Last week, Tamanu EMR went live across all inpatient departments at TTM Hospital in Samoa. Nineteen wards and departments came online on the same day, across more than 300 staff, with usage increasing for workflows across the week. It was the biggest single-day go-live for a digital health project in the Pacific this year, and the culmination of 12 months of work.

For anyone outside the digital health sector, a “go-live” might sound like flipping a switch. In reality, it is closer to a small, carefully choreographed operation, and one that looks very different depending on where in the world it happens.

The basics

The formula we aim for is always the same… after months of build-up, the Go-Live team arrives at the hospital in the week leading up to the big day. This is comprised of an international team from BES, put together with a local team of Change Champions from (in this case) Samoa. The team will wear high-vis vests throughout the week and be stationed throughout the hospital supporting staff, answering questions and identifying issues. Any problems are shared via messaging apps with the control room, staffed by a senior Project Manager and local IT staff. In Samoa, we supplemented this with a ‘virtual’ control room based in Suva, who were on rotation with one staff member allocated full-time to Samoa throughout the entire week.

The night before, the team meets for pizza and a final walk-through of the hospital, testing equipment and network access. Ribbons and balloons go up, the final ‘thumbs up’ decision is taken, senior staff are informed and at 6am on the day of Go-Live, Tamanu blinks to life.

Smooth sailing? Never! But with the right resourcing, calm staff and a lot of laughter, everything is manageable.

Go-lives are not the same everywhere

In a very highly-resourced hospital system, a go-live might come with a budget in the millions, a dedicated project office, and months of parallel running between old and new systems. In the Pacific, that model simply does not exist, and trying to force it rarely works.

At TTM, the same outcome – that is, a safe, complete transition to digital records across a 200-bed tertiary hospital – has to be achieved with a fraction of the resourcing. Every hour of planning has to count, every risk has to be understood in context and anticipated in advance and the people on the ground have to be trusted to make real-time decisions when something doesn’t go to plan. Lower-resourced does not mean lower-standard. It means different, and it means smarter.

People who understand context matter more than credentials

One of our clearest lessons from years of Pacific go-lives is that technical skill alone does not get you through the experience. The staff who make the biggest difference during a go-live are the ones who have worked in similar clinical and public health environments before, and who understand what it means to run a busy ward with limited staff, intermittent power, or a shared laptop between three nurses. Air conditioners leak, members of the public might have greater access to sensitive areas of the hospital, family priorities can pull crucial staff away at inopportune times, rain sometimes falls inside the building, rats chew through cables in the ceiling… none of those are a crisis, that’s just a normal Monday.

Placing people with that lived experience alongside local teams changes the dynamic of a go-live entirely. It is the difference between a support person who can only follow a script and one who can look at a ward at 7am and immediately understand what is about to go wrong, keep their sense of humour, roll up their sleeves and just help.

Local teams lead, we support

It is easy to frame a go-live around the visiting support staff. But the truth is that these projects are led by local health information and IT teams, not by BES. Our staff are there to back them up, not to run the show.

That distinction matters throughout the months of preparation before a go-live, and just as much on the day itself. The relationship with local teams, built over months of requirements gathering, configuration, testing, and training, is what allows a go-live to hold together under pressure. Without that trust, no amount of external support can make up the difference.

No half measures

Perhaps the hardest lesson to apply, and the easiest to get wrong, is that a go-live has to be all-in. If clinical data is allowed to move back and forth between paper and digital systems, even temporarily, the project is in trouble. Staff quickly learn which system is “real,” duplicate records creep in, and the continuity of care the system was meant to improve starts to break down instead.

At TTM, that meant a genuine cut-over. By the end of week one, every incoming patient had a fully digital experience, with handovers, triage, bed management, clinical care, and clinical administration all running through Tamanu. There is no fallback to paper.

Of course, there are still paper workflows in the hospital – forms that were missed, practices which are hard to break, departments which are more complex than others. That’s why the Go-Live isn’t the end of the story, it’s just the start. There are still months of work ahead, making sure that the new systems adapt to the workflows (and vice versa) until everything works seamlessly.

