Amsterdam-Oxford PSC Score calculator

The Amsterdam-Oxford PSC Score calculator can be used to determine the prognosis for PSC patients.

What you should know before using this tool - Disclaimer

The developers of the Amsterdam-Oxford PSC score have done their utmost to develop a score that reflects the current state of medical knowledge based on the Dutch and Oxford PSC cohorts as of July 2016. The model may be adapted as new information becomes available. It should be emphasized that the information derived from this model is an estimation only. Consumers who use the calculator do so at their own risk. For medical concerns, including decisions about medications and other treatments, users should always consult their physician or other qualified healthcare professional.

The model was developed based on a derivation cohort including patients of 18 years or older. The model was not validated for children, use of this model in patients aged < 18 years is discouraged.

The developers cannot assume any liability or responsibility for damage or injury (including death) to any user, other persons or property arising from the use of this website. The developers cannot and will not be held legally, financially, or medically responsible for decisions made using this calculator.

Please enter laboratory values from blood test at time of diagnosis, or laboratory values at follow-up. The model was validated for the use of laboratory values until three years of follow-up, after diagnosis.

When to use

The Amsterdam-Oxford prognostic model for PSC can be used to make an estimation of long-term transplant-free survival probabilities of PSC patients. The performance of the Amsterdam-Oxford prognostic model remains stable when laboratory values are used for prediction in the first few years after diagnosis.

Why to use

The Amsterdam-Oxford prognostic model for PSC is a broadly applicable model, consisting of readily available disease characteristics and biochemical variables. The model is based on a large, well phenotyped, predominantly population-based cohort from the Netherlands, and externally validated in a PSC cohort from the UK. It can be a useful tool for patient counseling and management. In addition, it may be explored as a tool for risk stratification in clinical trials.

Scientific background

The accuracy of the model to predict if an event will indeed happen somewhere in the future is measured by Harrell's c-statistic. In scientific terms, the c-statistic is the probability of concordance between observed and predicted outcome. If prediction is based on a Cox proportional hazards model, it is computed as the proportion of all pairs of subjects in which the event times can be ordered and are in correspondence with the predicted time-to-event curves. It is a generalization of the area under the receiver operating characteristic (ROC) curve for binary outcomes. A c-statistic of 0.5 means that there is no predictive power at all; 1.0 means perfect prediction. The c-statistic of the Amsterdam-Oxford PSC model is approximately 0.70. This means its predictive properties are fairly good.