Talk:Economic evaluation

From Opasnet
Revision as of 15:52, 6 September 2014 by Kauq (talk | contribs) (→‎Comments)
Jump to navigation Jump to search

Economic comparison method

How to read discussions

Fact discussion: .
Opening statement: The following comparison methods should be used in comparing the tehders:
  1. Incremental cost-effectiveness (ICER)
  2. Cost-effectiveness with a limit value harmonized in all health care
  3. Cost-benefit analysis
  4. Maximization of health benefits within total budget limit
  5. Price only

Closing statement: ICER should be used.

(A closing statement, when resolved, should be updated to the main page.)

Argumentation:

⇤--#: . There should be only one primary comparison method. Therefore all options should be rejected a priori and specific reasons should be presented to use a method. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: attack)

  1. Incremental cost-effectiveness (ICER)
    ←--#: . This is the preferred method, because the society has not set an exchange rate between money and health / QALY. This is necessary for cost-benefit and cost-effectiveness with limit value. --THL 3 July 2014 (type: truth; paradigms: science: defence)
    ←--#: . Implicit support from comment GSK1. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: defence)
    ----#: . This method was unpopular in a related questionnaire. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: comment)
  2. Cost-effectiveness with a limit value harmonized in all health care
    ⇤--#: . There is no agreed limit value in use in Finland. See comment Heini1. --Jouni (talk) 13:54, 5 September 2014 (UTC) (type: truth; paradigms: science: attack)
    ←--#: . See comment K2/Pfizer. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: defence)
    ←--#: . Implicit support from comment GSK1. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: defence)
    ----#: . This method was popular in a related questionnaire. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: comment)
  3. Cost-benefit analysis
    ----#: . This method was moderately popular in a related questionnaire. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: comment)
  4. Maximization of health benefits within total budget limit
    ----#: . This method was popular in a related questionnaire. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: comment)
  5. Price only
    ----#: . This method was unpopular in a related questionnaire. --Jouni (talk) 04:04, 5 September 2014 (UTC) (type: truth; paradigms: science: comment)

Comments

----K2: . Should willingness to pay/QALY be at the same level as applied in health care in Finland, i.e. 50 000-66 000 €/QALY? [1] [2] --Pfizer Oy, 2014-08-27 13:47:39 - R upload (type: truth; paradigms: science: comment)

⇤--Heini1: . In Finland, there is no threshold for the cost of an additional quality-adjusted life-year. In the light of recent experience of varicella vaccination programme the decision makers’ willingness to pay for a QALY was not 50 000-66 000 euros. From the health care payers perspective varicella vaccination programme was cost-effective (14 600 euros / QALY) and from the societal perspective it was cost-saving. There are favourable recommendations of experts (Expert group, National Advisory Committee for Vaccination) and even a decision of Ministry of Social Affairs and Health to propose Government funding for varicella vaccination as part of the national immunization programme. However, Varicella vaccination has not yet been approved in the Government's budget proposal. --Heini (talk) 12:35, 2 September 2014 (UTC) (type: truth; paradigms: science: attack)

Question:

  • "The health benefit (effectiveness) of the pneumococcal infant immunisation programme is assessed by the expected gain in Quality-Adjusted Life Years (QALYs), corresponding to the expected reduction in the annual number of invasive pneumococcal disease in the whole Finnish population."
⇤--GSK1: . PCV programmes have been demonstrated to deliver much broader public health and economic benefits which extend beyond reduction of the invasive disease.
While we acknowledge that the it may be difficult to adjust the epidemiological model to assess the impact and benefits of prevention of respiratory infection episodes, including pneumonia and acute otitis media related endpoints, which effectively represent majority of the vaccine preventable disease burden (e.g.Palmu, ESPID 2014, poster), it is imperative that the cost effectiveness analyses take into account these endpoints.
To consider only IPDs would distort the cost-effectiveness analysis and by ignoring the majority of cost savings would violate the procurement law requirement that the selection criteria must have economic relevance to the contracting entity.
The applicants should have the opportunity to submit their proposal for assumptions including scientific justification and/or to provide alternative models for consideration. -- GSK 13:17, 3 September 2014 (UTC) (type: truth; paradigms: science: attack)
⇤--1: . We refer to our replies to the same question on pages ‘Evaluation criteria’ and Epidemiological modelling. It should be noted that the tendering criteria are based on comparing different vaccine products. Critical assumptions about some serotype’s expected efficacy or effectiveness have been assessed in sensitivy analyses. Kauq, Sep 6, 2014 (type: truth; paradigms: science: attack)


⇤--GSK2: . ::Further, as indicated in the section “Comparison criteria” the assumption for serotype specific vaccine effectiveness for vaccine and vaccine-related types should be based on available clinical evidence, as far as possible.
In consequence, some parameters explored in the sensitivity analysis should be rather considered as a base case assumptions.
It is proposed to invite applicants to address such issues in the tender submissions. -- GSK 13:17, 3 September 2014 (UTC) (type: truth; paradigms: science: attack)
⇤--1: . The open discussion on these pages provides the forum for all stakeholders to express their views. Kauq, Sep 6, 2014 (type: truth; paradigms: science: attack)

