GiveBetter x USA

End Overdose

Naloxone access, training and overdose prevention

Research time: ~47 min on GPT-6 Astra Light + GPT-6 Astra Medium
  • Research — reviewed programs, finances and impact evidence.

Updated: 2026-09-14

Donate

Summary

What do they do? End Overdose teaches people to recognize opioid overdose and respond with naloxone, rescue breathing and emergency assistance. It distributes prevention supplies through online certification, schools, college chapters, music events and community partners, including rural outreach in Arkansas. The main health benefit occurs when its work equips an otherwise unprotected witness to provide effective help early enough to change a person's outcome.

Why we’re interested in this organization:

  • Naloxone has a direct clinical mechanism, and the organization delivers tangible training and supplies rather than reporting awareness alone.

  • The original quarterly report and independent public-health partners document substantial real activity, including publicly supplied medication distributed through End Overdose.

  • The current giving page identifies the nonprofit and offers both a general fund and program designations, making a future, specifically costed community-delivery proposal practical to investigate.

Our main reservations:

  • No public evidence identifies how many additional high-risk recipients an ordinary gift would reach with usable medication, or how often they would add a rescue beyond existing protection.

  • Manufacturer donations, public supplies and funded partnerships may complement private delivery spending, but their inventory, restrictions, remaining funding and substitution effects are not reconciled.

  • Follow-up and financial reconciliations are incomplete; our estimate depends heavily on targeting and additional delivery, and several filing discrepancies need clarification.

What do you get for your dollar? $1.8M per better life: ten additional quality-adjusted life years in USA. Conditional naloxone-pathway estimate charging the full donation; other program effects are not estimated.

Nominal training-and-kit offers; additional offers; completed delivered two-dose kits; doses; timely usable access; unique witnessed overdose opportunities; additional timely administrations with unique-patient credit; additional three-day survivors; finite retained health. Reference per $1,000 of total gift: 15 nominal offers, 6 additional delivered kits, 12 doses and 0.00375 additional credited administrations. These are conditional model outputs, not measured services per dollar.

1. What do they do?

End Overdose combines education, direct supply distribution and outreach. Its 2026 campaign supports digital instruction, instructor certification, chapters and state operations as well as naloxone. Ordinary unrestricted support therefore cannot be modeled as a purchase of medication alone. This review counts health accruing to residents of the 50 states and DC; attendance at a US event does not by itself establish residence.

A July 2026 Arkansas announcement describes 12 grant-funded rural cabinets and reports 89,000 doses distributed in that state to date. Claims that cabinets were rapidly claimed show demand for placement, not independently confirmed installation, continuing restocking or extra rescues. The Arkansas bulk form offers free shipping and asks applicants how supplies will be used. This is a different delivery channel from the national store, where certified recipients are asked to pay shipping.

The official donation checkout identifies END OVERDOSE as a 501(c)(3). Its default is General Fund Project; other designations include Community Organizations & Treatment Facilities, Middle & High Schools, and Nightlife & Music Festivals. The receiving organization's original tax return identifies EIN 83-0696963 and endoverdose.net. This review remains a general-fund analysis: a menu designation alone does not establish an enforceable restriction or an unfunded delivery opportunity.

2. Monitoring and information sharing

The original July–September 2024 report reports 44,952 people trained and 71,733 naloxone doses distributed. It separately describes a school pilot with 1,797 kits; those kits are not added to the quarter's doses or treated as unique rescues. The 2024 filing reports over 230,000 trainees for the year. These measures have different periods and units and do not establish a matched cohort of trained people receiving and retaining kits.

Public partners corroborate distribution activity. The Overdose Response Strategy account says Southern Nevada Health District supplied 30,000 doses for EDC 2025. Los Angeles County's announcement describes joint festival outreach. Neither provides an experimental estimate of End Overdose's incremental mortality impact.

