1,000+ Patients at Home: How Penn State Health Built a High-Performing Hospital-at-Home Program
Guest: Kamia Thakur, MD - System Medical Director, Hospital at Home Program, Penn State Health
Penn State Health’s Hospital at Home (HaH) program led by Dr. Kamia Thakur started as a throughput solution (reduce ED boarding, improve flow in a consistently over-capacity hospital) and evolved into a broader strategy that pairs high-quality inpatient-level care at home with strong patient experience and better outcomes. The program also expanded to improve access to inpatient-level care for rural patients who may lack reliable primary and specialty care. (00:01:46–00:03:30)
Operationally, Penn State runs a three-pronged model: (1) a CMS waiver-based acute HaH inpatient model, (2) a value-based “ambulatory” model that delivers hospital-level care with extended monitoring (30 days), and (3) a post-acute “SNF-at-home” pathway with an acute phase plus 60-day monitoring. They emphasize early admissions, often moving patients’ home within ~12 hours and admitting patients directly from the ED - supported by tight screening workflows, consistent visit intensity, and a strong virtual nursing layer. (00:04:10–00:07:04)
Outcomes discussed include strong patient satisfaction (“top box” scores and unusually high survey response rates), short acute length of stay (about 2.6 days), and very low readmissions in the ambulatory model (reported ~4–5% during monitoring, with “almost zero” for early-to-home patients). In post-acute, the program readmissions down from ~50% early on to ~20% by improving monitoring and treating evolving acute issues at home (e.g., IV diuresis, labs). Financially, they report meeting budget goals consistently and achieving profitability over the last year, enabling growth in nursing capacity. (00:36:33–00:40:57, 00:28:10–00:29:32)
What is Penn State Health’s “North Star” for Hospital at Home, and how does the program support the health system overall?
(Timestamp: 00:01:46)
The program began to improve ED throughput and reduce boarding in a hospital running >105% capacity, then expanded as outcomes and patient experience proved strong. It now serves multiple institutional priorities: throughput, quality, shorter LOS, lower readmissions, and increased rural access to hospital-level care.
“The goal was to decrease ED boarding time, improve throughput.” (00:01:57)
“We run at a capacity more than 105% most of the times.” (00:02:08)
“Patients are doing really well… length of stay is short… outcomes are great.” (00:02:34)
“We’re also providing access to high-level inpatient care in rural areas.” (00:03:08)
“Patients and their outcomes has become kind of our priority… especially with the readmission rates and length of stays.” (00:03:30)
What does the program look like today - what are the key models/pathways and how do they work together?
(Timestamp: 00:03:51)
Penn State operates three integrated pathways managed by the same team: waiver-based acute HaH, a value-based/“ambulatory” acute monitoring model, and post-acute “SNF-at-home.” They aim for a balanced mix (roughly “30-30-30”), are often limited by nursing capacity, and consistently hit census goals. Their distinctive operational focus is early admissions (moving patients home fast, often from the ED) rather than primarily “early discharge.”
“We have actually three prongs of our program… waiver… value-based… and… post-acute, which is replacing SNF at home.” (00:04:15–00:04:50)
“It’s the same team that manages all three programs.” (00:05:03)
“We kind of do a 30-30-30%.” (00:05:03–00:05:26)
“The limitation is nursing capacity.” (00:05:31–00:05:42)
“We are able to take patients out of the hospital within the first 12 hours… we are more focused on early admissions.” (00:06:02–00:06:40)
“We are screening the ED like hawks… and… tried to capture patients from clinics as well.” (00:06:40–00:07:04)
“Now… we get referrals out of… specialty clinics, ED physicians, and… inpatient physicians.” (00:07:21–00:07:36)
What changed over time to improve internal buy-in and referrals?
(Timestamp: 00:07:04)
Early confusion was common; sustained education and demonstrated results shifted perceptions, leading to broad referral sources and smoother adoption.
“When we first began, there was a lot of confusion about the model within the system.” (00:07:04)
“At this point… the entire system is working towards making this program a success.” (00:07:21–00:07:36)
How do you decide which patients fit in the waiver vs. value-based/ambulatory model - and what are the benefits of having both?
(Timestamp: 00:08:05)
When both options apply, the team often lets patients choose - most pick the ambulatory/value-based option because it includes longer monitoring and transition support (30 days). Clinically, Penn State largely standardized intensity across waiver and non-waiver (e.g., two daily visits if needed, daily physician touch, virtual monitoring, specialty access), because non-waiver patients can be “equally sick, if not more.” Post-acute differs fewer physician touches (often weekly), heavier PT/OT, and longer monitoring (60 days).
“We decided we would give both the options to the patient and let them decide.” (00:08:50–00:09:04)
“For a value-based model… monitor… 30 days… helping with care transitions… closing… care gaps.” (00:09:05–00:09:27)
“99% of the patients would end up choosing the model where they’re monitored… 30 days.” (00:09:58–00:10:14)
“We pretty much do the same thing with our ambulatory model too.” (00:10:14–00:10:38)
“Every patient gets a visit a day with the physician… virtual monitoring for all.” (00:10:55–00:11:04)
“Post-acute is different… physician rounding call once a week… more PT and OT.” (00:11:17–00:11:52)
“We don’t differentiate… specialty care… waiver patients… that’s provided for our ambulatory patients as well.” (00:11:52–00:12:06)
How does the value-based/ambulatory program get paid, and what are the eligibility criteria?
