. 2021 Dec 13;37(6):1463–1474. doi: 10.1007/s11606-021-07106-8

Table 2.

Provider Interventions to Manage Hospital Capacity Strain

Interventions	Perceptions	Exemplar quotes
Provider team adaptations
Team caps (i.e., maximum number of patients on a given team), safety thresholds (i.e., targets for teams, not necessarily caps) Census sharing agreement (ways that teams may share patients, e.g., cardiology takes certain patients) Level loading (i.e. patients distributed across a variety of teams such that no one team is given more patients than another) Emergency census protocol/back up plans/surge coverage/moonlighters Private patients (i.e., attending patients that are separate from teaching teams); addition of non-teaching teams	Caps protect providers, patients with safe patient numbers Interventions can drive up costs Discontinuity can occur with moonlighting or unexpected staffing needs which can lead to increased length of stay, readmissions, and decreased patient satisfaction Lower census allows for more time spent with patients	And even with that, we have a soft cap, wherein we don’t give two or three more patients than they are supposed to see for the day. This is being done to prevent burnout, and also to protect the patients to ensure the quality of care is not compromised. So what this does to us is this will drive up our moonlighting cost. Every time we have a higher census, we have to bring in moonlighters. So, that will drive up cost for the division and this is like unplanned cost. [moonlighting] …it leads to discontinuity in care with the moonlighters who are coming in and probably coming in for a day or two, and they are not here on a regular cycle. So the discontinuity again leads to a lot of things, including increased length of stay, increased readmissions, and also poor patient satisfaction. (Participant 103b, hospitalist leader)
Provider rounding styles adaptations
Discharges first/discharge by “X” time Conditional discharges (discharge once “X” occurs)	May lead to longer lengths of stay Hard to sustain Good for certain patient types	Well, we’ve done a few things like any other hospital, like discharge 2 patients by 2 pm or 2 patients by 12 pm initiative. I don’t think they have any great impact, because that’s culture change, and it has to happen over time. And if you expect something to change only on high capacity, it doesn’t work; it usually won’t work for this type of intervention. (Participant 103b, hospitalist leader) I do go on a working philosophy that you’re going to get to a saturation point for discharge before noon, because if you’re going to be able to discharge before noon, say, greater than 30%, my argument is your excess days are probably too high. (Participant 105a, hospital leader)
Increasing APP support
Increasing APP roles on teams	Challenges when integrating APPs into teams (at first) Lots of gains with APPs	We’re relatively new to using nurse practitioners on our service. We’ve tried a few things to figure out what’s the best way to have the nurse practitioners help us with these—these flow surges, like focus on discharges and taking care of patients who are expected to go home that day or before noon. We’ve tried to have them pitch in with complex discharge, a lot of things along those lines. I think the balancing thing here is that we want the job to be satisfying for the nurse practitioners. So having them focus on just one specific type of patient, ultimately the feedback we got from them was this isn’t what I signed up for. (Participant108b, hospitalist leader)
Novel team types
Social admitting team Barriers team Long stay units Complex discharge team Transitional care unit	May be able to increase the census for providers with patients who are less acutely medically complex Large discharge barrier for patients; teams can get really good at this type of care	We have like a sizable population of social admissions and geriatric/psychiatric patients. We try to cohort those patients on to one provider and increase the census on this provider. We are a little methodical on how we assign patients—one provider does not get all the sick patients in the hospital. That way everybody has an equal opportunity to work on discharges and get people out. (Participant 103b, hospitalist leader) One of the biggest discharge barriers certainly is housing and security. Part of the reason our census is so high at baseline is because we probably have about 15% of our service consists of patients who do not actually require hospital level care. Some of them are patients that they need their six weeks of intravenous antibiotics but they’re homeless and they don’t have anywhere to go. A larger proportion of them are patients who are cognitively impaired either due to dementia or psychosis or some other reason, and they don’t have a surrogate decision-maker, and/or they’re homeless and so they came into the hospital for some acute reason, but now they have nowhere to go. (Participant 108b, hospitalist leader)
Innovative care models Admitter rounder models Comprehensivist/extensivist Low-risk chest pain Hospital-at-home	Novel care models can help with specific populations Can gain efficiency Costs money and the financial gains may be more indirect	It is an extensivist model where you have a certain cohort of patients that we found out were responsible for the large number of admissions, and we do have a special team that will see them in the hospital when they’re in the hospital and then see them in the clinic when they’re out of the hospital and develop care plans for those patients to help make the hospitalizations consistent if they’re coming in for similar reasons. (Participant 104b, hospitalist leader) We are partnering with some local organizations and other healthcare providers to work on what we—it is a hospital at home model to try to have some patients who can be safely managed at home. We’ve also done work in expanding our urgent care, trying to divert patients from getting so sick and need to be in the hospital. (Participant 106a, hospital leader)

