What score means you should have a mammogram? Recommendations, the government tool's five percentile bands, three risk groups and what four routes cost, 2026

TL;DR Hong Kong does not tell you to have a mammogram simply because you have reached a certain age; it tells you to be assessed first. And "increased risk" has a published line: the Government's risk assessment tool prints five percentile bands, and only the top two, at the 76th percentile or above, return "mammography screening every two years is recommended". At the same time, local incidence and mortality have both risen over the past decade, and both rises are statistically significant.

---

#Hong Kong does not call you in by age; it tells you to be assessed first

This is the most fundamental difference between the Hong Kong system and those of the United Kingdom, Australia and the United States: those three invite you by age, and Hong Kong tells you to have a risk assessment first.

The Cancer Expert Working Group has adopted risk-based screening: women aged 44 to 69 first assess themselves with the Government's risk assessment tool, and only if the assessment shows increased risk do they discuss with a doctor whether to have a mammogram every two years.

A Legislative Council written reply of 5 June 2024 puts the threshold in a sentence: these women's risk is "higher than 75 per cent of Hong Kong women of the same age"; and the second note of the same reply states that about 25 per cent of women aged 44 to 69 have certain combinations of personalised risk factors that raise their risk of breast cancer [Note 1].

---

#Four easily confused ideas, and ten symptoms from three different sources

Four pairs to keep apart first; they are the foundation of the whole article.

Screening is not diagnosis — screening looks for disease in the absence of symptoms. Having a risk factor is not having the disease — the tool speaks of probability, not fate. A positive is not a diagnosis — an abnormal mammogram only means more investigation. Detection is not clinical significance — a cancer can be found that would never have harmed you, which is overdiagnosis.

And anyone with symptoms is not on the screening route. The Government's risk assessment tool lists seven symptoms of its own [Note 21]:

Three more are not in that list of the tool's but are given by two other official documents. This article marks the source of each and does not merge them:

• A lump in the armpit — in both the Expert Working Group's English text and the English of the Department of Health's press release of 3 October 2025.
• Skin of the breast with an orange-peel appearance and bleeding from the nipple — only in the Department of Health's press release of 3 October 2025.

⚠️ This article does not attribute the orange-peel skin to the Expert Working Group; the reason and the method of checking are at [Note 21].

The same press release adds a sentence: the symptoms of early breast cancer are often not obvious.

⚠️ With any of the above, see a doctor. Do not self-assess, and do not wait for the next screen.

---

#The tool's five bands: the 76th percentile is where it tells you to go

"Increased risk" has two official statements: a definition in words, and the five percentile bands the tool actually returns. The second is the rule that decides whether you get the "recommended" sentence.

The tool's scoring logic comes from its own scoring file, executed by the browser and never appearing in the page text, but publicly readable. The percentile bands, the risk descriptions and the screening advice are all in the source's own Chinese:

So "what score means you should go" does have an answer: the 76th percentile or above. Only the top two bands return "screening ... is recommended"; the lower three all return "please pay attention to breast health".

That threshold and the Legislative Council's "risk higher than 75 per cent of Hong Kong women of the same age" are two statements of the same line.

Two things worth spelling out. First, the tool's Chinese output says 「同齡香港婦女」 while the Chinese of the Legislative Council reply says 「同齡香港女性」 — both are official wording and neither may be substituted for the other. Second, on the same page the body text says women should consider having a mammogram every two years while the sentence its own tool returns says one is recommended every two years; the gap between "should consider" and "is recommended" appears inside a single official page.

Put another way, the threshold is the 75th percentile of the risk distribution among women of the same age (the tool implements it from the 76th band up), and by definition a quarter of women at every age sit above it. The University of Hong Kong's estimate of "about 25 per cent" and this threshold correspond to one another; they are not two independent facts.

Practice runs close to it too: as at 31 March 2024, 30 048 women had been assessed and 8 484 referred for mammography, which the Chinese of the Legislative Council reply itself puts at 「約百分之二十八」 (28.2% on this article's calculation); and across 2024, a Department of Health press release itself writes 「2 689名婦女（約29%）」 (29.1% on this article's calculation).

⚠️ A few percentage points above the model's expected quarter — but those assessed come forward themselves and are not a random sample.

---

#Seven factors, only two of which you can change

The tool uses seven factors, with age first: age, breast cancer in an immediate family member, a past diagnosis of benign breast disease, never having given birth or a late age at first birth, early age at menarche, a high body mass index, and lack of physical activity [Note 2].

Age is not one factor among equals: the tool's limits of 44 to 69 and its expression of the output as a percentile among women of the same age both come from it.

The thresholds on the input side are public too, and need not be guessed at:

So "a high body mass index" above means 23 or above in this tool.

Weight and physical activity are the only two of the seven factors you can change yourself.

Behind the tool is the Hong Kong Breast Cancer Study (3,501 cases / 3,610 controls). The Expert Working Group's paper of 2022 states that its area under the receiver operating characteristic curve in Hong Kong is 0.60 — it separates higher from lower risk at the level of a population, but it cannot say which individual will fall ill.

The tool draws its own limits too: it cannot accurately estimate the risk of those classified as high or moderate risk, and is designed only for women aged 44 to 69.

---

#Three risk groups, with different official advice for each

⚠️ The English version of the same page distinguishes three levels; the Chinese distinguishes two. The English page says of high risk "have mammography screening every year", of moderate risk "are recommended to have mammography every two years", and of increased-risk women aged 44 to 69 in the general population "are recommended to consider mammography screening every two years"; the Chinese page uses the same 「建議」 construction for both of the latter two groups [Note 3].

Both texts tell the reader to discuss it with a doctor first; the difference is that the English marks two levels with "consider" and "have" while the Chinese uses 「建議」 for both. Both pages were revised on the same day and neither is the more recent.

So the difference is a matter of wording, version by version and place by place, and not "the official Chinese is uniformly stronger" — two constructions sit side by side within a single version of the cancer.gov.hk page. On what to do next, the texts all agree: assess with the tool first, then discuss the benefits and risks with a doctor.

---

#Administrative eligibility is not clinical high risk

The breast cancer screening pilot is a subsidy programme, not a population screening programme.

Eligibility for its second phase (from 10 June 2025) requires all of: a woman aged 35 to 74 (those with a family history of breast or ovarian cancer may start from the age at diagnosis of the youngest affected relative minus 10 years, but not before 30); holding a Hong Kong Identity Card or Certificate of Exemption; registered with eHealth; being at high risk of breast cancer, that is, confirmed by genetic testing to carry a specified gene variant — the full gene list being BARD1, BRCA1, BRCA2, CDH1, NF1, PALB2, PTEN, STK11, TP53 — or having a strong family history of breast or ovarian cancer; no symptoms of breast cancer; no history of breast cancer; and no screening mammogram in the past year, unless one was already due on a doctor's advice at enrolment.

The programme also has a full list of ineligibility criteria: chest radiotherapy received between the ages of 10 and 30; a past breast cancer; a past atypical ductal hyperplasia or atypical lobular hyperplasia [Note 4].

⚠️ This is the easiest thing in the article to misread: those three groups are precisely the ones the risk stratification places at clinical high risk and advises to have annual mammography, and they do not meet the pilot's administrative eligibility. The programme refers them back to their own doctors. The same words "high risk" mean different things in the two places.

Three organisations take part: the Hong Kong Breast Cancer Foundation, the United Christian Nethersole Community Health Service and the Haven of Hope Christian Service. Ultrasound is arranged only for those confirmed to carry a specified gene variant.

And one group is not covered by the official texts at all: women under 44 or over 69 who are not at high or moderate risk have no official screening advice of any kind.

---

#Why is there no population screening? Three reasons, and a counter-sentence in the same paragraph

It was not an omission; it was a deliberate choice of another route after reviewing the evidence.

The Centre for Health Protection's overview of June 2018 said there was insufficient evidence to support or refute population-based mammography screening for asymptomatic women at general risk in Hong Kong; the revised version reads that a risk-based approach to breast cancer screening is recommended, a revision endorsed at the 15th meeting of the Cancer Coordinating Committee on 19 June 2020 [Note 5].