(Timestamp: 00:12:33)
Payment is a bundled rate determined per DRG/diagnosis; payers require patients meet inpatient criteria (they run criteria checks and avoid short 24-hour-type admits early in program maturity to stay financially viable). She argues that strong outcomes and fewer readmissions support the business case; they estimate they effectively “added” beds without building infrastructure.
“It’s a rate decided per DRG… and it’s a bundled rate.” (00:12:53–00:13:18)
“They need to meet inpatient criteria… make sure we’re not just providing outpatient… or observation.” (00:13:18–00:13:32)
“You want to make sure the startup is financially viable.” (00:13:32–00:13:48)
“We have technically added 12 more beds… without creating any walls.” (00:14:13–00:14:32)
Do you see opportunities to taper intensity over time rather than delivering the same cadence for 30 days?
(Timestamp: 00:15:06)
Penn State hasn’t fully explored “less visits,” but sees potential for a subset of stable patients who don’t need a nurse physically in the home. For now, standardization simplifies scheduling/workflows; in the future, multiple physician teams could enable a lower-acuity sub-track within HaH.
“I don’t think I have really pondered… providing less visits…” (00:15:57–00:16:08)
“There are certain patients… sick but stable… don’t need direct care from a nurse being present.” (00:16:08–00:16:30)
“Right now… easier… to do the same thing… scheduling easier.” (00:16:30–00:16:46)
“Possibility… a team just managing patients that need… a lesser level of care.” (00:16:46–00:17:05)
How do you identify patients every day - and how do you present the program options, so patients understand what they’re signing up for?
(Timestamp: 00:17:06)
Patients buy-in is usually not the main barrier; messaging is often straightforward. Operationally, they use nurse navigators, EHR screening lists, and – critically- dedicated frontline screening by nurses and a physician. They first confirm operational eligibility (distance, insurance, which arm), then do clinical chart review before engaging the primary team (often ED). Two team members (nurse + physician) speak with the patient to explain the program, consent, safety evaluation, caregiver expectations, informed consent risks, and escalation back to facility. The end-to-end process is designed to take about two hours from identification to transport home, and they report >95% acceptance.
“Presenting the program to the patient was one of the easier things we do.” (00:18:23–00:18:42)
“We do have nurse navigators… screening lists… Cerner… transitioning to EPIC.” (00:18:42–00:18:56)
“Just having flags wasn’t enough… we decided… more upfront screening.” (00:19:06–00:19:31)
“If the physician screens, we do get a more higher quality patient.” (00:19:31–00:19:52)
“Operational eligibility… how far they live… insurance… which arm… then we… run through the clinical care.” (00:20:15–00:20:21)
“Two team members are physically seeing the patient… and… a video call… for sites without a physician physically.” (00:21:50–00:22:14)
“Our acceptance rate has been more than 95% most of the time.” (00:22:14–00:22:22)
“We absolutely discuss… escalation back to the facility.” (00:22:44–00:23:04)
“It takes us about 2 hours from first finding the patient… till… transported out of the hospital.” (00:23:05–00:23:26)
What changed perceptions among ED and inpatient teams over time?
(Timestamp: 00:20:21)
Early skepticism (e.g., oxygen requirements) was tied to confusion with home health; repeated education (including residents) improved trust and streamlined approvals.
“We were always misconstrued as home health service.” (00:20:45–00:20:55)
How do you handle medications, IV therapy, and logistics across the waiver vs ambulatory vs post-acute models?
(Timestamp: 00:23:26)
In the waiver model, the patient remains an inpatient “off facility,” so the hospital supplies medications; they initially lacked IV/pharmacy bandwidth and used a third-party pharmacy for IV meds across models. In Pennsylvania, ambulatory patients can’t remain inpatient, so they’re discharged and re-admitted under an outpatient encounter; they may use home meds or outpatient pharmacy fills, which can be inconvenient- so nurses/couriers often pick up meds. Post-acute similarly uses home meds/outpatient workflows, with IV meds still supported via the outpatient/third-party pathway; a near-term goal is bringing IV fulfillment back to Penn State retail pharmacy with a dedicated pharmacist.
“For the waiver program… considered inpatients off the facility… hospital… provide all of the medications.” (00:23:59–00:24:20)
“We… utilized… a third-party pharmacy… for IV medications… for both ambulatory and non-ambulatory.” (00:24:20–00:24:44)
“For the ambulatory model… in PA… they could not be considered inpatient… discharged… and readmitted under an outpatient encounter.” (00:24:27–00:24:44)
“Our nurses will pick up the medications… or a courier will.” (00:24:52–00:25:14)
“Biggest goals… dedicated pharmacist… bring it back to our own retail pharmacy.” (00:25:14–00:25:41)
What does the post-acute “SNF-at-home” clinical care model look like, and how did you improve readmissions?
