Sao Paulo, 18 March 2015 1 |

Cost-effectiveness and budget impact analysis of quadrivalent (QIV) versus trivalent influenza

vaccines (TIV) in low and middle income countries

Preliminary Report

Jan Hendriks, WHO/EMP/PHI/TTi

Sao Paulo, 18 March 2015 2 |

Policy Question

Under which scenario’s will QIVs be cost-effective compared to TIVs in low and middle income

countries from a societal perspective?

Sao Paulo, 18 March 2015 3 |

Study Teams

University of Western Australia – George Milne, Nilimesh Halder, Joel Kelso

National Institute of Communicable Diseases, South Africa – Cheryl Cohen, Jocelyn Moyes

University of Groningen, The Netherlands – Maarten Postma – Pieter de Boer

WHO Initiative for Vaccine Research (IVR) – Raymond Hutubessy

Sao Paulo, 18 March 2015 4 |

Part I: Community model – Mimic real communities by individual-based, multi strain influenza simulation models – No vaccination vs TIV vs QIV

• Run simulator without vaccination and with vaccination strategy activated

• Difference determines effectiveness of vaccination as reduction in illness attack rate

• Infection profile of each infected individual then input into economic analysis model to determine cost and cost-effectiveness.

Methods (1/2)

Sao Paulo, 18 March 2015 5 |

Methods (2/2)

Part II: Cost-effectiveness analysis model – Perspective:

• societal public health – Costs:

• vaccine procurement and delivery costs; • treatment • household, including productivity losses due to influenza

– Benefits: • influenza-related loss of Quality of Life (QoL)

Part III: Budget impact analysis

Sao Paulo, 18 March 2015 6 |

Health outcomes decision process

Parameters associated with

Disease Burdens

GP visits, Hospitalisations, ICU Admissions,

Deaths

Quantification of resulting Disease

Burdens

Economic Analysis

Overall Costs and Vaccination

Cost-effectiveness

Census Data Population Contact Network

Influenza Transmission

Procedure

Vaccination strategies

Individual-based simulation models

Simulation outcomes

Cost parameters

Vaccination Effectiveness

Disease parameters

Vaccination parameters

Integrated simulation model and economic analysis methodology

Sao Paulo, 18 March 2015 7 |

8 Vaccination Scenario’s Simulated

Vaccine: TIV or QIV

Distribution: choice target group – Randomized

• (unrealistic) baseline

– Transmitters-first • first to highest age-specific attack rate

– Vulnerable-first. • first to those aged 65+, followed by those under 5, followed by adults

– HIV and Vulnerable-first • first to HIV positive, and remainder to vulnerable-first option as above

Sao Paulo, 18 March 2015 8 |

Communities

Australia South Africa Vietnam

Community Albany Agincourt Thai Nguyen

Population 30000 50000 74000

GDP/capita (2013 US$, WB) 67,458 6,618 1,911

Type (WB classification) High-income relatively isolated

Upper-middle-income rural

Lower-middle – income urban

Seasonal vaccination coverage 20% 2% 0%

Sao Paulo, 18 March 2015 9 |

Results

Vaccination scenario: – “Transmitters-first” most cost-effective scenario for all 3 locations

for both vaccines, – QIV gives a health advantage over TIV in all scenarios for all 3

locations

Cost effectiveness: – QIV over TIV: more benefits but at more costs !! – Results depend on setting

• In Australia picture is less clear due to a “better B-strain match” in 2003-2013 and higher hospitalization rates and higher treatment costs compared to SA and Vietnam

Sao Paulo, 18 March 2015 10 |

Albany, Australia Cost Effectiveness

QALY loss

Costs (I$)

No vaccination

TIV

QIV20 QIV50 QIV80

Vaccination scenario: Transmitters First. QIV20: Price Premium of QIV over TIV= 20%

0

100

200

300

400

500

600

0 1000000 2000000 3000000 4000000

Sao Paulo, 18 March 2015 11 |

Agincourt, South Africa Cost Effectiveness

QUALY loss

Costs (I$)

No vaccination

TIV

QIV20 QIV50 QIV80

0

100

200

300

400

500

600

0 200000 400000 600000 800000 1000000 1200000 1400000 1600000 1800000

Vaccination scenario: Transmitters First. QIV20: Price Premium of QIV over TIV= 20%

Sao Paulo, 18 March 2015 12 |

Thai-Nguyen, Vietnam Cost Effectiveness

QALY loss

Costs (I$)

No vaccination

TIV

QIV20 QIV50 QIV80 0

50

100

150

200

250

300

350

400

0 500000 1000000 1500000 2000000

Vaccination scenario: Transmitters First. QIV20: Price Premium of QIV over TIV= 20%

Sao Paulo, 18 March 2015 13 |

Budget Impact

% of National Health Budget

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Australia South Africa Vietnam

TIVQIV

Sao Paulo, 18 March 2015 14 |

Conclusions Preliminary analysis shows that influenza vaccination with

TIV compared to no-vaccination is cost effective

Cost-effectiveness of QIV vs TIV is setting-dependent:

– In low-income settings such as SA and Vietnam: • Potential benefits but at significant additional costs • Major budget impact implications !!

– In high-income setting such as Australia:

• Marginal benefits to gain • Minor budget impact implications

Sao Paulo, 18 March 2015 15 |

CAVEAT !

This is modelling – the value of the results is dependent on the data put into the model and the assumptions that are made.

These are very preliminary results – analysis is continuing.

Three small communities analysed – NOT PREDICTIVE of national, regional or global situation.

Sao Paulo, 18 March 2015 16 |

Alternative model

Price increase about 35%

Efficacy increase – about 5%

You choose !

Sao Paulo, 18 March 2015 17 |

For Discussion

Local decision makers need to decide whether the additional investments for QIV are worth the resources required compared to TIV

What are the opportunity costs?

Sao Paulo, 18 March 2015 18 |

Thank you for your attention