Interventions

Perceptions

Exemplar quotes

Provider team adaptations

Team caps (i.e., maximum number of patients on a given team), safety thresholds (i.e., targets for teams, not necessarily caps)

Census sharing agreement (ways that teams may share patients, e.g., cardiology takes certain patients)

Level loading (i.e. patients distributed across a variety of teams such that no one team is given more patients than another)

Emergency census protocol/back up plans/surge coverage/moonlighters

Private patients (i.e., attending patients that are separate from teaching teams); addition of non-teaching teams

Caps protect providers, patients with safe patient numbers

Interventions can drive up costs

Discontinuity can occur with moonlighting or unexpected staffing needs which can lead to increased length of stay, readmissions, and decreased patient satisfaction

Lower census allows for more time spent with patients

And even with that, we have a soft cap, wherein we don’t give two or three more patients than they are supposed to see for the day. This is being done to prevent burnout, and also to protect the patients to ensure the quality of care is not compromised. So what this does to us is this will drive up our moonlighting cost. Every time we have a higher census, we have to bring in moonlighters. So, that will drive up cost for the division and this is like unplanned cost. [moonlighting] …it leads to discontinuity in care with the moonlighters who are coming in and probably coming in for a day or two, and they are not here on a regular cycle. So the discontinuity again leads to a lot of things, including increased length of stay, increased readmissions, and also poor patient satisfaction. (Participant 103b, hospitalist leader)

Provider rounding styles adaptations

Discharges first/discharge by “X” time

Conditional discharges (discharge once “X” occurs)

May lead to longer lengths of stay

Hard to sustain

Good for certain patient types

Well, we’ve done a few things like any other hospital, like discharge 2 patients by 2 pm or 2 patients by 12 pm initiative. I don’t think they have any great impact, because that’s culture change, and it has to happen over time. And if you expect something to change only on high capacity, it doesn’t work; it usually won’t work for this type of intervention. (Participant 103b, hospitalist leader)

I do go on a working philosophy that you’re going to get to a saturation point for discharge before noon, because if you’re going to be able to discharge before noon, say, greater than 30%, my argument is your excess days are probably too high. (Participant 105a, hospital leader)

Increasing APP support

Increasing APP roles on teams

Challenges when integrating APPs into teams (at first)

Lots of gains with APPs

We’re relatively new to using nurse practitioners on our service. We’ve tried a few things to figure out what’s the best way to have the nurse practitioners help us with these—these flow surges, like focus on discharges and taking care of patients who are expected to go home that day or before noon. We’ve tried to have them pitch in with complex discharge, a lot of things along those lines. I think the balancing thing here is that we want the job to be satisfying for the nurse practitioners. So having them focus on just one specific type of patient, ultimately the feedback we got from them was this isn’t what I signed up for. (Participant108b, hospitalist leader)

Novel team types

Social admitting team

Barriers team

Long stay units

Complex discharge team

Transitional care unit

May be able to increase the census for providers with patients who are less acutely medically complex

Large discharge barrier for patients; teams can get really good at this type of care

We have like a sizable population of social admissions and geriatric/psychiatric patients. We try to cohort those patients on to one provider and increase the census on this provider. We are a little methodical on how we assign patients—one provider does not get all the sick patients in the hospital. That way everybody has an equal opportunity to work on discharges and get people out. (Participant 103b, hospitalist leader)

One of the biggest discharge barriers certainly is housing and security. Part of the reason our census is so high at baseline is because we probably have about 15% of our service consists of patients who do not actually require hospital level care. Some of them are patients that they need their six weeks of intravenous antibiotics but they’re homeless and they don’t have anywhere to go. A larger proportion of them are patients who are cognitively impaired either due to dementia or psychosis or some other reason, and they don’t have a surrogate decision-maker, and/or they’re homeless and so they came into the hospital for some acute reason, but now they have nowhere to go. (Participant 108b, hospitalist leader)

Innovative care models

Admitter rounder models

Comprehensivist/extensivist

Low-risk chest pain

Hospital-at-home

Novel care models can help with specific populations

Can gain efficiency

Costs money and the financial gains may be more indirect

It is an extensivist model where you have a certain cohort of patients that we found out were responsible for the large number of admissions, and we do have a special team that will see them in the hospital when they’re in the hospital and then see them in the clinic when they’re out of the hospital and develop care plans for those patients to help make the hospitalizations consistent if they’re coming in for similar reasons. (Participant 104b, hospitalist leader)

We are partnering with some local organizations and other healthcare providers to work on what we—it is a hospital at home model to try to have some patients who can be safely managed at home. We’ve also done work in expanding our urgent care, trying to divert patients from getting so sick and need to be in the hospital. (Participant 106a, hospital leader)

APP Advanced Practice Provider