Issues in TB care and financing Gesine MeyerRath

Issues in TB care and financing Gesine Meyer-Rath Health Economics and Epidemiology Research Office (HE 2 RO) University of the Witwatersrand/ Boston University BEMF consolidation meeting 15 July 2011

Why TB financing? • HIV care and treatment is better planned and funded than ever before in South Africa • Lots of cost and epi data is known • HIV care has ear-marked, national-level funds (Conditional grant) • HIV care has own vertical system including own distribution channels for drugs • TB diagnosis and treatment uses outdated tools and drugs and produces shoddy outcomes • Very few cost data available, not much epi data • TB care is funded at provincial level via Equitable share, like all other provincial programmes • TB care is highly localised and integrated into horizontal systems

TB issues PREVENTION Uninfected DIAGNOSIS Suspect Case TB TREATMENT Cured MDR-TB TREATMENT Cured XDR

TB prevention PREVENTION Uninfected Suspect

TB prevention • Early HIV diagnosis and ART initiation – HCT: 11. 9 million people tested, but no link to TB diagnosis and treatment, not enough link to ART – ART initiation at >200 CD 4 cells/µl • <350 for patients with TB since April 2010 • Initiation at <500 or irrespective of CD 4 cell count? • CMX and INH prophylaxis for HIV+ves – IPT: INH prophylaxis for 6 mts irrespective of CD 4 cell count after excluding active TB – INH stock-outs as a result of badly planned demand – INH/CMX combination possible but still in development

TB diagnosis DIAGNOSIS Uninfected Suspect Case

TB diagnosis: Issues 1. Coverage: Case finding – Intensified Case Finding (ICF) campaign: visited 41, 000 families and screened 112, 000 contacts since Feb 2011 2. Sensitivity of current diagnostics (smear microscopy) 3. Time to result (culture, LPA, DST) 4. Cost of newer diagnostics (Xpert MTB/RIF, lineprobe assay, drug-susceptibility testing)

TB diagnosis: Current situation • Current algorithm starts with smear microscopy • Smear microscopy of sputum technology more than 125 years old • Sensitivity of smear microscopy (ability to find TB when it is there) is approximately 70% • Sensitivity of smear microscopy in HIV+ population drops to 35 -45% • Current algorithm depends on culture • Culture highly sensitive (considered ‘gold standard’) • However, takes 2 -6 weeks for results • High rates of contamination, missing results

Possible solution: Xpert MTB/RIF • The Xpert MTB/RIF (Cepheid) test detects TB and rifampicin resistance (indication of MDR-TB) in less than 2 hours • The sensitivity of Xpert MTB/RIF has been shown to be 86% in a SA demonstration study • HIV status does NOT affect sensitivity of Xpert

Patients with suspected pulmonary. MTB Current guidelines Patients with TB history VISIT 1 if negative TB culture + LPA +/- DST Stop Sputum if positive microscopy if negative TB diagnosis VISIT 2 Patients without TB history Sputum microscopy if positive Antibiotics Chest X-ray if negative Stop Gene. Xpert if negative Antibiotics if resolved Stop if failed HIV positive/ status unknown Antibiotics Chest X-ray TB culture + LPA +/- DST if negative HIV status HIV negative Antibiotics if resolved Stop if failed Chest X-ray RIF resistant/ inconclusive TB culture + LPA +/- DST TB diagnosis if positive TB diagnosis Stop Sputum microscopy Antibiotics TB culture + LPA +/- DST if unsuccessful Chest X-ray if positive Repeat Gene. Xpert if negative TB diagnosis Stop if negative Sputum microscopy TB culture + LPA +/- DST if positive HIV positive/ status unknown (MTB+/RIF+) Stop VISIT 3 if negative if unsuccessful if positive TB diagnosis HIV status HIV negative All patients TB diagnosis HIV status HIV positive/ status unknown Xpert algorithm if positive TB diagnosis Antibiotics HIV status HIV positive HIV negative TB culture + LPA Chest X-ray Stop if negative Stop if positive TB diagnosis

Xpert roll-out: Cost input Cost item Cost in 2011 ZAR TB diagnostics Xpert MTB/RIF test Sputum microscopy (fluorescent microscopy) TB culture (liquid medium, growth) TB culture (liquid medium, no growth) Line probe assay (LPA) for all positive cultures Drug susceptibility test (DST) (first-line drugs only) Chest x-ray Antibiotic trial Clinic visit: Nurse Clinic visit: Physician TB treatment - First-line treatment (non- resistant) - Second-line treatment (non-resistant) - RIF monoresistance - INH monoresistance - Multi-drug resistance (outpatient care only) R 190 -220 R 26 R 114 R 87 R 189 R 519 R 110 R 19 R 71 R 130 R 3, 441 R 6, 806 R 29, 677 R 6, 275 R 29, 677