A genuinely regional effort

Last week’s go-live was also a reminder of how connected the Pacific digital health workforce has become. Staff from New Zealand, Fiji, Australia, and Samoa worked together on the ground and remotely, including through a virtual control room based at our regional support centre in Suva, supporting the Samoa Ministry of Health IT team in Apia.

This regional model matters. It means expertise built in one country’s health system can be shared with another. TTM becomes the sixth national hospital in the Pacific to implement Tamanu across all departments, and the second largest hospital to use the system. The cohort of staff that is building are mutually supporting and represent a very real peer-to-peer network which can eventually take BES out of the loop.

Twelve months in the making

Last week did not happen overnight. It followed 12 months of preparation, with the last four in particular given over to intensive requirements gathering, configuration, testing, and training, all underpinned by a broader change management strategy. Over the same period, the team also delivered go-lives at more than ten smaller facilities and outpatient departments across Samoa, each one building the experience and confidence that made today possible.

On the ground in Apia, the day itself scaled up gradually. Final hardware inspections were completed on Sunday night. On Monday morning, staff worked side by side with local teams to migrate current inpatients ahead of each critical shift change, and ward rounds were digitised progressively across the first 48 hours. Support will continue through July, starting with five BES staff on the ground and reducing progressively to one project manager as the local team takes full ownership.

After that, the focus shifts from go-live support to something arguably more important: using the data now being captured for evidence-based decision making and improved clinical care.

What the data already shows

Even before today, Tamanu’s impact in Samoa was measurable. Since the system was first introduced in 2021, more than 1.2 million encounters and 4.6 million clinical events have been recorded across Samoa’s public health facilities. More than 210,000 patients have had more than one encounter in the system, and for 74,000 of those patients, encounters spanned different facilities, a strong signal that longitudinal, connected care is genuinely happening.

More than 900 users now work using Tamanu, from clinicians to public health and health information staff. TTM’s go-live adds a national tertiary hospital to that picture, with 24/7 digital records now underpinning emergency care, paediatrics, ICU, and surgical theatres.

Thank you

None of this would be possible without our donors and partners, including the Australian Department of Foreign Affairs and Trade through the Global Health Division and Partnerships for a Healthy Region (PHR), the World Bank Group, and the Tautua Human Development for All Program, delivered by Palladium.

Congratulations to the local health information and IT teams in Samoa who led this project, and to everyone across Samoa, Fiji, New Zealand, and Australia who supported them to get here.

The financial implications are equally significant. Duplicate records inflate administrative costs, create billing complications, and reduce operational efficiency across healthcare organizations. Staff spend valuable time reconciling conflicting information, while the integrity of health data analytics and population health initiatives suffers from inaccurate patient counts and scattered clinical data.

Recognizing these challenges, the Tamanu EMR has developed a comprehensive, multi-layered approach to patient deduplication that addresses both prevention and remediation. This systematic strategy ensures that healthcare providers can maintain clean, accurate patient databases while delivering safe, coordinated care.

At the centre of any patient management strategy of course, are good processes and procedures, implemented by diligent staff. Tamanu can only help – we recognise though that it is the people at the heart of the system who determine its success.

The five Tamanu strategies for de-duplication

Strategy 1: Patient Master Index Integration

The foundation of effective duplicate prevention lies in establishing a unified patient identity management system through a Patient Master Index (PMI) integration. A PMI serves as the authoritative source for patient identities across an entire healthcare network, maintaining a single, comprehensive registry of all patients served by the organization.

Tamanu’s PMI integration offers remarkable flexibility to accommodate diverse healthcare environments. In small to medium sized healthcare systems or those starting fresh, Tamanu can function as the primary PMI, becoming the central repository for patient identities across all connected systems and departments. This approach ensures seamless coordination between different clinical areas, from emergency departments to specialty clinics or allied health, with each patient maintaining a single, consistent identity throughout their care journey.

