Assessing the reliability of self-reported weight for the management of heart failure: application of fraud detection methods to a randomised trial of telemonitoring

BMC Medical Informatics and Decision Making

Research Article

Harlan M. Krumholz¹  Zhenqiu Lin¹  Jennifer A. Mattera¹  Adam Steventon²  Sarwat I. Chaudhry³ 

[1] Center for Outcomes Research and Evaluation, Yale School of Medicine, New Haven, USA;Data Analytics, The Health Foundation, 90 Long Acre, WC2E 9RA, London, UK;Yale School of Medicine, New Haven, USA;

关键词: Telemedicine;    End-digit preference;    Self-report;    Alert fatigue;   

DOI  :  10.1186/s12911-017-0426-4

 received in 2016-06-02, accepted in 2017-03-16,  发布年份 2017

来源: Springer

【 摘 要 】

BackgroundSince clinical management of heart failure relies on weights that are self-reported by the patient, errors in reporting will negatively impact the ability of health care professionals to offer timely and effective preventive care. Errors might often result from rounding, or more generally from individual preferences for numbers ending in certain digits, such as 0 or 5. We apply fraud detection methods to assess preferences for numbers ending in these digits in order to inform medical decision making.MethodsThe Telemonitoring to Improve Heart Failure Outcomes trial tested an approach to telemonitoring that used existing technology; intervention patients (n = 826) were asked to measure their weight daily using a digital scale and to relay measurements using their telephone keypads. First, we estimated the number of weights subject to end-digit preference by dividing the weights by five and comparing the resultant distribution with the uniform distribution. Then, we assessed the characteristics of patients reporting an excess number of weights ending in 0 or 5, adjusting for chance reporting of these values.ResultsOf the 114,867 weight readings reported during the trial, 18.6% were affected by end-digit preference, and the likelihood of these errors occurring increased with the number of days that had elapsed since trial enrolment (odds ratio per day: 1.002, p < 0.001). At least 105 patients demonstrated end-digit preference (14.9% of those who submitted data); although statistical significance was limited, a pattern emerged that, compared with other patients, they tended to be younger, male, high school graduates and on more medications. Patients with end-digit preference reported greater variability in weight, and they generated an average 2.9 alerts to the telemonitoring system over the six-month trial period (95% CI, 2.3 to 3.5), compared with 2.3 for other patients (95% CI, 2.2 to 2.5).ConclusionsAs well as overshadowing clinically meaningful changes in weight, end-digit preference can lead to false alerts to telemonitoring systems, which may be associated with unnecessary treatment and alert fatigue. In this trial, end-digit preference was common and became increasingly so over time. By applying fraud detection methods to electronic medical data, it is possible to produce clinically significant information that can inform the design of initiatives to improve the accuracy of reporting.Trial registrationClinicalTrials.gov registration number NCT00303212 March 2006.

【 授权许可】

CC BY   
© The Author(s). 2017