discussions on how AEFI case investigation and causality assessment could be improved in countries and in the region. As part of a quality improvement exercise, each participant was then offered the opportunity to comment and suggest modifications on each on the four sections of the WHO Causality Algorithm using a 17 item questionnaire (see Appendix). 2. Observations and recommendations from the workshop Of the 11 member states in SEAR, 9 (Bangladesh, Bhutan, India, Indonesia, Maldives, Nepal, Sri Lanka, Thailand, Timor-Leste) sent experts to participate. The Democratic People’s Republic of Korea and Myanmar did not participate. One country from the Western pacific region (Vietnam) also sent delegates. These 10 countries sent 25 participants (see authorship list) with a wide range of backgrounds, experience and expertise in pharmacovigliance.
6903
Table 1 Summary of conclusions and recommendations from the workshop. 1. Workshop facilitated much learning from each other. -Hold yearly/regional inter-country workshops on CA of AEFI 2. Review WHO Algorithm, CIOMS AEFI definition based upon suggestions from workshop and questionnaire results 3. Request a formal review of Brighton definitions to develop definition levels that will work in settings where no equipment and limited health worker training (i.e. outside of a trial, i.e. in real life settings in LMIC where serious AEFI happen) 4. Form working group to adapt WHO verbal autopsy to fit this AEFI case clinical need in SEA, WP regions 5. Work with regional Pediatric Societies to develop practical screening questions/observations for detection of clinically significant congenital heart disease (CHD) for referral BUT not let presence of CHD impede immunization 6. Develop a SEA region survey to identify to what extent National Regulatory Authorities are involved in AEFI in each of the participating countries and determine strategies used to increase participation in vaccine pharmacovigilance 7. Explore development of a practical guiding tool – AEFI definition, time interval, rates AEFI with different vaccines used in the region 8. Convene a meeting where SEA region countries could review compensation schemes–pros, cons; assess what might work in region; not just extrapolate from high-income country experiences 9. Explore how immunization and AEFI can be better incorporated into medical and nursing school curriculum in SEA region countries 10. Document workshop method, findings, and recommendations for submission to a peer review journal
The overall recommendations from the workshop are summarized in Table 1. Of the 10 participating countries, Timor-Leste and the Maldives had no national AEFI committees. AEFI surveillance capacity varied widely (Table 2) ranging from less than 20 cases per year to over 6000. The rate was not related to population size e.g. India 0.03/100,000, Bangladesh 1–3/100,000 and Sri Lanka 31.8/100,000. Despite these wide differences in reported country rates of serious AEFI, all 10 countries agreed on the need for and importance of good AEFI surveillance, timely quality investigations of serious AEFI and rational causality assessment conclusions. The 10 countries also reported varied relationships between their Immunization Program’s AEFI surveillance and causality assessment and their National Regulatory Authority. Reported constraints on strengthening vaccine-pharmacovigliance included poor quality of data (4 countries); high staff turnover rates (3 countries) and deficits in AEFI investigation training and lack of timely investigation (6 countries). All 10 agreed that the quality and timeliness of the data collected in an AEFI investigation had a great impact on the ability to determine if the implicated vaccine did or did not cause the adverse event. Inadequate data often lead to causality assessment categorization as ‘unclassifiable’, the least useful outcome. All agreed that pragmatic guidelines to better assist national immunization programs on planning for field investigation of serious AEFIs and the ability to more promptly respond were needed. With respect to data gathered for causality assessment, an important gap noted by several countries was the general lack of an autopsy following the death of an infant or child post immunization. Reasons given were the culture in the country, cost and the lack of access to local quality paediatric forensic expertise even if an autopsy was done. In the discussion session, no easy solutions were brought forward. The 2012 World Health Organization Verbal Autopsy form and manual might be helpful as it covers suitable age groups [8]. However, none of the countries reported routine use of this instrument in their AEFI investigations when a death occurred. Regular use of this tool, if adapted for AEFI use in the region, might improve the quality and quantity of information collected, facilitating causality assessment. Furthermore, there are no published reports of use of this instrument by AEFI investigation teams in
6904
Table 2 Summary country profile, status of AEFI and performance of vaccine pharmacovigilance systems.
