New inpatient insulin pump protocols: what hospitalists need to know

What hospitalists need to know about the new insulin pump protocols for inpatients is straightforward: align the device, insulin regimen and ward workflow with the patient’s condition, then document safe escalation and discharge. A pump provides adjustable basal insulin and timed boluses, but it cannot prevent occlusion, hypoglycaemia or ketosis by itself. The Hospital Medicine 2017 programme offered a forum for pairing clinical evidence with safer systems of care.

For Australian teams, a protocol must fit a mixed public and private market, state-specific formularies and variable local device support. Pump hardware and continuous glucose monitoring (CGM) are not routinely funded through the Pharmaceutical Benefits Scheme (PBS), so state programmes, charitable assistance and private cover can shape access. Large Melbourne and Sydney services may have dedicated pump educators, while regional and remote hospitals often rely on telehealth and outreach teams.

Why protocols matter at the bedside

A useful protocol separates planned pump continuation from a new inpatient start. Most hospital pump use is continuation for people with type 1 diabetes or established pump therapy, rather than routine initiation during an acute admission. Starting a pump requires a clear indication, an insulin regimen, training and a named clinician; otherwise, prescribed basal-bolus injections may be safer and simpler.

Newer protocols should define eligibility, expected monitoring, ward ownership and the point at which pump therapy is stopped or converted. They should also distinguish planned changes, such as a temporary basal increase during steroid treatment, from emergency treatment of diabetic ketoacidosis (DKA). This prevents a device adjustment from delaying fluids, insulin or senior review.

Orders, settings and medication reconciliation

On admission, an authorised prescriber should reconcile the pump against the medication chart, discharge summary and, where helpful, the patient or carer. The order needs to identify the active insulin, basal rate, correction factor, target glucose, bolus type and any maximum dose. It should also state when the pump is used as basal-only, when it is suspended and which insulin regimen provides backup.

At every shift handover, the receiving team needs the last confirmed settings, reservoir and infusion-set details, recent glucose pattern and the next planned review. Battery status, date and time settings, reservoir contents and occlusion alarms should be checked against the patient’s usual device. Newer pumps and software can transfer data electronically, but an electronic record does not replace bedside verification.

Choosing basal, correction and monitoring plans

Basal rates, correction ratios and glycaemic targets should be prescribed individually rather than copied from a general protocol. Temporary basal rates can address predictable periods such as morning cortisol elevation, reduced appetite or steroid treatment, but each change needs a review time and a fallback plan. Correction doses should account for active insulin, recent meals and the possibility of insulin-on-board.

Selection should follow the patient’s previous regimen, cognition, mobility, length of stay and available support. Acute illness may make a familiar pump regimen unsafe even when the pump remains functional. In that situation, the treating team should explain the change, document the exact cutover time and ensure that no basal insulin is duplicated.

Monitoring glucose and ketones safely

Monitoring frequency should reflect insulin type, illness severity, prior hypoglycaemia and whether the patient can self-manage. Continuous readings require a plan for confirming unexpected results, particularly when symptoms and sensor readings disagree. Pump users still need access to capillary testing when a CGM is unavailable or unreliable. Colloquially, Australian ward staff may call a low reading a “hypo” and refer to the blood glucose record as the “sugar chart”.

Persistent hyperglycaemia, nausea, vomiting, abdominal pain, altered breathing or increasing thirst require ketone testing even if the pump appears to be working. An occlusion can cause rapid glucose rise, while pump interruption may cause ketones to develop before severe hyperglycaemia appears. Suspected DKA is an emergency requiring senior review, fluids and insulin according to the hospital’s emergency pathway, not simply repeated correction boluses.

Troubleshooting device failure and clinical deterioration

Device failure requires a rapid, standardised sequence: check the alarm, battery, reservoir, tubing and infusion site; review the glucose and ketone trend; and determine whether backup insulin is required. If the pump cannot be used safely, clinicians should preserve access to insulin rather than waiting for a device replacement. The plan must specify when the pump is suspended, who starts the alternative regimen and how the two insulin sources are reconciled.

Behavioural escalation is separate from device escalation. If a pump wearer is confused, distressed or refusing care, staff should use a graduated response rather than restraining the device or obscuring its alarm. The conference resource on safe de-escalation skills offers a useful framework for reducing stimulation and maintaining clear communication while a replacement plan is arranged.

Discharge planning and outpatient ownership

Discharge planning begins at admission, not on the day of transfer. The patient should leave with an updated pump and medication plan, sufficient reservoirs, infusion sets, batteries and compatible supplies. A backup insulin prescription and clear instructions for equipment failure are also required. Documentation must include the final basal rate, correction settings, glucose targets, training completed and the clinician responsible for follow-up.

In Australia, continuity may involve a general practitioner, endocrinologist, diabetes educator, pharmacist and pump clinic, with telehealth linking rural and remote patients to specialist services. Private hospitals may use different suppliers and replacement arrangements from public services, so discharge planning should not assume a standard device pathway. Aboriginal and Torres Strait Islander patients may also need coordinated support from an Aboriginal and Torres Strait Islander health practitioner to make the plan workable at home.

A ward-ready safety checklist for Australian teams

Before rollout, local governance should compare state requirements, hospital capabilities and procurement arrangements. Public and private services need compatible observation, laboratory access and after-hours cover. The final protocol should be tested with pump users, diabetes educators, nursing staff, pharmacists and community teams before routine use.

• Complete medication reconciliation and a two-person device check before the first hospital dose.
• Prescribe basal rates, correction settings, targets, suspension criteria and backup insulin in one visible order.
• Define ketone testing, glucose-monitoring frequency and senior escalation for device or metabolic failure.
• Record pump interruptions, severe hypoglycaemia, DKA and time to restoration of safe insulin delivery for regular audit.

Training should include pump data review, recognition of occlusion, management during fasting or procedures, and safe handover between the emergency department, ward and intensive care unit. It should also cover how to support patients who cannot operate the device because of sedation, delirium, disability or cognitive impairment. Families and carers can be valuable partners, but consent and responsibilities should be documented.

The practical takeaway is that a pump is a delivery device; safety comes from a verified medication plan, active monitoring and a reliable route back to subcutaneous or intravenous insulin whenever the device, patient or clinical condition changes.