A discharge is not an event but a queue of handoffs, and most hospitals only schedule the first one. The physician writes the order, and everything that follows, the medications, the transport, the ride, the paperwork, the cleaning, waits its turn behind work that has a scheduled slot. That is why hospital patient flow stalls in the afternoon while the emergency department holds patients for beds that are technically empty. The bottleneck is not the order but the unowned gaps behind it. Ask a charge nurse on a Sunday which patients are going home and you will get the list by nine in the morning. Ask her at four in the afternoon how many have actually left and you will get a shorter number and a shrug. Two are waiting on the pharmacy queue. One is waiting on a daughter who finishes work at five. One left an hour ago and the room has not been cleaned, because cleaning did not know it was empty. She is right about all of it, and nothing in the building acts on what she knows before it becomes a problem. I run accounts and internal operations, not a ward. I am not a clinician, and none of this is clinical advice. It is an operations argument about handoffs, owners, and deadlines, and the ward version looks a lot like the version I have watched stall a shift change in a plant. Why does hospital patient flow stall at discharge? Hospital patient flow is the movement of patients through admission, treatment, and discharge, and the rate at which beds are freed and refilled. Look at what a hospital day is made of. The operating room has a schedule. Imaging has slots. Labs have runs. Meals have times. Rounds have a start. Every one of those activities has a place on a calendar, a person expecting it, and a visible consequence when it slips. Discharge has a target. Usually something like “before noon,” which is a target with no owner attached. When a target competes with a scheduled appointment, the appointment wins every time. So the porter takes the scheduled imaging transfer first. The pharmacy fills the scheduled cart before the one-off. The cleaner does the rooms on the list before the room that just came free. A hospital schedules the work that fills beds and improvises the work that frees them. That imbalance is the whole problem, and it explains why adding another occupancy screen never fixes it. The screen reports the backup. It does not put the discharge on anyone’s calendar. An empty bed and an available bed are different things There are three states worth separating, and most flow reporting collapses them into one. Occupied. The patient is in the room. Vacated. The patient has left. Nothing has happened yet. Often nobody outside the ward knows this state exists, because it is created by a person walking out of a door, not by a transaction in a system. Available. The room is cleaned, marked ready, and assigned. This is the only state the emergency department can use. The distance between vacated and available, bed turnaround time on a flow report, is where most of the delay happens. It is also the distance nobody owns, because the ward believes it handed the room off and the cleaning team believes it has not received it. Both are telling the truth about their own record. There is no shared record of the moment in between. How to measure discharge delays with timestamps you already have This is a half-day of work and you can do it with the data you already have. Nothing to buy, nothing to install. Pick ten discharges from last week. Mix the days, and include at least one weekend day, because weekends are where the informal fixes stop working. For each one, write down every timestamp you can find, in order: order written, medication list ready, family notified, transport requested, transport arrived, patient left the room, cleaning requested, cleaning complete, bed marked available, next patient assigned. Then do the part that matters. Measure the gaps between steps, not the duration of the steps. Almost every step is quick once it starts. The waiting between them is what makes the discharge take all day. Rank the gaps by median rather than average. One eight-hour outlier pulls the average up and covers over the ninety-minute wait that happens every single day, and the ninety-minute one is the one you can fix. Now write a name beside each gap. Not a department. A person who is accountable for closing that specific wait on a Tuesday. If the honest answer is a department, or two departments, you have found your bottleneck without needing any more analysis. Many of the timestamps will not exist. A step nobody records is a step nobody manages. Write down which timestamps are missing and keep going. That list is half your answer. The widest gap is almost never the clinical one. It sits between two teams who each believe the other one has it. Do one thing with the result this week. Take the single widest gap, give it one owner and one deadline, and tell both teams what the deadline is. One gap, one owner, one deadline. That is a change you can make on Monday without a project plan. Why bed boards and dashboards have not fixed patient flow The screen has usually been bought already. Bed boards, occupancy charts, a flow dashboard on the wall of the coordination room. The information is genuinely there, and the ward still backs up. The reason is that a screen waits for someone to look at it, and the person who needs to act is not looking. The cleaner is not watching a bed board. The porter is not watching a bed board. They are working a list, and the list came from somewhere else. A screen only speeds up discharge when the instruction reaches the person holding a mop, in the minute the room comes free, without anyone having to notice a colour change on a monitor. That is what the charge nurse does today, by memory and by phone, for as many rooms as one person can hold in her head. It works, and it stops when her shift ends. Someone is in a corridor right now waiting on a bed that was vacated at 14:20, cleaned at 15:05, and marked available at 16:40. Nothing clinical happened in that window. Nobody made a bad decision. A room was ready and nobody who needed to know was told. That is the part of hospital patient flow that gets treated as something nobody can control. It is a handoff. You can name it, put a deadline on it, and give it to somebody. This week: Pull the timestamps for ten recent discharges, rank the waits between steps by median, and give the widest one a single owner and a deadline.