The team at #TalkingTeeth started the year with an article discussing the issues of item number driven care. A clear problem in dentistry, but obviously spans across health too. Today, they return with “an alternate way”. Importantly, with the current focus on Artificial Intelligence, their article shows a new way that not only gives us more clarity in our data reporting, but can also enhance the machine learning systems of augmented health care into the future.
The #TalkingTeeth team also introduce a new author, Professor Heiko Spallek, an expert in data systems and dentistry. Welcome Heiko.
Heiko Spallek, Estie Kruger and Marc Tennant write:
In our last article we talked of the complex nature of measuring outcomes of healthcare, based on care provided. Procedure-driven “accounting” has the tendency to distort the care provided, as it is the primary funding mechanism of remuneration to healthcare professionals, including dentists. Today, we want to explore the “health care event” with the lens focused on diagnosis—observing the process from the “other end”.
At the population level, diagnosis provides a status update that describes the health of a nation and permits estimates that predict the necessary provision of care. In Australia and many other countries in the world, Diagnostic Related Groups (DRG’s), reflect this relationship in hospital care.
DRGs for dentistry?
However, ambulatory care settings, including dentistry, have not yet adopted this approach. Drivers for dentistry, by contrast, focus on service provision, resulting in ever expanding care delivery that favours more over better. Researchers and public health policy makers are affected by this procedure-oriented accounting that results in datasets on the care provided through item numbers (but hidden – do see our previous #TalkingTeeth on open data).
Australian dentistry, like most other dental care delivery systems, does not have officially sanctioned diagnostic codes, resulting in barriers to record and monitor diagnoses and subsequently analyse treatment outcomes, much like hospitals do.
Triggered by the desire to improve dental care outcomes and improve patient safety, work has advanced in the area of recording diagnosis in a logical and structured manner, see relevant research here and here. However, these innovations have been predominately in the United States and have not made their way across the Pacific to Australia.
There are standardised diagnostic dental terminologies available; in fact the SNO-DDS terminology is an American National Standards Institute (ANSI) approved standard, now in use in several countries, and supported by the American Dental Association . SNO-DDS is a collection of specific dental terms that is designed for the purposes of dental diagnosis documentation. It consists of about 1700 terms organised into 106 subcategories and 17 major headings.
If we want to track the health status of the population, we need standardised diagnostic terms applied to all conditions. Drawing conclusions from procedures performed is misleading, as we would not know if a filling was placed due to a lost filling, or a primary caries, or secondary caries, or a fractured filling.
The impact of computerisation
After transforming many parts of life, computerisation is also starting to transform health care delivery, also called the Second Machine Age. However, unlike humans, machine-learning algorithms require vast quantities of data to support clinicians in their work. While we disagree with proponents of the complete displacement of professionals from the care delivery process we must acknowledge that dentistry has already greatly benefitted from computerisation (e.g. sequential aligners and milled restorations), and is likely to benefit more from the adoption of technology (e.g. improved delivery of tobacco interventions).
However, all Computerised Clinical Decision Support Systems require detailed standardised, machine-readable inputs. Thus, the adoption of standardised diagnostic terminologies is on the crucial path to improved care. In dentistry, as advocated by Kalenderian et al we need to move to a diagnostic-driven profession to reap the benefits that medicine has derived from structured diagnostic documentation: improved quality of care and improved communication with patients, and between providers.
Population health benefits
At the population level, diagnostic data will assist governments, and more specifically the community, in identifying and monitoring high need or high risk groups, and establish a strong, near-real-time, disease surveillance system so as to assess trends in the nation’s health. And for researchers and policy makers it allows comparisons to be drawn across cultures/countries and focus efforts to the most effective, best practice innovations, to reduce the burden of disease to society.
The adoption of standardised diagnostic terminologies will also enable dentistry to collect data that will be the foundation of a Learning Health System (LHS). The LHS continuously, economically and routinely analyse all data, not just from electronic health record systems, but includes patient-generated health data, wellbeing data, environmental data and other social determinants of health.
When this all becomes part of the culture, cycles of learning that result in improvement can happen on an ongoing basis. Subsequently, patients and healthcare providers at all levels access the “learned” knowledge to proactively monitor and improve health anywhere, any time, and with any device on any platform.
Patient management systems must become learning systems
Patient management systems, that are now near universally used in the public sector, can play a critical role for these services in delivering their mission if they embrace the core principles of the Learning Health System: routine capture of all patient data, transformation of information into knowledge and its dissemination, and economical and continuous improvement as part of the culture.
Health systems, at any level of scale, become learning systems when they can continuously and routinely, study and improve themselves using data that are valid, up-to-date and easily accessible in a Health Information Technology Ecosystem.
Professor Heiko Spallek is Acting-Dean, Faculty of Dentistry, Sydney University. Professors Estie Kruger and Marc Tennant are from the International Research Collaborative Oral Health and Equity, School of Human Sciences, The University of Western Australia.