For larger healthcare networks or those with existing infrastructure investments, Tamanu can consume data from external PMI systems already in operation. This integration capability means organizations don’t need to abandon their current patient identity management investments. Instead, Tamanu synchronizes with these existing systems, ensuring consistency while leveraging established workflows and data governance practices.

Whether consuming data or making it available to other systems, Tamanu achieves this through an HL7 FHIR compliant API, which defines the resources and the business rules around which the data is exchanged.

The PMI approach creates a robust foundation for preventing duplicates at the source. By establishing unique patient identifiers that persist across all systems and encounters, healthcare organizations can eliminate the confusion that arises when patients are known by different identifiers in different departments or facilities.

Strategy 2: ID Card System

Physical patient identification represents a practical, user-friendly solution to many duplicate creation scenarios. Tamanu’s ID card system addresses the common registration challenges that lead to duplicate records—misspelled names, forgotten dates of birth, or variations in how patients provide their personal information.

The system supports both basic ID cards containing essential patient identifiers and more sophisticated photo ID cards that provide visual verification. When patients present their ID cards during registration or subsequent visits, healthcare staff can quickly and accurately retrieve their existing records without relying on potentially inconsistent verbal information. Each ID card contains a barcode, which is scanned in seconds across any part of the health system and all points of care.

This approach proves particularly valuable in busy clinical environments where registration staff may be working under time pressure or dealing with patients who may be distressed, confused, or unable to clearly communicate their details. The ID card system also benefits patients who frequently interact with the healthcare system, creating a streamlined experience that reduces wait times and administrative friction.

Furthermore, the physical ID card serves as a tangible connection between patients and their healthcare records, encouraging patients to maintain and present their identification consistently across different healthcare encounters and providers within the network.

ID cards can be configured for local context and printed directly from Tamanu

Strategy 3: Real-Time Duplicate Detection

Tamanu’s newly released real-time duplicate detection represents a significant advancement in proactive duplicate prevention. This intelligent system operates at the point of patient registration, analyzing incoming patient data against existing records using sophisticated matching algorithms that consider multiple data points simultaneously.

The system evaluates various criteria including name similarities (accounting for common spelling variations and phonetic matches), birth dates, contact information, and other identifying characteristics. This multi-factor analysis helps identify potential duplicates even when exact matches aren’t present, recognizing that patient information may be provided differently across various encounters.

One of the system’s most significant advantages is its offline capability. Many healthcare facilities, particularly in rural or resource-limited settings, face connectivity challenges that can interrupt their workflows. Tamanu’s offline duplicate detection ensures that patient safety measures remain active regardless of internet availability, making it an ideal solution for diverse healthcare environments.

When potential duplicates are identified, the system presents them to registration staff for review, enabling informed decisions about whether to link to an existing record or proceed with creating a new patient entry. This human-in-the-loop approach combines the efficiency of automated detection with the nuanced judgment that healthcare professionals bring to patient identity verification.

Strategy 4: Patient Merge Tool

Despite the best prevention efforts, duplicate records sometimes occur, particularly when integrating legacy data or managing complex patient populations. Tamanu’s patient merge tool provides a secure, comprehensive solution for consolidating duplicate records while preserving the integrity of all associated clinical data.

The merge process is designed with both safety and usability in mind. Authorized administrators can access the merge functionality easily through Tamanu’s admin panel, using a simple user interface. Here they can review duplicate records and complete the consolidation process. The system preserves all medical history, appointment records, clinical notes, laboratory results, and other patient data, ensuring that no information is lost during the merge operation.

This comprehensive data preservation is crucial for maintaining clinical continuity and meeting regulatory requirements for medical record retention. The merged record provides clinicians with a complete, unified view of the patient’s healthcare history, enabling more informed clinical decision-making and better care coordination.

The merge tool also maintains an audit trail of the consolidation process, documenting which records were merged and when, supporting quality assurance efforts and regulatory compliance requirements.