The report index still exposed the original summer PDF; the displayed annual-report label did not yield a separately retrievable 2024 annual document in this review. Public totals lack a reconciled stock-flow ledger, channel-specific costs, unique recipient follow-up, and linked rescue and survival records. Necessary monitoring would track certification, dispatch, receipt, usable carriage or access, witnessed overdose, other available rescuers, timely administration, three-day survival and subsequent mortality. Repeated trainees, shared scenes and repeat rescues of the same person require linked, privacy-preserving accounting.

Information quality also affects usable supply. The bulk-order page broadly says doses can remain effective 30 years after expiration. The study behind that familiar claim tested expired injectable solutions, not clinical effectiveness of this organization's intranasal inventory over 30 years. FDA guidance advises observing the printed expiry for the relevant product. This does not establish that End Overdose ships expired medicine; it means current product, storage and expiration records are needed before treating all stock as equally usable.

3. Qualitative assessment

The strongest case is for funding otherwise unavailable, reliable delivery to people likely to witness an overdose. Naloxone's clinical action is well established, but a drug's efficacy does not identify the value of another general-fund dollar. The Rhode Island simulation emphasizes saturation, targeting and witnessing; it is neither an End Overdose evaluation nor an estimate that transfers automatically to other states. The HEALing Communities randomized trial found no statistically significant primary mortality difference for its broader intervention bundle: adjusted rate ratio 0.91, 95% CI 0.76–1.09. Only 38% of implemented strategies had begun by the comparison year's start, so this is a caution about achieved implementation and timing, not a clean test of naloxone alone.

A Colorado festival survey found that 55.5% of 227 participants carried naloxone in some context and 8.4% had ever administered it. The selected, cross-sectional sample suggests both relevant exposure and substantial existing protection; its lifetime reports cannot be divided into an annual rescue rate. A Rikers visitor cohort reached 226 of 283 enrollees at six months; 40 respondents witnessed 70 overdoses. It supports the plausibility of higher-risk channels while illustrating repeated events and loss to follow-up. It is not a rate for the organization's students or event attendees.

Rescue also does not confer normal remaining life expectancy. In a Massachusetts EMS cohort, 7.5% of those alive three days after treatment died by one year. The related ED cohort found 5.5% one-year mortality, and a King County study reported 5.2 all-cause deaths per 100 person-years. These are different cohorts, endpoints and periods, not interchangeable exact hazards. They justify explicit continuing mortality and finite health horizons. Future mortality and utility in End Overdose's additional survivors remain unknown; patient-reported OUD utility research also cautions against assigning low quality of life solely from general-public stereotypes.

Our assessment is therefore conditional rather than an endorsement. Community and rural distribution could be worthwhile, especially where free stock does not reach likely witnesses. Broad awareness, reduced stigma, fentanyl-test use, care referrals and family effects may also matter, but are not valued here. Their omission is not evidence that they are zero, nor does it make the quantified pathway a proven lower bound on the whole organization's value.

4. What do you get for your dollar?

Our best-guess naloxone-pathway estimate is about $1.8 million per better life in the USA. It charges the full donation, not just the assumed naloxone allocation, but does not estimate other program effects. This is a judgment-based model, not a measured return or a verified offer of additional services.

The central assumptions allocate 75% of additional support to training and delivery, at $50 per offered two-dose package including paid staff, platform, fulfillment and overhead. After allowing for replacement funding and incomplete delivery, each $1,000 of total support enables six additional kits, or twelve doses: about $167 per additional kit. These are modeled services, not a ratio of historical expenses to reported doses. The calculation includes spending on unquantified programs and uses a larger normalization only to make rare downstream outcomes easy to inspect.

The model then assumes a 0.5% one-year probability of an opioid-overdose opportunity, 50% timely usable access, 50% additional help after other responders, and 50% unique-patient credit. The result is 0.375 additional administrations with unique-patient credit, or about $267,000 per additional credited administration. An assumed 20-percentage-point improvement in three-day survival produces 0.075 additional three-day survivors. A finite 20-year survival schedule, 7.5% first-year and 4% later annual mortality, utility 0.75, and 3% discounting gives 7.65 discounted QALYs per additional survivor. With an illustrative 99% USA-resident share, each $100,000 of modeled spending produces 0.568 USA QALYs: about $1.76 million per ten. None of the delivery or rescue assumptions is an End Overdose measurement.