Ajattelimme Karin kanssa, että nykyinen epidemiologinen malli-ohjelmakoodi toimii hyvin ja sen voisi jättää suunilleen nykyiselleen. Taloudellinen malli perustuu epidemiologisen mallin ennusteisiin, mutta siihen tarvitaan 101-luokkainen ikäluokitus. Taloudellisen arvioinnin sivulla voisi siis olla kokonaan oma koodinsa jossa tehdään uusi ennuste tällä tiheämmällä luokituksella. Tai sitten epidemiologista mallia kutsutaan talousmallin sivulta - miten vain.

Olen nyt ladannut Opasnettiin alla olevaan muuttujaan 101-ikäluokkaisen aineiston.

http://fi.opasnet.org/fi/Special:Opasnet_Base?id=op_fi4433.pneumokokki_vaestossa

Tämä data on myös liitteenä "IPD_data.dat" (101*27 taulukko).

IPD= tapausten lukumäärä Suomessa per vuosi. Huom 1: Tiedostossa "IPD_data.dat" on kokonaislukuja, jotka tulee jakaa 1000:lla, jotta saadaan tapausten lukumäärä Suomessa per vuosi. Huom 2: tässä aineistossa carriage on ikäluokkakohtainen serotyyppijakauma, jonka summa on joka ikäluokassa 1000 (siis kantajuuosuudet promilleina). Tämä ei haittaa, koska mallissa ratkaisevia ovat vain suhteelliset kantajuusosuudet.

Tämän datan pitäisi sopia nykyiseen epidemiologisen mallin koodiin yhtä lailla kuin nykyinen kahden ikäluokan datakin.

(sivuhuomautus: Jostakin kumman syystä onnistuin kyllä ajamaan vastaavanlaisen 7 luokan datan (http://fi.opasnet.org/fi_wiki/index.php?title=Markunkoesivu&oldid=25237) mutta tästä 101-luokkaisesta versiosta koodi ei pitänyt.)

Ovariable VacIPD tuottaisi siis taloudellista mallia varten 101-pituisen vektorin joka vastaa IPD tapausten määrää Suomessa kussakin yhden vuoden pituisessa ikäluokassa per vuosi valittua rokotetta käytettäessä (tai ilman rokotuksia).


ICER LASKENTA

Heiniltä saamme ensi viikon alussa 101-pituisen vektorin "QALYs_lost" (QALY=quality adjusted life year), joka vastaa kussakin ikäluokassa yhden IPD tapauksen keskimäärin tuottamaa QALY-menetystä ottaen huomioon kuoleman tai pysyvän haitan tms riskin ja näiden arvioidut elämänlaadun menetykset QALYina.

Samaten saamme vektorin "Costs_incurred", joka vastaa kussakin ikäluokassa yhden IPD tapauksen keskimäärin aiheuttamia terveydenhuollon kokonaiskustannuksia.


Siten rokotteen PCV_X jälkeen rokotusten ja IPD tapausten kustannukset ovat (sinisellä merkityt ovat 101-vektoreita)

Cost(PCV_X) = cost_of_vaccine(PCV_X) + sum(IPD(PCV_X) * Costs_incurred)

ja IPD tapausten haitat QALYina ovat

QALYs(PCV_X) = sum(IPD(PCV_X) * QALYs_lost)

R-koodi, jossa nämä lasketaan PCV10 ja PCV13 rokotteille on tiedostossa "Cost_Eff_model_run.R". Tässä käytetään funktiota "Vaccination", jota opasnetissä vastaa ovariable "VacIPD".


Koodissa käyttäjä antaisi siis inputtina kahden rokotteen, PCV_A ja PCV_B hinnat ja serotyypit ja saisi tuloksena taulukon:


       QALYs gained               incremental QALYs gained     incremental costs          ICER
       -------------------------  --------------------------   ---------                  ----------------   
PCV_0   0                          0                            Cost(PCV_0)                0
PCV_A   QALYs(PCV_0)-QALYs(PCV_A)  QALYs(PCV_0)-QALYs(PCV_A)    Cost(PCV_A)-Cost(PCV_0)    inc.costs/inc.QALYs.g
PCV_B   QALYs(PCV_0)-QALYs(PCV_B)  QALYs(PCV_A)-QALYs(PCV_B)    Cost(PCV_B)-Cost(PCV_A)    inc.costs/inc.QALYs.g


missä PCV_0 on tilanne, jossa ei rokoteta. Tässä oletetaan, että A ja B on järjestetty siten, että QALYs(PCV_A)>QALYs(PCV_B).

References

  1. Soini et al 2012
  2. Leussu 2011.