Table 2.

Formal and informal criteria applied in the direct admission process, presented thematically with representative quotations from respondents.

Theme 1: Conditions and populations appropriate for direct admissions (n=17, 17%)*

Most direct admissions are for patients with chronic illness who have a slow progression requiring admission …They are seen in clinic, and then directly admitted, rather than going through a needless ED evaluation. Ideally, I would like all patients with respiratory problems to be evaluated within the ED prior to transfer to the floor. These are the patients that we most commonly have to transfer emergently to the PICU within a short time from admission. We can get quite busy with delivery attendance, well newborn service, circumcisions, NICU coverage so it makes me nervous when I can't get right to the unit to see a patient. We therefore never take respiratory patients, or patients who need a whole sepsis evaluation and urgent antibiotics.

Theme 2: Requisite pre-hospital assessments (n=27, 26%)

There must be a medical evaluation to determine the need for hospitalization. In our rural location, we are an hour and a half away from the nearest ICU (which is not a pediatric ICU). Someone needs to assess the patient to determine if our hospital can take care of the patient, whether transfer to a larger facility is necessary and what resources there are if our hospital isn't the optimal place for the patient. Nothing need be fixed with physicians with whom we have a working relationship. In those instances, the patients are generally as described. In other instances, it is possible to have a patient other than as described. In instances of this being an ongoing problem with referring physicians, we have elected to have patients evaluated in the ED before transfer to the ward.

Theme 3: Clinical stability/need for emergent care (n=27, 26%)

No urgent intervention required in first 60 minutes of arrival to floor. Stable without intervention (except nebs or oxygen) for two hours after arrival. Patients should be ok without any intervention, study or medication for 4–6 hours.

Theme 4: Triage procedures (n=21, 20%)

All of our pediatric direct admits are instructed to enter through our emergency department entrance where they are assessed quickly for stability by a triage nurse. If they are deemed stable they are then directed to proceed to the admissions desk and then to the pediatric unit. If they are recognized to need immediate or urgent intervention then they are sent through the ED queue for treatment and stabilization. Primary physicians and/or affiliates will call the hospitalist attending physician to discuss an admission and an MD to MD hand-off will be provided with demographics and clinical information about the patient. PEWS [Pediatric early warning system] score, vitals, history and PE [physical exam] reported by referring physician.

Theme 5: Role of physician judgment (n=25, 24%)

Within our policy there are guidelines for what should not come directly to the floor. But they are guidelines and are interpreted and reviewed with each clinical situation. Our group has gotten very good at asking questions of our callers to determine if the child should first go to our ED for immediate and rapid care, and we will choose that route if need be. Hospitalists know a direct admit when they hear it? I am sort of joking but it is true. I find it hard to have hard rules on who can and cannot be a direct admit…maybe guidelines but not a policy?

Theme 6: Availability of adequate staffing and beds (n=7, 7%)

In our busy season, if we do not have a bed that will be open within a 30 minute window on our pediatric unit, the patient will also be referred for initial treatment in the ED. [We] will refuse a direct admission if no bed immediately available---patient is then is sent to the ED and is unhappy.

Theme 7: Time of day (n=8, 8%)

We are in-house from 7 AM to 7 PM and sometimes later. If in-house and [the] patient sounds stable, we will admit to the floor. Once home, all come through our ED. We only admit when our hospitalists are in house.

^*number/percent of respondents discussing theme