Country profile Total pop. (million) Child. < 1 ( 000) New vaccine introduced in 2005–2013 Immunization coverage-2013 Source JRF WHO/UNICEF
Vaccine supply sources
AEFI guideline dates and update
BHU
IND
INO
MAL
NEP
SRL
THA
TLS
VTN
153.9
0.7
1200
244
0.3
28.6
21.1
64.2
1.2
90
3478 Penta (2009) MR (2012) Measles 2nd dose (2012) BCG: 99% OPV3: 92% PVV3: 92% MCV1: 86%
14.6 Penta (2009) HPV (2010)
26,000 Penta (2011 – in phased based)
4600 Td (2011) Penta (2013–14)
5.5 Penta (2012)
700 Penta (2009)
350 Penta (2008) MMR (2012) JE LAV (2012)
796 N/V
42.1 Penta (2012)
1700 Penta (2010) Measles2 (2011) DPT booster (2011)
BCG: 94% OPV3: 97% PVV 3: 97% MCV1: 95% MCV2: 89%
BCG: 94% HepB0:83.4% HepB3:95.6% OPV3: 95.3% MCV1: 93.5%
Government funded. Procurement through UNICEF
BCG: 99.7% DPT3: 99.4% HepB3: 99% Hib3: 99.5% MCV1: 99.4% MCV2: 98.6% Government funded. Procurement through UNICEF
BCG: 97% DPT3: 92% HepB3: 92% Hib3: 92% OPV3: 92% MCV1: 88% Government funded, except PVV through UNICEF. All vaccine are from WHO PQ sources.
BCG:99% OPV3: 99% PVV3: 99% MCV1:97% MCV 2: 96% JE: 98% Government funded, direct procurement from WHO PQ sources. Only PVV supplied by UNICEF with GAVI funds
BCG: 100% DPT3: 99.4% OPV3:99.4% HepB3:99.4% MCV1: 98.7% JE3: 89.3% Government funded, direct procurement of both locally produced and imported vaccines
BCG: 81.6% OPV 3:76% PVV3: 76.3% MCV1: 69%
Government funded. Procurement through UNICEF. Local vaccine manufactures available
BCG: 92% DPT3: 76% HepB3:67% Hib3: NA MCV1: 88% MCV2: 42% Government funded. Direct procurement from local producers mostly WHO PQ
2011 Not fixed, when necessary
2009 Twice/year or when necessary
No committee
2008 Not fixed, when necessary
2008 Twice/year, when necessary
2010 Not fixed, When necessary
No committee
Academia, paediatriians, microbiologist, virologists, pathologists, immunologist, epidemiologists/EPI, NRA/MCH, cardiologist, neurologist, pharmacologist, forensic expert Guideline 2007 updated in 2012
Academia, epidemiologist, pediatric of infectious diseases, pediatric neurology, neonatology, forensic expert, neuropathology, EP, NCLI
N/A
2003 and updated in 2008
Not available
2003 and updated in 2013 and 2014
6455 (2012)
1359 (2012)
0 (2012) 1 (2013)
30 (2012) 31 (2013)
2008 Quarterly
Government funded. Direct procurement of local producer and are WHO PQ
2012 Quarterly, when necessary
N/A Pediatrician, Academia, Academia, epidemiologists, clinicians, EPI, infectiologist, public health NRA/DDA, WHO, pediatricians, obstetriUNICEF specialist, cian/gynecologist, pediatrician, forensic NRA, specialist, EPI microbiologist, managers, NRA neurologist, technicians, pathologist, pharmacists, forensic experts, infectious legal experts. diseases surveillance, Guideline (2005) Guideline (2005), National Guideline (2007) Guideline 2003 guideline under revised in 2010 AEFI technical in Nepalese and SOP guideline (2012); development language in 2008 developed in AEFI MOH 2011. regulation (2013) All countries follows WHO criteria’s of serious AEFI: Death, hospitalization, cluster of reportable AEFI, persistent or significant disability/incapacity or life threatening condition, parental/community concerns Academia, EPI, NRA, virologist, pathologist, pediatricians, epidemiologists, neurologists
Categories of serious AEFI to be reported Vaccine PV performance 2298 (2012) Number of reported AEFI 2288 (2013) cases to National responsible unit 2012–2013
Pediatrician, medical specialist, microbiologist, EPI, clinical lab NRA and pharmacist
3 (2012) 16 (2013)
386* (2012) 536* (2013) (only serious are reported)
10,469 (2012) 18,621 (2013)
19 (2012) 22 (2013)
Funded and procured by UNICEF.