The reasoning has three layers:

One, a lower baseline incidence than the West — but "lower" is a comparison across places, not a trend over time. The professional guidance of June 2020 runs two sentences together in the same paragraph: after adjustment for population ageing, the age-standardised incidence of female breast cancer showed a rising trend from 1991 to 2018 while the age-standardised mortality showed no significant change from 1991 to 2019; and both Hong Kong rates remain lower than those of a number of Western developed economies and of neighbouring Singapore [Note 6].

⚠️ The two sentences have to be read together: the Government cites "lower" to explain why there is no population screening, but the same paragraph of the same document also says incidence is rising. "Lower than the West" and "rising locally" can both be true, but giving only the first leaves a reader thinking the rate is flat or falling.

Two, local modelling found that risk-based screening buys much the same mortality reduction for less money — but the next sentence in the same paragraph excludes two groups. The text states that this conclusion does not apply to the high-risk group as defined by the Expert Working Group (BRCA1/2 variant carriers, for example) or to the moderate-risk group [Note 7]. Which is to say that the cost-effectiveness conclusion holding up "why risk-based" applies only to the general-population group of the three.

The University of Hong Kong's 2021 study produced three incremental cost-effectiveness ratios, of which the third is the actual decision:

That is, going from no screening to risk-based is good value, and going from risk-based on to population screening costs a great deal more. (The original separates thousands with a space, as in "US$18 151", and this article follows the original's format.)

Three, allocation of resources. A Legislative Council written reply states that excessive screening under a public health programme not only wastes resources for public health overall but may also pose unnecessary health risks to individuals [Note 8].

---

#Incidence and mortality are both rising, and both significantly

"Risk-based" does not mean the problem has got smaller.

The Cancer Registry's statistics for 2023 state that the age-standardised rates (against the Segi 1960 world standard population, per 100,000) are 68.9 for incidence and 8.7 for mortality; and that the average annual percentage change over the past ten years is +2.5% for incidence and +0.3% for mortality, both marked with an asterisk denoting statistical significance at the 0.05 level [Note 9].

⚠️ With population ageing taken out, both the incidence and the mortality of breast cancer in Hong Kong women have risen over the past decade, and neither rise is noise. The Department of Health said as much itself in a press release of 29 May 2025: over the past 10 years, both the age-standardised incidence and the age-standardised mortality of breast cancer have risen.

Even the mortality cell now reads differently from the "no significant change" of 2020.

The 2023 baseline: 5,585 new invasive breast cancers in women, 28.9% of new cancers in women and the commonest cancer among them; a further 831 carcinomas in situ of the breast (non-invasive, officially also called stage zero breast cancer); and breast cancer third among causes of cancer death in women, with 834 deaths, 13.1% of the total.

The same document notes that age-standardised rates of this kind can only be compared with figures computed against the same standard population.

---

#Incidence does not peak in the forties

Incidence of invasive breast cancer per 100,000 women of the same age:

The rate at 40–44 is 44% of the rate at 70–74.

Nor does the curve rise monotonically: 50–54 (189.6) is slightly lower than 45–49 (191.3), the only dip before 85, and this article prints it rather than passing over it.

⚠️ Case numbers and incidence rates are two different things: the largest number of invasive cases in 2023 was at 60–64 (807 cases), against 406 at 40–44, but the population base of each age band differs.

"Breast cancer strikes younger in Hong Kong" means the age curve sits earlier than in the West, not that incidence peaks in the forties.

⚠️ For anyone who has felt a lump and is thinking of waiting for the next screen: the Expert Working Group cites a five-year survival 7% higher where treatment began within 3 months of symptoms appearing than where it began at 3 to 6 months (the original says "7% higher", without distinguishing an absolute from a relative difference).

---

#A positive is not a cancer: locally, about 1 in 8 of those screened is referred on

The same machine at the same accuracy, used in a population where the disease is rarer, yields a lower proportion of true disease among the positives. That is not a question of machine quality; it is arithmetic.

The WHO's guide uses mammography as its example: where incidence is high the positive predictive value is higher and false positives fewer; where incidence is low the predictive value is lower and false positives more numerous [Note 10].

Hong Kong has two sets of first-hand figures.

The pilot's first phase (September 2021 to December 2023): 27,807 people had a risk assessment and 7,785 (about 28%) were referred for and had mammography; of the 409 cases referred to specialists for abnormal mammograms and followed up in public hospitals, 68 (about 16.6%) were diagnosed with breast cancer; the detection rate was 8.7 per thousand mammograms, which the Government describes as comparable to international standards (about five per thousand); and 97% of the 68 diagnoses were stage II or below.

Work it through: 68 ÷ 7,785 = 8.7 per thousand mammograms, agreeing with the press release. Of the 409 cases followed up in public hospitals, 68 were diagnosed, so only about 1 in 6 women called back for further investigation was ultimately diagnosed.

⚠️ Note the denominator: 409 is the number followed up in public hospitals, which is not the number referred, so 16.6% cannot be taken as the whole programme's positive predictive value.

The second set is the recall rate. In 2024, 9,230 women aged 44 to 69 had a risk assessment and 2,689 (about 29%) went on to mammography on the strength of it (the press release separates thousands with a space, printing 「9 230」 and 「2 689」); and of the women who had a mammogram, 321 (about 12%) were referred to a specialist for further investigation and management.

⚠️ The denominator of that 12% is the 2,689 women aged 44 to 69 who went for a mammogram after a risk assessment, in the single year 2024 — it is not a general recall rate for "all women in Hong Kong who have had a mammogram", and it cannot be applied to other age bands or other mammography routes.

It is nonetheless a first-hand, local, official figure for the harm side. The corresponding United Kingdom figure is 4%.

The United Kingdom also publishes its own positive predictive value: of every 100 people screened, 96 need no further tests and 4 do; and of those 4, 1 will be diagnosed with breast cancer [Note 11] — a positive predictive value of one in four.

⚠️ The "68 ÷ 409 ≈ 16.6%, about one in six" above cannot be set against the United Kingdom's one in four, because the denominators are defined differently. At most it can be said to point the same way as the mechanism that a lower incidence gives a lower positive predictive value.

---

#Benefits and harms: locally, only half the figures exist

Both sides together. Every mortality-reduction and overdiagnosis figure below is from overseas research — Hong Kong has never published a local estimate of overdiagnosis.

The United Kingdom's independent review (a meta-analysis of 11 randomised controlled trials): the relative risk of breast cancer death in the invited group against the control was 0.80 (95% confidence interval 0.73–0.89), a relative reduction of 20%. The other side: for every 10,000 United Kingdom women aged 50 invited to screening over the following 20 years, 43 breast cancer deaths are prevented while 129 breast cancers (invasive and non-invasive) are overdiagnosed — about three overdiagnoses for every death prevented [Note 12]. The report states of these estimates that they should be treated only as a rough guide.

The US Preventive Services Task Force's 2024 model medians (per 1,000 women):

⚠️ That "14" is not a precise figure — the source footnotes it in the same sentence, the six models' estimates ranging from 4 to 37, the widest span of anything in the table. For overdiagnosis on the 50–74 row the source gives no separate range, and this article reproduces it without estimating one.

And the table's own heading states that the modality is 3D tomosynthesis, not the ordinary 2D screening discussed below.

For starting at 40 rather than 50, the increment the source itself prints is 1.3 more deaths averted.

⚠️ That 1.3 is not the difference between the table's two rows — 8.2 − 6.7 = 1.5, and the source adds a note in the same paragraph expressly warning that the two are not the same thing: the 1.3 is the median of each of the six models' own differences, which is not the same as subtracting two medians [Note 13].

The harm side follows the source's own presentation too: for overdiagnosis the source prints the increment itself, 2 more (range 0 to 4); for false positives the source prints no absolute increment, saying only that it rises by about 60%. The 1,376 − 873 = 503 is this article's own subtraction of two medians, and by the source's own note above it is not the same as "the median of each model's difference".

