GiveBetter x SF

San Francisco Community Health Center

Proposed expansion of integrated hepatitis C treatment

Research time: ~15 min on GPT-5.6 Sol Medium
  • Research — organization and evidence review.
  • Modeling — cost-effectiveness analysis.
  • Historical estimate for research done before time tracking.

Published: 7 September 2026.

Donate

Summary

What do they do? SFCHC’s Wellness Clinic provides hepatitis C testing, counseling and treatment. Its donation identity is Asian and Pacific Islander Wellness Center, Inc., EIN 94-3096109. This review tests expanding integrated clinical access for people otherwise facing referral-only care, not the whole organization or a confirmed new project. More

Why this approach interests us

  • Flexible on-site treatment can reach people who do not complete referrals, preventing future liver disease.

Our main reservations

  • The center already provides integrated care. Extra navigation may add little, and many patients may get treatment later without this gift.

What do you get for your dollar?

Our best guess is $1,997,838 per better life: 10 additional QALYs for an integrated clinical-access expansion. A favorable scenario reaches about $36,500, but depends on durable additional cures and inexpensive delivery. We would investigate a tightly priced expansion, not recommend an unrestricted gift on this estimate alone. Inspect the model →

our best guess
$1,997,838 — per 10 QALYs; hypothetical integrated-access expansion
positive scenarios
$36.5K–$309M — Joint analyst cases, not confidence bounds; null/harm possible
evidence
Very low for local price — External randomized cure effect; uncertain lifetime-health transfer
funding room
Unverified — Existing public services are not an additional funding offer

1. What do they do?

SFCHC’s Wellness Clinic provides hepatitis C testing, counseling and treatment. Its donation identity is Asian and Pacific Islander Wellness Center, Inc., EIN 94-3096109. This review tests expanding integrated clinical access for people otherwise facing referral-only care, not the whole organization or a confirmed new project.

Reach otherwise unserved patients

Establish a real barrier to getting clinical treatment, beyond a referral or test.

Deliver and confirm treatment

Flexible prescribing, medication access and follow-up lead to confirmed sustained viral response 12 weeks after treatment, not merely an appointment.

Count additional durable health

Subtract usual-care cures, discount later treatment and reinfection, and separate the gift from publicly financed medication.

Scope of this review. City contract objectives list 96 enrollments, 60 initial visits, 48 starts, 40 completed courses and 32 SVR4 results. These are targets, not actual outcomes. SVR4 is not the modeled SVR12 endpoint; the FY2025–26 file contains stale internal headers. No corresponding marginal dollar quote was verified.

2. Monitoring and information sharing

CTN-0064; HCV RNA-positive subgroup

Navigation-focused randomized counterweight. SVR12 occurred in 6/41 care-facilitation patients versus 5/53 controls.

Our assessment. Small, uncertain secondary cure results despite better care-cascade progression. This does not justify borrowing a 44-point integrated-care effect for navigation alone.

165 eligible adults with recent injection drug use; follow-up one year

Randomized trial, New York. SVR12: 55/82 integrated care versus 19/83 facilitated referral. The comparator already included navigation.

Our assessment. Not a navigation-only effect or a local measurement. Our expansion must reach people outside equivalent existing care.

84 intervention enrollees; national 2020 prices

Primary trial microcosting. $6,035 per enrollee, plus $3,030 physician downtime if charged and $4,677 program startup; drugs excluded.

Our assessment. We retain costs of two excluded enrollees using 84/82. Full-package costs are not a marginal SF quote or an incremental societal ICER.

UK reference patients; age and fibrosis scenarios

Lifetime disease simulation. Age-40 discounted marginal cure gains imply 2.3–6.4 QALYs.

Our assessment. Current local disease mix, later care and competing mortality are unknown. Our timing adjustments are explicit analyst approximations.

3. Qualitative assessment

Flexible on-site treatment can reach people who do not complete referrals, preventing future liver disease.