Government funded direct procurement from local suppliers, except PVV by UNICEF
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
Status AEFI Year AEFI committee established and meeting frequency Membership of AEFI committee
BAN
All N/A 163 (2012) 156 (2013) 1 (2012) 3 (2013) 15 (2012) 17 (2013) # of cases reviewed by committee
BAN= Bangladesh, BHU = Bhutan, IND= India, INO= Indonesia, MAL= Maldives, NEP= Nepal, SRL= Sri Lanka, THA= Thailand, TLS= Timor Leste, VTN=Vietnam. * Only serious are reported
1107 (2012) 8 (2012) 2 (2013) 19 (2012) 22 (2013)
30 (2012) 31 (2013) 0 (2012) 0 (2013) 1107 (2012) 2414 (2012) 894 (2013) 19 (2012) 22 (2013)
32% of case timely investigated with first investigation report available (2012) 30% (2013) 86 (2012) includes backlog previous years 89 (2013) 1 (2012) 3 (2013) 54 (2012) 62 (2013)
163 (2012) 156 (2013)
386 (2012) 536 (2013) 1 (2012) 3 (2013)
Number of serious AEFI reported 2012–2013 # of cases investigated
54 (2012) 73 (2013)
123 (2012) 120 (2013)
No AEFI investigated
19 (2012) 22 (2013)
2414 (2012) 894 (2013)
1359 (2012)
0 (2012) 1 (2013)
30 (2012) 31 (2013)
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
6905
low- and middle-income countries. Death following immunization led to other concerns. When these deaths and their investigation were managed by local police, serious problems often arose with the death being blamed on the vaccine with the health care worker being then held accountable without proper causality assessment being done. Lack of adequate country financial support for AEFI committees and investigation teams was raised as a constraint. Of note, countries producing vaccines were far more likely to have established a country “vaccine pharmacovigilance centre” with financial support from their Ministry of Health as AEFI data are needed by the local manufacturers. This information led other countries to speculate on how they too might improve their efforts to garner country resources for their AEFI program. All 10 countries acknowledged that a lack of planning for regular reviews of AEFI cases through a committee of experts was a concern. Several countries showed listings with AEFI deaths being due to vaccine when in fact these were coincidental events. Countries with very small populations commented that they simply did not have enough local expertise in causality assessment review and sufficient volume of serious AEFI cases to generalize conclusions on vaccine safety performance, particularly on potential signal detection. Suggestions to address this included small countries forming an arrangement to work with a larger one, or several small ones working together to share work and expertise. With respect to the causality assessment CIOMS definitions [9], participants from several countries noted that the revision to the definition of AEFI has not been that helpful in the field in their countries as the nuances were hard to explain in some languages. For example, the terms describing sign (“The adverse event may be any unfavorable or unintended sign, abnormal laboratory finding, symptom or disease.”) have proven too difficult to translate. What is the difference between an unfavorable or unintended sign? How do these differ from the adjective abnormal used to describe the other items? Review of each step of the WHO Revised Causality Algorithm methods during the causality assessment case plenary presentations and discussions as well as review of the comments on each of the four sections provided by the 20 participants who completed the quality assurance questionnaire highlighted several modifications that might facilitate ease of use. For example, the term “valid diagnosis” was unclear for 11 of the 20 questionnaire respondents. Adding an explanation of the meaning of this term in the use of the Algorithm might help for clarification. Similarly, five participants noted that “indeterminate” classification needed more explanation. In the plenary session, more than half of the participants noted challenges in their setting when trying to apply the Brighton definitions [10] for some AEFI. Participants suggested that a review of the CIOMS and Brighton definitions taking into account AEFI in lowand middle-income field settings could be helpful. Perhaps an additional glossary of explanations pertinent to these settings could be developed bearing in mind the need for simplicity given that these may need to be translated into local languages and applied with a minimum of resources. Similarly, a listing of AEFI