Data quality in healthcare: A report of practical experience with the Canadian Primary Care Sentinel Surveillance Network data

Abstract

Data quality (DQ) is the degree to which a given dataset meets a user’s requirements. In the primary healthcare setting, poor quality data can lead to poor patient care, negatively affect the validity and reproducibility of research results and limit the value that such data may have for public health surveillance. To extract reliable and useful information from a large quantity of data and to make more effective and informed decisions, data should be as clean and free of errors as possible. Moreover, because DQ is defined within the context of different user requirements that often change, DQ should be considered to be an emergent construct. As such, we cannot expect that a sufficient level of DQ will last forever. Therefore, the quality of clinical data should be constantly assessed and reassessed in an iterative fashion to ensure that appropriate levels of quality are sustained in an acceptable and transparent manner. This document is based on our hands-on experiences dealing with DQ improvement for the Canadian Primary Care Sentinel Surveillance Network database. The DQ dimensions that are discussed here are accuracy and precision, completeness and comprehensiveness, consistency, timeliness, uniqueness, data cleaning and coherence.

Chronic Disease Surveillance in Alberta’s Tomorrow Project using Administrative Health Data

“…Epidemiology studies and surveillance programs have shown the potential of using administrative health data to identify cases with chronic diseases in Canada [ 2 – 4 , 8 , 13 , 30 – 34 ] and many other countries [ 35 – 44 ]. A few studies have also linked administrative health data to large population-based cohorts, to carry out disease surveillance beyond the primary self-report data possible for these cohorts [ 45 – 48 ].…”

Tackling COVID-19 through responsible AI innovation: Five steps in the right direction

“…A second issue raised by Wynants et al has to do with data impact, viz. the need, in diagnostic, prognostic, and policy-level prediction models, for complete, consistent, accurately measured, relevant, and timely data that is of sufficient quantity to produce reliable out-of-sample generalization (Ehsani-Moghaddam, 2019;Hienrich et al, 2007;Wang et al, 1995). Wynants et al highlight the fact that the studies that they appraise face the common hazard that small sample sizes (drawn from scarce and geographically limited patient populations) will lead to overfitting and compromised generalizability (Foster et al, 2014;Riley 3 The problem here was that the designers of the model chose health care costs as the measurable proxy for the target concept of ill health.…”

A novel decentralized federated learning approach to train on globally distributed, poor quality, and protected private medical data

“…Data quality can also pose a challenge if there is no way to assess quality of individual datasets that are distributed. For many real-world problems, data can be inherently poor-quality due to uncertainty, subjectivity, errors, or subjected to adversarial attack 11 – 13 . This problem is exaggerated when private data at each locality cannot be manually seen or verified.…”