A 60-year-old patient needs a cardiology review. In a paper hospital this is a phone call, a card index and a clerk with a pen. Here the request becomes a record: people, hardware, software, data and procedures working as one system — and the earliest slot is resolved against the live schedule in about three seconds.
Slot held with 3.2 s of clerical work — and the same record is now readable at every terminal in the building.
One tap on the terminal has to move four records at once — patient balance, hospital receivables, receipt archive and pharmacy stock. Either all of them move or none do. Flip the network failure switch to watch the database refuse to leave a half-finished bill behind.
Idle · terminal ready
A networked monitor samples the patient continuously and streams the numbers straight into the clinical database. No nurse transcribes them by hand and no chart travels down a corridor — and every wheeled asset on the floor is tagged, so the crash cart is a search result instead of a hunt.
The monitor streamed the drop to the clinical database in 0.4 s and raised an alert in the physician's queue. Nothing was written on paper, and nothing waited for a shift handover.
Six tags broadcasting on the ward network. Staff can see every wheeled asset on the floor map instead of hunting corridors.
A 46 bpm reading has fired. The decision support system has already pulled the history, the medication timeline and the lab results into one panel — but it does not decide. That is still the physician's job, and this is where the system earns its keep.
The DSS did not generate these facts. It retrieved and ranked them — the model management component scoring arrhythmia risk against the knowledge base's guidance on beta-blocker initiation.
Weights recompute the moment the physician records a decision — the system learns from what was actually done, not from what was predicted.
The journey above is not one clever app. It is four classes of information system cooperating on a single longitudinal record — and the record is what makes a hospital smart, not the sensors.
CBIS · people, hardware, software, data, procedures
TPS · OLTP · atomicity, consistency, isolation, durability
DSS + IoT · data, model, knowledge base, interface
Same patient, same hospital, two operating models. The gap is not effort — it is the cost of moving information between systems that do not talk to each other.
The monitor that streams an ECG is cheap. The hard part is that the same reading must be legible to the nurse at the bedside, the physician in the queue, the billing system that charges for the interpretation, and the auditor who reads the record two years later — all without a single retyped number. Every layer in this simulation is really the same question: who is allowed to write to the record, and what happens when they do.