definitions with time intervals and rates for the vaccines used locally based upon local data would be very helpful to all especially when trying to classify the events using the Algorithm. Participants noted that death was often the reported serious AEFI (see Table 3; 6 of 8 cases for review were deaths following immunization) they had to address but there are no evidence based articles or Aide-Memoire from WHO that succinctly present how a vaccine can cause death in the first 24 h post immunization e.g. anaphylaxis, toxic shock syndrome, immunization error related with examples, etc. Such information would be very helpful for those collecting AEFI case information in the field and for the causality assessment committees. Participants also commented that in these
6906
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
Table 3 Causality assessment outcomes of 8 selected cases reviewed. Age and sex
Vaccine
AEFI investigation preliminary findings
Causality assessment outcome
1
6 wk, male
Death
Contents lists available at ScienceDirect
Vaccine journal homepage: www.elsevier.com/locate/vaccine
Strengthening of causality assessment of adverse events following immunization in the WHO South East Asia and Western Pacific regions: Lessons from the 2014 SEAR inter-country workshop Noni E. MacDonald a,∗ , Stephane Guichard b , Ananda Amarasinghe c , Madhava Ram Balakrishnan d , 2014 Inter-country SEAR Workshop Participants1 a
Department of Paediatrics, Dalhousie University, IWK Health Centre, Halifax, Canada Vaccine Supply and Quality, IVD, SEARO, Bangkok, Thailand c Epidemiology Unit, Ministry of Health, Colombo, Sri Lanka d Safety and Vigilance, Regulation of Medicines and Other Health Technologies, Department of Essential Medicines and Health Products, Health Systems and Innovation, WHO HQ, Geneva, Switzerland b
a r t i c l e
i n f o
Article history: Received 7 October 2014 Received in revised form 7 January 2015 Accepted 12 January 2015 Available online 21 January 2015 Keywords: AEFI Adverse events following immunization Causality assessment World Health Organization South East Asia Region
a b s t r a c t Background: Poorly managed AEFI undermine immunization programs. Improved surveillance in SEAR countries means more AEFIs but management varies. SEAR brought countries together to share AEFI experiences, and learn more about causality assessment. Methods: Three day 10 country workshop (9 SEAR; 1 WPR). Participants outlined county AEFI experiences, undertook causality assessment for 8 AEFIs using WHO methodology, critiqued the process by questionnaire and had a discussion. Results: All 10 valued AEFI monitoring and causality assessment, and praised the opportunity to share experiences. Participants determined a range of AEFI and causality assessment needs in SEAR such as adapting WHO Algorithm, CIOMS/Brighton definitions, WHO verbal autopsy to fit context, requesting a practical guide – AEFI definition, time interval, rates of AEFI for different vaccines and evidence for vaccine related causes of death under 24 h. Conclusions: LMIC need WHO AEFI tools adapted to better fit LMIC. Learning from each other builds capacity. Sharing AEFI experiences, case reviews help LMIC improve practices. © 2015 Published by Elsevier Ltd.
Serious or unexpected adverse events following immunization (AEFI) can raise concerns about the safety of vaccines for local health workers, communities and families leading to an undermining of public confidence in a country’s immunization program [1,2]. As vaccine adverse events differ by age group, population health status, concurrent illnesses, program delivery as well as by specific vaccine and/or combination and manufacturer, every country needs to have good AEFI surveillance systems and review causality assessment of AEFIs as part of their vaccine pharmacovigilance program. To address this need in the South East Asia Region (SEAR), the WHO Regional Office in 2003 established a Global Training Network Centre in Colombo, Sri Lanka to provide training on AEFI monitoring to SEAR Member States and to other WHO regions.
∗ Corresponding author at: Department of Paediatrics, Dalhousie University, IWK Health Centre, 5850/5980 University Avenue, Halifax, Nova Scotia, Canada B3K 6R8. Tel.: +1 902 470 8799; fax: +1 902 470 7232. E-mail address: [email protected] (N.E. MacDonald). 1 List of all participants by country, name and position is given in Appendix A. http://dx.doi.org/10.1016/j.vaccine.2015.01.033 0264-410X/© 2015 Published by Elsevier Ltd.