(Timestamp: 00:25:43)
Post-acute patients transition into an outpatient encounter. Day 1 includes an in-home nursing visit; PT evaluates early and PT/OT occur most days of the week, with additional nursing visits for labs and the ability to arrange home imaging. After the acute post-acute phase (average reported ~17 days), patients enter a 60-day monitoring phase led by virtual nursing (med adherence, vitals, weights, equipment, follow-up care). They note these are often the sickest patients and initially had very high readmissions (~50%), improving to ~20% by learning to identify and treat evolving acute problems at home (e.g., managing heart failure exacerbations post-sepsis with IV diuretics, labs, and increased nursing support).
“First day… met by our nursing staff… next day… physical therapist evaluates… alternating PT/OT… 6 days a week.” (00:26:12–00:26:55)
“Length of stay… about 16 to 70 days… 17 days on an average.” (00:26:31–00:26:55)
“Then… enter into the monitoring phase… virtual nursing team… weights, vitals… equipment… PCP/specialists… for 60 days.” (00:26:58–00:27:30)
“Patients are the sickest of all three.” (00:27:30–00:27:40)
“Readmission rates were… 50%… we are down to 20% now.” (00:28:10–00:28:44)
“Now… treat them at home… IV diuretics… labs at home.” (00:28:57–00:29:32)
How do you staff the program - physician caps, nursing ratios, and the role of virtual nursing?
(Timestamp: 00:29:33)
Penn State currently run one physician team and a larger nursing team that has grown significantly (from ~3 early to ~9–10). Nurses typically complete ~4-5 visits/day and may flex between roles (screening, care coordination, in-home visits). Physician cap is ~8–10 patients when the physician is also screening/admitting/rounding/discharging; they’re planning additional physician coverage (including admissions-focused FTE during peak hours) and hope to add another full physician team to scale capacity toward ~20 patients across models. Virtual nursing is a separate team with daytime and overnight coverage, crucial especially for post-acute monitoring where patients have less in-person contact.
“Peak admitting hours are between 7 and 1 p.m.” (00:31:54–00:31:59)
“Starting 2026… hoping to get another physician team… could go… up to 20 patients.” (00:32:19–00:32:19)
“We have a separate virtual team… one or two virtual nurses… and… overnight coverage.” (00:32:59–00:33:45)
“If we didn’t have the virtual team… we wouldn’t be able to do what we’re doing.” (00:33:45–00:33:59)
Is the virtual nursing function part of the HaH program or an enterprise/joint-venture structure?
(Timestamp: 00:34:02)
The program is under a joint venture; nursing staff (including virtual nurses) are largely joint venture employees with Penn State input on hiring. Physicians remain Penn State-employed hospitalists.
How do you approach technology and monitoring - do different models use different equipment?
(Timestamp: 00:35:05)
They currently use the same equipment package for all patients and adjust monitoring intensity (continuous vs twice-daily) based on patient need. Kamia sees future opportunity to tailor equipment because not all patients need everything, but standardization simplifies operations today.
What outcomes and performance metrics matter most - and what results are you seeing clinically, operationally, and financially?
(Timestamp: 00:36:13)
They define success across three dimensions: patient outcomes/satisfaction, operational performance (meeting volume goals), and financial viability (budget adherence and profitability). They report meeting budget goals early for 1.5–2 years, profitability over the last year, and strong patient satisfaction (especially in the ambulatory model due to longer engagement). Clinically, acute LOS averages ~2.6 days; they report very low readmissions in ambulatory monitoring (~4–5%) and “almost zero” for patients moved home very early. They also highlight fewer hospital-acquired complications (no C. diff noted; rare pressure injuries), and improving falls performance in higher-acuity post-acute patients via standardized workflows. Patient survey response rates are notably higher than typical inpatient response rates.
What are the biggest remaining opportunities as you scale to the next thousand patients?
(Timestamp: 00:42:36)
Key challenges include regulatory volatility (waiver dependency and shutdown-related pauses), payer participation limits in value-based models, and ongoing institutional buy-in that requires continuous education due to staff turnover. Additional hurdles include nurse staffing constraints tied to home health licensing/partner availability, plus legal/compliance complexity, EHR limitations, and the cost/availability of enabling technology—described as the program’s backbone.
Where is Hospital at Home headed, and what advice would you give teams building programs now?
(Timestamp: 00:49:28)
Dr. Thakur expects HaH to become a standard of care for mild-to-moderate complexity patients, reserving hospitals for higher acuity. She emphasizes starting with a clear institutional “why” (the problem you’re solving), then designing a model that can be customized to local needs. She stresses that beyond C-suite buy-in, success depends on a passionate, dedicated core team—especially given hospital medicine burnout and the rarity of building something new.
“This is going to be the standard of care… mild to moderate… leave the hospitals for higher acuity patients.” (00:49:51–00:50:16)
“You can customize the program based on the needs… of the people you’re serving, as well as the institute.” (00:48:07–00:48:32)
“First question… why does your institution want to have this program?” (00:48:53–00:49:09)
“Your core team should be really passionate… our core team is so dedicated to this vision.” (00:50:37–00:51:00)