⚠️ A common misreading sets the "1.3 more" against the "1,376" — but 1.3 is an increment between two strategies and 1,376 is the full figure for the 40–74 strategy; the two are in different units and cannot be set side by side.

Why every two years, and why from 44? Start with the first sentence of the same paragraph from the Expert Working Group: no randomised controlled trial has ever directly compared annual with biennial mammography screening in any age group [Note 14]. Every comparison below sits under that limitation.

Cumulative ten-year false-positive rates: 61% with annual screening and 42% with biennial; cumulative biopsy rates 7% against 5%.

⚠️ The source's next sentence cannot be dropped: these rates are likewise higher in women aged 40 to 49 and in those with dense breasts. Readers aged 44 to 49 particularly need that sentence.

The University of Hong Kong's 2021 study found biennial screening starting at 44 to be the most cost-effective, with screening from 44 to 69 reducing breast cancer mortality by 25.4% (95% credible interval 20.5%–29.4%, the original stating that this applies across all risk strata). The same paper also computed that with a lifetime risk threshold of 8.4%, the proportion eligible for screening is 25.8% — a third independent source for the "about a quarter" figure.

Two biases to remember. Screening can make someone live longer from diagnosis without changing when they die (lead-time bias); and screening more readily finds slower-growing, less aggressive cancers (length-time bias). So Hong Kong's five-year relative survival for breast cancer of 84.0% (cases from 2010–2017) is a measure of treatment results, not evidence that screening works.

---

#Ultrasound, 3D and self-examination: the official position differs for each

Four things are routinely treated as substitutes for or upgrades on mammography, and the official position on each of them is different.

One, ultrasound is not an alternative for women generally. The professional guidance states that there is insufficient evidence to recommend clinical breast examination or ultrasound as a breast cancer screening tool for asymptomatic women. The pilot's examination page states that ultrasound is done when clinically indicated, not as "an X-ray substitute for those who would rather not have one".

The Group also notes that dense breast tissue affects imaging, and that ultrasound as an adjunct may find small cancers a mammogram misses while also increasing false-positive recalls. The concluding sentence cannot be skipped: several systematic reviews consider the evidence insufficient to show that ultrasound as an adjunct to mammography screening reduces breast cancer mortality, and the Canadian task force goes further and recommends against ultrasound for screening women at general risk [Note 15].

The United States has likewise issued an I statement on adding ultrasound or MRI after a normal mammogram (irrespective of breast density) — the evidence is insufficient to weigh the benefits and harms.

Two, the position on breast self-examination has to be stated precisely. The official position is not to use self-examination as a screening tool, but it is official advice that every woman stay breast aware and see a doctor at once on any suspicious symptom [Note 16].

⚠️ That cannot be shortened to "the Government tells people not to check themselves". The reason the Group gives is that randomised controlled trials in Shanghai and Russia showed no significant difference in the size or stage of breast cancers or in the number of deaths, while producing more false positives and unnecessary biopsies.

Three, 3D mammography is not an officially endorsed better screening tool. A Legislative Council written reply gives three reasons, not one: overseas research provides no evidence that 3D screening is more cost-effective than 2D in identifying breast cancer and reducing breast cancer mortality; the cost involved is higher; and specialists take longer to read the images. The Group's paper adds that although 3D lowers the false-positive recall rate and finds more cancers, its radiation dose is higher, and international guidelines do not currently support its use as a screening tool.

Four, MRI. The official Chinese wording is that MRI is not recommended for breast cancer screening in women generally.

---

#A normal mammogram has four limits

The official documents say so plainly.

Interval cancers. Breast cancer still arises between one screen and the next, which is why regular observation and self-palpation of the breasts remain necessary. The WHO's guide states that interval cancers will certainly occur, because no screening programme is 100% sensitive [Note 17].

False negatives. What the WHO writes is that false negative (normal) results will certainly occur in a screening programme, because no programme is 100% sensitive. "Not entirely accurate either" is a softened way of putting it; what the WHO says is that they will certainly occur, and that the reason is sensitivity.

Sensitivity is lower in younger women and those with dense breasts. Mammography is generally less sensitive in younger women and those with a genetic predisposition, because denser breast tissue obscures the imaging features of early breast cancer in premenopausal women [Note 18]. Readers aged 44 to 49 particularly need that sentence; it is not only about ultrasound.

2D mammography itself involves radiation. The Group quantifies the boundary: the cumulative risk of breast cancer death caused by mammography screening is estimated at 1 to 10 per 100,000 women depending on age, frequency and duration, and that figure is at least 100 times smaller than the estimated breast cancer deaths mammography screening prevents. So the radiation risk exists, is quantifiable, and is far smaller than the benefit — and it is not the case that "only 3D involves radiation".

One more thing the official texts do say: the mammogram hurts. The pilot's examination page states that compressing the breast may cause discomfort or pain, but generally only for a few seconds. The Group's quantification: the proportion of women reporting pain during mammography ranges from 1% to 77%, and among those the proportion refusing future screening because of pain ranges from 11% to 46%.

After an abnormal result, the next step is a diagnostic procedure, not treatment. The pilot's examination page states that all screening has limitations and potential risks and is not completely accurate; the potential risks include false positive and false negative results, overdiagnosis and consequent unnecessary treatment.

The WHO adds: every positive result requires further investigation to distinguish true from false positives, so those with a false positive undergo unnecessary tests and bear their complications, and the anxiety, low mood and sleep problems a false positive causes can last as long as three years or more in some people.

⚠️ The other direction has to be stated too: a normal result does not mean nothing is wrong. The WHO notes that people with a normal result may overlook important symptoms and so delay diagnosis, which is why staying breast aware and seeing a doctor at once on any suspicious symptom has nothing to do with when you last had a mammogram.

---

#What do others say? The most-cited documents, more than a decade apart

The disagreement is real, but leaving out the dates misleads.

⚠️ The College of Radiologists document predates the revision of June 2020; the Anti-Cancer Society's is almost identical to the 2018 sentence, so it reflects the former position rather than a live objection to the current risk-based one.

Three advanced economies all run age-based population screening, with different ages and different intervals — so "do what other countries do" is not by itself an answer.

The United States' 2016 version was every two years from 50 to 74 with an individualised decision at 40–49, and an I statement for 75 or above. So the United States has moved from individualised towards uniform by age while Hong Kong has moved from uniform by age towards risk-based: opposite directions.

The Hong Kong system should be judged against the criteria for targeted screening, not those for population screening. The United Kingdom National Screening Committee has, besides its 20 population criteria, seven criteria for targeted screening programmes, and states expressly that the evidential thresholds differ because randomised controlled trials "may not be feasible or ethical" in some high-risk groups; the same page also cautions that a looser threshold may increase the number of targeted programmes [Note 19].

The fifth is the most apposite: the criteria for identifying those eligible should be robust, inclusive and evidence-based. Hong Kong's rule for identifying them is that model and the threshold of "risk higher than 75 per cent of Hong Kong women of the same age".

And against the fourth — that a screening programme's benefits overall should outweigh its harms — Hong Kong has never published local measured data on overdiagnosis or mortality reduction. What exists locally is the detection rate, the stage distribution and that recall rate of about 12%, not those two.

Judging Hong Kong's risk-based system against population criteria is a category error; judged against targeted criteria, the question becomes "how sound is the rule for identifying those eligible" — and that question currently has a model answer but no local effectiveness answer.

---

#What the four routes cost

Four public routes, each with first-hand prices or charging rules that can be checked: the pilot programme (high-risk group, currently $0), the Women Wellness Satellite co-payment, the Hospital Authority, and self-payment. The table below has six rows — the Hospital Authority column splits in two by situation, and the self-payment column has two organisations.

The nature of the Women Wellness Satellite charges needs stating first. Its service charge page states that it uses a co-payment model: eligible persons receive part of the cost as a government subsidy and must pay a specified co-payment. So the Women Wellness Satellite figures below are co-payments, not the full cost of the service — the amount of the government's share is not published on the charge page.