Key reservations

  • SFCHC already offers integrated care and has public support. Adding navigation inside that existing service cannot inherit the entire integrated-care-versus-referral effect. Deduplicate patients and shared health gains with SFAF, DOPE, HYA and medication-access programs; do not add independent lifetime gains to the same person without a joint model.
  • The 20-year timing curve, 70% later cure and excess mortality are judgment calls, not fitted local parameters. They drive the difference between a promising scenario and our central estimate.
  • The trial reinfection rate comes from only four events over 57.9 person-years. It does not establish a permanent local risk or predict reinfection for every patient.
  • Actual incremental drug starts and costs could differ from the transferred trial effect. Treating publicly financed drugs as zero resource cost would overstate cost-effectiveness.
  • A favorable $36.5K scenario is not the ranking estimate or proof of an available opportunity. Donation restrictions, capacity and patient outcomes require provider diligence.

Benefits not included in our estimate

  • Reduced onward transmission
  • Avoided future healthcare expenditure
  • Family welfare and income
  • Retreatment benefits and costs
  • Irreversible disease prevented before later usual-care cure

4. What do you get for your dollar?

Our best estimate: $1,997,838 per 10 QALYs

The external trial’s cure difference is 55/82 minus 19/83, or 44.2 percentage points. Retaining half locally and assigning half the expansion to the gift gives 0.110 additional one-year cures per eligible enrollee. Our uncertain health bridge retains 0.476 QALY per added cure, yielding 0.0526 QALY per enrollee. Clinical cost is $10,504: published 2020 costs, half physician downtime, eligibility-denominator reconciliation, startup spread over 100 people and a judgmental 1.35 cost multiplier. No usual-care clinical cost offsets or reimbursement are deducted.

A better life is our comparison unit of 10 additional quality-adjusted life years (QALYs), potentially spread across people. These are uncertain estimates, not measured returns or verified donation offers.

How we calculate the estimate

DOLLARS PER BETTER LIFE: 10 × donor cost ÷ [(55/82 − 19/83) × access transfer × funding additionality × net QALYs/cure − harms]
$1,997,838 per 10 QALYs

