Covid-19 Impact on Standard Coding Systems Update

Elena Cardillo

, Maria Teresa Chiaravalloti

and Erika Pasceri

Institute of Informatics and Telematics, National Research Council, Rende, Italy

Department of Culture, Education and Society, University of Calabria, Rende, Italy

Keywords: Coding Systems, Covid-19, Standards Updates, Knowledge Organization Systems, Healthcare

Interoperability.

Abstract: The outbreak of Covid-19 pandemic has sped up many healthcare processes and practices. Both stakeholders

and standard organizations and authorities had to quickly implement new guidelines and codes to uniquely

identify the disease and all the related healthcare data. The object of this work is to study the impact of the

Covid-19 pandemic on clinical coding systems, in terms of updates and introduction of new specific codes

for the identification of the SARS-CoV-2 virus, with the aim of allowing a better description of the disease

and interoperability of the clinical data. The analysis is focused on ICD, SNOMED CT, LOINC, ATC as

coding systems either included into the Italian EHR regulation or widely used internationally. Results show

that coding systems that created a plenty of new codes for Covid-19 have: i) a flexible structure; ii) a speed

process for updates; iii) a large user community for inputs. Others instead demonstrated in this circumstance

that they are limited by hierarchical structures or excessively cumbersome updating processes, which conflict

with the flexibility required to standards to represent the evolution of clinical knowledge. This is especially

true in exceptional situation like the pandemic one.

1 INTRODUCTION

Systems for organising information and knowledge

are essential to reduce possible semantic conflicts

(ambiguities) and issues related to the specialization

of a domain terminology. They are commonly known

as Knowledge Organization Systems (KOSs).

“Knowledge is the lifeblood of modern society, but

without organization knowledge is dead. We could

even say that without organization knowledge is not

knowledge at all. Knowledge must be organized in

order to be used, be it by people or by machines”

(Dagobert, 2009). To this end, KOS in clinical

domain are used to classify, represent and encode

diseases and other types of data (symptoms, medical

procedures, drugs, etc.) in a unique way, with the

purposes of indexing and retrieving information of

interest, supporting epidemiological studies and

decision-making, and ensuring semantic

interoperability during data exchange between

different information systems. In addition, they

https://orcid.org/0000-0001-5003-205X

https://orcid.org/0000-0003-4695-2026

https://orcid.org/0000-0001-9917-2184

support the digitization process of clinical settings

and enable physicians to access and process relevant

data to support diagnosis, define risk profiles, and

facilitate statistical and epidemiological studies.

Some clinical KOSs are reference standards for the

semantic area they represent. They differ mainly in

structure, relationships with other standards, and

updating process.

The pandemic was an exceptional testing ground

in many respects, as it gave the opportunity to test the

functioning of systems under conditions of stress and

urgency. Globally, the Coronavirus Disease 2019

(Covid-19) had a significant impact not only on

public health, but also on the economy and society,

involving millions of people and causing serious

health crises, forcing many countries to implement

restrictive measures to reduce contagion, such as

lockdown, social distancing and the use of masks.

The Covid-19 confirmed cases growth put pressure

on healthcare systems around the world, with

hospitals and intensive care units often overcrowded.

490

Cardillo, E., Chiaravalloti, M. and Pasceri, E.

Covid-19 Impact on Standard Coding Systems Update.

DOI: 10.5220/0012387600003657

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 17th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2024) - Volume 2, pages 490-497

ISBN: 978-989-758-688-0; ISSN: 2184-4305

Additionally, Covid-19 left many patients with

serious long-term symptoms, known as “long-

haulers”. From the point of view of clinical coding

systems, this meant keeping up with the expressive

needs of the Covid-19 and everything related to it

(e.g. biological studies) to correctly record the

pandemic data, share them without ambiguity, be able

to aggregate them for surveillance and forecasts.

Therefore, this paper aims to carry out an analysis of

how much and how some clinical KOSs have updated

their contents in relation to the Covid-19 disease to

verify whether and how timely they have been and if

they have covered the semantic areas of the new

concepts, determined by the contingent situation. The

analysis was carried out on KOSs which are either

included in the Italian regulations to manage EHR

data or widely used internationally and with a well-

known effort made to adapt to the pandemic

emergency.

