The CCF Case Nobody Escalated

Bed 7. 62-year-old man, CCF with reduced ejection fraction, admitted for diuresis. Day 3. The morning round notes him as "comfortable, no acute issues."

Time	HR	BP	SpO₂	RR	Weight	NEWS2
10 a.m.	78	124/78	96%	18	76.4 kg	0
2 p.m.	84	118/72	94%	22	77.1 kg	3
6 p.m.	92	110/68	92%	24	77.8 kg	6
10 p.m.	98	102/64	90% (RA)	28	78.4 kg	8

NEWS2 aggregated from the four recorded parameters, assuming he stayed afebrile and alert throughout — nobody charted temperature or AVPU after the morning round.

Every single reading in that table is one a busy ward will sign off on. None of them, in isolation, is a number that makes a resident stop walking. That is precisely the problem.

Three trends, one direction

Three things are happening at once: he is gaining weight, his respiratory rate is climbing, and his oxygen is falling.

The weight. Two kilograms in twelve hours, on a diuresis admission, is not a diet. It is roughly two litres of retained fluid in a man whose ventricle already cannot move what it has. Heart-failure guidance has treated a gain of 2 kg over three days as a trigger to reassess diuretic dose for two decades. He did it in half a day, and he is still on the dose he was admitted on.

The respiratory rate. RR is the earliest of the vitals to move and the most reliably mis-recorded — the ward literature on this is unambiguous, and every resident knows the sheets that read "18, 18, 18, 18" for a full week. Here it is genuinely rising: 18 → 22 → 24 → 28. A respiratory rate that climbs while the patient is lying still, in a man retaining fluid, is pulmonary oedema announcing itself several hours before it is audible from the foot of the bed.

The saturation. 96 → 94 → 92 → 90 on room air. On its own, 92% invites a shrug. Arriving from 96% in eight hours, in the same patient who has gained 1.4 kg over those same eight hours, it is not a shrug. It is the third instrument in the same chord.

By 10 p.m. he is in decompensated CCF.

The 6 p.m. inflection point

Look at the NEWS2 column rather than any individual cell. Zero at 10 a.m. Three at 2 p.m. Six at 6 p.m. — which is the urgent-response threshold that most NEWS2 escalation policies write down in bold, the score at which the protocol asks for a registrar review and a change of plan. Eight by 10 p.m.

The catch should have happened at 6 p.m. Everything needed to make the call was already on the sheet: RR up six from baseline, sats down four, weight up 1.4 kg, a score that had crossed the line the hospital's own policy defines. What was missing was not data and not knowledge. It was the arithmetic, performed at the bedside, in the eighty seconds the resident actually had.

Why the handover didn't carry it

The night-shift resident arrives at 8 p.m. and is told, in handover, that Bed 7 is "stable, comfortable, day 3 CCF, no acute issues." He glances at the obs sheet, sees the 2 p.m. numbers, agrees with the handover, and moves on.

Nothing about that is negligent. He was handed a label, and the label was accurate at the moment it was formed — that patient was stable at 10 a.m. Labels are what handovers carry when they are spoken from memory at the end of a fourteen-hour shift. Trends are what get dropped, because reciting four sets of numbers takes longer than saying "stable" and the listener has eleven more patients to hear about. This is the failure mode structured handover exists to catch, and it is exactly why the structure has to be in the tool rather than in the resident's discipline.

At 11:40 p.m. the patient is acutely short of breath. He needs BIPAP. He needs another 80 mg of frusemide. He needs an ABG. None of this is catastrophic — the patient recovers — but it is the kind of avoidable middle-of-the-night escalation that exhausts the night-shift team and prevents the patient from having a quiet night.

What the workspace has to do

None of these four is a clever feature. All four are the difference between the 6 p.m. catch and the 11:40 p.m. crash.

1. Show the slope, not the snapshot. A vitals row that displays the last value is a bedside obs sheet with worse handwriting. The card has to show direction and delta over the shift, in the same glance.
2. Do the arithmetic before the resident asks. The aggregate score, the weight delta, the trend across the last four sets — computed on write, not on request. A score the resident has to open a page to find is a score that gets computed after the patient is already in trouble.
3. Change colour when the slope is wrong. Amber on the card at 6 p.m., not a page of numbers that has to be read to be understood.
4. Carry the trend into handover, not the label. If the auto-composed handover says "RR 18 → 24, weight +1.4 kg since morning, NEWS2 6" instead of "stable," the night resident is looking at Bed 7 before he has taken his coat off.

This is why the bed card matters. Not because the doctor does not know to check trends — every resident on that ward could recite the pathophysiology faster than I can write it. Because at the 6 p.m. round, with the senior pushing to leave for the night, the resident has eighty seconds at the patient's bedside, and the obs are on a chart on the wall that nobody is going to read forty numbers from.

If the card on the resident's phone had been amber at 6 p.m. — because the slope was wrong — the resident would have stopped. The patient would have gone on BIPAP at 7 p.m. The night-shift team would have had a quieter shift. The patient would have slept.