.png)
A large Brazilian diagnostic or hospital network accumulates, over years of operation, one of the most valuable assets in the health system: millions of laboratory results, from millions of patients, over time. This set is, in principle, a real-world data base, with direct value for clinical research, evidence generation, industry partnerships, and improvement of the operation itself. In practice, most of this value remains locked, and the reason is rarely a lack of data. It is a lack of comparability.
The network has the data. What it doesn't have is the ability to treat it as a coherent set, and without that the real-world asset is reduced to a file too large to throw away and too disorganized to use.
What makes real-world data usable
Real-world data is not synonymous with accumulated raw data. A set of laboratory results only becomes usable real-world data when the values it contains are comparable to each other: when the same analyte, measured in different units by different methods at different operational units over years, can be brought together into a coherent series that faithfully represents the clinical phenomenon it aims to describe.
This is exactly where most bases fail. A network formed by successive acquisitions carries the legacy of multiple laboratory information systems, each with its nomenclature, its units, its methods, and its reference ranges. The same test appears with dozens of distinct designations throughout the base. The same analyte appears measured in units that require non-trivial conversion. The base contains everything, but it contains it in a form that prevents reliable aggregation.
Muñoz Monjas and colleagues, in recent work on real-world data interoperability in healthcare, demonstrate precisely this point: harmonization of laboratory units is a methodological prerequisite for real-world data to be interoperable and analytically usable, and the absence of that harmonization compromises the validity of any analysis built on the base.¹ The data exists, but it isn't in a condition to be used as evidence.
The value that stays locked
The cost of an unusable real-world base doesn't appear in any expense line, because it is an opportunity cost. It is the value of everything the network could do with its data and doesn't.
On the clinical research and evidence generation axis, harmonized real-world bases enable observational studies, effectiveness analyses, and population characterization that fragmented data cannot sustain. The literature on the economic value of laboratory medicine recognizes that aggregated laboratory information has value extending far beyond the individual diagnostic episode, feeding population health decisions and outcome evaluation.² A network that can't reliably aggregate its own data is leaving that value on the table.
On the institutional partnership axis, harmonized real-world data is a negotiable asset. The pharmaceutical industry, research centers, and evidence-generation initiatives seek real population bases, and Brazilian bases have particular value for the scale and diversity of the population. But what is negotiated in these arrangements is not the raw volume of records, it is the quality and comparability of the data. A base that can't be treated as a coherent set has drastically reduced negotiation value, regardless of its size.
On the operation axis, the inability to reliably aggregate data limits internal benchmarking, population trend analysis, and the identification of patterns that only emerge at scale. The network that doesn't see its own data in unified form makes strategic decisions based on partial slices, when it could make them based on the complete set.
Why integrating isn't the same as harmonizing
The most common response of large networks to this problem is investment in a corporate data lake or data warehouse that physically brings together data from all units. This investment resolves the most visible part of the problem, the physical dispersion of the data, and is therefore frequently treated as if it resolved the whole problem. It doesn't.
Bringing the data together in the same repository is technical integration. Making it comparable to each other is semantic harmonization. A data lake with perfect technical integration and absent semantic harmonization is a repository where all the data is physically present and still analytically incomparable, because nomenclature, units, methods, and reference ranges remain heterogeneous inside the unified repository. The network invested in bringing the data together and discovers, at the moment of analyzing it, that bringing it together didn't make it usable.
Vest and Gamm, in one of the main reviews on health information exchange, observe that technical integration resolves only the simplest layer of the interoperability problem, and that semantic fragmentation persists even when the data is physically brought together.³ Standards like LOINC for laboratory test identification exist precisely to resolve that semantic layer, mapping heterogeneous designations to a common identity that makes aggregation reliable.⁴
The layer that turns accumulation into an asset
The difference between a real-world base that generates value and an accumulation of records that doesn't is in the harmonization layer applied over the data. When each result in the base is semantically identified, with its analyte mapped to a recognized standard, its unit normalized, its method tracked, and its reference range documented, the base stops being a collection of heterogeneous records and becomes a coherent set on which valid analyses can be built.
This transformation has a retroactive and prospective effect at the same time. Retroactively, it unlocks the value of the historical data the network has already accumulated and that is today trapped in incomparability. Prospectively, it ensures the data that continues to be generated enters the base already in a usable condition, without accumulating one more liability of non-harmonized data to be dealt with in the future. The network stops generating data debt and starts building an asset that grows in value as it grows in volume.
Plebani, in his analysis of the role of laboratory medicine as a producer of data and not just results, argues that the strategic value of laboratory data depends on its structuring and its ability to sustain decisions beyond the individual diagnosis.⁵ A harmonized real-world base is the materialization of that strategic value: the moment when the data the network always produced stops serving only the point-of-care encounter and starts serving research, evidence, and strategy.
This is where OpenHealth Technologies operates. The platform automatically correlates multiple data streams with rigorously validated logical layers of laboratory tests, semantically harmonizing the real-world data accumulated by a network, with analytes mapped to LOINC, units normalized, methods tracked, and reference ranges documented, across over 3,500 biomarkers. For large diagnostic and hospital networks, this means the historical base that is today trapped in incomparability transforms into a usable real-world data asset for research, evidence generation, partnerships, and strategy, and the data that continues to be generated enters the base already in a usable condition.
Learn how your network can transform years of accumulated, incomparable data into a harmonized real-world asset, usable for research, evidence, and strategic partnerships.
References
- MUÑOZ MONJAS, A. et al. Enhancing real world data interoperability in healthcare: A methodological approach to laboratory unit harmonization. International Journal of Medical Informatics, v. 193, p. 105665, 2025. DOI: 10.1016/j.ijmedinf.2024.105665.
- BERGER, D.; DOBAN, V. The economic value of laboratory medicine. Clinical Chemistry and Laboratory Medicine, v. 52, n. 12, p. 1737–1744, 2014.
- VEST, J. R.; GAMM, L. D. Health information exchange: persistent challenges and new strategies. Journal of the American Medical Informatics Association, v. 17, n. 3, p. 288–294, 2010.
- McDONALD, C. J. et al. LOINC, a Universal Standard for Identifying Laboratory Observations: A 5-Year Update. Clinical Chemistry, v. 49, n. 4, p. 624–633, 2003.
- PLEBANI, M. Clinical laboratories: production industry or medical services? Clinical Chemistry and Laboratory Medicine, v. 53, n. 7, p. 995–1004, 2015.
- GREENBERG, N. Update on current concepts and meanings in laboratory medicine: Standardization, traceability and harmonization. Clinica Chimica Acta, v. 432, p. 49–54, 2014.