The paper is structured as follows: Section 2 gives

an overview of the Background related to the

pandemic and its impact on healthcare services;

Section 3 focuses on the analyzed coding systems and

their updates related to Covid-19; Section 4 is

dedicated to Discussion and Conclusions,

highlighting future perspectives of the study.

2 BACKGROUND AND

SIGNIFICANCE

Covid-19 is an infectious disease caused by the

SARS-CoV-2 virus, belonging to the family of

coronavirus, firstly identified in Wuhan, China, in

December 2019 and quickly becoming a pandemic.

The outbreak of the Covid-19 pandemic was

impactful from many points of view as it introduced

a new way of thinking and acting for previously

unknown aspects and, at the same time, required a

review and readjustment of the known ones. The role

of telehealth, for example, was enhanced and became

crucial to handle ordinary medical activities, when

hospitals’ departments and outpatient care facilities

were “closed” because of the infections. The

provision of medical care remotely allowed – in many

cases - to manage the pandemic more safely and

efficiently, by reducing the risk of virus transmission

and ensuring safe access to care for patients who

would otherwise have struggled during lockdown or

isolation. Furthermore, telemedicine allowed remote

monitoring of patients with Covid-19 and other

chronic conditions, facilitating collection of data and

treatment management.

Although Covid-19 had an unprecedented impact on

the world with disastrous health and economic

consequences and highlighting significant gaps in

healthcare systems all over the world, it led to a series

of advances in terms of scientific and technological

improvements. This posed new challenges and

opportunities for innovation in healthcare, like the

new vaccine development approach (it has been done

at unprecedented speed, with the usual sequential

steps done in parallel), recognized now as being

successful (Buchy et al., 2021).

Throughout the pandemic, many clinical

terminologies / classification and coding systems

were quickly updated to introduce new specific codes

for the SARS-CoV-2 virus, with the aim of allowing

a better understanding and description of the disease.

The intent was also to uniquely identify it ab origine

to distinguish it from existing diseases of the

respiratory system and correctly classify it. It allowed

a more detailed registration and monitoring of

epidemiological data, guaranteeing improved

analyses and researches related to the Covid-19

pandemic. As an example, the code U07.1 (Covid-19,

Virus Identified) was added among the ICD-10 codes

shortly after the pandemic began. During the

pandemic it was particularly difficult for health

systems to keep up with evolving diagnostic and

procedural coding recommendations, but the new

codes had been extremely decisive in strengthening

observational studies, characterising the disease

phenotype, and for responding to other important

epidemiological questions (such as disease

prevalence) (Marwaha et al., 2021).

Among the

various types of clinical data, diagnostic tests for

Covid-19 played a crucial role as the main tool for

identifying confirmed cases of Covid-19. To quickly

respond to the pandemic urgency, the LOINC

Committee released a set of nearly 1,300 new codes

to identify new clinical and laboratory observations

related to the Covid-19 pandemic to guide the

uniform coding of these concepts (Dong et al., 2020).

A similar study was previously conducted by

(Zeng et al., 2020), but while recognizing the

fundamental role played by KOSs in critical moments

when the correct identification of information

becomes crucial, it has the limitation of having been

published in May 2020. Therefore, it begins to trace

the changes to coding systems in relation to the

pandemic but cannot provide full recognition of them

for obvious temporal reasons.

Covid-19 Impact on Standard Coding Systems Update

491

3 CLINICAL STANDARDS’

UPDATES RELATED TO

COVID-19

During the Covid-19 pandemic, medical coding

systems played a crucial role in the standardization

process, organization of clinical information and for

the definition of pandemic epidemiological aspects,

being able to uniquely identify the type of pathogen

involved and other relevant connected data (e.g., its

transmissibility, incubation times and duration of the

disease). They had also been used for several

purposes, including identifying at-risk groups,

developing care plans, supporting the institutional

authorities' planning of both primary and secondary

prevention measures, such as containment strategies

and the need for the supply of protective equipment.