Initially, the course focused on AEFI due to programmatic errors and systems for early detection of serious AEFI cases and their management. By 2005, with strengthened National Regulatory Authority (NRA) capacity to regulate vaccines, the course was expanded to reach out to sub-national programme managers and immunization service providers. Since 2008, WHO SEARO has provided training support to NRA, National Immunization Programme and members of the national AEFI committee to strengthen capacity to detect, report, investigate and to carry out causality assessment for serious AEFI [3]. For some member countries, SEAR also provides support for development of infrastructure and small scale pilot projects to test and validate procedures, skills needed for quality AEFI reporting and causality assessment. These combined efforts have lead to a marked increase in AEFIs being reported in SEAR countries [4] including serious AEFIs. However, the causes of the reported AEFIs have not always been well understood leading in some instances to local health care worker, public and/or political concerns about vaccine safety that have caused disruptions in routine immunization program in many countries [5]. To help address this AEFI potential for loss of public confidence in vaccines, SEAR countries have
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
expanded the expertise on their national AEFI causality assessment committees. Recognizing that most of the countries in the regions acquire their vaccines from the same manufacturers and acknowledging that AEFI causality assessment is an important need for LMIC and that a lack of formal opportunities exist for countries to share experiences and concerns about AEFI surveillance and causality assessment, SEAR organized an inter-country workshop in February 2014. The main objective was to enhance regional capacity to evaluate investigated AEFI and carry out causality assessment of serious AEFI previously assessed by country committees. Use of harmonized terminology, standardized AEFI investigation processes, and causality assessment by countries would promote data aggregation from several countries for signal detection as many of the countries use vaccines from regional manufacturers. This is important for quality regional pharmacovigilance. We report here the lessons learned from this workshop based upon the plenary discussions, small group reports and outcomes of a participant questionnaire that reviewed each step in the WHO Causality Algorithm [6]. These findings may be of interest to immunization programs, National Regulatory Authority and vaccine pharmacovigilance groups in other LMIC, and also to WHO Regional offices, WHO Headquarters, vaccine safety educators and manufacturers. 1. Methods The WHO country offices were instrumental in mobilizing National AEFI Committee members to identify AEFI case reports that might benefit from inter-country discussion. Six of the 10 participating countries provided a total of 23 serious AEFI case reports from which 8 cases were short-listed for review at the workshop. The cases were anonymized as to patient, locale and country prior to review. A standard template was sent to the countries requesting that they provide details of the AEFI surveillance systems and the performance for the last 2 years. During the workshop each invited country shared an overview of their country’s AEFI program, specific AEFI causality assessment processes, experiences and concerns. The revised WHO causality assessment methodology (http://www.who.int/vaccine safety/ publications/aefi surveillance/en/) was then reviewed, discussed and critiqued. The participants then worked in three teams of 7–10 members with heterogeneous backgrounds and countries, and formally reviewed and assessed the 8 selected anonymized serious AEFI cases using the WHO Causality Algorithm methodology [6,7]. Following causality assessment of each case by one of the three teams, the logic and conclusions were then presented at a plenary session. This was followed by interactive plenary discussions on how AEFI case investigation and causality assessment could be improved in countries and in the region. As part of a quality improvement exercise, each participant was then offered the opportunity to comment and suggest modifications on each on the four sections of the WHO Causality Algorithm using a 17 item questionnaire (see Appendix). 2. Observations and recommendations from the workshop Of the 11 member states in SEAR, 9 (Bangladesh, Bhutan, India, Indonesia, Maldives, Nepal, Sri Lanka, Thailand, Timor-Leste) sent experts to participate. The Democratic People’s Republic of Korea and Myanmar did not participate. One country from the Western pacific region (Vietnam) also sent delegates. These 10 countries sent 25 participants (see authorship list) with a wide range of backgrounds, experience and expertise in pharmacovigliance.
