A clinical decision unit, or CDU, is a dedicated hospital area where patients who are not sick enough to be admitted but not well enough to go home can be monitored, tested, and treated for a short period, typically under 24 hours. These units sit organizationally between the emergency department and a full inpatient stay, giving doctors a structured space to figure out whether a patient’s condition is going to improve or worsen before committing to a costly hospital admission.1Annals of Emergency Medicine. Predictors of Observation Failure in a Clinical Decision Unit The concept has grown rapidly across hospitals in North America, the United Kingdom, and beyond, and the evidence behind CDUs touches on everything from emergency department crowding to insurance billing rules.
How a CDU Differs From the ER and a Regular Hospital Stay
If you arrive at an emergency department with chest pain, the initial workup might take a few hours. At that point, the ER doctor faces a fork: send you home, or admit you to the hospital. A CDU creates a third path. You stay in a monitored bed, often in a space physically adjacent to or embedded within the emergency department, and the clinical team follows a specific protocol for your condition. Blood draws might be repeated at set intervals, imaging may be scheduled, and a specialist could weigh in, all within a compressed time frame designed to reach a clear answer quickly.
The key distinction from a regular inpatient admission is time and intent. An inpatient stay assumes you will need at least one or two midnights in the hospital. A CDU stay assumes the opposite: that your problem can be resolved, ruled out, or clearly identified within roughly 24 hours. The clinical protocols are diagnosis-specific, built around conditions where a short observation window is enough to make a safe call.2PubMed. Continuous quality improvement for the clinical decision unit From the patient’s perspective, you are in a hospital bed being monitored by nurses and doctors. The difference is largely behind the scenes: different billing codes, different staffing models, and a much shorter expected stay.
What Kinds of Conditions Get Managed in a CDU
CDUs are not one-size-fits-all holding areas. They work best for conditions that have well-established short-term observation protocols. Chest pain is the classic example and was one of the earliest applications. A landmark trial compared patients with unstable angina who were randomly assigned to a chest pain observation unit versus standard hospital admission. The observation group had no cardiac events among those who completed the protocol and were discharged, and overall the two groups had similar rates of heart attacks and other complications. Meanwhile, resource use over six months was significantly lower in the observation group.3PubMed. A clinical trial of a chest-pain observation unit for patients with unstable angina
Beyond chest pain, common CDU diagnoses include asthma exacerbations, syncope (fainting episodes), skin infections like cellulitis, and kidney infections. One study of a hospitalist-run 10-bed CDU found that overall length of stay for these conditions dropped after the unit opened, with the biggest decreases in asthma (from about 2.2 days down to 1.2 days) and cellulitis (from 2.4 days to 1.9 days).4PubMed. Implementation of a hospitalist-run observation unit and impact on length of stay (LOS): a brief report The pattern is the same across all of them: these are conditions where most patients either improve quickly with treatment or declare themselves as needing something more intensive.
Some CDUs have expanded into less obvious territory. A recent study tracked patients treated for copperhead snake bites in a CDU and found that 92% were safely discharged from observation without needing a full inpatient admission. The small number who were converted to inpatient stays mostly had uncontrolled pain rather than worsening envenomation.5PubMed Central. Outcomes of Copperhead Snake Envenomation Managed in a Clinical Decision Unit That breadth hints at the flexibility of the model: if you can write a safe observation protocol for a condition, a CDU can probably handle it.
The Impact on Emergency Department Crowding
One of the biggest selling points of CDUs is that they are supposed to unclog the emergency department. When patients who need several more hours of monitoring are moved out of ER beds and into CDU beds, those ER spots open up for new arrivals. The evidence supports this, though the effect is more modest than hospital administrators sometimes hope.
A study across seven Canadian emergency departments found that sites with pilot CDUs saw small but real reductions in how long lower-acuity patients spent in the ER. For nonadmitted patients, the reduction was about seven minutes on average. The units also slightly reduced the absolute admission rate for higher-acuity patients by roughly 0.8 percentage points. Those are not dramatic numbers, but when applied across thousands of patients per year at a busy hospital, the cumulative effect on flow is meaningful.6PubMed. Evaluating the effect of clinical decision units on patient flow in seven Canadian emergency departments
A more striking finding comes from looking at what happens downstream. One analysis compared patients managed by a CDU with observation patients managed through other pathways. Among CDU patients, only about 18% ended up being converted to a full inpatient admission, compared with nearly 64% of observation patients managed outside the CDU.7Annals of Emergency Medicine. The Halo Effect of an Emergency Department-Run Observation Unit on Hospital-Wide Throughput That gap is enormous. It suggests that the structured protocols and focused staffing of a CDU are genuinely better at identifying who can go home safely, rather than defaulting to a full admission out of caution.