The rapid and unexpected nature of the pandemic

required careful monitoring of the progress of

infections and enshrined the importance of the Health

Surveillance as «the continuous, systematic

collection, analysis and interpretation of health-

related data», as defined by the WHO. According to

a study published in Journal of Hospital Infection

(Lin et al., 2020), thanks to a health surveillance

algorithm, based on the data codified and collected in

electronic medical records, it was possible to identify

patients admitted to hospital whose “pneumonia” did

not show a clear improvement with antibiotic

treatment. This had produced daily alerts addressed to

general practitioners. The surveillance algorithm,

thanks to the data registered within the EHR,

demonstrated the utility of the information

technology to facilitate infection control.

To achieve the objectives of this study, a double

survey was carried out regarding the contents related

to Covid-19 in the coding systems considered. For

each of them, the official web page, guidelines and

other resources provided by the respective Standards

Development Organizations (SDOs) were

investigated, and an additional check through the

UMLS Metathesaurus Terminology Services (UTS)

allowed to evaluate a possible expansion of the

coding systems / value sets to be considered. The

specific keywords used, in this case, were: “Covid-

19”, “SARS-CoV-2”, “Covid-19 Vaccine”, “SARS-

CoV-2 vaccine”, “Covid19 (disease)”, “Suspected

Covid-19”, while the semantic groups applied were:

“Disorders”, “Chemical & Drugs, “Living beings”;

the semantic types applied were: Disease or

Syndrome, Pharmacologic Substance Immunologic

Factor, Laboratory Procedure.

3.1 Covid-19 in ICD

Because of the advent of the new pathogen SARS-

COV-2, WHO introduced in February 2020 two

emergency codes in the ICD-10 classification. These

codes are: i) U07.1 – Covid-19, virus identified (for

confirmed cases), ii) U07.2 – Covid-19, virus not

identified (for suspected cases). Initially, U07.1 was

referred to as “acute respiratory disease 2019-nCoV”,

and this code was not made available for use until

April 2020. Subsequently, it was labelled as “Covid-

19, virus identified” (De Lissovoy, 2020). The codes

for Covid-19 disease, defined between 2020 and

2021, have been integrated into “Chapter XXII -

Special purpose codes”, block “U00-U49 Assignment

provisional of new diseases of uncertain aetiology or

for emergency use” of the 2019 edition of ICD-10.

Although it was generally suggested to assign U07.1

to cases with a lab-confirmed diagnosis and U07.2 to

cases with a clinical diagnosis, rules for recognizing

Covid-19 diagnoses and thus for using these codes

were not that easy. In fact, in US some flowcharts

were provided and used to guide the proper

assignment of diagnostic codes related to Covid-19

encounters (Varghees, 2020). The introduction of

new codes was based on requests of Member States

to ensure accurate reporting of Covid-19 related

conditions. The need to distinguish between acute

disease, long-term effects, and complications led to

the adoption of the neutral term “post-Covid”. This

term implies no specific etiologic relationship,

allowing any condition to be linked to a previous

acute Covid-19 infection. The term “post-Covid”

refers to persistent, recurrent or new symptoms (e.g.,

fatigue, short of breath, insomnia, etc.) and other

health effects that occur after the acute phase of

Covid-19 infection. Despite the early detection of this

condition, a specific code was not made available for

clinical use until 2021 in ICD-10 (the suggestion was

to use U09.9, “Unspecified post-Covid-19

condition”). From January 2021, new codes had been

introduced on immunization to prevent Covid-19 and

adverse reactions to vaccines against Covid-19.

These codes are: U11.9 “Need for vaccination against

Covid-19, not specified” and U12.9 “Vaccines

against Covid-19 that cause adverse effects in

therapeutic use, unspecified”. Figure 1. summarises

the Covid-19 coding flowchart using ICD,

considering both ICD-10 and ICD-11, the last

revision of the classification, and distributing the new

codes according to the purpose (e.g., for diagnoses,

prevention).

HEALTHINF 2024 - 17th International Conference on Health Informatics

492

Figure 1: Use of ICD codes for Covid-19 during the disease

outbreak.

The promptness in the introduction of the new

ICD codes has been extremely helpful in accurately

recording and documenting the diagnoses and

conditions of patients with Covid-19, for monitoring

and surveillance on the diffusion of the pandemic and

for collecting epidemiological data that help in

identifying clusters of cases, monitoring the spread of

the virus and assess the impact of control measures.