Donated stock changes cash financing, not clinical exposure. Reducing the assumed package cash cost from $50 to $31 produces about 968 extra delivered kits and an illustrative $1.09 million per ten USA QALYs. The $19 difference is a CalRx twin-pack benchmark, not an invoice or a nationally available procurement contract. The manufacturer contribution's remaining usable quantity is unknown. Conversely, a joint higher-risk scenario yields about $65,000 per ten; it requires materially different targeting and additionality and is not a forecast.

All staff, administration, fundraising, acquisition and fulfillment must be paid somewhere. In addition to the gift, a diagnostic resource envelope values unpaid support at $10 per nominal offer, donated stock at $19 per nominal offer where applicable, and extra emergency response at $1,000 per credited administration. It is $115,375 in the reference and about $170,766 with donated stock, before unknown recipient time, uncovered shipping, future medical care and other downstream effects. This envelope deliberately covers nominal activity and is not a net incremental societal-cost estimate. Full resource costs cannot yet be estimated. The model does not call contributed drugs free resources or credit medical savings without evidence.

Zero added delivery produces zero benefit. A harm-only scenario sets survival benefit to zero and includes an illustrative acute loss of 0.001 present-value QALY per added administration; it yields negative net health, without a positive cost-effectiveness price. CDC's formulation comparison supports monitoring withdrawal and vomiting, but does not estimate harm from End Overdose or establish that it uses the higher-dose product. Delayed emergency care and false reassurance are unquantified risks, not observed organizational harms.

Model, assumptions and sensitivity

Whole ordinary general-fund gift from the donor's perspective. All G remains in the denominator; a is an explicitly hypothetical marginal program allocation, not the Form 990 program share. Cash cost c includes paid staff, overhead, fundraising allocation, training, stock where purchased, and fulfillment. No restriction or current $100,000 capacity is assumed. Nominal-activity resource envelopes separately add contributed medicines, unpaid support, recipient cash and emergency response. Future medical care, recipient time, exact postage receipts, opportunity costs and unpriced portfolio effects remain unknown; no full societal return is claimed.

USA means residents of the 50 states and DC. Domestic program and partner records support a predominantly domestic footprint but do not measure patient residence. All scenario shares g, including reference 0.99, are analyst judgments. Territorial and international attendee benefits fall outside edition QALYs and are not separately identified. No California, Bay Area or population-share allocation is inherited. Do not sum this report with nested state or metro benefits.

N=G*a/c nominal offered packages. A=N*b additional offers after funding replacement. D=A*q additional completed delivered two-dose kits; doses=2*D. R=D*e*k*t*d additional timely administrations with unique-patient credit. Z=R*f additional three-day survivors. S(0)=1; S(1)=1-m1; S(y)=S(y-1)*(1-m) thereafter. L=sum(y=1..T)[u*(S(y-1)+S(y))/2/(1.03^(y-0.5+delay))]. Q_all=Z*L-R*h; Q_USA=g*Q_all. Donor price for ten USA QALYs=10*G/Q_USA only when Q_USA>0. Resource envelope=G+N*(stock+unpaid+recipientCash)+R*ems, with omitted future net care and time explicitly unknown.