6903
Table 1 Summary of conclusions and recommendations from the workshop. 1. Workshop facilitated much learning from each other. -Hold yearly/regional inter-country workshops on CA of AEFI 2. Review WHO Algorithm, CIOMS AEFI definition based upon suggestions from workshop and questionnaire results 3. Request a formal review of Brighton definitions to develop definition levels that will work in settings where no equipment and limited health worker training (i.e. outside of a trial, i.e. in real life settings in LMIC where serious AEFI happen) 4. Form working group to adapt WHO verbal autopsy to fit this AEFI case clinical need in SEA, WP regions 5. Work with regional Pediatric Societies to develop practical screening questions/observations for detection of clinically significant congenital heart disease (CHD) for referral BUT not let presence of CHD impede immunization 6. Develop a SEA region survey to identify to what extent National Regulatory Authorities are involved in AEFI in each of the participating countries and determine strategies used to increase participation in vaccine pharmacovigilance 7. Explore development of a practical guiding tool – AEFI definition, time interval, rates AEFI with different vaccines used in the region 8. Convene a meeting where SEA region countries could review compensation schemes–pros, cons; assess what might work in region; not just extrapolate from high-income country experiences 9. Explore how immunization and AEFI can be better incorporated into medical and nursing school curriculum in SEA region countries 10. Document workshop method, findings, and recommendations for submission to a peer review journal
The overall recommendations from the workshop are summarized in Table 1. Of the 10 participating countries, Timor-Leste and the Maldives had no national AEFI committees. AEFI surveillance capacity varied widely (Table 2) ranging from less than 20 cases per year to over 6000. The rate was not related to population size e.g. India 0.03/100,000, Bangladesh 1–3/100,000 and Sri Lanka 31.8/100,000. Despite these wide differences in reported country rates of serious AEFI, all 10 countries agreed on the need for and importance of good AEFI surveillance, timely quality investigations of serious AEFI and rational causality assessment conclusions. The 10 countries also reported varied relationships between their Immunization Program’s AEFI surveillance and causality assessment and their National Regulatory Authority. Reported constraints on strengthening vaccine-pharmacovigliance included poor quality of data (4 countries); high staff turnover rates (3 countries) and deficits in AEFI investigation training and lack of timely investigation (6 countries). All 10 agreed that the quality and timeliness of the data collected in an AEFI investigation had a great impact on the ability to determine if the implicated vaccine did or did not cause the adverse event. Inadequate data often lead to causality assessment categorization as ‘unclassifiable’, the least useful outcome. All agreed that pragmatic guidelines to better assist national immunization programs on planning for field investigation of serious AEFIs and the ability to more promptly respond were needed. With respect to data gathered for causality assessment, an important gap noted by several countries was the general lack of an autopsy following the death of an infant or child post immunization. Reasons given were the culture in the country, cost and the lack of access to local quality paediatric forensic expertise even if an autopsy was done. In the discussion session, no easy solutions were brought forward. The 2012 World Health Organization Verbal Autopsy form and manual might be helpful as it covers suitable age groups [8]. However, none of the countries reported routine use of this instrument in their AEFI investigations when a death occurred. Regular use of this tool, if adapted for AEFI use in the region, might improve the quality and quantity of information collected, facilitating causality assessment. Furthermore, there are no published reports of use of this instrument by AEFI investigation teams in
6904
Table 2 Summary country profile, status of AEFI and performance of vaccine pharmacovigilance systems.
Country profile Total pop. (million) Child. < 1 ( 000) New vaccine introduced in 2005–2013 Immunization coverage-2013 Source JRF WHO/UNICEF
Vaccine supply sources
AEFI guideline dates and update
BHU
IND
INO
MAL
NEP
SRL
THA
TLS
VTN
153.9
0.7
1200
244
0.3
28.6
21.1
64.2
1.2
90
3478 Penta (2009) MR (2012) Measles 2nd dose (2012) BCG: 99% OPV3: 92% PVV3: 92% MCV1: 86%
14.6 Penta (2009) HPV (2010)
26,000 Penta (2011 – in phased based)
4600 Td (2011) Penta (2013–14)
5.5 Penta (2012)
700 Penta (2009)
350 Penta (2008) MMR (2012) JE LAV (2012)
796 N/V
42.1 Penta (2012)
1700 Penta (2010) Measles2 (2011) DPT booster (2011)
BCG: 94% OPV3: 97% PVV 3: 97% MCV1: 95% MCV2: 89%
BCG: 94% HepB0:83.4% HepB3:95.6% OPV3: 95.3% MCV1: 93.5%
Government funded. Procurement through UNICEF
BCG: 99.7% DPT3: 99.4% HepB3: 99% Hib3: 99.5% MCV1: 99.4% MCV2: 98.6% Government funded. Procurement through UNICEF
BCG: 97% DPT3: 92% HepB3: 92% Hib3: 92% OPV3: 92% MCV1: 88% Government funded, except PVV through UNICEF. All vaccine are from WHO PQ sources.
