Three Technology Questions Healthcare Leaders Should Be Asking Now

Healthcare organizations are under pressure to improve financial performance, expand clinical capacity, and care for patients in settings that cost less than the hospital. Technology is beginning to offer credible answers to each of these problems. The relevant questions for leadership are becoming more specific.

How far can AI automate the revenue cycle?

Revenue-cycle management has long been an attractive target for automation because the work is expensive, repetitive, and dependent on large amounts of structured and unstructured information. The current development is different from earlier rules-based automation. Agentic AI can potentially carry work across multiple steps rather than completing a single isolated task.

Denial management illustrates the opportunity. A system could identify authorization requirements before care is delivered, retrieve supporting documentation, detect deficiencies, prepare submissions, monitor payer responses, and initiate an appeal when necessary. That begins to resemble management of a workflow rather than assistance with one part of it.

The economics are substantial. McKinsey estimates that AI-enabled revenue-cycle transformation could reduce cost to collect by 30 to 60 percent. In its 2025 survey of care-delivery organizations, denial management and appeals were the most frequently identified priorities for automation, cited by 57 percent of respondents. (mckinsey.com)

For senior leaders, the decision is whether to use AI to improve individual revenue-cycle tasks or to redesign selected workflows around substantially greater automation.

The leadership issue is therefore no longer whether AI has applications in revenue cycle. It is how much of the function can be redesigned around automation, which activities still require human judgment, and whether organizations are prepared to change the underlying process rather than layer AI onto existing work.

A practical starting point is to identify one high-volume workflow, define the baseline cost to collect and denial burden, and measure whether automation reduces labor, cycle time, and preventable denials.

Can virtual nursing materially increase bedside capacity?

The nursing workforce problem is usually discussed in terms of supply: recruitment, retention, training and burnout. Virtual nursing introduces another possibility. Some nursing work may not need to be performed physically at the bedside.

Admissions, discharge education, medication reconciliation, patient observation and portions of care coordination can be performed remotely. Hospitals can then reserve bedside nursing capacity for work that requires physical presence, clinical examination, procedures and direct patient care.

This model is already moving beyond experimentation. Virtual care, including virtual nursing and patient monitoring, remains a major priority among nursing leaders, and health systems are increasingly integrating remote nurses with bedside teams rather than treating virtual nursing as a separate service. (aha.org)

The decision is whether virtual nursing should remain a supplemental service or become part of the core inpatient staffing model.

The important measure will not be the number of virtual encounters. It will be whether the model changes staffing requirements, reduces nonessential bedside workload, improves throughput, or allows a given nursing workforce to safely care for more patients.

A useful pilot should therefore be evaluated against staffing requirements, nurse time at the bedside, discharge efficiency, and throughput rather than adoption metrics alone.

That is a more consequential proposition than using technology to make individual nurses somewhat more efficient.

Which hospital patients should no longer be in the hospital?

Hospital-at-home has moved far enough that the question is no longer primarily theoretical. As of February 2026, 366 hospitals had received approval to provide acute hospital care in the home. Current models combine in-person nursing visits with physician telehealth, remote monitoring, diagnostics, pharmacy services and escalation pathways. (aha.org)

This creates an opportunity to reconsider the physical hospital itself.

Many patients need hospital-level care without needing continuous physical access to the infrastructure of a hospital building. If technology can provide reliable monitoring and communication while mobile clinical teams deliver the necessary in-person services, some inpatient capacity can effectively be created without adding beds.

The question to answer is which diagnoses and patient populations should be shifted into home-based acute care, and how much inpatient capacity the organization should plan around that shift.

The strategic question is therefore not simply whether a health system should operate a hospital-at-home program. It is which diagnoses and patient populations can safely move out of the hospital, what proportion of inpatient demand could eventually be managed this way, and how the organization should redesign capacity if that proportion becomes meaningful.

Early programs should track conversion from eligible inpatient care, escalation back to the hospital, outcomes, cost per episode, and the amount of physical capacity released.

These three questions have something in common. None asks leaders simply to acquire a new technology. Each asks whether a new technological capability allows an existing healthcare function to be redesigned.

The largest gains from health technology may not come from making the current system incrementally more efficient. They may come from reducing the work required to collect revenue, changing how scarce clinical labor is deployed, and reconsidering which patients require a hospital bed in the first place.