Xpert roll-out: Results of National TB Cost Model Xpert leads to an increase of • 30% in TB cases diagnosed • 76% in MDR cases identified • 39% in number of patients treated • 55% in the cost of the TB diagnostic programme • 32% in the outpatient cost of the TB treatment programme if scaled-up fast Under the Xpert scenario, • 87% of diagnosed patients are diagnosed by Xpert • 83% are diagnosed after the first visit

TB treatment Uninfected Suspect Case TB TREATMENT Cured MDR

Patients with diagnosed pulmonary. MTB Patients without TB history (New cases) Patients with TB history (Retreatment cases, old guidelines only) Regimen 1 (RHZE for 2 months, RH for 4 months) Regimen 2 (RHZES for 2 months, RHZE for 1 month, RHE for 4 months) Patients with INH mono-resistance Patients with RIF mono-resistance Patients with multi-drug resistant TB (MDR-TB) RHZE for 2 months, RH for 4 months + OFL for 6 months KNM, ETH, PZA, OFL+ TZ for 6 months + ETH, OFL+ TZ for 18 months KNM, ETH, PZA, OFL + TZ for 6 months + ETH, OFL + TZ for 18 months + Inpatient care until culture -

TB treatment: Non-resistant TB • Community-based treatment • Good coverage, but low quality: – Low completion rate – Low cure rate – High MDR and XDR rate • 2008 NHLS data: 6, 219 MDR cases and 576 XDR cases • Community-based system needs restrengthening – PHC revitalisation – Community health-worker framework – Intensified case-finding (ICF) campaign

MDR/ XDR-TB treament Uninfected Suspect Case Cured MDR-TB TREATMENT Cured XDR

MDR-TB treatment: Issues 1. Cost of drugs and monitoring of side effects 2. Centralised vs. decentralised care; inpatient vs. community care – – – Infection control Patient autonomy Operational challenges • • Drug supply chains Administration of daily injectibles for 6 months 3. Improving outcomes – Cure rates between 33% and 45% – Default rates up to 70% – Death rate at Tugela Ferry after 1 year 75% (2005 -2007)

MDR-TB treatment: Current situation • MDR-TB care differs vastly by province – KZN: mostly community care; GP: exclusively centralised care • Current guidelines prescribe inpatient care until culture negative (2 cultures taken 30 days apart) – currently at specialised MDR/ XDR hospitals (9 specialised, 11 decrentralised units) – 73% of known MDR/ XDR cases in 2008 were hospitalised – on average 4 months – not enough capacity (1, 854 beds in Feb 2010), waiting list – at average cost per patient-day equivalent (R 1, 069) 4 months cost R 163, 000

MDR-TB treatment: Decentralised care • Decentralisation plan for MDR care approved by National Health Council • Rationale: – Low capacity and infection risk while waiting – Nosocomial transmission of MDR/XDR-TB – Patients abscond due to lack of recreational facilities, family responsibilities and lack of income – Even if ambulant, monthly trips to centralised hospitals are difficult – Decentralised care is uncoordinated and chaotic

MDR-TB treatment: Decentralised care • Planned guidelines: – MDR-TB treatment at MDR units in TB hospitals (district and sub-district level) and at PHC clinics – Inpatient treatment for 8 weeks or until 2 consecutive sputum smears negative – Follow-up at community level by PHC staff or mobile units and DOTS plus supporters

Summary: Cost issues I • Very few data available – Cost of diagnosis modelled – Cost of treatment: • 1 study of community-based treatment (Sinanovic et al Int J TB Lung Dis 2003) • 1 study on cost of care for HIV-co-infected patients (Cleary et al Cost Effect Resource Alloc 2006) – No data on cost of MDR care • Studies on real cost of Xpert roll-out and on cost of MDR-TB inpatient care planned

Summary: Cost issues II • ICF: more suspects (target: 10% growth every year) • Xpert: 30% more cases, 76% more MDR cases, 39% more patients on treatment Higher cost • ICF: lower smear sensitivity in suspects • Xpert: more cases treated, lower transmission, less new cases (including of MDR-TB) • MDR/XDR-TB treatment decentralised with less inpatient care Lower cost