Future direction · not part of the first pilot

Six in ten referred women never reach colposcopy. The screening already found them.

Bangladesh’s national programme screened 1.56 million women in five years and recorded every positive result. Then it loses most of them at the next step. The Prevention Adoption Initiative is a weekly routine for that step: find who is overdue, rank by urgency, act with a call and a booking and the fare, close when it is done. This page shows how it could run here, first on cancer screening and then on the chronic conditions a corporate check-up finds, who would do what, and who would pay. The design is published; the first pilot is in Michigan; this is the Bangladesh adaptation, ready for a first conversation.

Talk to me, 30 minutes →The one-page brief (PDF) →

40.4%

of VIA-positive women reached colposcopy, 2018–2023. Six in ten did not.

32%

of newly diagnosed hypertension and diabetes patients at NCD corners did not return within three months.

This week’s list · invented rows2 hours · the provider’s nurse

1Rina B.VIA-positive · 70 days, no colposcopyDone, day 76
2Nasima K.VIA-positive · 48 days · missed one bookingBooked, fare sent
3Shirin A.VIA-positive · 23 daysCalled in Bangla

Ranked by days overdue, days since the positive result and prior no-shows. Three rules, no model. The factory never sees this list.

How it moves · every stage ends with its own number

Michigan, one clinicThree months of feasibility, then the clinic decides whether to go to month twelve. An independent evaluator, once one is confirmed, judges against the clinic’s own trend; without one, the result is the clinic’s own before-and-after.
Bangladesh, one providerModel A: one provider, one factory client, three months, then the same decision Michigan makes: stop, or continue to twelve months. The evaluator here is a university or research institute, named before the start (see 06).
Then the programmeModel C goes to the national programme on model A’s number, not before. Each stage is judged by the number it set out to move.

01 · The gap

The system finds people. It loses them after the result.

One national cohort, Jan 2018 to May 2023: 1,557,002 women screened by VIA at 497 facilities and 6,398 community clinics. What happened to the women with a positive result:

VIA-positive3.3% of women screened

51,913

Attended colposcopythe step this page is about

20,95440.4%

Treatedonly those who needed it

6,205

Histopathology on recordthe registry cannot store it

3,079of 6,205

Nessa et al., BMC Global and Public Health 2025;3:34, from the national DHIS2 case-based registry, which already sends each positive woman an SMS. The same shape appears outside cancer: 32% of new NCD-corner patients lost within three months, mostly to medicine stock-outs, transport cost and queues (BMC Health Services Research, Jan 2026); BRAC’s NCD360 could track 22% of the patients it referred into government facilities (Daily Waadaa, Sep 2026); factory VIA clinics treat about half of abnormals on site and refer the rest, with nothing published on the referred half (Jhpiego, Jan 2026).

Nobody publishes a completion rate for factory or check-up referrals.The national programme reports one five-year figure, 40.4% to colposcopy. For women referred from a factory camp or a corporate check-up, how many reached the next step is not published. That is the number this routine produces.

02 · The routine

Two hours a week, one short list, four steps.

The provider’s own nurse runs it with a worklist. HSREP keeps the list and the rules. An independent evaluator judges the number. No software to buy, no new staff.

01 · Find
Everyone who was told to get a test and has no result recorded.
02 · Rank
Days overdue, days since a positive result, prior no-shows. Three rules, no model.
03 · Act
A call in the patient’s language, a kit, a booking, help with transport.
04 · Close
The test is done, and the days from referral to result are counted.

Invented patient · what one row looks likeRina B. · VIA-positive at a factory screening camp · 70 days, no colposcopy → called in Bangla, booked at the medical college colposcopy clinic, fare sent by bKash, reminder the day before, a paid morning off agreed with the factory → done on day 76.

Paid per completion, never per referral.Doctors paid a commission for each test have no reason to chase a patient who already paid, and patients who suspect padded tests have a reason not to come back. In this routine the outreach nurse is paid for the list, and the provider is paid when the woman finishes.

03 · Three ways it could run

Same routine, three places to put it.

In order. The first is small enough to start with one provider and one factory client, and it is where the money for screening is already being spent.

A · Start here

The referral tail of factory screening

A private provider already screens garment workers in the factory and refers the women it cannot treat on site. The routine follows those women to the public colposcopy clinic.

The list: The provider’s own register of referred women. The factory never sees a result.
Who acts: The provider’s outreach nurse, two hours a week. The factory doctor or welfare officer releases the right women on the right days.
The test: Free at the public colposcopy clinic. Private only where the public clinic is far or slow, at a fixed price.
Who pays: The sponsor that already pays for the screening, for example a brand programme, UNFPA, or the factory as a corporate client. A small fee per completed follow-up, plus the fare.

MeasureShare of referred women reaching colposcopy within 90 days, against the national 40.4%.