The structure matters too: 2D mammography is a "basic service"; breast ultrasound and 3D mammography are "value-added services", charged separately item by item.

One route is in the middle of moving, and its prices are already published. Both press releases say the move is not yet complete: women's health services are "being integrated into the Primary Healthcare Commission's district health networks", the Commission "will progressively begin integrating women's health services within this year", and the three Women Wellness Satellites "replace" the Department of Health's women's health service, of which only the Chai Wan one has a published start date. No announcement says that all three are operating.

There is a transitional arrangement, and it is the most useful sentence in this section: during the transition, those who have paid the annual fee and are still within their service period will continue to receive services at the Department of Health.

On the division by age: women aged 64 or below may receive the relevant screening services at the Women Wellness Satellites, while women aged 65 to 69 continue to be served at the Department of Health's Elderly Health Centres.

A worked sum (using the prices already quoted, not anyone's actual bill). The Women Wellness Satellite route runs, on the official announcements, only to 64, so every two years from 44 means 44, 46, 48 … 64, 11 occasions in all.

• Women Wellness Satellite co-payments: 11 × $250 = $2,750 (for the 2D mammography alone; the $150 "doctor assessment and consultation" and $80 "dedicated nurse clinic" each time are separate, the actual number of them depending on how the service is arranged, and this article assumes nothing).
• The occasions at 66 and 68 are not in that sum: 65 to 69 stays at the Elderly Health Centres, and this article could find no published charge for those two, so it will not put a figure on them.
• Hong Kong Breast Cancer Foundation self-paid bilateral 2D mammography (again on 11 occasions from 44 to 64): 11 × $1,500 = $16,500; at 3D each time ($2,800), 11 × $2,800 = $30,800, a difference of $14,300.
• The pilot's $225 is the co-payment cap per examination (not per year and not once for all): a high-risk woman screened annually from 35 to 74 has 40 occasions, so the total at the cap is 40 × $225 = $9,000.

⚠️ These figures cannot be compared directly. The $2,750 and $9,000 are co-payments with a government subsidy behind them, and the pilot's three organisations currently charge $0; the $16,500 is paid entirely by the patient; and the Hospital Authority general outpatient route is $0 for underprivileged people exempted from medical fees. The routes also differ in interval (every two years against every year), age (44–69 against 35–74), eligibility (increased risk against genetic or family history high risk) and entry conditions (the Women Wellness Satellites require District Health Centre or Express membership first).

The programme's FAQ page adds: health care vouchers may not be used to pay the co-payment; and if a participant is referred to another healthcare institution for further investigation or treatment, that is not subsidised.

---

#What comes first officially is not screening but primary prevention

The Chinese of a Legislative Council written reply: to reduce the risk of cancer, primary prevention — reducing exposure to carcinogenic risk factors — is of the first importance; the Department of Health has consistently promoted healthy living habits, including avoiding alcohol, taking regular physical activity and maintaining a healthy weight and waist circumference, and treats this as its foremost strategy for preventing cancer [Note 20].

"Foremost strategy" is the source's own ordering. The Department of Health's press release of 3 October 2025 says the same thing: a healthy lifestyle is primary prevention and screening is secondary prevention.

And there are numbers to follow. Of the seven risk factors above, body mass index and physical activity are the only two you can change yourself, and the Government's risk assessment tool hands you targets directly on its results page:

• Maintain a healthy weight — an adult's body mass index should be between 18.5 and 22.9.
• Physical activity — adults should do at least two and a half hours of moderate-intensity, or one hour 15 minutes of vigorous-intensity, physical activity a week.

Neither of those needs a referral or a queue; both can be started today.

---

#What to do next

1. Be assessed first, rather than screened first. The official position is not that every woman aged 44 to 69 should be screened; it is that you should use the tool first.
2. Know where the line is: the 76th percentile. Only the top two bands return "screening ... is recommended".
3. If you are in the high-risk or moderate-risk group, do not use that tool. It states itself that it cannot accurately estimate the risk of those two groups, which have their own advice.
4. If you have had breast cancer, chest radiotherapy, or atypical hyperplasia: you are clinically at high risk but cannot join the pilot programme, and the only route is to arrange it with your own doctor.
5. Two sentences matter particularly at 44 to 49: mammography is less sensitive in this age band, and the false-positive and biopsy rates are higher.
6. Feeling a lump means seeing a doctor, not waiting for the next screen. Staying breast aware has nothing to do with when you last had a mammogram.
7. The two changeable risk factors can be started on today: a body mass index of 18.5 to 22.9, and two and a half hours of moderate-intensity or one hour 15 minutes of vigorous-intensity physical activity a week.

---

#Frequently asked questions

I am 45 with no family history. Do I need a mammogram?

The official position is not that every woman aged 44 to 69 should be screened; it is that you should assess yourself first with the Government's risk assessment tool. The tool's line is public: only if you fall in the top two bands, at the 76th percentile or above, does it return "mammography screening every two years is recommended, together with a discussion with a doctor about its benefits and risks." The lower three bands all return "please pay attention to breast health at all times, and consult your family doctor with any questions." The Legislative Council written reply puts it as "risk higher than 75 per cent of Hong Kong women of the same age".

Why does Hong Kong not follow the United Kingdom and Australia and screen everyone?

The Government publishes three reasons: local incidence and mortality remain lower than in Australia, the United Kingdom and Singapore; local modelling shows risk-based screening to be more cost-effective; and excessive screening wastes public health resources and may pose unnecessary health risks to individuals. Note that the same paragraph of the same document also states that local incidence is rising, and that the cost-effectiveness conclusion expressly does not apply to the high-risk or moderate-risk groups.

Does checking my own breasts count as screening?

No. The official position is not to use breast self-examination as a screening tool, while at the same time advising every woman to stay breast aware and see a doctor at once on any suspicious symptom. Those two sentences have to be read together and cannot be shortened to "the Government tells people not to check themselves".

Is 3D or ultrasound better?

Officially, no. On 3D, a Legislative Council reply gives three reasons for not supporting it as a screening tool: no evidence that it is more cost-effective than 2D, higher cost, and longer reading time; its radiation dose is also higher. On ultrasound, the official position is that the evidence is insufficient to recommend it as a screening tool for asymptomatic women, and several systematic reviews likewise consider the evidence insufficient to show that it reduces breast cancer mortality as an adjunct. MRI is expressly not recommended for screening women generally.

If the mammogram is normal, can I stop worrying for two years?

No. Breast cancer still arises between one screen and the next; false negatives will always occur in any screening programme, because no programme is 100% sensitive; and mammography is generally less sensitive in younger women and those with dense breasts. Staying breast aware and seeing a doctor at once on any suspicious symptom has nothing to do with when you last had a mammogram.

Does a mammogram involve radiation? Does it hurt?

2D mammography does involve radiation: the cumulative risk of breast cancer death caused by mammography screening is estimated at 1 to 10 per 100,000 women depending on age, frequency and duration, and that figure is at least 100 times smaller than the estimated breast cancer deaths it prevents. As for pain: the official texts state that compressing the breast may cause discomfort or pain, generally for a few seconds; the proportion of women reporting pain ranges from 1% to 77%, and among those the proportion refusing future screening because of pain ranges from 11% to 46%.

Can I join the pilot programme?

The pilot is a subsidy programme, not a population screening programme, and eligibility requires several conditions to be met at once, the core of them being a specified gene variant confirmed by genetic testing or a strong family history. Note particularly that three groups — those who have had breast cancer, those who received chest radiotherapy between the ages of 10 and 30, and those who have had atypical hyperplasia — are at clinical high risk under the risk stratification and are advised to have annual mammography, yet do not meet the pilot's administrative eligibility; the programme refers them back to their own doctors.

What does one mammogram cost?