Strategy 5: Database-Wide Reporting

Maintaining long-term database integrity requires periodic assessment and cleanup of existing patient data. Tamanu’s comprehensive reporting tool addresses this need by analyzing entire patient databases to identify potential duplicates that may exist within legacy data or have been missed by real-time detection systems.

The reporting system employs advanced pattern recognition algorithms to examine patient data across multiple dimensions, identifying subtle similarities that might indicate duplicate records. Like the real-time duplicate detection, this analysis goes beyond simple name matching to consider combinations of demographic information, contact details, and other identifying characteristics that could suggest the same individual is represented multiple times in the database. The system is also designed to improve as it goes. We leverage machine learning to enhance the identification of potential duplicate patient records. Our models are trained to account for regional naming variations and cultural nuances—for instance, recognizing that “Jenny” and “Jennifer” may refer to the same individual. This is done by labelling pairs of potential matches as duplicate, different or unsure. This approach improves accuracy across diverse datasets and geographies.

The reports provide actionable insights, presenting suspected duplicates in a format that enables efficient review and decision-making by administrative staff. This systematic approach to database maintenance helps healthcare organizations conduct regular data hygiene audits, ensuring ongoing database quality and supporting accurate healthcare analytics and reporting.

These periodic cleanups are particularly valuable for organizations transitioning from paper-based systems or consolidating data from multiple legacy electronicsystems, where duplicate creation may have occurred during data migration processes.

The Importance of Integrated Systems and Processes

While Tamanu’s built-in deduplication strategies provide robust protection against patient duplicates, the most effective approach to patient identity management can involve integration with complementary systems and processes that strengthen identification accuracy from multiple angles.

Biometric Identification Systems

Biometric identification represents a powerful complement to traditional patient identification methods, offering unique advantages in accurately distinguishing between individuals. Simprints, an open-source biometric identification platform, is one such example which exemplifies how these technologies can enhance patient identity management in healthcare settings.

Biometric systems capture unique physiological characteristics—such as fingerprints, iris patterns, or facial features—that remain consistent throughout an individual’s lifetime. When integrated with EMR systems like Tamanu, biometric identification can provide definitive patient verification that transcends the limitations of traditional identifiers like names, addresses, or even identity documents, which can be forgotten, lost, or shared between family members.

Simprints has demonstrated particular effectiveness in challenging healthcare environments, including refugee settings, rural communities, and populations with limited documentation. The system’s open-source nature makes it accessible to healthcare organizations with varying resource levels, while its offline capabilities ensure functionality in areas with limited connectivity—a critical consideration for many healthcare deployments.

Biometric identification is not a panacea – in larger populations across broad age and ethnic spectrums, biometrics can help to massively reduce patient duplication and improve patient identification, but the systems still need to be backed by strong patient management systems and high quality policies and procedures.

Civil Registration and Vital Statistics (CRVS) Integration

Civil Registration and Vital Statistics systems provide another crucial layer of patient identity verification and lifecycle management. These systems maintain authoritative records of births, deaths, marriages, and other vital life events that establish and track individual identities throughout their lives. EMRs should interact with CRVS systems at moments including birth and death to ensure that people have official identities that benefit them throughout all facets of their lives, not just healthcare. The integration of CRVS systems with EMRs like Tamanu creates a comprehensive identity ecosystem where healthcare records are anchored to officially recognized vital events.

Birth registration data helps establish definitive patient identities from the earliest stages of life, providing healthcare systems with authoritative information about patient demographics and family relationships. Death registration integration ensures that deceased patients are appropriately flagged in healthcare systems, preventing inappropriate record creation and supporting accurate population health statistics

OpenCRVS represents a leading example of modern CRVS implementation, working globally to improve the management of vital events and strengthen the foundation of identity management systems. Also a free and open-source platform, OpenCRVS are working in the Indo-Pacific region with partners including BES and agencies such as SPC. When integrated with healthcare systems, CRVS data provides several critical benefits for patient identity management.