Who Runs the Unit Matters
CDUs are not all structured the same way, and the staffing model makes a real difference. Some are run by emergency physicians as extensions of the ER. Others are run by hospitalists, the internal medicine doctors who specialize in caring for hospitalized patients. A few hospitals use a hybrid model where emergency medicine oversees the early hours and hospitalists take over if the stay extends.
The hospitalist-run model has shown strong results in terms of cost. An economic analysis of a hospitalist-staffed acute medical unit found roughly a 30% reduction in total medical costs compared to a non-hospitalist comparison group. The benefit-cost ratio worked out to about 1.33 per patient admission, meaning every dollar spent on the hospitalist model returned about $1.33 in value.8PubMed Central. Economic evaluation of the hospitalist care model in an acute medical unit: a benefit-cost analysis That study looked at an acute medical unit rather than a CDU specifically, but the staffing principle is similar: having dedicated physicians whose workflow revolves around short-stay patients produces efficiency gains that general ward coverage does not.
Emergency physician-led CDUs have their own advantages, particularly for conditions that overlap heavily with ER presentations. If a patient with chest pain needs serial cardiac enzyme tests over several hours, keeping them under the same emergency medicine team avoids a handoff. But when the clinical question requires more of a medicine perspective, such as managing a flare of inflammatory bowel disease or adjusting medications for heart failure, the hospitalist model tends to be a better fit. Many hospitals are still figuring out which approach works best for their patient mix.
Does Skipping Admission Actually Stay Safe
The obvious concern with CDUs is that sending people home faster could mean sending them home too soon. The data on this point is reassuring overall, though not without caveats. In the Canadian multicenter study, the introduction of CDUs did not change emergency department revisit rates.6PubMed. Evaluating the effect of clinical decision units on patient flow in seven Canadian emergency departments A separate study of a hospitalist-run CDU found that 30-day revisit rates for CDU patients were comparable to those for pre-CDU observation patients.9PubMed. Caring for patients in a hospitalist-run clinical decision unit is associated with decreased length of stay without increasing revisit rates And a UK-based study of a CDU that used alternatives to emergency hospitalization found no consistent trend toward higher unplanned return visits.10European Journal of Emergency Medicine. Can an Emergency Department-based Clinical Decision Unit successfully utilize alternatives to emergency hospitalization?
In some settings, CDUs have actually improved safety metrics. After one hospital implemented a mandatory CDU pathway for patients undergoing total hip and knee replacement, the readmission rate dropped substantially. Before the CDU, 40 of 141 emergency visits by these surgical patients resulted in readmission; afterward, only 13 of 224 visits did.11PubMed. Use of a Mandatory Clinical Decision Unit Reduces Readmission Rates following Total Joint Arthroplasty The CDU gave the orthopedic team a place to evaluate post-surgical complaints like swelling or pain without reflexively readmitting patients who turned out to be fine.
CDUs for Children
Pediatric CDUs have their own story. Children present a unique clinical challenge because they tend to get sick and recover faster than adults, and parents are understandably anxious about sending a sick child home. A CDU lets the clinical team watch a child with croup, an asthma flare, or dehydration from a stomach bug for several hours, deliver treatments like nebulizers or IV fluids, and discharge the child once there is clear evidence of improvement.12Clinical Pediatric Emergency Medicine. Pediatric clinical decision units: Observations past, present, and future
One interesting benefit has shown up in the area of pediatric head injuries. After one hospital opened an observation unit, the rate of CT scans performed on children with minor head injuries dropped from about 5.7% to 4.0%, a relative reduction of roughly 30%.13Pediatric Emergency Care. The Effect of an Observation Unit on Pediatric Minor Head Injury That matters because CT scans expose children to radiation, and the vast majority of minor head bumps in kids do not involve a brain bleed. Having a place to watch a child for a few hours gives doctors the confidence to skip the scan when the child looks clinically stable, rather than ordering imaging defensively because the only other option is to send the family home with uncertainty.
The Insurance and Billing Tangle
If you have been placed in a CDU, your hospital stay is almost certainly classified as “observation status” rather than an inpatient admission, and that distinction has real financial consequences for you. Under Medicare’s rules, observation patients are considered outpatients even though they are sleeping in a hospital bed. This affects cost-sharing: outpatient copays apply instead of inpatient deductibles, and crucially, observation days do not count toward the three-midnight inpatient stay required to qualify for Medicare-covered skilled nursing facility care afterward. For an older adult recovering from a fall who spends 36 hours in a CDU and then needs rehab, the observation classification can mean paying thousands of dollars out of pocket.