FY2022 recipient spending
516362 USD per calendar year (observed). Original 2022 Part IX; same EIN, 12 months, cash accounting. [eo-99022]
FY2023 recipient spending
942519 USD per calendar year (observed). Original 2023 Part IX; same EIN, 12 months, cash accounting. [eo-99023]
FY2024 recipient spending
970440 USD per calendar year (observed). Original 2024 Parts I and IX agree. Part III is not used for program allocation. [eo-99024]
Kit content
2 intranasal doses per national store kit (observed). Not a definition of the campaign's manufacturer units or all bulk order packaging. [eo-kit] [eo-bulk]
Current remaining usable inventory
null doses, by source and expiry (unknown). Campaign reports a past contribution; current stock-flow and fulfillment capacity are not published. [eo-campaign] [eo-kit]
CalRx acquisition benchmark
19 USD per twin-pack (observed). Current California procurement benchmark. It is neither an End Overdose invoice nor the opportunity cost of a donated medicine. [calrx-price]
Actual additional completed delivery per gift
null kits per USD (unknown). Requires allocation, funding replacement, inventory and labor constraints, shipment completion and recipient charges. [eo-bulk] [eo-ors]
G
100000 USD hypothetical ordinary gift (judgment). An arithmetic scale, not verified funding room or a requested transaction.
a
0.75 share of hypothetical gift (judgment). Assumed marginal allocation to the modeled pathway; all other spending remains in G. Do not infer it from the Part IX program ratio.
c
50 cash USD per nominal offered package (judgment). Illustrative comprehensive paid-delivery budget; $31 with contributed stock is a sensitivity, not an observed marginal cost.
b
0.5 share of nominal offers additional to other funding (judgment). Current partnerships and funded activity demonstrate a relevant replacement counterfactual; they do not identify this fraction. [eo-ors] [eo-arkansas]
q
0.8 additional offers yielding completed kit delivery (judgment). Explicit completion stage; certification alone is not receipt. [eo-kit] [eo-bulk]
e
0.005 one-year overdose-opportunity probability per additional delivered kit (judgment). Broad-audience reference. Neither lifetime festival experience nor six-month jail-visitor events establishes this rate. [naloxone-festival] [naloxone-rikers]
k
0.5 probability of timely usable access conditional on opportunity (judgment). Combines retention, carriage or accessible storage, correct use and product usability. Do not apply another adherence factor. [naloxone-festival]
t
0.5 share of potential responses adding timely help beyond other responders (judgment). Clinical counterfactual distinct from replacing money or medicines; remains unmeasured. [naloxone-zang]
d
0.5 unique-patient credit per additional response (judgment). Overlap proxy for repeat witnesses, scenes and patients. Actual deduplication data are missing. [naloxone-rikers] [naloxone-king-survival]
f
0.2 absolute probability of three-day survival gained per credited administration (judgment). Not the survival fraction among treated people. This cannot be estimated from reported reversals.
Observed EMS survivor mortality benchmark
0.075 risk of death by one year conditional on three-day survival (observed). Massachusetts 2013–15 cohort: 668 deaths among 8,927 three-day survivors, rounded. Not an End Overdose parameter. [naloxone-ems-survival]
m1
0.075 first post-rescue year mortality risk (judgment). Transfers the EMS benchmark illustratively to additional survivors; f already covers early three-day outcome. [naloxone-ems-survival] [naloxone-ed-survival] [naloxone-king-survival]
m
0.04 annual mortality after year one (judgment). Long-term extrapolation including recurrent overdose and other mortality; not identified by one-year cohorts.
T
20 post-rescue years (judgment). Finite reference horizon, with 10 and 30 years tested. Does not assert a 20-year life expectancy.
u
0.75 health utility while alive (judgment). Not measured in additional survivors. Patient-reported OUD research provides context but not a constant applicable to this population. [oud-utility]
g
0.99 share of health for residents of 50 states and DC (judgment). Domestic footprint supports predominant USA reach but exact residence remains unmeasured. [eo-quarter] [eo-ors]
delay
0.5 years from gift to rescue (judgment). Midyear rescue assumption; post-rescue annual-cycle health uses half-cycle survival and discounting.
h
0 gift-time present-value QALYs lost per added administration (judgment). Reference leaves acute harm unquantified; harm-only scenario uses 0.001. Neither is an observed harm estimate. [naloxone-harms]
unpaid
10 resource USD per nominal offered package (judgment). Diagnostic unpaid labor/support value; not observed opportunity cost.
stock
0 contributed-stock resource USD per nominal offered package (judgment). Zero means purchased stock is already inside c; donated-stock scenario adds $19 separately. [calrx-price]
recipientCash
0 recipient USD per nominal offered package (judgment). Reference assumes no extra receipts outside its paid fulfillment budget. Actual recipient shipping and time need reconciliation. [eo-kit] [eo-bulk]
ems
1000 resource USD per additional credited administration (judgment). Diagnostic emergency-response allowance, not an observed incremental transport/hospital cost.
Unpriced downstream net costs
null USD (unknown). Patient time, future net medical costs, alternative uses of contributed stock and other portfolio consequences remain unmeasured.

