The quantitative truth hidden behind a binary test The quantitative truth hidden behind a binary test
糞便潛血檢驗(FIT)長年以來被視為「陽性 / 陰性」的二分判讀:超過閾值就轉介大腸鏡,沒超過就回家等下一輪篩檢。
但若我們仔細看 f-Hb 的具體數值,會看見一個被忽略的訊號 —— 它是一條連續的風險軌跡,不是一個開關。
過去 20 年,橫跨亞洲、歐洲與全球的研究,都在說同一件事。
For years, the fecal immunochemical test (FIT) has been read as a binary signal — positive or negative: above the threshold means colonoscopy, below means waiting for the next round of screening.
But when we look closely at the actual f-Hb value, a signal long overlooked emerges — it is a continuous risk trajectory, not a switch.
Studies spanning Asia, Europe, and beyond — over two decades — all point to the same conclusion.
f-Hb 愈高,長期追蹤期內大腸腫瘤累積發生率整體呈上升趨勢。 Higher f-Hb predicts a rising long-term cumulative incidence of colorectal neoplasia.
f-Hb 愈高,長期累積大腸癌死亡率愈高。最高組相較未檢出組,死亡率可達約 17 倍。 Higher f-Hb maps to higher long-term CRC mortality. The top tier shows roughly 17-fold mortality compared to undetected.
f-Hb 升高的訊號不只反映大腸癌風險,同時對應更高的全死因死亡率 ── 一個更廣泛的健康指標。 Elevated f-Hb signals not only CRC risk, but also higher all-cause mortality — a broader marker of systemic health.
f-Hb 不是一個開關,
而是一條連續的風險軌跡。
這正是精準篩檢的起點。
f-Hb is not a switch.
It is a continuous risk trajectory.
And this is where precision screening begins.
依基線 f-Hb 分層 · 隨追蹤年數的累積發生率 By baseline f-Hb tier · cumulative incidence over follow-up
每一條線代表一群人,依他們第一次篩檢時的 f-Hb 數值分成不同區間;縱軸是這群人後來累積被診斷出大腸腫瘤的比率,愈往右是追蹤愈久。整體可以看到:基線 f-Hb 愈高的族群,曲線爬升得愈快、愈高;最低與未檢出的組(下方)則長期維持平緩。 Each line follows a group of people, sorted by their f-Hb level at the first screening. The vertical axis is how many in that group were later diagnosed with colorectal neoplasia; further right means longer follow-up. Overall, the higher the baseline f-Hb, the faster and higher the curve climbs; the lowest and undetected groups (below) stay flat over time.
以這條線為例 Follow one line 看最上面那條紅線(≥100):它從第一年就比其他線高,而且一路愈拉愈開。再看最下面那條灰色虛線(未檢出):六年下來幾乎貼著底部。同樣追蹤六年,起點的一個 f-Hb 數字,就讓兩群人走向完全不同的曲線。 Look at the top red line (≥100): it sits higher from year one and keeps pulling away. Now the grey dashed line at the bottom (Undetected): six years on, it barely lifts off the floor. Same follow-up, but a single f-Hb number at the start sends two groups onto completely different curves.
在第一次篩檢就「陰性」的人之中,f-Hb 接近切點(80–99 ng/mL)那組的後續罹病風險,是最低組的 約 3.4 倍 ── 這條「愈高愈陡」的梯度,正說明 f-Hb 是一條連續的風險軌跡,而不是非陽即陰的開關。 Even among those who screened "negative", the group nearest the cutoff (80–99 ng/mL) carries about 3.4 times the risk of the lowest group — this widening gradient is exactly why f-Hb is a continuous risk trajectory, not an on-off switch.
依基線 f-Hb 分層 · 隨追蹤年數的累積死亡率 By baseline f-Hb tier · cumulative mortality over follow-up
每一條線代表一群人,依他們篩檢時的 f-Hb 數值分成不同區間;縱軸是這群人後來因大腸癌死亡的比率,愈往右是追蹤愈久。基線 f-Hb 愈高的族群,死亡率曲線爬升得愈快、愈高;最低與未檢出的組則長期維持平緩。 Each line follows a group of people, sorted by their f-Hb level at screening. The vertical axis is the rate of death from colorectal cancer in that group; further right means longer follow-up. The higher the baseline f-Hb, the faster and higher the mortality curve climbs; the lowest and undetected groups stay flat over time.
以這條線為例 Follow one line 看最上面那條紅線(≥150):它很早就脫離其他線往上衝,代表這群人不只更容易死於大腸癌,而且來得更早。對照最下面的灰色虛線(未檢出):多年下來仍貼近底部。線的高低,就是不同 f-Hb 族群面臨死亡風險的差距。 Look at the top red line (≥150): it breaks away upward early — this group not only dies from colorectal cancer more often, but sooner. Compare the grey dashed line at the bottom (Undetected), which stays near the floor for years. How high a line climbs is the gap in mortality risk between f-Hb groups.
f-Hb 愈高的族群,大腸癌死亡率明顯愈高;最低與未檢出的組則長期貼近底線 ── 同一個數字,不只預測「會不會得到」,更預測「會不會因此死亡」。 The higher the f-Hb, the higher the colorectal cancer mortality; the lowest and undetected groups stay near the baseline — the same number predicts not just who develops cancer, but who dies from it.
依基線 f-Hb 分層 · 隨追蹤年數的累積全死因死亡率 By baseline f-Hb tier · cumulative all-cause mortality over follow-up
每一條線代表一群人,依他們篩檢時的 f-Hb 數值分成不同區間;縱軸是這群人後來「不分死因」的整體死亡率,愈往右是追蹤愈久。基線 f-Hb 愈高的族群,死亡率曲線爬升得愈高;最低與未檢出的組則相對平緩。 Each line follows a group of people, sorted by their f-Hb level at screening. The vertical axis is their overall death rate from any cause; further right means longer follow-up. The higher the baseline f-Hb, the higher the curve climbs; the lowest and undetected groups stay relatively flat.
以這條線為例 Follow one line 這張圖有個跟前兩張不同的地方:連最低的那條(未檢出)也會緩緩上升 ── 因為「所有死因」本來就包含每個人都會面對的各種離世原因。再看 f-Hb 最高的那條(≥450),它幾乎全程都爬得最高最快;只有到了第 8 年,才被 250–449 那條()些微追過。重點不是「會不會上升」,而是 f-Hb 高的族群,整體死亡的步調明顯更急。 This chart differs from the previous two in one way: even the lowest line (Undetected) drifts upward — because "all causes" includes the many reasons everyone eventually faces. Now the highest-f-Hb line (≥450) climbs highest and fastest for almost the entire follow-up; only by year 8 is it edged out by the 250–449 line (). The point isn't whether a line rises, but that higher-f-Hb groups face a markedly steeper overall pace of death.
注意這裡量的是「所有死因」,不只大腸癌。f-Hb 升高的訊號,連扣掉大腸癌之後仍對應較高的整體死亡率 ── 換句話說,同一個篩檢數字,可能是一個更廣泛的健康與壽命指標。 Note this counts death from any cause, not just colorectal cancer. The signal in a rising f-Hb still tracks higher overall mortality even after CRC deaths are set aside — suggesting the same screening number may be a broader marker of health and life expectancy.