3.2 Covid-19 in LOINC

LOINC is the most widely used coding system for

laboratory and clinical observations (McDonald et al.,

2003). After the pandemic outbreak, a specific

LOINC subset was created for facing the emergency,

during that time, new term requests related to Covid-

19 from worldwide users had a preferential route in

the submission process. Furthermore, it was created a

specific webpage on the standard website to give a

quick and direct access to LOINC content related to

the disease. The mentioned subset contains 562 terms.

Some of them belong to pre-pandemic LOINC

versions (existing codes), therefore they were not

created for the specific purpose, but their semantics

adapt to it, while others are codes created starting

from LOINC version 2.68 (June 2020), so specifically

required to respond to the needs posed by Covid-19

outbreak. They are further divided according to their

purpose of use in the following subcategories:

▪ SARS-CoV-2 lab tests: all the 160 codes of this

category are newly created after pandemic

outbreak;

▪ SARS-CoV-2 AOE questions: 11 codes,

including 4 existing codes;

▪ Convalescent plasma: 2 newly added codes;

▪ LOINC terms related to public health case

reporting: 110 codes, and 44 of them released

before June 2020;

▪ Covid-19 and Telehealth documents: 120 codes

(48 belonging to previous LOINC versions, and

72 created after the Covid-19 outbreak);

▪ Covid-19 Survey terms: 159 codes, including 1

already existing LOINC code.

The Covid-19 pandemic produced effects also on

the types of clinical documents produced by the

healthcare stakeholders, because new ones had been

introduced and the existing ones needed to be better

organized. Clearly identifying clinical document

types is a non-trivial task because often behind

identical names there is not the same content and,

viceversa, there are plenty of different names for

documents related to the same semantic area. With

the aim of standardizing this field, in the first 2000s

the joint effort of the LOINC committee and HL7

created the LOINC Document Ontology (Frazier et

al., 2001). This section of the standard models the six

main LOINC axes to uniquely identify clinical

document types based on different metadata (e.g.,

role, subject, type of service, etc.).

LOINC document type (DT) codes are one of the

metadata used in the registries of the Italian EHR

infrastructure to index clinical documents. Admitted

values are specified into the Affinity Domain (AD)

document, which regulates interoperability services

among regional EHR systems. Usually AD specifies

LOINC DT which have a corresponding CDA2 HL7

Implementation Guide (IG) already defined or in case

of special needs, as those posed by the pandemic. This

latter was the case of the two new LOINC codes

proposed by LOINC Italy to address the urgency to

have these kinds of documents indexed into the

patients’ EHR:

▪ LOINC 97500-3 Proof of Covid-19

immunization or negative status certificate,

which is the so-called Green Pass, a document

in digital or paper format that certifies

vaccination against the Covid-19 disease or,

alternatively, the recent negative test result or

recent recovery from the infection;

▪ LOINC 97499-8 Proof of Covid-19 recovery

certificate, which is a document certifying the

recovery from the disease and can be around

others. It is issued by the GP.

LOINC DT codes related to Covid-19 are 120: 48

belonging to previous LOINC versions, and 72 new

codes created ad hoc for the scope. Many of the new

codes have the value Telehealth in the “System” axis,

reflecting the fact that the pandemic was really a

boost for remote healthcare services. There is no

prevalence of new codes assigned to a specific

medical specialty, demonstrating how the impact has

been truly global and has had, despite the very high

costs in terms of human lives, also important positive

implications for the progress of the healthcare sector

Covid-19 Impact on Standard Coding Systems Update

493

towards full digitalization. Figure 2 shows the trend

of LOINC Covid-19 related codes introduced over the

last three years.

Another aspect to consider is the need highlighted

by the pandemic to have aggregable and immediately

available comparable data, both for reporting and for

clinical / scientific analyses and studies, which in

those months underwent an extraordinary

acceleration to contain the virus and propose

solutions (e.g., vaccines).

Figure 2: Covid-19 related LOINC terms over the years.

