News
From Guidelines to Personalised Follow-up: How LOG-after Brings Clinical Knowledge to the Point of Care
10/09/2026
What began as a solution for long-term follow-up after childhood cancer in France is now expanding. Through the e-QuoL project, LOG-after has been adopted by a centre in Bosnia and Herzegovina and is being implemented in Belgium and several centres in Hungary. This international implementation is an important next step. It will help explore how a shared knowledge-engineering approach can support personalised survivorship care across different healthcare systems while allowing appropriate adaptation to national contexts.
Surviving cancer during childhood, adolescence or young adulthood is only the beginning of a much longer journey. Years — and sometimes decades — after treatment, survivors may remain at increased risk of health problems related to their cancer, previous treatments or individual risk factors.
This is why long-term, risk-adapted follow-up is so important. International organisations have developed detailed recommendations to help healthcare professionals identify which examinations and preventive actions are appropriate for each survivor. Long-term follow-up is widely recognised as an essential part of survivorship care, and European initiatives have demonstrated the value of person-centred, guideline-based approaches.
But there is a practical challenge: how can hundreds of pages of evolving recommendations be translated into the right follow-up plan for one individual survivor?
To address this challenge, digital survivorship passports such as LOG-after and SurPass have been developed.
The challenge: many guidelines, one patient
Recommendations for cancer survivorship come from multiple international and national organisations, including the International Late Effects of Childhood Cancer Guideline Harmonization Group (IGHG), PanCare, the Children’s Oncology Group (COG), PENTEC and other specialist groups.
These sources do not always use the same definitions, levels of evidence or level of detail. Recommendations also evolve as new scientific evidence becomes available.
For a healthcare professional, applying all this knowledge means considering not only a survivor’s original cancer, but also their age, treatments and cumulative doses, radiotherapy exposures, genetic predispositions, family history and other individual characteristics.
LOG-after was developed to make this complex body of knowledge computable, transparent and usable in clinical practice.
Building LOG-after: knowledge engineering before software
LOG-after is more than a database of guidelines.
Its development required a process known as knowledge engineering: clinical recommendations are analysed, compared and transformed into structured pieces of knowledge that a computer can evaluate while retaining their clinical meaning.
The process can be summarised in four steps:
Evidence → Harmonisation → Computable knowledge → Personalised recommendation
First, recommendations from different scientific and clinical sources are reviewed and compared. When recommendations differ, experts analyse the target population, eligibility criteria, evidence, surveillance strategy and publication date to establish an operational recommendation while keeping alternative positions and their rationale traceable.
Next, each actionable recommendation is converted into a modular knowledge object. This object contains much more than a simple rule: it can include eligibility and exclusion criteria, the recommended surveillance action, when it should start, how often it should be repeated, the supporting evidence and bibliographic sources, professional comments, patient information and its version history.
The result is clinical knowledge that can be updated without rebuilding the entire system whenever guidelines change.
Finding the most appropriate recommendation with the information available
One particularly important feature of LOG-after is its hierarchical inference approach.
Information about treatments received many years ago is not always equally precise. For one survivor, detailed radiation doses to a specific organ may be available; for another, only the irradiated field or prescribed dose may still be known.
LOG-after therefore evaluates recommendations from the most specific to progressively broader levels.
If highly detailed information is available, the system can use it. If it is not, it can move to an appropriate fallback rule based on the information that is available.
The objective is simple: use the most specific applicable knowledge for each survivor, without pretending that unavailable information is known.
A decision-support tool — not an automated doctor
The result is a proposed personalised care plan bringing together relevant surveillance, prevention, patient information and contextual recommendations.
Importantly, LOG-after is a clinical decision-support system. It does not replace the healthcare professional. The generated recommendations support clinical decision-making, while responsibility for the final care plan remains with the treating physician.
This distinction is also important for trust: recommendations are based on explicit rules rather than an opaque prediction. Healthcare professionals can trace why a recommendation was produced and identify the evidence behind it.
From France to European implementation through e-QuoL
What began as a solution for long-term follow-up after childhood cancer in France is now expanding.
The knowledge base currently described in the project work contains 193 computable knowledge objects implementing 61 surveillance recommendations, with hierarchical groups containing up to 14 decision levels depending on the clinical domain.
LOG-after is deployed in 15 French paediatric oncology centres and is also being extended within the French paediatric radiation oncology network (LOG-EOL-PEDIA project), with the support of the French Society of Childhood Cancer (SFCE).
Through the e-QuoL project, this experience is now crossing borders: LOG-after has been adopted by a centre in Bosnia and Herzegovina and is being implemented in Belgium and several centres in Hungary.
This international implementation is an important next step. It will help explore how a shared knowledge-engineering approach can support personalised survivorship care across different healthcare systems while allowing appropriate adaptation to national contexts.
Building knowledge that can evolve with survivorship care
LOG-after illustrates a broader ambition of e-QuoL: digital innovation in survivorship is not simply about putting existing guidelines on a screen.
It is about making evidence actionable, explainable, maintainable and personalised.
As evidence changes, individual knowledge objects can be reviewed and updated while their history remains traceable. Feedback from clinicians can also feed into continued refinement. This creates a living knowledge system rather than a static digital guideline.
For survivors, the ultimate objective is straightforward: the right surveillance, for the right person, at the right time.
If you want to know more, a scientific article has been submitted:
Amandine Bertrand, Isabelle Pellier, Marie-Dominique Tabone, Véronique Christophe, Isabelle Ray-Coquard, Christine Rousset-Jablonski, Les Aguerris, Pascal Veillon, Charlotte Demoor-Goldschmidt, Survivorship care plan: correlation between practitioners and a digital tool (LOG-after), The Journal of Pediatrics: Clinical Practice, 2026, https://doi.org/10.1016/j.
Schweizerische Eidgenossenschaft