Conditional naloxone-pathway best guess: Cost: $100K; USA QALYs: 0.568348260281638; all-population QALYs: 0.5740891517996344. {"G":100000,"a":0.75,"c":50,"b":0.5,"q":0.8,"e":0.005,"k":0.5,"t":0.5,"d":0.5,"f":0.2,"m1":0.075,"m":0.04,"T":20,"u":0.75,"g":0.99,"delay":0.5,"h":0,"stock":0,"unpaid":10,"recipientCash":0,"ems":1000}

Available donated stock; otherwise the reference assumptions: Cost: $100K; USA QALYs: 0.9166907423897387; all-population QALYs: 0.9259502448381199. {"G":100000,"a":0.75,"c":31,"b":0.5,"q":0.8,"e":0.005,"k":0.5,"t":0.5,"d":0.5,"f":0.2,"m1":0.075,"m":0.04,"T":20,"u":0.75,"g":0.99,"delay":0.5,"h":0,"stock":19,"unpaid":10,"recipientCash":0,"ems":1000}

Added delivery reaches higher-risk witnesses: Cost: $100K; USA QALYs: 15.345403027604222; all-population QALYs: 15.500407098590124. {"G":100000,"a":0.75,"c":50,"b":0.8,"q":0.9,"e":0.05,"k":0.5,"t":0.75,"d":0.5,"f":0.2,"m1":0.075,"m":0.04,"T":20,"u":0.75,"g":0.99,"delay":0.5,"h":0,"stock":0,"unpaid":10,"recipientCash":0,"ems":1000}

Weak targeting, delivery and funding additionality: Cost: $100K; USA QALYs: 0.000027925212626205838; all-population QALYs: 0.00002939496065916404. {"G":100000,"a":0.5,"c":100,"b":0.1,"q":0.5,"e":0.001,"k":0.25,"t":0.1,"d":0.25,"f":0.05,"m1":0.1,"m":0.08,"T":10,"u":0.65,"g":0.95,"delay":0.5,"h":0,"stock":0,"unpaid":10,"recipientCash":0,"ems":1000}

Gift replaces funding and adds no delivery: Cost: $100K; USA QALYs: 0; all-population QALYs: 0. {"G":100000,"a":0.75,"c":50,"b":0,"q":0.8,"e":0.005,"k":0.5,"t":0.5,"d":0.5,"f":0.2,"m1":0.075,"m":0.04,"T":20,"u":0.75,"g":0.99,"delay":0.5,"h":0,"stock":0,"unpaid":10,"recipientCash":0,"ems":1000}

No survival gain and a small acute harm burden: Cost: $100K; USA QALYs: -0.00037125; all-population QALYs: -0.000375. {"G":100000,"a":0.75,"c":50,"b":0.5,"q":0.8,"e":0.005,"k":0.5,"t":0.5,"d":0.5,"f":0,"m1":0.075,"m":0.04,"T":20,"u":0.75,"g":0.99,"delay":0.5,"h":0.001,"stock":0,"unpaid":10,"recipientCash":0,"ems":1000}

Other portfolio effects: not estimated: Cost: $100K; USA QALYs: unknown; all-population QALYs: unknown. Education, referrals, test strips and other program effects are not identified. The central calculation values only the modeled naloxone pathway while charging full donation cost; it is not a proven lower bound.