To this end, the LOINC codes of the types of

clinical documents served to trace the flow of

documents in the federated EHRs, as in the case of

the Italian one (namely FSE), especially about

laboratory reports and vaccination cards. We have

above mentioned the two new types of documents

introduced during the pandemic, but it was also

required a further specification of the existing ones so

that their tracking in the FSE can immediately lead

back to Covid-19. To give an example, to track the

Covid swabs carried out daily in Italy, values for the

EventCodeList metadata have been inserted in AD to

specify that a LOINC DT Laboratory report contains

the result of the execution of an antigen or molecular

Covid-19 swab test or the outcome of a serological

Covid-19 test. They are LP418019-8 Covid-19

antigen swab test; LP417541-2 Covid-19 molecular

swab test; 96118-5 Qualitative Serological Test and

94503-0 Quantitative Serological Test. Similarly, the

DT identifying the Single vaccination card could be

combined with the values of the metadata

EventCodeList to specify the ATC code of the

administered vaccine type (J07BN Covid-19 vaccines

for the specific pandemic purpose, but admissible

values for this metadata include those belonging to

the WHO Anatomical Therapeutic Chemical

classification (ATC) coding system). This combined

use of the metadata of the AD made it possible to have

daily reports on the progress of the Country's

immunization process.

3.3 SNOMED CT Updates

As the global terminology for health, SNOMED CT

can serve as a common language for recording,

sharing, integrating and analyzing Covid-19 related

data elements, such as symptoms, risk factors and test

results. Following the advent of the pandemic,

SNOMED International quickly published a

comprehensive version of the Covid-19 concepts

including updated descriptions and mappings from

SNOMED CT to ICD-10. This version provided

physicians, researchers and administrators the most

up-to-date terminology needed to properly code,

analyse and address Covid-19 (Sutton, 2020). Some

of these concepts are illustrated in Table 1.

Table 1: Examples of SNOMED-CT Covid-19 related

codes.

A guide presenting concrete examples of subsets

of SNOMED CT codes that can be used to code

different types of Covid-19 data, including

symptoms, risk factors and test results, has been

published on the official SNOMED International

website. These subsets can be adapted according to

the specific needs of different health communities to

ensure health service delivery, pandemic monitoring,

international cooperation and retrospective data

analysis, as seen for the other coding systems.

SNOMED CT subsets have been organized into a

number of categories, based on groupings of data that

can be recorded together, for example the “Provider

and Facility Details” (i.e., Health profession, place of

care, personal protective equipment), “Clinical

Assessments”, “Test and Investigations”, etc.

HEALTHINF 2024 - 17th International Conference on Health Informatics

494

During the last three years also specific SNOMED

CT value sets related to Covid-19 have been created

by different organizations to be used in clinical

documents and for interoperability purposes (i.e., use

in EHRs, HL7 CDA documents). These includes the

“LIVD SARS-CoV2 Test Result Codes value set” or

the “COVID_19 (Antibody Substance in Lab

Results)” created respectively in May 2020 and in

May 2021 as Extensions. From a terminological point

of view, it is important to make clear the definition

and thus the purpose of use of these subsets /value

sets. As stated in (Rossander et. al., 2021) a subset is

“a collection of components from a terminology.

SNOMED CT subsets presented in RF2-format are

simple reference sets”. These subsets can be called

value sets in some use cases. A SNOMED CT subset

can include either SNOMED CT concepts, which can

be represented by any of the descriptions linked to

them, or specified descriptions. The use of SNOMED

CT Covid-19 related subsets and value sets was

important in the context of data exchange so in HL7

messages, where we can see the complementary use

with LOINC codes, LOINC used for coding the

testing method, and SNOMED CT used for coding

non-numeric answers. For example, in the case of the

LOINC code 94500-6 “SARS-CoV-2 (COVID-19)

RNA [Presence] in Respiratory system specimen by

NAA with probe detection”, SNOMED CT answer

codes indicated are 10828004 “positive”, 260385009

“negative”, 455371000124106 “invalid result”). This

use is motivated by the fact that both standards are

already commonly in place within laboratory

information systems.

