Validation of an Online Knee Pain Map and Questionnaire: A Probabilistic Diagnostic Tool

Terminated

• Conditions: Knee Pain

• Registration Number: NCT01492244
• Lead Sponsor: Brock Foster

Brief Summary

"Blank" has designed a medical diagnostic system in the form of an unvalidated online questionnaire and drawing tool used to describe and identify the location of knee pain, respectively. A component of the survey includes the patient inputting their diagnosis as the etiology of their knee pain. Dr. Ivo Dinov's team has used the data from 100,000 patient surveys to construct a probabilistic model to diagnose those who fill out the questionnaire and knee pain map but do not have a diagnosis. However, the validity of the online survey and the accuracy of the probabilistic model has not been confirmed in patients with known diagnoses. Therefore, the purpose of this study will be to recruit patients with knee pain at UCLA orthopedic clinics to complete the online survey which will then be applied to the probabilistic model to output possible diagnoses. The results will be compared to the actual diagnosis assigned to that patient in the clinic. If validated, the online survey may serve as a tool for diagnostic and research purposes.

Detailed Description

Not available

Study Timeline

• Study Start: 2011-12
• Study First Submitted: 2011-12-06
• Study First Submitted QC: 2011-12-12
• Study First Posted: 2011-12-14
• Status Verified: 2012-03
• Last Update Submitted: 2012-03-26
• Last Update Posted: 2012-03-27

Recruitment & Eligibility

• Status: TERMINATED
• Sex: All
• Target Recruitment: 1000

Inclusion Criteria

• patients with knee pain and a known diagnosis for their pain
• patients older than 18 years old

Exclusion Criteria

• patients that are unable or unwilling to complete the online survey.
• patients who do not have a diagnosis for their knee pain

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
The ability of the UCLA modeling software to predict diagnosis based on questionnaire answers	One year	UCLA has developed modeling software that may be accurate at predicting diagnoses depending on the answers given by patients to an online questionnaire and knee pain drawing map. The accuracy of the software has not been tested or validated. This study will determine the accuracy of this software by comparing UCLA orthopedic surgeon input diagnosis to that output by the modeling software following completion of the questionnaire by study participants.

Secondary Outcome Measures

Name	Time	Method
Accuracy of patient input diagnosis compared to orthopedic surgeon diagnosis	One Year	Patients may inaccurately input "known diagnoses" into the online questionnaire because they were diagnosed inaccurately by their doctor, input the wrong diagnosis into the questionnaire by accident, or were never diagnosed with a condition but they input a diagnosis. Therefore, because the modeling software is contingent on accurate patient input diagnoses, determining if patients accurately input their diagnoses into the questionnaire by comparing surgeon input diagnosis to patient input diagnosis may be helpful in elucidating modeling inaccuracy.

Trial Locations

Locations (1)

UCLA; 🇺🇸 Los Angeles, California, United States