Modelling early iInitiation of HAART to forestall AIDS-related lymphoma in South Africa: a cost-effectiveness analysis.

Date
2017-03
Journal Title
Journal ISSN
Volume Title
Publisher
Stellenbosch : Stellenbosch University
Abstract
ENGLISH SUMMARY: The burden of cancer in low and middle income countries has been projected to shift from 59% (in 2012) to 65% (by 2030) of all cancer cases globally. Since the introduction of highly active antiretroviral therapy (HAART) in South Africa, there has been a marked decline in AIDS-related illnesses and premature death. However, this has not been the case with AIDS-related lymphoma which was observed to be on the rise despite the large HAART roll-out programme. It was discovered that 37% of all lymphoma cases diagnosed in 2009 in the Western Cape province, were HIV-related, indicating a remarkable increase from 5% in 2002. The ineffectiveness of HAART in reducing the incidence of lymphoma is largely attributed to late commencement of treatment. However, increasing HAART coverage is still a major challenge to many developing countries in Africa as well as South Africa due to limited resources. Therefore, increasing early HAART initiation has a cost implication that needs to be investigated in resource limited settings. In this study we used a deterministic compartmental model to investigate the potential impact of initiating HAART at different CD4 cell count levels on the incidence of lymphoma in HIV-infected individuals. We also developed a linked transmission and health state transition (Markov) model in TreeAge Pro to determine the cost-effectiveness of early HAART initiation from the public healthcare payer perspective. The CD4 count driven transmission model predicted lymphoma incidence in HIV-infected adults (aged 15 to 80 years) over a period of ten years. The Markov model predicted the health outcomes and costs. Data on transmission, transition probabilities, CD4 count thresholds, life expectancy, effectiveness and costs were obtained from literature. We compared early initiation of HAART at a CD4 cell count of greater than 500 cells= L to initiation at less than 500 cells= L, the current standard of care. Our primary outcomes were Quality-adjusted life years (QALYs), expected costs, net monetary benefit and the incremental cost-effectiveness ratio (ICER). Health outcomes and costs were discounted at a rate of 5% per annum as recommended in Southern Africa. We also performed deterministic sensitivity analyses to assess parameter uncertainty. Results indicated that early HAART initiation prevents lymphoma cases and related deaths, translating to 3.76 QALYs gained over the 10-year time horizon (6.47 vs. 2.71 expected QALYs with HAART initiation at greater than 500 cells= L and less than 500 cells= L, respectively). The incremental cost of early initiation was $14,613 compared to the alternative. The net monetary benefit (NMB) of early initiation was $90,581 and $30,063 for the alternative. HAART initiation at greater than 500 cells= L was therefore cost-effective with an ICER of $3,890/QALY gained. Sensitivity analysis showed outcomes were sensitive to the effectiveness of HAART in preventing lymphoma, with early initiation being more sensitive than the base case. Therefore, early HAART initiation at CD4 greater than 500 cells= L would not only be effective in forestalling AIDS-related lymphoma but also cost-effective in resource limited settings.
AFRIKAANSE OPSOMMING: Die las van kanker in ’n lae en middel-inkomste lande is geprojekteer vanaf 59% (in 2012) te skuif na 65% (in 2030) van alle kankergevalle wêreldwyd. Sedert die bekendstelling van hoogs aktiewe antiretrovirale terapie in Suid-Afrika, daar was ’n merkbare afname in vigs - verwante siektes en voortydige dood. Maar dit is nog nie die geval met vigsverwante limfoom wat waargeneem om op te gewees die opkoms ten spyte van die groot HAART uitrol program. Daar is vasgestel dat 37% van al limfoom gevalle gediagnoseer in 2009 in Wes-Kaap was MIV-verwante wat dui op ’n merkwaardige toename van 5% in 2002. Die ondoeltreffendheid van HAART in die vermindering van die voorkoms van limfoom is grootliks toegeskryf word aan die laat aanvang behandeling. Maar, die verhoging van HAART dekking is nog steeds ’n groot uitdaging om baie ontwikkelende lande in Afrika, insluitend Suid-Afrika as gevolg van beperkte hulpbronne. Daarom, die verhoging van die vroeë HAART inisiasie het ’n koste-implikasie wat gevolg moet word ondersoek in hulpbron beperk instellings. Ons gebruik ’n deterministiese kompartementele model om die potensiële impak van die inisiëring HAART op verskillende CD4-seltelling vlakke op die voorkoms van limfoom in MIV-geïnfekteerde individue te ondersoek. Ons het ook ’n gekoppelde transmissie en gesondheid toestand oorgang (Verborge) model in TreeAge Pro om die koste-effektiwiteit van die vroeë HAART inleiding bepaal uit die openbare gesondheidsorg betaler perspektief. Die CD4-telling gedryf transmissie model voorspel limfoom voorkoms van MIV-geïnfekteerde volwassenes (15-80 jaar oud) oor ’n tydperk van tien jaar. Die Markov model voorspel dat die gesondheid uitkomste en koste. Data oor die oordrag , oorgang waarskynlikhede, tel CD4 drempels, lewensverwagting, doeltreffendheid en koste is verkry uit die literatuur. Ons vergelyk vroeë aanvang van HAART op ’n CD4-seltelling van meer as 500 selle= L na inisiasie teen minder as 500 selle= L, die huidige standaard van sorg. Ons primêre uitkomste was lewe jaar Kwaliteit-aangepaste (QALYs), verwagte koste, netto monetêre voordeel en die inkrementele koste-effektiwiteit verhouding (ICER). Gesondheid uitkomste en koste word verdiskonteer teen ’n koers van 5% per jaar soos aanbeveel in Suider-Afrika. Ons het ook uitgevoer deterministiese sensitiwiteitsanalises om parameter onsekerheid te evalueer. Resultate het aangedui dat vroeë HAART inisiasie verhoed limfoom gevalle en verwante sterftes, vertaal na 3.76 QALYs wat oor die 10-jarige tydhorison (6,47 teen 2,71 verwag QALYs met HAART inleiding op groter as 500 selle= L en teen minder as 500 selle= L , onderskeidelik). Die inkrementele koste van vroeë aanvang was $14,613 in vergelyking met die alternatiewe . Die netto monetêre voordeel (NMB) van vroeë aanvang was $90,581 en $30,063 vir die alternatiewe . HAART inleiding op groter as 500 selle= L was dus kostedoeltreffende met ’n ICER van $3,890/QALY opgedoen. Sensitiwiteitsanalise het uitkomste was sensitief vir die doeltreffendheid van HAART in die voorkoming van limfoom, met vroeë aanvang om meer sensitief as die basis geval. Daroom, vroeë HAART inleiding op CD4 groter as 500 selle= L sou nie net effektief in forestalling vigsverwante limfoom, maar ook koste-effektief in beperkte hulpbronne instellings wees.
Description
Thesis (MSc)--Stellenbosch University, 2017.
Keywords
AIDS malignancies -- Treatment, Highly active antiretroviral therapy (HAART) -- Effectiveness -- Mathematical models, Highly active antiretroviral therapy -- Cost -- Mathematical models, Lymphomas -- Treatment -- Mathematical models, AIDS (Disease) -- Treatment -- Mathematical models, HIV infections -- Complications -- Mathematical models, Cancer -- Treatment -- Mathematical models, UCTD
Citation