Model inputs and assumptions
Clinical cost per trial enrollee, 2020 USD
6035 (range: 6035 / 6035 / 6035). Kapadia microcosting: $6,035 excludes startup, physician downtime and DAA medication. Multiply by 84/82 to retain excluded participants' resources against the eligible efficacy denominator. Calibration or explicit analyst judgment; see basis.
Program startup, 2020 USD
4677 (range: 4677 / 4677 / 4677). Published $4,677; spread over an assumed 100 additional eligible people, not a verified expansion size. Calibration or explicit analyst judgment; see basis.
People sharing startup costs
100 (range: 100 / 100 / 100). 100-person hypothetical tranche; never equated with local contract targets. Calibration or explicit analyst judgment; see basis.
Full physician downtime per trial enrollee
3030 (range: 3030 / 3030 / 3030). Published $3,030; multiplied by scenario share and 84/82. Flexibility may require idle capacity; it is not automatically free. Calibration or explicit analyst judgment; see basis.
Downtime charged to this package
0.5 (range: 0 / 0.5 / 1). 0/50/100% judgment: integrated clinics may use some availability for other services. No local allocation verified. Calibration or explicit analyst judgment; see basis.
Local/current cost multiplier
1.35 (range: 1 / 1.35 / 1.7). 1/1.35/1.7 judgment, not an inflation series or SFCHC quote. Calibration or explicit analyst judgment; see basis.
Fraction of trial incremental access retained
0.5 (range: 1 / 0.5 / 0.2). Trial 55/82 minus 19/83 SVR12 difference. 100/50/20% judgment for otherwise referral-only people; could be zero inside existing integrated care. Calibration or explicit analyst judgment; see basis.
Share of expansion caused by gift
0.5 (range: 0.9 / 0.5 / 0.25). 90/50/25% judgment; current public service and replacement funding can make this zero. Calibration or explicit analyst judgment; see basis.
External lifetime QALY-per-cure anchor
4 (range: 6.4 / 4 / 2.3). MONARCH age-40 discounted bounds 2.3 (F0) to 6.4 (F4), algebra from 0.23–0.64 QALY per ten-percentage-point SVR improvement. Central 4 is a severity-mix judgment, not measured SF health. Calibration or explicit analyst judgment; see basis.
Effective health-benefit allocation window
20 (range: 20 / 20 / 20). 20-year heuristic, not a rerun of the lifetime Markov model. External total is allocated across this window only to stress-test timing and durability. Calibration or explicit analyst judgment; see basis.
Timing weight exponent
1 (range: 0 / 1 / 2). Annual weight year^shape: uniform / linearly later / quadratically later. Source does not provide this timing distribution; it is a major model judgment. Calibration or explicit analyst judgment; see basis.
Reference health discount rate
0.035 (range: 0.035 / 0.035 / 0.035). 3.5% matches MONARCH. Normalize the annual weights to the already-discounted anchor, avoiding a second discount of its total. Calibration or explicit analyst judgment; see basis.
Annual reinfection hazard approximation
0.069 (range: 0.02 / 0.069 / 0.12). Central .069 from 4 events/57.9 post-cure person-years in the external trial; 0.02/0.12 scenarios. Treating this short follow-up rate as a sustained hazard is speculative; no retreatment benefit included. Calibration or explicit analyst judgment; see basis.
Excess competing mortality hazard
0.03 (range: 0.01 / 0.03 / 0.06). 1/3/6% annually by judgment, additional to general mortality embedded in the reference model, not total mortality. Calibration or explicit analyst judgment; see basis.
Counterfactual eventual cure probability
0.7 (range: 0.2 / 0.7 / 0.9). 20/70/90% among people counted as additional one-year cures; judgment, not the trial control cure rate already subtracted. Calibration or explicit analyst judgment; see basis.
Years until counterfactual cure
3 (range: 5 / 3 / 1). 5/3/1-year judgment. No incremental health credit thereafter for that fraction. Does not model irreversible disease prevented before later cure. Calibration or explicit analyst judgment; see basis.
Resource cost per additional DAA start
15000 (range: 10000 / 15000 / 25000). $10K/$15K/$25K analyst assumptions, not current net Medi-Cal prices. Start difference 64/82 minus 22/83, scaled by access transfer; do not multiply by cure success again. This is a conservative gross-package resource sensitivity before funding displacement, not gift-attributable net resource use. Calibration or explicit analyst judgment; see basis.
Donor fraction of medication costs
0 (range: 0 / 0 / 0). Zero central conditional on public/assistance payment, unverified locally. Setting one charges full drug resources without changing health. Calibration or explicit analyst judgment; see basis.
Additional implementation harm
0 (range: 0 / 0 / 0). Zero central; explicit stress parameter for adverse effects and model omissions, not a measured harm estimate. Calibration or explicit analyst judgment; see basis.

Illustrative $100,000: 0.501 QALYs; not available funding room

  • Optimistic positive: $36,476 per 10 QALYs. 0.398 additional cures and 1.71 net QALYs per eligible enrollee.
  • Central analyst estimate: $1,997,838 per 10 QALYs. 0.11 additional cures and 0.0526 net QALYs per eligible enrollee.
  • Pessimistic positive: $309,143,568 per 10 QALYs. 0.0221 additional cures and 0.000513 net QALYs per eligible enrollee.

Uncertainty. No additional confirmed cures, zero funding additionality, or immediate universal counterfactual cure produces no positive benefit. Reinfection and later cure are not known locally. Treatment or implementation harms can outweigh gains; the central model assumes no extra harm.

Why not assign every cure its full lifetime benefit?

