Mapping hospital discharge data to close follow-up gaps for stroke survivors
Stroke is one of Australia's leading causes of long-term disability, and the months after a patient leaves hospital often determine how well they recover. In Queensland, where vast distances separate tertiary stroke units from smaller regional hospitals, the handoff from inpatient care to community rehabilitation is a pressure point for clinicians, families, and carers. The gap between discharge and the first outpatient review is where many patients quietly fall off the clinical radar.
Hospital discharge summaries contain a wealth of information that rarely gets analysed at a system level. Length of stay, discharge destination, medication changes, allied health referrals, and follow-up plans are all recorded in structured fields. When these records are pooled across health services and linked back to outpatient attendance, a clear picture emerges of where the continuum of care is breaking down. Brisbane Diamantina Health Partners has been championing this cross-sector data work as part of its translational mission, connecting researchers, clinicians, and health services across south-east Queensland.
Routinely collected discharge information can be turned into a practical tool for identifying where stroke survivors miss out on timely reviews, rehabilitation, and secondary prevention. Building the partnerships and governance frameworks needed to make that work ethically sound and clinically useful across metro and regional settings is just as important as the analytics themselves.
Where the discharge to clinic journey breaks down
Many stroke survivors leave hospital with a folder of referrals, a medication list, and a follow-up appointment that may or may not happen within the recommended four to six weeks. In rural and remote Queensland, the nearest outpatient neurology clinic can be a full day's drive away, and patient travel subsidies often fall short of covering a return trip for someone with mobility or cognitive impairments. The result is that a scheduled review becomes a missed appointment, and a missed appointment quietly becomes a preventable readmission.
GPs are usually the first port of call once a patient is home, but the transfer of information from the discharging hospital to the local practice is patchy. Summaries sometimes arrive days after the patient has already presented at the clinic with new symptoms, or they arrive incomplete. Allied health follow-up, including speech pathology, occupational therapy, and physiotherapy, is more fragmented, particularly for patients discharged to residential aged care rather than home.
The social gradient shows up as clearly as the geographic one. People from lower-income suburbs, culturally and linguistically diverse communities, and First Nations patients often wait longer for community rehabilitation. Tracking these patterns through discharge data allows services to direct outreach resources where they will make the biggest difference, rather than relying on anecdotal reports from individual clinicians.
What discharge data can tell us about continuity
Every time a stroke patient leaves a Queensland Health facility, a small set of variables is captured almost by default. Admitting diagnosis, ward, discharge date, discharge destination, reconciled medications, and flagged appointments are all part of the standard record. Individually these fields tell a modest story; collectively they describe the patient's trajectory through the system.
When the records are pooled, two kinds of signals tend to emerge. One tracks what the system intended to do for the patient, and the other tracks what actually happened after they walked out the door. Comparing the two quickly reveals where the gap sits.
Key indicators worth tracking in discharge records:
- Time from symptom onset to admission, by hospital and Local Hospital Network
- Discharge destination coded as home, rehabilitation, residential aged care, or other
- Whether a follow-up appointment was scheduled before discharge, and the time to that appointment
- Referrals to community allied health and whether contact was made within 28 days
Patterns that signal a follow-up gap:
- High rates of discharge home without a scheduled outpatient review
- Long intervals between discharge and first contact with community allied health
- Readmission for stroke-related complications within 30 days of leaving hospital
- Low rates of medication reconciliation with the patient's usual GP after discharge
These signals do not diagnose a problem on their own, but they show where audit or service redesign is most likely to pay off.
Building a regional registry from routine records
Turning routine discharge information into a usable registry is less glamorous than it sounds, and it lives or dies on the quality of the underlying documentation. Clinicians need to know that the time they spend filling in fields actually feeds into a system that will change practice. Building pipelines to extract structured information from free-text notes and feeding the cleaned registry back into multidisciplinary team huddles on a quarterly basis is part of the bargain. Particular attention is needed in rural sites such as those in the Wide Bay region, where small sample sizes make outlier events easier to miss.
Governance is the other non-negotiable. Any work that reuses patient data across services has to clear ethics review, comply with the Australian Privacy Principles, and respect cultural protocols that apply to Aboriginal and Torres Strait Islander communities. Putting Indigenous health leaders at the centre of decisions about how their communities' data are used is increasingly seen as best practice across Australia, and it sets a template other groups can learn from.
Funding is rarely glamorous either. Building the pipeline requires software engineers, data scientists, clinical leads, and project managers working across multiple organisations with different finance systems. Pooled funding models that draw contributions from hospitals, universities, and grant bodies are far more likely to survive a change of government than any single line-item allocation.
Linking datasets across Queensland's health services
The real power of discharge data is unlocked when it is linked to other sources, such as Medicare Benefits Schedule claims, Pharmaceutical Benefits Scheme dispensing records, and state-level emergency department presentations. With appropriate ethics and governance, these linkages reveal whether a patient filled their anticoagulant prescription, attended a GP medication review, or bounced back into the emergency department within 30 days of discharge.
For stroke survivors specifically, the linkage can surface uncomfortable truths. It might show that only a third of patients referred to community physiotherapy actually started therapy within the recommended window, or that readmission rates for recurrent stroke are concentrated in a handful of postcodes around Logan and Ipswich. Findings like these justify outreach clinics, telehealth-enabled rehabilitation, and culturally safe community models of care.
Data linkage work in Australia is supported by state data linkage units and the Australian Institute of Health and Welfare, but the application process can be slow and the legal agreements complex. Researchers who have navigated the system often describe it as a marathon rather than a sprint, and they tend to advocate for streamlined master agreements that can be reused across projects.
Turning findings into better outpatient pathways
A dataset on its own does not change a patient's life. The value comes when findings are fed back to clinical teams who can act on them, and when health service executives are willing to redesign pathways. That might mean creating a stroke nurse navigator role that proactively contacts patients in the first week after discharge, or funding a part-time physiotherapist in a regional outreach clinic that previously had no allied health presence at all.
Patient and carer voices should sit at the centre of any redesign. Co-design workshops with stroke survivors and their families often surface simple, low-cost fixes, such as a single phone number to call after discharge, translated information packs, or transport assistance booked before the patient leaves the ward. Combining these lived-experience insights with the patterns visible in linked discharge data creates a much stronger case for change than either source alone.
If you are working on stroke care, health data linkage, or translational research more broadly, explore the data integration work documented through Tazkra Network and connect with the collaborative projects underway at Brisbane Diamantina Health Partners to see how these models are reshaping follow-up care for stroke survivors across the region.