BCG:99% OPV3: 99% PVV3: 99% MCV1:97% MCV 2: 96% JE: 98% Government funded, direct procurement from WHO PQ sources. Only PVV supplied by UNICEF with GAVI funds
BCG: 100% DPT3: 99.4% OPV3:99.4% HepB3:99.4% MCV1: 98.7% JE3: 89.3% Government funded, direct procurement of both locally produced and imported vaccines
BCG: 81.6% OPV 3:76% PVV3: 76.3% MCV1: 69%
Government funded. Procurement through UNICEF. Local vaccine manufactures available
BCG: 92% DPT3: 76% HepB3:67% Hib3: NA MCV1: 88% MCV2: 42% Government funded. Direct procurement from local producers mostly WHO PQ
2011 Not fixed, when necessary
2009 Twice/year or when necessary
No committee
2008 Not fixed, when necessary
2008 Twice/year, when necessary
2010 Not fixed, When necessary
No committee
Academia, paediatriians, microbiologist, virologists, pathologists, immunologist, epidemiologists/EPI, NRA/MCH, cardiologist, neurologist, pharmacologist, forensic expert Guideline 2007 updated in 2012
Academia, epidemiologist, pediatric of infectious diseases, pediatric neurology, neonatology, forensic expert, neuropathology, EP, NCLI
N/A
2003 and updated in 2008
Not available
2003 and updated in 2013 and 2014
6455 (2012)
1359 (2012)
0 (2012) 1 (2013)
30 (2012) 31 (2013)
2008 Quarterly
Government funded. Direct procurement of local producer and are WHO PQ
2012 Quarterly, when necessary
N/A Pediatrician, Academia, Academia, epidemiologists, clinicians, EPI, infectiologist, public health NRA/DDA, WHO, pediatricians, obstetriUNICEF specialist, cian/gynecologist, pediatrician, forensic NRA, specialist, EPI microbiologist, managers, NRA neurologist, technicians, pathologist, pharmacists, forensic experts, infectious legal experts. diseases surveillance, Guideline (2005) Guideline (2005), National Guideline (2007) Guideline 2003 guideline under revised in 2010 AEFI technical in Nepalese and SOP guideline (2012); development language in 2008 developed in AEFI MOH 2011. regulation (2013) All countries follows WHO criteria’s of serious AEFI: Death, hospitalization, cluster of reportable AEFI, persistent or significant disability/incapacity or life threatening condition, parental/community concerns Academia, EPI, NRA, virologist, pathologist, pediatricians, epidemiologists, neurologists
Categories of serious AEFI to be reported Vaccine PV performance 2298 (2012) Number of reported AEFI 2288 (2013) cases to National responsible unit 2012–2013
Pediatrician, medical specialist, microbiologist, EPI, clinical lab NRA and pharmacist
3 (2012) 16 (2013)
386* (2012) 536* (2013) (only serious are reported)
10,469 (2012) 18,621 (2013)
19 (2012) 22 (2013)
Funded and procured by UNICEF.
