A Focus on Standards - Pinga Clinical Coding and the Impact of Standardisation
Pinga - Clinical Coding and the Impact of Standardisation
ICD-10 is a popular international standard for diagnosis in healthcare. In this article we present a story of the importance of using standards to drive to a more standard lexicon to help health data interoperability.
ICD-9, ICD-10 and SNOMED are all international coding standards for diagnosis in healthcare. ICD-10 in many jurisdictions is being replaced by SNOMED but still widely used internationally. SNOMED is an all-encompassing clinical terminology standard as a standard for quality information. It is increasingly being deployed in New Zealand to improve the interoperability of clinical data and diagnosis so we can compare data across systems.
Dalibor from Aceso Health discusses launching their platform Pinga into the Republic of North Macedonia and the impact of mass adoption of ICD-10 within this healthcare system. Standardisation has and will lead to improved interoperability allowing health systems to work together.
Dalibor’s team, implemented the Pinga platform (powered by Sorsix technology) in Macedonia between 2012 and 2013. Pinga is a single EHR and workflow platform combining list management, patient administration, clinic management, practice management, real time digital referral, e-prescription, billing integration and clinical enterprise scheduling. An essential part of the Pinga deployment in Macedonia is the enforcement of ICD-10 codes as a standard that underpins all clinical activity. Pinga supports all standard interoperability frameworks (HL7/FHIR and others) and coding regimes (DRG’s (Diagnosis-related group’s), ICD and SNOMED) but in Macedonia only DRGs and ICD-10 codes are used.
A key tenet of this deployment of Pinga is that all clinical activity must be associated with one, or more, ICD-10 codes (a primary diagnosis, and optional secondary diagnoses). Training and working with doctors were necessary to ensure a high quality of data in the system; this took approximately 12 months. Now, with almost a decade of statistics on the performance of this system, both operationally and clinically, the impact is striking.
Pinga allows GPs, specialists, and their administrative teams to refer and book appointments directly via a central hub and spoke model, giving patients access to priority slots for urgent care. By basing all clinical activity on ICD-10 codes, and integrating these referrals electronically, the amount of time spent ‘chasing information’ is dramatically reduced. This improvement in efficiency caused the capacity of the Macedonian health system to almost double over 4 years as adoption grew more widespread.
The figure indicates that the amount of care delivered in the Republic of Macedonia functionally doubled over 4 years. This includes care delivered across all specialities and in all regions. During this time there was no increase in the resources available to the system: doctors, machinery, and beds.
Implementation of this system in Macedonia abolished waiting lists for care in 6 months, due to increased supply from this jump in efficiency. Most patients referred to secondary or tertiary care are seen within one day. The longest wait time is approximately 2 months for care or hospitalization – but there is no ‘waiting list’; patients are allocated a timeslot and know when they will be seen.
The link between standardisation of coding standards and waiting times is indirect but powerful. By establishing a standard coding for diagnoses:
1. Referral triage is greatly simplified since diagnoses guide referrals.
2. Clinical standards for triage can be implemented, such as requiring referrals for certain conditions to have prerequisites met.
a. One example is ICD-10 code N20, kidney conditions; consultation for this condition requires a KUB ultrasound and blood and urine tests. Process triage is enabled when diagnoses are standardised, enabling clerical staff to ensure prerequisites are completed before a specialist consultation.
3. A population-level statistical assessment can be made of the distribution of demand and supply for types of conditions.
a. For example, depending on water treatment, regions of the country might have different ratios of kidney stones to osteoporosis (these are generally inversely proportional).
b. This enables reallocation of clinical resources to ensure that supply more closely matches demand in the health system.
4. Mandating a clinical standard reduces time spent chasing clinical records before a patient comes to a consultation.
5. The mandate of one standard empowers the value provided by other standards. For example, HL7-driven systems suffer from a lack of ‘standards within the standard’ of messaging; such that systems do not share adequately structured information. Implementation of a diagnostic coding standard ensures better standardisation of information throughout the system and in turn make the HL7 payload more useful and interoperable.
Altogether the experience of standardisation was remarkably positive. From a change management perspective, despite some initial resistance from the health work force across North Macedonia, effective system solutions helped to encourage uptake and effective compliance.
Ensuring end users have a well-formed search module to facilitate coding simplicity and ensuring user interface design integrates coding into the workflow as seamlessly as possible is paramount. A major advantage in successful implementation of coding is to start with simple and powerful standards. By enforcing standardisation across health systems, interoperability will be improved, and health systems nationally will be able to work together to create a better health experience for all.
New Zealand’s path towards rationalising our coding standards from READ and ICD codes used differently across providers and health systems today, to standardising on SNOMED is going to have a dramatic impact on our ability to build a more interoperable New Zealand health system with better insights and analytics and reduce the burden on clinical workflows that can be more logically staged.