Appendix: Content Summaries of Best Papers for the Health Information Management Section
of the 2021 IMIA Yearbook
Powell KR, Deroche CB, Alexander GL
Health data sharing in US nursing homes: a mixed methods study
J Am Med Dir Assoc 2021 May;22 (5):1052-59
Efforts to achieve nationwide interoperability in the US have been ongoing for several
years in part due to federal legislation and regulation. Some providers, such as long
term and post-acute care providers did not receive incentive payments for implementing
electronic health records. The authors studied nursing homes' capability for data
sharing and nursing home leaders' perceptions of data sharing with other health care
facilities and with residents and family members. The authors explore longstanding
challenges to improve data access by patients and their caregivers as well as provider-to-provider
data sharing and exchange across sites of care. This was an exploratory mixed methods
study. The authors performed a secondary analysis of data from a national survey of
815 nursing home administrators in the United States and used a survey developed to
measure nursing home information technology (IT) adoption. The authors used descriptive
statistics and logistic regression models to examine the relationship between health
data sharing capabilities and nursing home characteristics such as location, bed size,
and type of ownership. Additionally, between November 2018 and December 2019, researchers
conducted qualitative interviews with nursing home administrators. Interviews included
questions about processes for sharing data with residents and family members and perceptions
of data sharing with other clinical partners (e.g., hospitals and other entities).
Perceived barriers to data sharing included privacy and security concerns, transparency
and control, fear of lawsuits, and organizational factors which slowed the uptake
of technology. These organizational factors included an overregulated and punitive
environment, perceptions about lack of interest among residents and providers, available
and accessible resources, and other constraints (including financial, workflow, and
technical issues), such as lack of patient portal availability through the health
IT vendor. Perceived benefits of data sharing included improved communication and
care planning and being able to anticipate future demand. The authors found that nursing
homes varied greatly in their technological capabilities and perceptions about what
and with whom they could share information. The COVID-19 pandemic highlighted the
need for data collection and reporting about quality-of-care in nursing homes, testing
and immunization of patients and staff. The pandemic also demonstrated the need for
health information technology and electronic health records to help address infection
prevention, and mitigation. Powell and colleagues include recommendations for additional
research and public policies to address health IT gaps and challenges. Addressing
these gaps and shortcomings will improve the secure exchange of information during
public health emergencies, as well as on a more routine basis.
Cappetta K, Lago L, Potter J, Phillipson L
Under-coding of dementia and other conditions indicates scope for improved patient
management: A longitudinal retrospective study of dementia patients in Australia
Health Inf Manag 2020:1833358319897928
The importance of the quality, accuracy, and consistency of clinical coding of medical
information in hospitals is important for various use cases including payment, resource
allocation, surveillance, epidemiological research, prevention, treatment, and to
inform policy. This paper by Cappetta and colleagues was a longitudinal study of coding
quality of dementia after the initial diagnosis to examine the implications for patient
management and quality of care when dementia was not coded (“under coding”) given
a prior confirmed diagnosis of dementia. The researchers sought to inform future intervention
studies to improve identification and management of dementia in hospitals. The population-based
retrospective cohort study was conducted in a regional local health district of New
South Wales, Australia with five hospitals. This study described rates of dementia
coding over the 5 years (the study period was from 1 July 2006 to 30 June 2015 with
a 5-year lookback period from 1 July 2001) after an initially coded admission for
dementia. The study also identified unintended consequences related to lack of clinical
coding (such as the potential under-management of dementia) and identified patient
subgroups at risk of having inaccurate or incompletely coded diagnoses. The diagnoses
were recorded using the International Classification of Diseases, Australian modification
(ICD- 10-AM Ninth Edition). The researchers found that dementia was coded in 63.9%
of admissions in the 12 months following the index admission for dementia and that
the coding of dementia decreased to 53.7% after 5 years. They also reported that patients
were 20% more likely to have dementia actively managed when it co-occurred with delirium.
The paper highlights the relationship of data accuracy and clinical documentation
completeness. Coding accuracy relies on robust clinical documentation and the absence
of documentation (and related coding) raises questions about gaps in care delivery,
patient management, quality of care and patient safety. The authors offer recommendations
to address under-coding of chronic conditions and improve identification and management
of dementia through dementia-specific care, enhanced clinical protocols, and other
interventions.
Sheriffdeen A, Millar JL, Martin C, Evans M, Tikellis G, Evans SM
(Dis)concordance of comorbidity data and cancer status across administrative datasets,
medical charts, and self-reports
BMC Health Serv Res 2020;20(1):858
As Sherffdeen and colleagues note, risk-adjustment for co-morbidities often requires
data from sources that were originally designed for other purposes; therefore it is
important to understand the underlying foundational principles and definitions of
the datasets and the reliability of different data sources. In Australia, prostate
cancer represents the second leading cause of cancer-related mortality in males. The
Prostate Cancer Outcome Registry-Victoria (PCORVic) was developed in 2009 as a clinical
quality registry, to measure and report on quality of care, using benchmarking of
performance at a clinician and hospital level. The researchers used a retrospective
cohort study design to study the completeness and accuracy of co-morbidity documentation
as reflected in different data sources. The authors studied the level of concordance
for same-patient comorbidity data extracted from administrative data sets (coded from
ICD-10 AM), from the medical record, and data self-reported by men with prostate cancer
who had undergone a radical prostatectomy between January 2017 and April 2018 at one
of six convenient hospitals. The authors analyzed diseases based on the conditions
included within the Charlson Comorbidity Index and compared comorbidities across the
three data sources. Concordance was calculated using percentage agreement and the
kappa statistic. The paper reports on the level of statistical concordance between
the medical chart and administrative datasets; the medical chart and patient self-report
data; and administrative data and patient self-report data. The researchers also included
a summary of the concordance of comorbidity data across the three data sources. The
study notes that agreement between comorbidity data collected by the Victorian Admitted
Episodic Dataset, medical charts, and self-reports by men who have undergone a radical
prostatectomy varied across the analyzed comorbidities. The authors identified discrepancies
between (coded) administrative data sets and the medical charts and noted the need
to further explore the impacts of coding guidelines and practices. They also found
discrepancies between patient self-reported data and the other datasets which might
highlight a need for better patient education or improved communication between patients
and providers. Comorbidity data are important for accurate monitoring of risks and
understanding the accuracy of data sources is critical to data use. The findings about
the data quality of various sources are highly relevant to the HIM domain and functions.
HIM is concerned with the reliability of coded data and the completeness and accuracy
of the documentation to support coding. There are unintended consequences of incomplete
and inaccurate data sources including documentation and coding. Recognizing that data
completeness, accuracy, definitions, and formats may vary by sources is also relevant
to the COVID-19 pandemic since multiple data sources (such as medical records, laboratory
test results, case reports, immunization registries, patients, and caregivers) across
sites of care are critical for public health surveillance, care delivery, disease
management and analytics. Recognizing and reconciling discrepancies in data reporting,
data and interoperability standards, and definitions is also essential to information
sharing and exchange between health care and public health use cases.