Hospitalist-Run Code Sepsis Teams: Lessons From HM17

Sepsis care depends on speed, coordination and clinical judgement under pressure. A patient may arrive with vague symptoms, an abnormal respiratory rate and no obvious source of infection, while the team must decide whether to activate a sepsis response before the diagnosis is fully settled. Hospitalist-run code sepsis teams place an experienced inpatient physician at the centre of that decision, linking emergency, nursing, pharmacy, pathology, intensive care and antimicrobial stewardship.

The model featured in education and discussion around Hospital Medicine 2017, held at Mandalay Bay Resort and Casino in Las Vegas from 1–4 May 2017. The conference brought hospitalists and other healthcare professionals together to examine research, patient safety, practical innovations and continuing medical education. Its conference resources remain useful for understanding how hospital medicine was approaching rapid response systems at the time.

For Australian hospitals, the question is not whether to copy an American protocol word for word. Public hospitals in Sydney, Melbourne, Brisbane and Perth operate within different staffing structures, state-based escalation policies and rural referral patterns. The more useful question is whether a hospitalist-led sepsis team can produce reliable recognition and treatment while fitting Australian governance, workforce and antimicrobial requirements.

Why A Hospitalist At The Centre Can Help

Code sepsis systems often begin with a screening tool, but screening alone does not provide a treatment plan. Fever, hypotension, confusion, tachypnoea and elevated lactate can arise from many conditions, including pulmonary embolism, myocardial infarction, adrenal crisis and medication toxicity. A hospitalist can interpret the pattern, examine the patient and decide whether a sepsis pathway is appropriate.

Hospitalists are also positioned between departments. They understand ward capacity, escalation routes, local prescribing rules and the practical limits of intensive care access. That position can reduce the handover gap between the emergency department and inpatient units, particularly when a deteriorating patient is first noticed by a ward nurse during an evening shift.

The strongest argument is therefore organisational as much as clinical. A named physician leader can give nurses a clear escalation contact, help junior doctors distinguish urgent sepsis from a false-positive screen and ensure that reassessment occurs after fluids, antibiotics and source-control decisions. The team becomes a clinical service rather than an alert on an electronic observation chart.

What HM17 Made Visible

The HM17 discussion placed emphasis on outcomes that matter at the bedside: earlier recognition, timely antibiotics when indicated, appropriate fluid management, faster senior review and fewer avoidable transfers to intensive care. These outcomes should be assessed together. A shorter time to antibiotics is not automatically a success if it produces unnecessary broad-spectrum prescribing or delays source control.

A hospitalist-run response also creates accountability across the whole episode. Someone must review the working diagnosis, confirm whether cultures were obtained appropriately, check renal dosing and document the plan for reassessment. That continuity is particularly valuable after the first hour, when the initial emergency response has ended but the patient may still be unstable.

The case for the model should not be overstated as proof of a single superior staffing arrangement. Conference material can generate a practical hypothesis without replacing a controlled evaluation. HM17’s value lies in showing how hospitals could connect rapid response, education, measurement and interdisciplinary teamwork rather than treating sepsis as a single medication-timing target.

Measuring Outcomes Beyond The First Hour

A credible evaluation should combine clinical outcomes, process measures and balancing measures. Mortality is important, but it is affected by illness severity, comorbidities, treatment limitations and referral delays. A hospital should report risk-adjusted results where possible and review individual cases instead of relying on a headline percentage.

Process measures reveal whether the team is functioning as intended. These may include time from recognition to senior review, blood culture collection where clinically indicated, antibiotic appropriateness, lactate reassessment, fluid documentation, source-control review and completion of a structured handover. Patient-centred measures include length of stay, functional decline, readmission and communication with families.

A useful dashboard should be reviewed by clinicians who understand the limitations of the data. For example, an electronic timestamp may record when an order was entered rather than when a medicine reached the patient. In Australia, different state systems and hospital information platforms can make comparison difficult, so definitions should be agreed before performance is published.