A UK lifetime model implies 2.3–6.4 discounted QALYs per extra cure at age 40 across fibrosis stages; we choose 4 before adjustments. This is not a San Francisco measurement. To test timing, we allocate that already-discounted total across 20 years using weights t/(1.035^t). We then reduce each year for reinfection, excess competing mortality and counterfactual cure: 70% are assumed cured elsewhere after three years. Retained weight is 11.9%, leaving 0.476 QALY. The annual allocation is a heuristic, not a reproduced disease-state model.

HEALTH RETAINED: 4 × Σ[w(t) × exp(−(0.069 + 0.03)t) × (t ≤ 3 ? 1 : 0.30)] ÷ Σw(t)
0.476 QALYs per additional one-year cure

QALY conversion assumptions
  • No reinfection, excess mortality or later cure: $237,751 per 10 QALYs. Keeps central costs and access assumptions but credits the full external 4-QALY anchor. This is deliberately favorable, not our estimate.
  • Immediate universal counterfactual cure: No finite positive price. If the same patients are cured just as soon elsewhere, the proposed access expansion adds no modeled health.
  • Maximum donor cost at central health effect: $526. Per eligible enrollee to beat $100K/10 QALYs. A threshold, not an available $526 offer; cutting services may also cut the effect.

The source already discounts health at 3.5%; normalizing weights preserves that total instead of discounting it twice. Timing is allocated from cure, without an extra delay from donation to cure. General mortality is already embedded; the 3% hazard is additional. Reinfection removes later credited benefit without retreatment. We give no credit after counterfactual cure and assume no additional harm thereafter; this approximation can miss both irreversible prevention and adverse differences.

5. Funding and previous grants

No HCV-specific priced expansion, verified funding gap or current incremental cure results were found.

Gross additional clinical package cost, with no deduction for usual-care clinical costs or assumed reimbursement. Donor pharmacy share is zero under an unverified public/assistance-payment scenario; gross package DAA resources before funding displacement are reported separately. Not a complete societal incremental cost-effectiveness analysis. Adding $15,000 per incremental medication start as a cost assumption produces $2,733,071 per 10 QALYs. This is not a verified drug price and excludes later healthcare, patient time and retreatment.

This review does not establish a verified marginal funding offer or a complete history of grants.

Donate

Opens the organization’s giving page. A general donation may not fund the specific activity modeled here; confirm allocation with the recipient.

6. Sources

  1. SFCHC Wellness Clinic. San Francisco Community Health Center. Provider disclosure. Published: Undated; retrieved: 2026-09-07.
  2. SFCHC donation identity. San Francisco Community Health Center. Provider disclosure. Published: Undated; retrieved: 2026-09-07.
  3. Eckhardt et al.: accessible HCV care randomized trial. JAMA Internal Medicine. Primary research. Published: 14 March 2022; retrieved: 2026-09-07.
  4. Kapadia et al.: co-located HCV microcosting. Drug and Alcohol Dependence Reports. Primary research. Published: 7 October 2022; retrieved: 2026-09-07.
  5. McEwan et al.: incremental SVR and lifetime QALYs. PLOS ONE. Primary research. Published: 2015; retrieved: 2026-09-07.
  6. SFDPH FY2025–26 CHEP program objectives. San Francisco Department of Public Health. Public contract objectives. Published: FY2025–26 filename; retrieved: 2026-09-07.
  7. CTN-0064 randomized HCV care facilitation trial. Open Forum Infectious Diseases. Primary research. Published: 2021; retrieved: 2026-09-07.

Annual expenses: years and sources

Average annual expenses (three consecutive fiscal years): $19,011,536. Organization size is separate from the modeled cost-effectiveness of a donation.

Asian and Pacific Islander Wellness Center, Inc. dba San Francisco Community Health Center

Form 990 reported whole-entity expenses; includes program, administration and fundraising costs, but excludes costs netted against revenue.

Fiscal years ended March 31. FY2024 uses amended return 202502239349300000 ($17,198,992), corroborated by FY2025 prior-year expense comparison; supersedes earlier FY2024 return showing $17,581,008.