It depends on the route. The pilot's co-payment is capped at $225 per examination, and none of the three service organisations currently charges one; the Women Wellness Satellite 2D mammography co-payment is $250 each time; for eligible persons the Hospital Authority classes mammography as an advanced item at $250 each, with no charge for underprivileged people exempted from medical fees; and self-paid, the Hong Kong Breast Cancer Foundation charges $1,500 for a bilateral 2D mammogram. These figures are different in kind — a co-payment has a government subsidy behind it and self-payment does not — and cannot be compared directly.

---

#What this article does not state

• A local estimate of overdiagnosis, and local measured data on mortality reduction. The official documents reviewed on 1 and 2 August 2026 — the professional guidance of June 2020 (retrieved 2 August 2026), the pages of the breast cancer screening pilot website (retrieved 1 August 2026), the Department of Health press releases of 29 May 2025 and 3 October 2025, and the Legislative Council written reply of 5 June 2024 — none carries a Hong Kong estimate of overdiagnosis or measured post-screening mortality reduction. It is not that there are no local figures for the harm side at all: the press release of 29 May 2025 gives the specialist referral rate of 321 women, about 12%, quoted above; what is not stated here is confined to overdiagnosis and mortality reduction. Every figure of those two kinds cited in this article comes from United Kingdom and United States research; locally there is only the modelled estimate of 2021.
• Local measured mortality reduction. The reduction in breast cancer mortality since the risk-based system began has not been evaluated; what exists locally is the detection rate, the stage distribution and that recall rate of about 12% in 2024.
• The pilot programme's overall positive predictive value. The denominator of the 68 ÷ 409 ≈ 16.6% in the text is only the cases followed up in public hospitals, which is not the number referred, so it cannot be taken as the programme's positive predictive value or set against the United Kingdom's one in four.
• The charge for the two mammograms at 65 to 69 at the Elderly Health Centres. This article could find no published charge, so it puts no figure on them, and the lifetime sum above therefore runs only to 64.
• The amount of government subsidy behind the Women Wellness Satellite co-payments. The charge page publishes only the co-payment and not the subsidised portion, so this article does not state the full cost of the service.
• Whether all three Women Wellness Satellites are operating. Both press releases say the integration is in progress, and only the Chai Wan one has a published start date; no announcement says all three are operating.
• The source's figure for the increment in false positives. The source prints no absolute number, saying only "about a 60% increase"; the 1,376 − 873 = 503 is this article's own subtraction of two medians, which by the source's own note is not the same as "the median of each model's difference".
• The across-model range for overdiagnosis on the 50–74 row. The source gives no separate range for that age band, and this article reproduces the median without estimating one.
• Whether "7% higher" is an absolute or a relative difference. The original does not distinguish, and this article reproduces it as it stands.
• How the Women Wellness Satellites combine charges for "basic" and "value-added" services. The charge page prices each item and does not say whether ultrasound or 3D mammography must be charged together with a basic service, so this article lists each price and adds none of them together.
• The risk assessment tool's formula from inputs to percentile (the regression coefficients). The tool's output thresholds are set out above (five percentile bands, with "screening ... is recommended" from 76% up) and so are the boundaries of its input options; what this article does not state is the actual weighting from input to percentile, because that formula is not published in any publicly checkable file.
• Comparisons of crude incidence by site from the Cancer Registry. The denominator is printed in the source itself — the axis label of the female breast cancer page's chart states that the rate is per 100,000 female population (that page prints the English and Chinese axis labels on two lines, and this article does not merge them into one quotation). This article makes no cross-site crude rate comparison because crude rates are not age-standardised and the sex denominators differ between sites, so comparing them directly would mislead — not because the denominator could not be found.
• The item name for mammography on the Hospital Authority's itemised fee schedule (that schedule lists only three tiers and does not name items individually). But the tier and the price are not unpublished: the Authority's public healthcare fee reform FAQ states that mammography is an advanced item at $250 for eligible persons, and that is in the charge table above.
• The figures in the Hong Kong Breast Cancer Registry's Report No. 17. On 9 August 2026 the eleven PDFs on that report's public page (hkbcf.org/zh/our_research/main/1315/) were downloaded and each tested: all eleven returned HTTP 200 with %PDF as their first four bytes, running to 222 pages in total, from which PyMuPDF extracted a total of 16 non-whitespace characters — that is, almost the whole report is full-page scanned images with no text layer to check word by word, so this article cannot verify its content and therefore cites none of its figures. What is stated here is a measured result, not "could not obtain".
• Hong Kong's policy on notifying breast density. On 9 August 2026 the professional guidance of June 2020, the breast cancer risk assessment tool page and the scoring script file that page loads, and the pages of the pilot website were all re-checked, and none sets out an official arrangement for notifying those examined of their breast density — this item was re-verified in the manner appropriate to "content may be emitted by a script and absent from the page text".
• The position of surgical specialty bodies on breast cancer screening. This article states nothing about these three bodies, and expressly does not state that they have no position paper. Of the three, only the College of Surgeons of Hong Kong's official domain could be confirmed (cshk.org): re-verified on 9 August 2026, that domain resolves to 103.11.100.15 and is reachable (HTTP 200), and a site-wide search for 「篩查」, "screening" and 「立場」 returned nothing. The official domains of the other two bodies have never been confirmed by this article. The two hostnames used previously were guessed from the bodies' Chinese names; queried against 8.8.8.8 on 9 August 2026, neither had any DNS record, while in the same query the comparable Hong Kong medical body domains cshk.org, hkcr.org and hkacs.org.hk all resolved — so the failure to resolve is not an artefact of this article's environment, but neither does it prove that those two hostnames are those bodies' official websites; it proves only that they do not exist. Not having searched a site never confirmed to exist is no basis for saying a body has no published position.
• This article assesses no doctor, clinic, organisation or programme, and judges nobody's need for screening — that step is for you and a doctor together.

---

#Notes: the official wording

Where a source publishes in English, the wording below is the source's own. Where a passage exists only in Chinese, the rendering is ours and is marked as such, with the original set out alongside so it can be checked.

[Note 1] Legislative Council written reply (5 June 2024). The English says that women aged 44 to 69 with certain combinations of personalised risk factors are:

at a higher risk of breast cancer (viz. risk higher than that of 75 per cent of Hong Kong women of the same age)

The Chinese of the same reply, in our rendering:

showing that their risk of breast cancer is elevated (that is, a risk higher than that of 75 per cent of Hong Kong women of the same age), and they should consider mammography screening every two years (our translation from the Chinese original)

Chinese original:

顯示她們罹患乳癌的風險偏高（即風險高於七成半同齡香港女性），應考慮每兩年接受一次乳房X光造影篩查

And the second note of that reply:

According to research by the University of Hong Kong, about 25 per cent of women aged 44 to 69 have certain combinations of personalised breast cancer risk factors that raise their risk of breast cancer. (our translation from the Chinese original)

Chinese original:

根據香港大學的研究，44至69歲的婦女中，約有百分之二十五有某些組合的個人化乳癌風險因素，令她們罹患乳癌的風險增加。

The original writes the proportion in Chinese numerals, 「約有百分之二十五」; where the body of this article writes 25% it is this article's own conversion. The five output bands' percentile ranges, risk descriptions and screening advice come from cancer.gov.hk/js/questionnaire.js (the recomarr, risktext* and listtxt* arrays), loaded by that page as src="../../js/questionnaire.js", retrieved 9 August 2026.