B · Alongside A

A completion contract on corporate check-ups

A network already sells annual check-ups to employers. Every round produces high blood pressure readings, HbA1c over the line, abnormal creatinine. Then nothing happens. The routine follows each abnormal result to its next step inside the network.

The list: Abnormal results from the network’s own corporate check-ups with no next step recorded. See the finish lines below.
Who acts: Its outreach staff, paid for the list. No doctor is paid for the referral. One transparent price for the next step.
The test: In-house: a repeat reading, a confirmatory test, a doctor visit, the yearly eye and kidney tests.
Who pays: The network, and the employer through the check-up contract. It is buying completion from people it already tested, and a product every employer can buy: completion, not a check-up.

MeasureShare of abnormal results that reached their next step within the window, per employer.

C · After A

A follow-up cell inside the national programme

Two or three upazilas use the registry’s existing list of VIA-positive women and add what the SMS cannot do.

The list: Already in the DHIS2 tracker, with a national ID.
Who acts: The community clinic’s own provider: a phone call, a booked slot, the fare; colposcopy and histopathology written back.
The test: Public, with contracted private capacity at a fixed price where the public clinic is overloaded.
Who pays: It could come from programme money already committed for 2027–2031, the EIB Women’s Health Initiative with UNFPA technical assistance. None of it is agreed for this.

MeasureThe national 40.4% in those upazilas, a year later, and the share of treatments with histopathology on record.

Each stage ends with its own number, judged by an independent evaluator named before the start, and the next stage begins on that number.

03b · Beyond cancer screening

The same routine on a corporate check-up: three conditions, six finish lines.

Chronic disease is where the corporate check-up finds the most abnormal results and does the least about them. Each line below is a recommended step that people start and do not finish, with the size of the condition in Bangladesh and the finish line the pilot would count. Stress and lifestyle are real causes and bad finish lines: they belong in the Act step, never in the measure.

01 · Hypertension · first
21% of adults have raised blood pressure; nationally about 14% of those with it are controlled, and in the public programme that tracks every patient, 58% of those registered were controlled at their last visit.
Finish lines: a repeat reading on another day and a doctor visit, within the month; a repeat reading at 90 days after treatment starts.
02 · Diabetes · second
13% of adults, about 13.9 million people, and 39% of them do not know. HbA1c is already in the check-up.
Finish lines: a second test to confirm, with a doctor visit, within the month; HbA1c at 90 days after treatment starts; for everyone already diagnosed, the eye exam and the urine-albumin and eGFR check, at least yearly.
03 · Kidney · rides on the panel
About 22% of adults in pooled studies, almost always found late. Creatinine is already in the panel, so an abnormal result costs nothing extra to find.
Finish line: a repeat creatinine and urine-albumin test to confirm; chronic means abnormal for three months or more.
For factories · anaemia
About eight in ten female garment workers are anaemic in published samples. Haemoglobin is cheap and pairs with the VIA camp.
Finish line: a repeat haemoglobin after three months of iron.

Raised blood pressure 21.0% and raised blood glucose 8.3%, adults 18–69: WHO STEPS Bangladesh 2018. Control about 14% nationally, and 58% of registered patients at their most recent visit in the Bangladesh Hypertension Control Initiative, 328 facilities, March 2026: National Heart Foundation. Diabetes 13.2% of adults 20–79 (age-standardised), 13.9 million, 39.1% undiagnosed, 2024: IDF Diabetes Atlas. CKD 22.48% pooled: Banik et al., Int Urol Nephrol 2021. Anaemia in female garment workers: Hossain et al., Nutrients 2019. Finish lines follow the Bangladesh National Hypertension Guideline, the ADA Standards of Care and KDIGO 2024; the windows (the month, 90 days) are the pilot’s own and would be set with the clinical lead.

Declined for now: lipids on their own (the next step is a risk decision, not a test), stress and mental health (no finish line a nurse can close), fatty liver, tobacco.

04 · Who puts in what, who gets what

Why each party would say yes, for model A.

Nobody is asked to do a job that is not already theirs, and nobody pays for a test the public clinic gives for free.

Party	Puts in	Gets
Private provider	Its nurse, two hours a week on the worklist	The colposcopies, biopsies and pathology on referred cases where private, by contract; a per-completion fee; a completion rate to sell to the next corporate client
Factory	The worker’s paid morning; a room on screening day; fifteen minutes a week from the welfare officer	Compliance evidence for buyers; fewer women lost to long illness
Sponsor (brand, UNFPA or the factory as client)	The screening it already funds; a small fee per completed follow-up; the fare	A completion rate per factory instead of a screening count
Public programme	Nothing new. The colposcopy clinics exist	Its 40.4% moving; results written back to the registry when the camp runs under the programme
The woman	Nothing	Reaches the next step, with the morning paid and the fare covered
HSREP	The worklist, the three rules, the training, the written routine. The sponsor or the provider pays for the founder’s time; HSREP adds no fee on top	The first published completion number for factory and check-up referrals, and the toolkit

05 · In plain words

What I would say in the room.