Adapting The Model To Australian Hospitals

Australian implementation needs to account for the public-private divide and the uneven distribution of specialist staff. A tertiary hospital in Melbourne may have an in-house intensive care registrar, infectious diseases support and rapid pathology, while a regional hospital outside Perth may depend on retrieval advice and telehealth. The team design should reflect that reality rather than assume every site can provide a 24-hour hospitalist presence.

State and national standards provide a useful framework. The National Safety and Quality Health Service Standards support recognition and response to acute deterioration, while state services such as Safer Care Victoria, Clinical Excellence Queensland and the NSW Clinical Excellence Commission publish guidance relevant to escalation and sepsis care. Local policy must also align with the Therapeutic Guidelines, antimicrobial stewardship requirements and each hospital’s medicines committee.

An Australian code sepsis team should include nurses as equal operational partners. Nurses often identify subtle changes before a formal trigger appears, especially during routine observations or medication rounds. A simple escalation phrase, visible response criteria and a reliable callback number may be more valuable than a complex scoring algorithm that staff do not use consistently.

Point-of-care assessment can strengthen this response when clinicians are trained and governance is clear. The ultrasound checklist associated with HM17 illustrates how bedside imaging can support questions about cardiac function, lung pathology, volume status and alternative diagnoses. It should complement, not replace, senior examination, formal imaging or critical care consultation.

Building A Safe And Usable Response

The operational playbook should begin with activation criteria that are sensitive enough to catch deterioration but specific enough to avoid constant alarm fatigue. A nurse concern, rapid respiratory change or new confusion may justify review even when a score is below the formal threshold. The response should define who attends, how quickly they attend and what happens when the patient does not improve.

A practical team might include the hospitalist, an experienced nurse, a junior medical officer, a pharmacist or antimicrobial stewardship representative, and an intensive care contact for escalation. Pathology and radiology need clear priority rules, while ward managers must know how to obtain equipment and additional staff. The team should document a working diagnosis, treatment priorities, review time and ceiling of care.

Features That Support Reliability

• A single escalation number or electronic trigger
• Standardised review and reassessment documentation
• Senior antimicrobial and fluid decisions
• A daily audit with feedback to ward teams

Risks That Require Active Control

• Treating every abnormal observation as bacterial sepsis
• Delaying source control while repeating screening tests
• Giving fluids without considering heart or kidney failure
• Measuring speed while ignoring appropriateness and patient goals

Education should use short simulations, case reviews and bedside coaching rather than annual online modules alone. A hospital can rehearse a deteriorating patient during a quiet weekday and then test the same pathway during an overnight shift. This matters in Australia, where fatigue, rotating junior staff and reliance on after-hours medical cover can expose gaps that are invisible during daytime training.

From Conference Learning To Ward Practice

The most defensible business case combines patient safety with responsible resource use. A hospitalist-led team may reduce duplicated reviews, improve admission decisions and identify deterioration before an intensive care transfer becomes unavoidable. It may also increase demand for senior clinicians, pathology, pharmacy input and education. Those costs should be made explicit rather than hidden inside existing workloads.

Privacy and documentation requirements also matter. Clinical notes should record the reasoning behind an activation, antibiotic choice, fluid plan and treatment limitation discussion. Information sharing must comply with the Privacy Act 1988 and local health-service policy, particularly when cases are discussed across emergency, inpatient and retrieval teams. If digital dashboards use identifiable data, access controls and retention arrangements should be agreed with governance staff.

The next step is a 90-day pilot on one medical ward, using agreed activation criteria, a named hospitalist lead, nurse education and a baseline dashboard covering review time, appropriate antibiotics, ICU transfer, acute kidney injury and patient-centred outcomes. After the pilot, the multidisciplinary team should review consecutive cases and decide which elements are ready for wider implementation.