[Note 2] The seven factors, in the Expert Working Group's English:

a list of risk factors including age, presence of family history of BC among first-degree relatives, history of benign breast disease, age at menarche, age at first live birth, BMI and physical activity level

The first field of the tool's input form is age itself, with options from 44 to 69. To be precise: the University of Hong Kong's 2021 model builds 288 risk strata from six risk factors, age entering the model as the "starting age" parameter; while the Group, describing the assessment tool, lists age alongside the other six. The two accounts do not conflict, but "six" cannot be labelled the Group's complete list. The official Chinese uses 「直系親屬」 and 「良性乳腺疾病」 throughout, and this article follows it. The physical activity question and the three body mass index labels exist only in Chinese; the renderings are ours:

Over the past 10 years, have you taken part in vigorous exercise on average once a week or more (for example lifting heavy objects, aerobic exercise of various kinds, cycling fast and so on)? (our translation from the Chinese original)

Chinese original:

在過去10年，你是否平均每星期參與劇烈運動一次或以上（例如提重物、各類帶氧運動、快踩單車等）？

Body mass index under 18.5 / Body mass index between 18.5-22.9 / Body mass index greater than or equal to 23 (our translation from the Chinese original)

Chinese original:

體重指標少於18.5　體重指標介乎18.5-22.9　體重指標大過或等於23

[Note 3] The official advice for the three risk groups comes from the breast cancer screening pilot's page on breast cancer screening recommendations in Hong Kong (English "Last Revision Date: 15 June 2026"; Chinese 「最後更改日期: 2026年6月15日」). The English version's three levels: high risk, "have mammography screening every year"; moderate risk, "are recommended to have mammography every two years…"; increased-risk women aged 44–69 in the general population, "are recommended to consider mammography screening every two years…". The Chinese version uses 「建議」 for both of the latter two. The Chinese summary table on the same page is softer still, saying to use the breast cancer risk assessment tool to assess personal risk and to refer to the screening advice it gives. The Hong Kong College of Family Physicians' FP Links of April 2022 reports the stronger version, "are recommended to have mammography screening every 2 years", but it plainly predates the current page.

[Note 4] The second phase's ineligibility list, in the original:

History of radiation therapy to the chest for treatment between age 10 and 30 / History of breast cancer / History of atypical ductal or lobular hyperplasia of the breast

The gene list page states "Last updated: [26/02/2025]" of itself.

[Note 5] The change of position. The Centre for Health Protection's overview of June 2018:

There is insufficient evidence to recommend for or against population-based mammography screening for asymptomatic women at average risk in Hong Kong.

The revised version:

It is recommended that risk-based approach should be adopted for breast cancer screening.

That revision was "endorsed by the Cancer Coordinating Committee at its 15th meeting on 19 June 2020", and the document gives its own publication date as "June 2020". An entry in that PDF's reference list is marked "[Accessed January 2021]", so the file now in circulation was plainly finalised after its own stated date; this article follows the document's stated date and the date of endorsement and treats both as June 2020.

[Note 6] The "lower" paragraph, whose two sentences have to be read together:

After adjusting for population ageing, the ASIR of female BC had an upward trend for the period between 1991 and 2018 while the ASMR between 1991 and 2019 did not change significantly. The ASIR and ASMR of female BC in Hong Kong remained low when compared with the rates reported by a number of developed economies in the West (e.g. Australia, United Kingdom [UK]) and the neighbouring Singapore.

[Note 7] The scope of the cost-effectiveness conclusion, the next sentence in the same paragraph:

Having said that, this conclusion does not apply to the high risk group (e.g. BRCA1/2 mutation carriers) and moderate risk group as defined by the CEWG.

[Note 8] The resource allocation reason, Legislative Council written reply (5 June 2024):

Excessive screening under a public health programme not only wastes the resources for overall public health … and may pose unnecessary health risks to individuals.

[Note 9] The Cancer Registry's statistics for 2023. The document prints, in its own order, English first and then Chinese, "Average Annual Percent Change of ASR over the past 10 years2" followed by 「年齡標準化比率在過去十年內的平均每年百分比變化2」 (the "2" at the end of each line being the document's own footnote marker, pointing to its second note defining the measure). The document's own note:

* represents statistically significant at the 0.05 level

The Department of Health's press release of 29 May 2025 says the same in Chinese; the rendering is ours:

Over the past 10 years, both the age-standardised incidence rate and the age-standardised mortality rate of breast cancer have risen. (our translation from the Chinese original)

Chinese original:

過去10年，乳癌的年齡標準化發病率和年齡標準化死亡率均有所上升

On the 831 carcinomas in situ, "stage zero breast cancer" is official usage — the Chinese of the Legislative Council written reply of 5 June 2024 writes 「原位乳癌（註一）（又稱零期乳癌）女性新症」; this article does not attribute that term to the Cancer Registry's overview, which does not use it.

[Note 10] Positive predictive value against incidence, from the WHO Regional Office for Europe's Screening programmes: a short guide (2020):

Countries with high prevalence will have higher positive predictive value and fewer false positives. Countries with lower prevalence will have lower positive predictive value and more false positives.

[Note 11] The United Kingdom's recall rate and positive predictive value, from the GOV.UK invitation leaflet (updated 23 February 2026):

For every 100 people having breast screening, 96 do not need further tests and 4 will need further tests. Of those 4 people, 1 will be diagnosed with breast cancer.

The same page also states that the main risk of screening is finding a cancer that would never have harmed you, and that a normal result "does not guarantee that you do not have breast cancer or will not develop it in the future". The NHS invites women from 50 to just before their 71st birthday every three years, and GP patients registered as female are invited automatically. The description of the Australian programme is taken not from the live page but from an Internet Archive capture of that official page dated 13 December 2025, on which the page states of itself "Date last updated: 18 April 2023":

invites women aged between 50 and 74 years to have a mammogram every 2 years

[Note 12] Both sides of the United Kingdom's independent review:

for every 10,000 UK women aged 50 years invited to screening for the next 20 years, 43 deaths from breast cancer would be prevented and 129 cases of breast cancer, invasive and non-invasive, would be overdiagnosed

The report states of its estimates that they "should be regarded only as an approximate guide".

[Note 13] Two warnings the United States' 2024 model table carries with it. On the overdiagnosis median of 14:

although with a very wide range of estimates (4 to 37 cases) across models

On the increment from starting at 40:

note that the 1.3 deaths averted is the median of the differences in each of 6 models, which is not the same as the difference of the medians noted above and in the table

The source prints the overdiagnosis increment itself: "2 additional overdiagnosed cases of breast cancer (range, 0 to 4)"; for false positives it says only "about a 60% increase in false-positive results". The table's own heading states the modality as "Biennial Screening Mammography With Digital Breast Tomosynthesis", that is, 3D tomosynthesis.

[Note 14] The premise of the interval question, the first sentence of the same paragraph from the Expert Working Group:

There has not been any RCTs identified directly comparing annual to biennial MMG screening in women of any age.

After the cumulative false-positive and biopsy rates, the source's next sentence:

these rates were also higher among women aged 40 to 49 years and women with dense breasts

[Note 15] The position on ultrasound:

There is insufficient evidence to recommend clinical breast examination or ultrasonography as a screening tool for breast cancer for asymptomatic women.

The concluding sentence:

concluded that there is insufficient evidence that ultrasonography as an adjunct to MMG screening can decrease BC mortality while the Canadian Task Force on Preventive Health Care recommends not using ultrasonography to screen for BC in women at average risk

[Note 16] The position on self-examination, whose two sentences have to be read together:

Breast self-examination is not recommended as a screening tool for breast cancer for asymptomatic women. Women are recommended to be breast aware … and seek medical attention promptly if suspicious symptoms arise.

[Note 17] Interval cancers and false negatives. The GOV.UK leaflet:

Breast cancer can also develop in the time between screening appointments. You still need to look at and feel your breasts regularly

The WHO:

False-negative (normal) results always occur in screening programmes, because no programme is 100% sensitive.

The WHO also states of interval cancers that they "will always occur because no screening programme is 100% sensitive".

[Note 18] Sensitivity and breast density, from the Expert Working Group:

MMG generally has lower sensitivity in younger women and those with a genetic predisposition to BC due to increased mammographic density obscuring the radiological features of early BC in premenopausal women

On radiation, the Group quantifies it at 1 to 10 per 100,000 women and writes that the figure is "smaller than the estimates of BC deaths prevented by MMG screening by a factor of at least 100".

[Note 19] The United Kingdom National Screening Committee's criteria for targeted screening (29 September 2022). The parent page states expressly that the two evidential thresholds differ because randomised controlled trials "may not be feasible or ethical in certain high-risk population groups" — the scope being certain high-risk groups, not the general case. The fifth criterion:

There should be a robust and inclusive evidence-based selection criteria for identifying those eligible for targeted screening.