Medicare’s “Two-Midnight Rule,” introduced in 2013, tried to simplify the inpatient-versus-observation distinction. The basic idea is that if a physician expects the patient to need care spanning at least two midnights, the stay should be classified as inpatient. Anything shorter defaults to observation. An analysis of long observation stays found that the policy does capture most cases that clinically deserve inpatient status, but it introduces its own problems. Physicians are asked to predict at the time of placement how long a patient will stay, and that prediction can be influenced by non-clinical factors like the time of day a patient arrives or which day of the week it is.14PubMed Central. Informing Medicare’s Two-Midnight Policy with an Analysis of Hospital-Based Long Observation Stays A patient who shows up at 11 p.m. is more likely to cross two midnights than one who arrives at 8 a.m. with the same condition.
For patients with private insurance, the rules vary by plan. Some insurers have adopted observation status definitions similar to Medicare’s; others handle it differently. The practical advice for anyone placed in a CDU is to ask directly whether you are being classified as inpatient or observation, and to understand what your plan covers under each classification. Hospitals are required to give Medicare beneficiaries a written notice (called the MOON, or Medicare Outpatient Observation Notice) when observation status exceeds 24 hours, but by that point decisions have already been made. Asking early gives you more time to advocate for yourself if the classification seems wrong.
Where CDUs Fall Short
CDUs are not a magic bullet, and one persistent challenge is making sure patients understand what to expect after they leave. A pediatric CDU study that initially appeared to have a high return-to-ER rate dug into the reasons and found that about 9% of discharged patients came back, and of those returns, roughly a third were deemed clinically unnecessary. The most common reason was a gap between how the illness was actually progressing and what parents understood to be normal. In 90% of those unnecessary return visits, the discharge paperwork had not included any guidance about what symptoms to expect over the next few days.15PubMed Central. Decreased Hospitalizations at the Cost of Increased Emergency Department Returns? Outcomes of a Clinical Decision Unit
That finding points to a structural weakness. CDUs are optimized for rapid throughput: get the patient in, run the protocol, make the decision, move them out. Discharge education, which takes time and is hard to standardize, sometimes gets shortchanged in that workflow. A parent who watches their child improve in the CDU, gets discharged, and then sees symptoms return at home may have no way to know whether the return of symptoms is expected or alarming. The fix is straightforward on paper, give better written and verbal guidance at discharge, but it requires a deliberate investment in the discharge process that not every CDU has built into its workflow.
Another limitation is capacity. CDUs work because they turn over beds quickly. When a hospital’s CDU fills up, patients who would benefit from observation end up staying in ER beds instead, which recreates the crowding problem the CDU was supposed to solve. And CDUs are poorly suited for conditions with unpredictable time courses. A patient with a psychiatric crisis, for example, may need more than 24 hours of stabilization, social work involvement, and placement planning. Trying to compress that into a CDU protocol can lead to premature discharges or protocol “failures” where the patient is admitted anyway after the observation window runs out.
How CDUs Are Evolving
The CDU model is still relatively young, and hospitals continue to experiment with how to make these units work better. Quality improvement programs have become a standard feature, with most CDUs tracking metrics like the percentage of patients who “fail” observation and require admission, average length of stay within the unit, and rates of return visits. The idea is that a well-run CDU should be hitting benchmarks that prove it is not just dumping patients out faster but actually making better clinical decisions about who needs to stay and who can go.2PubMed. Continuous quality improvement for the clinical decision unit
Some hospitals are expanding the range of conditions their CDUs handle, pushing into areas like transient ischemic attacks (mini-strokes), allergic reactions, and even certain low-risk trauma presentations. The guiding principle remains the same: if you can write a protocol with a clear decision point within 24 hours, the condition might be a good CDU candidate. Others are experimenting with virtual observation, where patients are discharged home with remote monitoring devices and a nurse checks in by phone or video at scheduled intervals. Whether that still counts as a “CDU” is a definitional question hospitals have not settled, but the underlying logic, structured short-term observation as an alternative to admission, is the same.
The growth of CDUs also reflects a broader shift in how hospitals think about capacity. Emergency departments across the developed world are struggling with overcrowding and boarding, where admitted patients wait in ER hallways for inpatient beds that are not available. CDUs do not solve that systemic problem, but they chip away at it by diverting a slice of patients who never needed an inpatient bed in the first place. For the patient, the experience ideally means a faster resolution and fewer days spent in a hospital gown. For the hospital, it means better use of expensive inpatient beds. The tension, as always, lies in making sure speed does not come at the cost of thoroughness.