Counterfactual: Without the gift, existing End Overdose funding, already funded cabinets, manufacturer and public stock, other distributors, trained witnesses and EMS remain. b represents the share of nominal offers made additional by the gift, after replacement funding. q represents actual completed delivery; k usable access at the opportunity; t clinical timing beyond other available response. These are distinct counterfactual steps. Public stock may enable donor-funded last-mile work rather than make it redundant, but the current overlap is unmeasured.

Attribution: Doses, packages, trainees, witnessed events, administrations and unique patients are separate units. d is only a scenario proxy for shared-scene and repeat-patient overlap, not a substitute for linked records. f is an absolute three-day survival advantage from an added credited administration, never the observed fraction surviving after naloxone. m1 applies after the modeled three-day endpoint, avoiding inclusion of early deaths in both f and later survival. Repeat rescues cannot each earn an independent lifetime gain; subsequent rescue and competing mortality are reflected in the finite survival schedule. No intervention-caused treatment retention or normal lifespan is assumed.

The central naloxone-pathway estimate is a subjective best guess, not a whole-portfolio expected return. Its inputs and joint stress tests are not probability-weighted estimates, confidence intervals or bounds. Older EMS and ED survivor cohorts do not identify the survival of otherwise-fatal cases rescued by this organization in 2026; future hazards and utility are explicit judgments. Endpoints, cohort composition and annual rates versus risks differ. Capacity constraints may make the linear gift model fail. Unpriced portfolio effects can be positive or negative, so the naloxone pathway is not a proven lower bound on the whole organization's value.

Sensitivity

  • Conditional naloxone-pathway best guess: $1,759,485 per ten USA QALYs. Holding the reference assumptions fixed, a $1 million price requires annual opportunity e=0.8797%; a $100,000 price requires e=8.7974%. No source establishes either threshold for marginal recipients.
  • Changing e from 0.5% to 5% reduces the reference price tenfold. Halving b, q, k, t, d or f separately doubles it; these are not six independent empirically estimated effects.
  • The donated-stock scenario lowers the donor-cash price to $1,090,880 per ten, but adds the medicine to the resource envelope. A historical donation does not establish recurring free inventory.
  • With all other reference inputs fixed, horizon, health utility and first-year mortality alter retained health; the accompanying calculation artifact reports the exact one-way sensitivities. A 30-year horizon is a scenario rather than normal life expectancy.
  • The joint higher-risk scenario changes e to 5%, t to 75%, b to 80% and q to 90%, producing $65,166 per ten. It requires a different implemented delivery channel; no probability is attached.
  • Zero added delivery or zero added clinical help produces zero quantified benefit. A negative net health scenario has no positive cost-effectiveness price. Very large finite downside prices are an arithmetic warning about approaching zero benefit, not an upper confidence bound.
  • USA share 95% rather than 99% changes a positive price by about 4.2%; geographic uncertainty is smaller than unmeasured targeting, completed delivery and rescue additionality in these examples.

Unresolved inputs

  • Costed, currently unfunded capacity and the share of unrestricted next dollars allocated to added high-risk delivery.
  • Manufacturer/public/purchased inventory reconciliation by dose, package, expiry, remaining stock, restrictions and distribution channel.
  • Actual fulfillment completion, recipient shipping receipts, paid and unpaid labor, and current workload constraints.
  • Marginal recipient exposure, usable access and other available witnesses or EMS; linked unique scene and patient follow-up.
  • Additional survival probability versus no added response, subsequent mortality, utility, adverse effects and future net costs.
  • USA-resident outcomes separated from territorial and foreign attendee spillovers.
  • Reconciled 2022–24 program totals and 2022–23 net-asset roll-forwards; current management accounts, debt, restrictions and funding commitments.
  • Effects of education, test strips, stigma, referrals and other portfolio activity, including possible harms; no whole-organization lower bound is established.