Conclusion

Effective patient duplicate management requires a comprehensive, multi-layered approach that addresses both prevention and remediation while integrating with broader identity management ecosystems. Tamanu EMR’s five-strategy framework—PMI integration, ID card systems, real-time detection, merge tools, and database-wide reporting—provides healthcare organizations with robust tools for maintaining clean, accurate patient databases.

By combining Tamanu’s built-in deduplication strategies with thoughtful integration of complementary identification systems, healthcare organizations can achieve the highest levels of patient data integrity. This comprehensive approach ensures that every patient record is unique, accurate, and complete—supporting the delivery of safe, coordinated, and effective healthcare services.

Tamanu EMR’s comprehensive approach to duplicate management provides healthcare organizations with the tools and strategies needed to achieve and maintain this essential data quality standard. Come and chat with us about how our Project Management and Data teams can help your health ecosystem achieve its goals.

Watch our video on the five Tamanu deduplication strategies here:

Every day, digital health systems built by BES help visualise thousands of data points across the Pacific and beyond. But data isn’t useful just because it exists. It has to be structured, validated, and presented in a way that helps people make real-world decisions. That’s the job of our data team.

In this post, we share a bit about how that happens – the people behind it, the processes that keep it running, and the work we do to turn raw data into something that matters. Our data team spans Melbourne, Sydney and Canberra, and brings together a mix of backgrounds across analytics, software engineering, and biomedical research.

From point of care to point of impact

Data in Tupaia comes from many sources. Our own data collection tools — MediTrak on Android and DataTrak in the browser — are widely used, especially for surveys and field reporting. But we also pull data from other systems including Tamanu, DHIS2, mSupply, Kobo, and Google Drive, among others. These integrations are designed to be flexible: we can ingest data through APIs, database exports, backups, or spreadsheets, depending on the context.

But just ingesting data isn’t enough. Once we have it, we need to make sure it’s clean, consistent, and safe to report on. That’s where our processing pipelines come in.

Built-in checks and balances

Our data pipelines process over 30 million data points every single day. For that to work reliably, we use automated checks to detect issues in real time, before they impact reports.

When new data arrives, it goes through a series of steps that automatically:

• Remove any deleted, outdated, or otherwise unneeded records
• Check for issues like missing values, data integrity failures, etc.
• Flag inconsistencies for our team to review

If something looks off, it doesn’t just sit there quietly. A validation error is raised and the team is notified. These early alerts are often what allow us to catch issues before they make it into a report or dashboard. It also means we can go back to the original data source, and make sure it doesn’t happen again – whether that’s through user training, or requesting a bug fix in the source system.

Making it easy to report on what matters

As data is ingested, validated, and cleaned, we simultaneously structure it into reporting-friendly database views. We also apply standard calculations at this database view layer, so that metrics are consistent from one visualisation to another. This helps ensure dashboards are fast and reliable, while also keeping the reporting logic clear and easy to follow. Tamanu reports use straightforward queries on these views, and the data from them flows through the rest of our data pipelines. Once in Tupaia’s Viz Builder, they can be combined, filtered, and visualised however needed.

A real-world example: forecasting medication shortages

In September 2023, the Samoan Ministry of Health carried out a mass drug administration for lymphatic filariasis and scabies. Doses were weight-based, and recorded each evening in Tamanu. These were visualised immediately in Tupaia.

After a few days, the dashboards showed something unexpected: stock levels were dropping faster than anticipated. Because of how we model and visualise data, the Ministry was able to quickly spot the trend and take action — placing an emergency order for more medication to ensure the campaign could continue.

That’s the kind of impact we want our data systems to have: helping people see the issue early, and act on it.

To read more about this case study and others like it, visit Data for decision making case studies from the Pacific.

Why it matters

Whether it’s a national outbreak dashboard, a forecasting tool, or a facility-level report, our goal is always the same: to make health data useful. Only then can it be used to improve patient safety, make better policy decisions, or strengthen public health programming.

It takes a solid system, a clear process, and a good team to get there. Fortunately, we have all three.