If you run a private networkYou already screen factory workers. Around six in a hundred have an abnormal result; some you treat that day, the rest you refer, and then nobody knows. Your nurse keeps the list of referred women and works it two hours a week: call, book, send the fare, remind, write down when it is done. The follow-up cases come to you by contract, and you get a number no one else here has. One network, three months, one factory client, then a decision: stop, or continue to twelve months.

If you fund factory screeningYour screening finds the women. The programme loses them at the next step: nationally, four in ten reach colposcopy. Add a small fee per completed follow-up and the fare. The test itself is free at the public clinic. You get, for the first time, a completion rate per factory instead of a screening count.

If you run the national programmeYour registry already lists every VIA-positive woman and sends her an SMS. In two or three upazilas, add a phone call from the community clinic, a booked slot, the fare, and the result written back, including histopathology. The measure is your own 40.4%, in those upazilas, a year later.

06 · Who could matter

Organisations that fit, none of them confirmed.

Named because of what they already do, not because they have agreed to anything. None has been approached on behalf of this page. Disclosure: I worked at Praava Health as Head of Marketing & Corporate Sales until October 2023. I have no current role or financial interest there.

Providers (model A, B)

The ten providers under BGMEA’s member-care MoUs of Nov 2025, including Praava, United, LabAid and Popular
Surjer Hashi Network
Grameen Kalyan

Screening and sponsors (A)

UNFPA Bangladesh and Jhpiego, whose factory clinics already screen
Brand worker-health programmes
BGMEA, as a door to member factories, not an operator

Public programme (C)

DGHS Non-Communicable Disease Control Programme
The national cervical and breast screening programme at Bangladesh Medical University
Community Clinic Health Support Trust

Evaluation (all)

Bangladesh Medical University
icddr,b
BRAC James P Grant School of Public Health

07 · Sources and status

Where the numbers come from.

Nessa A, et al. Cervical cancer screening programme in Bangladesh: the national registry, 2018–2023. BMC Global and Public Health 2025;3:34.
MOHFW, National Cervical & Breast Cancer Statistics of Bangladesh. cancer-dashboard.mohfw.gov.bd.
The e-HIS case-based registry and its gaps. BMC Public Health, May 2025.
Retention at NCD corners, 12 upazilas, 2023. BMC Health Services Research, Jan 2026.
Factory VIA clinics, five factories. Jhpiego, 21 Jan 2026.
EIB Women’s Health Initiative, 2027–2031. UNFPA, 10 Oct 2026.
BGMEA member-care MoUs with ten providers. UNB, 6 Nov 2025.
Commission-based referral, 66 hospitals and 50 diagnostic centres. TIB via Dhaka Tribune, 7 Feb 2018.
BRAC NCD360 referral tracking. Daily Waadaa, 23 Sep 2026.
Raised blood pressure and blood glucose, adults 18–69. WHO STEPS Bangladesh 2018.
Hypertension control, national and programme. National Heart Foundation, BHCI, Mar 2026.
Diabetes, adults 20–79, 2024. IDF Diabetes Atlas.
Chronic kidney disease, pooled prevalence. Banik et al., Int Urol Nephrol 2021.
Anaemia among female garment workers. Hossain et al., Nutrients 2019.
Next steps: Bangladesh National Hypertension Guideline; ADA Standards of Care 2025; KDIGO 2024.

The one-page brief (PDF)

Experience in one setting does not show that the routine will work in another. Bangladesh is where I already have standing and screening-service experience; that is a reason to start here one day, not evidence that it would work here. The first pilot in Michigan, its evaluation and its published toolkit come first. Figures above are from the sources named; none is HSREP’s own data.

Questions people ask

Who would pay for what in model A?

Model A is a proposed design and nothing is agreed: the design is published, and a Bangladesh partner, site and funding come next. The private provider puts in its nurse, two hours a week on the worklist. The factory gives the worker's paid morning, a room on screening day and fifteen minutes a week from the welfare officer. The sponsor that already pays for the screening adds a small fee per completed follow-up, plus the fare. The woman pays nothing; the colposcopy is free at the public clinic. The sponsor or the provider pays for the founder's time; HSREP adds no fee on top.

Is the Prevention Adoption Initiative running in Bangladesh now?

No. The Bangladesh page is future direction, not part of the first pilot. The first pilot is designed for one health center in Michigan, on colorectal cancer screening. For Bangladesh, the design is published; a Bangladesh partner, site and funding come next. Everything on the page is illustrative, and no patient data has been collected. The Michigan pilot, its evaluation and its published toolkit come first. Experience in one setting does not show that the routine will work in another.

All of this depends on the first pilot.
If you run a provider, fund factory screening, or sit inside the programme, and you want to talk about the step after the referral, that is the conversation to have.

Book 30 min →

An HSREP initiative in development. The design is published; a Bangladesh partner, site and funding come next. Everything on this page is illustrative, and no patient data has been collected.