5. Funding and previous grants

The current campaign reports a manufacturer contribution of 25,000 naloxone units, but does not define that unit or reconcile receipts, inventory, dispatches, returns and expirations. It must not be combined automatically with the older public donation or converted into 25,000 two-dose kits. The national kit page says two intranasal doses, recipient-paid shipping and a 7–14-day shipping notice. Its extracted HTML contains a generic Sold out badge, but its product structured data says InStock and the purchasing control is restricted by certification. We therefore retract the initial report's implication that this text demonstrated a stockout. Neither the badge nor the structured data verifies a fulfillable donor-funded queue.

Public supply is a concrete alternative and potential complement. California's NDP provides free medication and shipping to eligible organizations; it does not distribute directly to individuals. This can leave real last-mile work for private support. The Nevada partner's donation and Arkansas' funded cabinets similarly do not prove that a new private gift adds inventory or capacity. A useful next proposal would specify the intended higher-risk channel, committed stock and other funds, unmet orders, additional paid hours, completed-delivery target and follow-up plan. No current costed, unfunded proposal meeting those conditions was located. An old match announcement in the 2024 report is not treated as a current match.

Annual spending appendix: the original 2022, 2023 and 2024 returns are all 12-month calendar-year, cash-basis filings for EIN 83-0696963. Part IX reports total spending of $516,362, $942,519 and $970,440, averaging $809,774. The respective program / management / fundraising splits are $464,735 / $51,627 / $0; $786,575 / $155,944 / $0; and $730,276 / $207,274 / $32,890. These are consistent total-cost accounting figures, not a marginal allocation to high-risk naloxone delivery or full values of contributed medicines and labor.

There are unresolved financial discrepancies. Part III program totals are $272,610, $1,199,424 and $980,427, respectively, and disagree with Part IX. The 2022 and 2023 Part XI net-asset roll-forwards miss $3,277 and $9,081; the latter Schedule O does not explain the difference. These discrepancies do not prove wrongdoing, but require reconciliation before relying on reserve or program-efficiency claims. At year-end 2024 the return reports $20,042 cash, $36,709 liabilities and negative $16,572 net assets; the 2024 roll-forward does reconcile. The filing index and targeted search yielded no newer original return. These historical figures cannot establish September 2026 runway, restrictions or funding room. The 2023–24 returns report compiled or reviewed statements, not audited statements.

Annual expenses

Organization-level spending, including programs, administration and fundraising. The research list averages three comparable, consecutive full fiscal years when available.

  • FY 2022: $516K; End Overdose, EIN 83-0696963, 12-month period, Original Form 990 Part IX total expenses; cash basis; calendar year; excludes unpriced ecosystem resources. Source
  • FY 2023: $943K; End Overdose, EIN 83-0696963, 12-month period, Original Form 990 Part IX total expenses; cash basis; calendar year; excludes unpriced ecosystem resources. Source
  • FY 2024: $970K; End Overdose, EIN 83-0696963, 12-month period, Original Form 990 Part IX total expenses; cash basis; calendar year; excludes unpriced ecosystem resources. Source

