ABSTRACT

Background:

A principal reason for developing diagnostic software is to reduce diagnostic error and facilitate a more accurate diagnosis. The National Academies of Sciences, Engineering, and Medicine publication “Improving Diagnosis in Health Care”, has three main themes: reduce diagnostic errors, patient involvement and how to address “the increasing complexity of health care”.

Aim:

To describe in principle a computerized decision-tree software Integral Theory Diagnostic System (ITDS).

Materials and Methods:

A node system of individual symptoms guided placement into three zones of ligament damage, anterior, middle, posterior. Percentage probabilities guided diagnosis. The system was tested against a Bayesian Network method.

Results:

The ITDS works well as a diagnostic aid. The diagnosis of zone of damage is displayed pictorially, with deepening colours of red indicating probability of damage. A separate button describes the iteration to the diagnosis. The Bayesian method was tested only in a fairly small number of women, it was found to be equivalent or even superior to that made by the expert.

Conclusion:

The Integral Theory is holistic and its control non-linear. The contribution of the six main ligaments to pathogenesis and particular symptom causation may vary from patient to patient. A more developed version of the ITDS would be very helpful in assisting a more accurate diagnosis and reducing diagnostic error.