Government funded direct procurement from local suppliers, except PVV by UNICEF
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
Status AEFI Year AEFI committee established and meeting frequency Membership of AEFI committee
BAN
All N/A 163 (2012) 156 (2013) 1 (2012) 3 (2013) 15 (2012) 17 (2013) # of cases reviewed by committee
BAN= Bangladesh, BHU = Bhutan, IND= India, INO= Indonesia, MAL= Maldives, NEP= Nepal, SRL= Sri Lanka, THA= Thailand, TLS= Timor Leste, VTN=Vietnam. * Only serious are reported
1107 (2012) 8 (2012) 2 (2013) 19 (2012) 22 (2013)
30 (2012) 31 (2013) 0 (2012) 0 (2013) 1107 (2012) 2414 (2012) 894 (2013) 19 (2012) 22 (2013)
32% of case timely investigated with first investigation report available (2012) 30% (2013) 86 (2012) includes backlog previous years 89 (2013) 1 (2012) 3 (2013) 54 (2012) 62 (2013)
163 (2012) 156 (2013)
386 (2012) 536 (2013) 1 (2012) 3 (2013)
Number of serious AEFI reported 2012–2013 # of cases investigated
54 (2012) 73 (2013)
123 (2012) 120 (2013)
No AEFI investigated
19 (2012) 22 (2013)
2414 (2012) 894 (2013)
1359 (2012)
0 (2012) 1 (2013)
30 (2012) 31 (2013)
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
6905
low- and middle-income countries. Death following immunization led to other concerns. When these deaths and their investigation were managed by local police, serious problems often arose with the death being blamed on the vaccine with the health care worker being then held accountable without proper causality assessment being done. Lack of adequate country financial support for AEFI committees and investigation teams was raised as a constraint. Of note, countries producing vaccines were far more likely to have established a country “vaccine pharmacovigilance centre” with financial support from their Ministry of Health as AEFI data are needed by the local manufacturers. This information led other countries to speculate on how they too might improve their efforts to garner country resources for their AEFI program. All 10 countries acknowledged that a lack of planning for regular reviews of AEFI cases through a committee of experts was a concern. Several countries showed listings with AEFI deaths being due to vaccine when in fact these were coincidental events. Countries with very small populations commented that they simply did not have enough local expertise in causality assessment review and sufficient volume of serious AEFI cases to generalize conclusions on vaccine safety performance, particularly on potential signal detection. Suggestions to address this included small countries forming an arrangement to work with a larger one, or several small ones working together to share work and expertise. With respect to the causality assessment CIOMS definitions [9], participants from several countries noted that the revision to the definition of AEFI has not been that helpful in the field in their countries as the nuances were hard to explain in some languages. For example, the terms describing sign (“The adverse event may be any unfavorable or unintended sign, abnormal laboratory finding, symptom or disease.”) have proven too difficult to translate. What is the difference between an unfavorable or unintended sign? How do these differ from the adjective abnormal used to describe the other items? Review of each step of the WHO Revised Causality Algorithm methods during the causality assessment case plenary presentations and discussions as well as review of the comments on each of the four sections provided by the 20 participants who completed the quality assurance questionnaire highlighted several modifications that might facilitate ease of use. For example, the term “valid diagnosis” was unclear for 11 of the 20 questionnaire respondents. Adding an explanation of the meaning of this term in the use of the Algorithm might help for clarification. Similarly, five participants noted that “indeterminate” classification needed more explanation. In the plenary session, more than half of the participants noted challenges in their setting when trying to apply the Brighton definitions [10] for some AEFI. Participants suggested that a review of the CIOMS and Brighton definitions taking into account AEFI in lowand middle-income field settings could be helpful. Perhaps an additional glossary of explanations pertinent to these settings could be developed bearing in mind the need for simplicity given that these may need to be translated into local languages and applied with a minimum of resources. Similarly, a listing of AEFI definitions with time intervals and rates for the vaccines used locally based upon local data would be very helpful to all especially when trying to classify the events using the Algorithm. Participants noted that death was often the reported serious AEFI (see Table 3; 6 of 8 cases for review were deaths following immunization) they had to address but there are no evidence based articles or Aide-Memoire from WHO that succinctly present how a vaccine can cause death in the first 24 h post immunization e.g. anaphylaxis, toxic shock syndrome, immunization error related with examples, etc. Such information would be very helpful for those collecting AEFI case information in the field and for the causality assessment committees. Participants also commented that in these
6906
N.E. MacDonald et al. / Vaccine 33 (2015) 6902–6907
Table 3 Causality assessment outcomes of 8 selected cases reviewed. Age and sex
Vaccine
AEFI investigation preliminary findings
Causality assessment outcome
1
6 wk, male
Death