6. Sources

  1. 2022 original Form 990, calendar year, Parts I, III, IX, X, XI and XII. End Overdose / IRS, reproduced by ProPublica. Published: 2023-03-01; retrieved: 2026-09-14.
  2. 2023 original Form 990, calendar year, Parts I, III, IX, X, XI and XII. End Overdose / IRS, reproduced by ProPublica. Published: 2024-04-25; retrieved: 2026-09-14.
  3. 2024 original Form 990, calendar year, Parts I, III, IX, X, XI and XII. End Overdose / IRS, reproduced by ProPublica. Published: 2025-03-25; retrieved: 2026-09-14.
  4. 2023 original Form 990 Schedule O. End Overdose / IRS, reproduced by ProPublica. Published: 2024-04-25; retrieved: 2026-09-14.
  5. 2024 original Form 990 Schedule O. End Overdose / IRS, reproduced by ProPublica. Published: 2025-03-25; retrieved: 2026-09-14.
  6. Available filing index; latest return located is calendar 2024. ProPublica Nonprofit Explorer. Published: not stated; retrieved: 2026-09-14.
  7. Summer 2024 Impact Report, July–September, pages 4–8. End Overdose. Published: 2024-10-01; retrieved: 2026-09-14.
  8. Impact report index. End Overdose. Published: not stated; retrieved: 2026-09-14.
  9. August 2026 Overdose Awareness Month campaign. End Overdose. Published: not stated; retrieved: 2026-09-14.
  10. Naloxone Kit; visible description and product structured data. End Overdose. Published: not stated; retrieved: 2026-09-14.
  11. Official donation checkout: receiving organization and program designations. End Overdose / GoFundMe Pro. Published: not stated; retrieved: 2026-09-14.
  12. End Overdose Distributes Free Naloxone Boxes Throughout Arkansas. End Overdose. Published: 2026-07-02; retrieved: 2026-09-14.
  13. Arkansas bulk naloxone order, product units and fulfillment terms. End Overdose. Published: not stated; retrieved: 2026-09-14.
  14. Nevada ORS Team Facilitates Partnership for Naloxone Distribution. Overdose Response Strategy. Published: not stated; retrieved: 2026-09-14.
  15. County announces End Overdose partnership at 2025 music and cultural festivals. Los Angeles County Department of Public Health, issued through PR Newswire. Published: 2025-01-23; retrieved: 2026-09-14.
  16. Naloxone Savings; $19 twin-pack price effective January 1, 2026. California CalRx. Published: not stated; retrieved: 2026-09-14.
  17. Naloxone Distribution Project FAQ, including April 2026 update. California Department of Health Care Services. Published: not stated; retrieved: 2026-09-14.
  18. Evaluation of Strategies to Enhance Community-Based Naloxone Distribution Supported by an Opioid Settlement. Zang et al. / JAMA Network Open. Published: 2024-05-30; retrieved: 2026-09-14.
  19. Community-Based Cluster-Randomized Trial to Reduce Opioid Overdose Deaths; primary abstract in author institution repository. HEALing Communities Study Consortium / NEJM; Weill Cornell Medicine. Published: 2024-06-16; retrieved: 2026-09-14.
  20. Naloxone use, carriage and prior training in 227 Colorado festival attendees. Piercey, Pince and Karoly / Journal of Addiction Medicine; PubMed. Published: 2025-12-23; retrieved: 2026-09-14.
  21. Witnessed overdoses and naloxone use among trained visitors to Rikers Island. Huxley-Reicher et al. / Addictive Behaviors; PubMed. Published: 2017-11-16; retrieved: 2026-09-14.
  22. One year mortality of patients treated with naloxone for opioid overdose by EMS. Weiner et al. / Substance Abuse; PubMed. Published: 2020-04-03; retrieved: 2026-09-14.
  23. One-Year Mortality after ED Treatment for Nonfatal Opioid Overdose. Weiner et al. / Annals of Emergency Medicine; PubMed. Published: 2019-06-20; retrieved: 2026-09-14.
  24. Overdose and mortality after a nonfatal EMS-treated overdose in King County. Hood et al. / Drug and Alcohol Dependence; PMC. Published: 2023-10-31; retrieved: 2026-09-14.
  25. Patient-reported health-state utilities in NIDA treatment-study data. Quality of Life Research; author manuscript hosted by PMC. Published: not stated; retrieved: 2026-09-14.
  26. Comparison of 8-mg and 4-mg intranasal naloxone in New York law enforcement. Payne et al. / CDC MMWR. Published: 2024-02-08; retrieved: 2026-09-14.
  27. Quality Assessment of Expired Naloxone Products from First-Responders' Supplies. Research article / PubMed. Published: not stated; retrieved: 2026-09-14.
  28. FDA shelf-life extension announcement for newly manufactured NARCAN 4-mg nasal spray. US Food and Drug Administration. Published: not stated; retrieved: 2026-09-14.