3.4 ATC Updates

The Anatomical Therapeutic Chemical (ATC)

Classification System, a classification system that

classifies the active ingredients of drugs according to

the organ or system on which they act and their

therapeutic, pharmacological and chemical

properties, has the purpose to help monitoring drug

use and improving quality medication use. The ATC

is updated twice a year. With the advent of the Covid-

19 pandemic, ATC updates were provided for new

drugs and therapeutic indications associated with the

treatment of the virus. For example, codes were added

for Covid-19 vaccines, as well as for antiviral drugs

and other treatments used in the management of the

disease (it is the case of the code J07BN02 “Covid-

19, viral vector, non-replicating”.

Since the purpose of ATC is to codify the name

of the drug molecule, its additions relating to Covid-

19 have consequently had repercussions on specific

coding systems for the identification of drug packages

placed on the market (such as, for example, the AIC -

Autorizzazione all’Immissione in Commercio -

coding system for drugs on the Italian market or

FDA's National Drug Code for the USA). The

antivirals used for Covid-19 in ATC are: J05AE30

“nirmatrelvir and ritonavir” and J05AB16

“remdesivir”. The prescription and the use of

antivirals for the treatment of COVID-19 is subject to

close monitoring, thus enabling a rapid identification

of new possible safety information. Healthcare

professionals are required to report any adverse

reaction using the National Pharmacovigilance

Network.

4 DISCUSSION AND

CONCLUSIONS

Data coding plays a crucial role in managing and

analyzing complex information, especially in health

emergency situations such as those posed by the

Covid-19 pandemic outbreak. Data collection has

been essential to monitor the spread of the virus, track

cases, identify affected areas and evaluate the

effectiveness of control measures. Sometimes,

however, the effectiveness of the data depends on

multiple elements, including the completeness of the

information recorded and the accuracy with which it

was acquired. The analyzed material and the results

of the study, reported in Section 3, on one hand, made

immediately evident the considerable work done and

effort spent by the different working groups of each

SDO in the update process, above all in the case of

LOINC and SNOMED CT, and the contribution

given with the introduction of new codes to ensure an

accurate registration of the health conditions and a

continuity in tracking and monitoring relevant

information useful during the pandemic as well as for

their regular use in clinical documents.

On the other hand, results showed a very

heterogeneous trend of updates applied over the last

three years among the considered systems. It can be

seen, in fact, that classification systems with a

hierarchical structure (ICD ones and ATC) had a

slower and more complex updating process (implying

the introduction of new classes/codes) mainly for two

reasons: i) the branches of the classification have a

limited extensive possibility, so it is often difficult to

find further space in structured notational systems

(for example those who use numeric digits up to 9);

ii) in the event that notations are available, the

location of insertion of a new concept must be

Covid-19 Impact on Standard Coding Systems Update

495

carefully evaluated by the appointed commissions, in

order to respect the hierarchical ordering of the

concepts and preserve the association of the new

concept/class to the semantic category covered by

chapters/ranges; iii) it is important to avoid semantic

inconsistencies considering also the inclusions and

exclusions criteria. On the contrary, in the case of

coding systems which present a linear structure and

are not based on decimal system for code notation,

i.e., SNOMED CT and LOINC it was easier to

introduce new codes. To confirm these observations,

Cimino, listing the desiderata for controlled medical

vocabularies (Cimino, 1998), and (Harrison, 2021

and Awaysheh, 2018) highlighted how flexibility is

the requirement that most supports the diffusion and

use of a coding system.

As shown in Figure 3, it was however common

for some of the systems to update existing codes by

proposing an adaptation of them in terms of

integrating so-called 'unspecified' classes and

modifying inclusion criteria. The analysis confirmed

that to make data truly useful, it is essential that they

are managed according to shared norms and

standards. Among KOSs, clinical coding systems are

used to structure data and uniquely identify pieces of

information. Nonetheless, the analysis of coded data

may encounter obstacles and potential alterations due

to the improper application of these systems.

Regarding this issue, the study confirmed that the

choice of coding systems must be well thought out

according to the context and purposes for which they

have to be used.

Figure 3: Covid-19 related updates trend.