The fourth: "The overall benefits from the screening programme should outweigh the harms".

[Note 20] The ordering of primary prevention, from the Chinese of the Legislative Council written reply (5 June 2024). The rendering is ours:

To reduce the risk of cancer, the primary prevention strategy — that is, reducing exposure to carcinogenic risk factors — is of the first importance. The Department of Health has consistently promoted healthy living habits (including avoiding alcohol, taking regular physical activity, and maintaining a healthy weight and waist circumference) … and treats this as its foremost strategy for preventing cancer. (our translation from the Chinese original)

Chinese original:

要減低患癌的風險，第一級預防策略（即減少接觸致癌風險因素）至為重要。衞生署一直推廣健康的生活習慣（包括避免飲酒、恆常進行體能活動、保持健康體重和腰圍等）……並以此作為預防癌症的首要策略。

The Department of Health's press release of 3 October 2025 uses "primary prevention" for a healthy lifestyle and "Screening, as secondary prevention" for the second layer. The Chinese release of 29 May 2025 states that the Department has consistently encouraged the public to adopt a healthy lifestyle, including avoiding tobacco and alcohol, eating healthily, taking regular physical activity and maintaining a healthy weight, so as to reduce the risk of non-communicable diseases including cancer. The two targets on the tool's results page exist only in Chinese; the renderings are ours:

Maintain a healthy weight (an adult's body mass index should be between 18.5 and 22.9). Adults should do at least two and a half hours of moderate-intensity or one hour 15 minutes of vigorous-intensity physical activity a week. (our translation from the Chinese original)

Chinese original:

保持健康體重（成年人的體重指標應介乎18.5至22.9之間）。成年人應每星期最少做兩個半小時的中等強度或一小時15分鐘的劇烈強度體能活動。

[Note 21] The three sources of the symptoms. The Government's risk assessment tool, in our rendering:

If you notice symptoms of breast cancer — a lump in the breast, a change in the size or shape of the breast, an abnormality of the skin of the breast or nipple, a rash near the nipple, an inverted nipple or discharge from it, or new and persistent discomfort or pain in the breast or armpit — please see a doctor as soon as possible. (our translation from the Chinese original)

Chinese original:

如果你發現乳癌的徵狀，例如乳房出現硬塊，乳房的大小或形狀有所改變，乳房或乳頭皮膚出現異樣，乳頭附近出疹、內陷或有分泌物，乳房或腋下新出現持續的不適或疼痛等，請盡快求醫。

On the lump in the armpit, the Expert Working Group's English writes "breast or axillary lump" and the English of the Department of Health's press release of 3 October 2025 writes "lumps in the breast or armpit". The orange-peel appearance and the bleeding from the nipple appear only in that press release, as "dimpling or "orange-skin" appearance of the breast" and "abnormal discharge or bleeding from the nipple".

This article does not attribute the orange-peel skin to the Expert Working Group, and that is a measured result: on 9 August 2026 the PDF was downloaded again (590,934 bytes, 20 pages, 42,141 characters extracted) and searched in full after removing line-break hyphenation, and orange, peel and dimpl each returned 0 hits; the same search on the same document does hit other sentences quoted in this article (the change in skin texture of the breasts or nipple of ¶7, the There has not been any RCTs identified… of ¶19, the recommends not using ultrasonography… of ¶21), so that 0 is not a failed search. The press release also states that "the symptoms of breast cancer at an early stage are often subtle".