For example, for hospital discharge letters, Italy

is still adopting, by regulations, ICD in its ninth

revision. Giving this, the national guidelines for

coding Covid-19 related discharge letters (Italian

Ministry of Health, 2020) indicate to use this version

to identify pathological conditions associated with the

virus and possible complications related to it, while

the use of ICD-10 is preferred for expressing the

cause of the death in death certificates. The Italian

Institute of Statistics in the COVID-19: interim report

on definition, certification and classification of

causes of death. (ISS et al., 2021), highlighted some

errors in the classification of the causes of death: i)

Lack of specificity: the initial cause of death should

be specific, and for example, a viral infection may be

an initial cause, but specifying the infectious agent

and type subsequent disease (e.g., "Covid-19

pneumonia"); ii) Intermediate causes: for example,

“pneumonia” could be an intermediate cause of death

because it can be caused by different infectious agents

or by inhaling a liquid or chemical substance; iii)

Illogical sequences: e.g., “chronic obstructive

pulmonary disease” cannot cause an infection,

although it may increase its criticality, so Covid-19

should be reported as the initial cause and the disease

as a consequence (ISS, 2021).

The study allowed also to check some preliminary

alignment among the considered coding systems with

respect to some generic semantic categories. In

particular, searching in UMLS among all the Covid-

19 related subsets and codes, we found exact

mappings for just 4 concepts: 1) Covid-19 (in

SNOMED CT “840539006” and ICD-10-CM

“U07.1”); 2) Post-acute Covid-19 condition (in

SNOMED CT “1119303003” and ICD-10-CM

“U09.9”), 3) Covid-19 vaccine (in SNOMED CT

“28531000087107” and ATC “J07BN”); and finally

Unvaccinated for Covid-19 / Vaccination not done (in

SNOMED CT “591191000124106” and ICD-10-CM

“Z28.310”. Although more mappings were expected

between ICD-10 and SNOMED CT, for example, it

was confirmed that the joint initiatives taken by the

SDOs to face with the Covid-19 coding and

identification emergency promoted the

complementary use of the coding systems and thus of

most of the new codes. Despite these observations,

the approach has shown some limitations in

comparing Covid-19 related subsets and value sets,

requiring a more accurate analysis and the use of

additional tools (e.g., the VSAC comparison tool of

the NLM) to improve results, including the flow.

In the healthcare domain, the use of coding and

classification systems sometimes has limitations due

to several factors, often related to a lack of

information about the systems themselves.

Healthcare professionals responsible for filling out

medical records are not able to take full advantage of

these systems lacking the proper training. Sometimes

physicians fill out clinical documents having no

access to complete information about patients’

medical history, including the incidence of Covid-19

on their clinical course. Accuracy in differencing

HEALTHINF 2024 - 17th International Conference on Health Informatics

496

deaths caused by Covid-19 from those caused by

other diseases with the ongoing Covid-19 infection

can be complex and subject to many variables. This

is especially true in a period with high prevalence of

death causing infections.

In conclusion, the main results of this study

highlighted the need for proper information collection

and management to have data structured according to

shared rules to be quickly accessible to the

appropriate government agencies. Despite advances

in the digital health domain there are still gaps in

semantic interoperability of these tools. One of the

lessons learnt from the study is that no matter how

good and widely used were coding systems in the

past, if there isn’t the required flexibility, any new

outbreak will, inevitably, bring different challenges

for their update, so the systems will never be perfect.

An improvement of this study will be the

improvement of the comparative analysis of the

versioning of the considered systems and the

adaptation of mappings between standards.

ACKNOWLEDGEMENTS

This study moved from the thesis of Dr. Maria

Francesca Littera (University of Calabria, Master’s

Degree in Management and Preservation of Digital

Documents), tutored by the authors. A special thank

is due for providing some materials and analyses.

REFERENCES

Awaysheh A, Wilcke J, Elvinger F, Rees L, Fan W,

Zimmerman K. (2018) A review of medical

terminology standards and structured reporting. J Vet

Diagn Invest. Jan;30(1):17-25. doi:

10.1177/1040638717738276.

Buchy, P., Buisson, Y., Cintra, O., Dwyer, DE., Nissen, M.,

Ortiz de Lejarazu, R., Petersen, E. (2021) Covid-19

pandemic: lessons learned from more than a century of

pandemics and current vaccine development for

pandemic control. Int J Infect Dis. Nov (112):300-317.

doi: 10.1016/j.ijid.2021.09.045.

Cimino JJ. (1998) Desiderata for controlled medical

vocabularies in the twenty-first century. Methods Inf

Med. Nov;37(4-5):394-403.

Dagobert, S. (2009). “Knowledge Organization Systems.

Overview”. P.2.

De Lissovoy G. (2020). Codes, Coding, and Covid-19.

Medical care, Volume 58, Issue 12, pp.1035–1036.

Dong X., Li J., Soysal E., Bian J., DuVall SL., Hanchrow

E., Liu H, Lynch KE., Matheny M., Natarajan K., Ohno

Ohno-Machado L., Pakhomov S., Reeves RM., Sitapati

AM., Abhyankar S., Cullen T., Deckard J, Jiang X,

Murphy R, Xu H. (2020) “COVID Covid-19 TestNorm:

A tool to normalize COVID Covid-19 testing names to

LOINC codes”. J Am Med Inform Assoc., 27(9):1437-

1442. doi:10.1093/jamia/ocaa145.

Frazier, P., Rossi-Mori, A., Dolin, R.H., Alschuler, L.,

Huff, S.M. (2001). The creation of an ontology of

clinical document names. In Stud Health Technol

Inform. 84(Pt 1):94-8.

Harrison JE, Weber S, Jakob R, Chute CG. (2021) ICD-11:

an international classification of diseases for the

twenty-first century. BMC Med Inform Decis Mak.

Nov 9;21(Suppl 6):206. doi: 10.1186/s12911-021-

01534-6.

Istituto Superiore di Sanità - ISS (2021). “Covid-19: interim

report on definition, certification and classification of

causes of death”. Updating Rapporto ISS Covid-19 n.

49/2020. Version of April 26, 2021. Working groups

ISS Cause di morte Covid-19 and Sovrintendenza

sanitaria centrale – INAIL, ISTAT 2021, ii, 16 p.

Rapporto ISS Covid-19 n. 10/2021 (in Italian).

Italian Ministry of Health (2020), General Directorate for

Health Planning, Linee Guida per la codifica della SDO

per casi affetti da malattia da SARS–COV–2 (COVID–

19), published on March 20, 2020.

Lin C-Y., Cheng C C-H., Lu P P-L., Shih D D-C., Hung C

C-T., Lo H H-H., Tsai M M-J., Hung J J-Y.. (2020).

Active surveillance for suspected Covid-19 cases in

inpatients with information technology. In Journal of

Hospital Infection, Volume 105, Issue 2, pp.197-199.

doi: 10.1016/j.jhin.2020.03.027.

Marwaha, J. S., Halamka, J.D., Brat, G.A, Gordon, W.J.

(2021). “Repurposing Billing and Administrative

Terminologies as Instruments of Public Health:

Lessons From The Covid-19 Pandemic”, in Health

Affairs Blog.

McDonald CJ, Huff SM, Suico JG, Hill G, Leavelle D,

Aller R, Forrey A, Mercer K, DeMoor G, Hook J,

Williams W, Case J, Maloney P. (2003) LOINC, a

universal standard for identifying laboratory

observations: a 5-year update. Clin Chem.

Apr;49(4):624-33. doi: 10.1373/49.4.624.

Rossander A, Lindsköld L, Ranerup A, Karlsson D. (2021).

A State-of-the Art Review of SNOMED CT

Terminology Binding and Recommendations for

Practice and Research. In Methods Inf Med. Dec;60(S

02):e76-e88. doi: 10.1055/s-0041-1735167.

Sutton K. (2020). Keeping up with Covid-19. In Wolkers

Kluwer, Health, March 25..

Varghees S. and Magoon V. (2020). Covid-19 diagnosis

coding explained in a flowchart, in FPM Journal,

Quick Tips, Jul 05.

Marcia Lei Zeng, Yi Hong, Julaine Clunis, Shaoyi He, L.P.

Coladangelo. (2020) Implications of Knowledge

Organization Systems for Health Information Exchange

and Communication during the COVID-19 Pandemic,

Data and Information Management, 4(3):148-170,

doi.org/10.2478/dim-2020-0009.

Covid-19 Impact on Standard Coding Systems Update

497