---

#Sources and dates checked

• The three risk groups, the full high-risk definition, the English version's three levels and the Chinese version's two, the Chinese summary table, and MRI not being recommended for the general and moderate-risk groups: Breast Cancer Screening Pilot Programme, "Breast Cancer Screening Recommendations in Hong Kong", https://www.breastscreen.gov.hk/tc/breast-cancer-screening/breast-cancer-screening-recommendations-in-hong-kong/ (Chinese, last changed 15 June 2026) and https://www.breastscreen.gov.hk/en/breast-cancer-screening/breast-cancer-screening-recommendations-in-hong-kong/ (English, Last Revision Date: 15 June 2026), retrieved 1 August 2026
• The tool's limits, the symptom list in the original, the seven risk factors, the boundaries of the input options, the body text's "should consider" wording, and its application to ages 44–69: breast cancer risk assessment tool, https://www.cancer.gov.hk/tc/bctool/, retrieved 9 August 2026
• The five output bands' percentile ranges, risk descriptions and screening advice in the original, and the numerical targets for weight and physical activity: the scoring script that page loads, https://www.cancer.gov.hk/js/questionnaire.js (the recomarr, risktext*, listtxt*, keephealthtxt and txtarr* arrays), retrieved 9 August 2026. That file is executed by the browser and its content never appears in the page text
• 「即風險高於七成半同齡香港女性」, 「應考慮」, the second note's 「約有百分之二十五」, 「約百分之二十八」 (30 048 assessed / 8 484 referred), 「原位乳癌……又稱零期乳癌」, primary prevention as the foremost strategy for preventing cancer, and the three reasons on 3D mammography: Legislative Council eleventh question written reply (Chinese), https://www.info.gov.hk/gia/general/202406/05/P2024060500564.htm, 5 June 2024
• "risk higher than that of 75 per cent of Hong Kong women of the same age", excessive screening wasting resources, and the lack of evidence on 3D mammography: LCQ11 written reply (English), https://www.info.gov.hk/gia/general/202406/05/P2024060500558.htm, 5 June 2024
• The official position on self-examination, clinical breast examination and ultrasound, the risk-based approach, the rising incidence trend from 1991 to 2018, the comparison of Hong Kong incidence with the West, the cost-effectiveness conclusion, and endorsement at the 15th meeting of the Cancer Coordinating Committee on 19 June 2020: CEWG, Recommendations on Prevention and Screening for Breast Cancer – For Health Professionals, Centre for Health Protection, https://www.chp.gov.hk/files/pdf/breast_cancer_professional_hp.pdf, the document giving its own publication date as June 2020 (with a reference-list entry marked [Accessed January 2021]), retrieved 2 August 2026
• The former position of 2018: CEWG, Recommendations … An Overview for Health Professionals, https://www.chp.gov.hk/files/pdf/overview_of_cewg_recommendations_professional_hp.pdf, June 2018
• CEWG ¶17 on pain rates of 1%–77% and refusal of future screening at 11%–46%, ¶18 on radiation of 1 to 10 per 100,000 and "by a factor of at least 100", ¶19 on the absence of any randomised trial comparing annual with biennial, ¶21 on the ultrasound conclusion and the Canadian task force's position, ¶25 on lower sensitivity in younger women and those with dense breasts, ¶29 on the seven risk factors, and ¶30 on the cost-effectiveness conclusion not applying to the high-risk and moderate-risk groups: CEWG, Recommendations on Prevention and Screening for Breast Cancer – For Health Professionals (as above)
• The CEWG's reasoning, the annual-against-biennial false-positive comparison, the assessments of ultrasound and 3D, the self-examination trials, the local model AUC of 0.60, and symptom delay against survival: Tsang THF, Wong KH, Allen K, et al. Hong Kong Med J 2022;28(2):161-168 (PMID 35400644)
• Starting at 44, the biennial interval, the mortality reduction of 25.4% "for all risk strata", the three cost-effectiveness ratios (US$18 151 / 34 953 / 48 303), and the 25.8% coverage at the 8.4% threshold: Leung K, Wu JT, Wong IO, et al. JNCI Cancer Spectr 2021;5(4):pkab060 (PMID 34377936). Checked against the full text via PubMed Central (PMC8346705), not the abstract alone, retrieved 9 August 2026
• The relative risk of 0.80, and 43 deaths prevented against 129 overdiagnoses: Independent UK Panel on Breast Cancer Screening. Lancet 2012;380(9855):1778-86 (PMID 23117178)
• The benefits and harms of biennial screening at 40–74, breast density against false positives, and the comparison with the 2016 version: USPSTF, "Breast Cancer: Screening" (2024), https://www.uspreventiveservicestaskforce.org/uspstf/recommendation/breast-cancer-screening, retrieved 1 August 2026
• Positive predictive value against incidence, the further tests and psychological effects following a false positive, false negatives, and lead-time bias: WHO Regional Office for Europe, Screening programmes: a short guide (2020), https://iris.who.int/items/ba910aa7-3c5a-4cf4-97d8-b128ac712881, retrieved 1 August 2026
• Length-time bias, and five-year survival not being evidence that screening works: National Cancer Institute, "Crunching Numbers: What Cancer Screening Statistics Really Tell Us" (1 May 2012), https://www.cancer.gov/about-cancer/screening/research/what-screening-statistics-mean
• The 2023 incidence and mortality figures, incidence by age, the age-standardised rates of 68.9 and 8.7, the ten-year average annual percentage change of +2.5% and +0.3% (both marked statistically significant), the Segi 1960 footnote, and five-year relative survival: Hong Kong Cancer Registry, Female Breast Cancer in 2023, https://www3.ha.org.hk/cancereg/pdf/factsheet/2023/breast_2023.pdf, August 2025
• The pilot's eligibility and ineligibility lists, the gene list (marked updated 26 February 2025), and the identity card footnote: https://www.breastscreen.gov.hk/en/join-the-pilot-programme/eligibility-of-participants/, last changed 15 June 2026, retrieved 1 August 2026
• The limitations of the screening test, and ultrasound's position as "when clinically indicated": https://www.breastscreen.gov.hk/en/breast-cancer-screening/breast-cancer-screening-test/, last changed 15 June 2026, retrieved 1 August 2026
• The three service organisations currently charging no co-payment, their names, the scope of ultrasound arrangements, the first phase's results, the 2024 assessment and screening numbers, the wording that women's health services are "being integrated" / "will progressively begin" / "replace", the transitional arrangement for those who have paid the annual fee, the division between 64 or below and 65 to 69, the Hospital Authority general outpatient route for underprivileged groups (from March 2025, at "selected" clinics), the 321 women and about 12% specialist referral rate, and 「過去10年……均有所上升」: Department of Health press releases, https://www.info.gov.hk/gia/general/202505/29/P2025052900276.htm, 29 May 2025; https://www.info.gov.hk/gia/general/202510/03/P2025100300294p.htm, 3 October 2025
• The co-payment cap of $225: the pilot's co-payment page, https://www.breastscreen.gov.hk/en/join-the-pilot-programme/co-payment/, last changed 15 June 2026, retrieved 1 August 2026
• Health care vouchers not being usable for the co-payment, and investigations after referral not being subsidised: the pilot's FAQ page, https://www.breastscreen.gov.hk/en/faq/, last changed 15 June 2026, retrieved 1 August 2026
• Hospital Authority charges (non-emergency radiology for eligible persons: basic items free, advanced items $250 each, high-end items $500 each): https://www.ha.org.hk/visitor/ha_view_content.asp?Content_ID=10045&Lang=CHIB5, effective 1 January 2026, retrieved 1 August 2026
• Mammography being classed as an advanced item at $250, and the wording on imaging charges at family medicine clinics: Hospital Authority, public healthcare fee reform FAQ, https://www.ha.org.hk/ho/corpcomm/fncr/index-tc.html, retrieved 9 August 2026
• Women Wellness Satellite co-payments (2D mammography $250, doctor assessment and consultation $150, dedicated nurse clinic $80, breast ultrasound $500, 3D mammography $1,500 unilateral / $2,500 bilateral), the wording of the co-payment model, and the split between basic and value-added services: Women Wellness Satellite, "Service Charges", https://www.wws.org.hk/page/service-charge (the page stating a last-updated date of 6 July 2026, retrieved 9 August 2026)
• The three conditions on the Women Wellness Satellite "who we serve" page: https://www.wws.org.hk/page/service-target, retrieved 9 August 2026; and cancer screening being "determined according to individual need and after a doctor's assessment": "Scope of Services", https://www.wws.org.hk/page/service-scope, retrieved 9 August 2026
• Hong Kong Breast Cancer Foundation self-paid prices: https://www.hkbcf.org/zh/breast_health_centre/main/25/ (effective 13 April 2026, retrieved 1 August 2026); its screening position (the page carrying no date): https://www.hkbcf.org/zh/breast_health_centre/main/21/ (retrieved 1 August 2026); and its Chinese statement that at least 30 countries and territories run population breast cancer screening: https://www.hkbcf.org/zh/media_centre/main/1150/ (the page carrying no date, its content referring to the 2024 Policy Address, retrieved 2 August 2026)
• United Christian Nethersole Community Health Service self-paid prices: https://www.ucn.org.hk/zh-hant/health-check/women-health-check-plan/optional-test, retrieved 1 August 2026
• Hong Kong College of Radiologists, Mammography Statement (first version 9 May 2006; revised 25 August 2015): https://www.hkcr.org/lop.php/mammography_statement, retrieved 1 August 2026
• Hong Kong Anti-Cancer Society, Guideline on breast cancer prevention, detection & screening (revised February 2009): https://www.hkacs.org.hk/, retrieved 1 August 2026
• Hong Kong College of Family Physicians, FP Links, April 2022: https://www.hkcfp.org.hk/upload/FP_Links/FPLinks_202204.pdf, retrieved 1 August 2026
• Academic responses: Lui CY, Fong JCY, Wong MCS. Hong Kong Med J 2022;28(2):100-2 (PMID 35470801); Chien CPY, Ho G, Lam TPW. Hong Kong J Radiol 2023;26(4):260-5 (DOI 10.12809/hkjr2217565)
• The United Kingdom's positive predictive value "Of those 4 people, 1 will be diagnosed with breast cancer.", interval cancers, a normal result guaranteeing nothing, and mammography possibly causing discomfort or pain: GOV.UK, "NHS breast screening: helping you decide" (updated 23 February 2026), retrieved 9 August 2026
• The Chinese wording that compressing the breast "may cause discomfort or pain": Breast Cancer Screening Pilot Programme, breast cancer screening test (Chinese), https://www.breastscreen.gov.hk/tc/breast-cancer-screening/breast-cancer-screening-test/, last changed 15 June 2026, retrieved 9 August 2026
• The United Kingdom's screening arrangements and the risks set out in its invitation leaflet: NHS.UK, https://www.nhs.uk/tests-and-treatments/breast-screening-mammogram/who-breast-screening-is-for/ (last reviewed 9 October 2025); GOV.UK, "NHS breast screening: helping you decide" (updated 23 February 2026)
• The Australian national programme's provisions: Australian Government Department of Health, "About the BreastScreen Australia Program" (the page stating of itself Date last updated: 18 April 2023), https://www.health.gov.au/our-work/breastscreen-australia-program/about-the-breastscreen-australia-program (the sentences quoted here are taken from an Internet Archive capture of that official page dated 13 December 2025, not the live page; retrieved 1 August 2026)
• The seven criteria for a targeted screening programme: UK National Screening Committee, "Criteria for a targeted screening programme", GOV.UK, https://www.gov.uk/government/publications/evidence-review-criteria-national-screening-programmes/criteria-for-a-targeted-screening-programme, 29 September 2022, retrieved 1 August 2026
• The different evidential thresholds for targeted and population programmes, and the feasibility of randomised trials in certain high-risk groups: UK National Screening Committee, "Evidence review criteria: national screening programmes", GOV.UK, https://www.gov.uk/government/publications/evidence-review-criteria-national-screening-programmes, retrieved 1 August 2026

This article was written by the editorial team from official and academic sources, with information as at 9 August 2026; it is health information and not medical advice.

This article was written from the sources listed above. It is about the evidence and the official texts; whether you should be screened, it does not answer — that step is for you and a doctor together.

---

#Further reading

• The Government's breast cancer risk assessment tool: cancer.gov.hk/tc/bctool
• Breast Cancer Screening Pilot Programme: breastscreen.gov.hk
• On this platform: Whole-body MRI screening