Preparation for the Implementation of the International Classification of Diseases 11th Revision: A Survey on Clinical Coders' Knowledge of the International Classification of Diseases 11th Revision, Educational Needs, and Readiness to Accept New Roles: A Cross‐Sectional Study

ABSTRACT

Background and Aims

Studies have shown that the knowledge and skills of clinical coders impact the quality of clinical coding, and professional training for clinical coders is essential to ensure coding quality. Like other countries, Iran should establish the required infrastructure to promptly implement ICD‐11. One of the essential infrastructures is clinical coders' familiarity with and skills in utilizing ICD‐11. The purpose of this study was to investigate clinical coders' knowledge of ICD‐11, their educational requirements, and their willingness to assume new responsibilities following the implementation of ICD‐11.

Methods

This was a survey conducted during 2021–2022. A researcher‐made questionnaire including five parts was applied as an instrument. Ten experts confirmed the content validity of the questionnaire, and CVR and CVI were applied to measure content validity. In addition, test retest was used to assess the reliability. A Likert scoring scale was used to measure the responses. The research population consisted of all clinical coders working in Iranian health facilities. Data were analyzed using descriptive statistics, Pearson's χ ², and Independent Samples t‐test.

Results

Totally, 453 people responded to the questionnaire. Twenty‐nine percent of the participants were familiar with the ICD‐11. The relationship between the level of familiarity with ICD‐11 and age, gender, ICD‐10 coding skills, and work experience was not statistically significant. Moreover, the lack of training courses was the main reason for not being familiar with ICD‐11. In addition, the participants believed that skill in using ICD‐11 tools had more priority than other educational axes. According to the majority of clinical coders, they may not be ready to accept new roles if ICD‐11 is implemented.

Conclusion

This study revealed that familiarity with ICD‐11 was low. Therefore, a training program is necessary to meet the needs of the target groups for nationwide ICD‐11 education.

Abbreviations

CVI

content validity index

CVR

content validity ratio

ICD

international classification of diseases

ICD‐O

international classification of diseases for oncology

ICF

international classification of functioning, disability, and health

ICHI

international classification of health interventions

ICPC

international classification of primary care

1. Introduction

Clinical coding serves multiple purposes, including data standardization, facilitating the collection and summarization of diagnostic and procedural statistics, assessing the adequacy and quality of medical services, comparing disease and mortality statistics at both national and international levels, quality management, case management, resource planning, and supporting medical research [1, 2, 3, 4, 5]. The International Statistical Classification of Diseases and Related Health Problems (ICD) is the common classification system used globally to code diagnoses and causes of death [6]. The World Health Organization began developing the 11th revision of the ICD to address the shortcomings of the previous edition, ICD‐10, released in 1992.

ICD‐10 is scientifically and technologically outdated; it lacks content for various de facto applications of ICD, such as primary care or clinical decisions [7]. Finally, the draft version of ICD‐11 was released in 2018 [8, 9]. As the WHO plan seeks global implementation of ICD‐11, many countries have initiated programs to adopt it [10, 11]. For example, in Iran, an evaluation of the pilot implementation of ICD‐11 began in 2019 [12, 13, 14].

ICD‐11 reflects critical advances in science and medicine, aligning classification with the latest knowledge of disease treatment and prevention [7]. The ICD‐11 has undergone numerous changes in the content and structure compared to the ICD‐10 [15]. For example, postcoordination and cluster coding approach was used to add more details to the chosen stem code. Moreover, changes to some chapters include restructuring of the hierarchy, the inclusion of more current terminology, and specific groupings for some diseases and conditions. There is also more meaningful clinical content than ICD‐10 [9, 15, 16, 17]. Additionally, several new chapters have been included in the ICD‐11. The new core chapters include “Diseases of the immune system,” “Sleep–wake disorders,” “Conditions related to sexual health,” “Supplementary Chapter Traditional Medicine Conditions,” and “Supplementary Section for Functioning Assessment” [7]. Furthermore, accurately coding some new features in ICD‐11, such as quality and patient safety, necessitates a thorough understanding and the ability to interpret causal relationships [18]. A key feature of the quality and patient safety code set in ICD‐11 is that a cluster of codes is required to represent each case [8].

These differences make the transition from ICD‐10 to ICD‐11 challenging [11, 16, 19]. Some studies have shown that these differences increase the risk of clinical coding errors [19]. In addition, studies have shown that clinical coders' knowledge and skills affect the clinical coding quality [20, 21, 22]. Besides, clinical coding is a complicated and error‐prone process, and the complexity of modern classification systems, such as ICD‐11, poses challenges to implementation [23]. Therefore, the professional training of clinical coders is essential to ensure coding quality [20, 24, 25]. Furthermore, due to the lack of official ICD‐11 training, one of the most critical challenges of implementing ICD‐11 will be the level of familiarity and skill of clinical coders with it [12]. Improving the knowledge of clinical coders is essential, based on their needs and the outcomes of evaluating their familiarity with ICD‐11. Additionally, it is necessary to provide a training program tailored to these results to ensure the successful implementation of ICD‐11 [10, 11]. Not considering the familiarity levels of coders and their specific needs during the implementation of ICD‐11 can heighten the risk of resistance to its adoption [26].

Another key aspect to consider is the role of coders following the implementation of ICD‐11. The ICD‐11 is designed for use by a diverse array of stakeholders within the healthcare system. Its expanded scope of application may affect the traditional role of clinical coders [7]. In many developing countries, such as Iran, the process of locating codes in the ICD‐10 is often done manually using printed ICD‐10 books [27]. The WHO provides an electronic browser and coding tool for ICD‐11. This tool allows users to retrieve concepts by searching for specific terms, anatomy, or any other elements included in the ICD‐11 content. ICD‐11 is designed for digital health applications and is compatible with various IT environments, featuring a new API [8]. The electronic structure of the ICD‐11 enhances coding speed compared to the ICD‐10 [13]. This issue can also impact the role of coders in healthcare facilities [15, 16]. On the other hand, the electronic structure of the ICD‐11 and its ability to integrate with various health information systems and medical terminologies facilitate its application in automated coding systems [28]. Today, particularly with the advancement of artificial intelligence (AI), the integration of AI and natural language processing (NLP) capabilities into computer‐assisted clinical coding system has gained significant attention [29]. In recent years, numerous studies have been conducted on the use of AI in computer‐assisted clinical coding systems. AI tools have the potential to facilitate the adoption of more complex and detailed classification systems, such as ICD‐11 [30, 31, 32, 33, 34]. Integrating the electronic structure of ICD‐11 with AI in coding assistance systems is expected to have a significant impact on the coding profession in the future.

Therefore, transitioning roles and empowering coders to embrace new responsibilities are also important considerations. Despite the significance of this issue, few studies have addressed the training of clinical coders and the specific content of that training. For example, in a study, Canadian researchers mentioned the educational content provided for clinical coders [11]. In addition, a pilot study conducted in Iran reported on the training provided to clinical coders [12]. The widespread implementation of ICD‐11 necessitates comprehensive training for clinical coders. Moreover, the familiarity of clinical coders with ICD‐11 is essential for the successful implementation of ICD‐11 in Iran. Therefore, this study aimed to examine the familiarity level of Iranian clinical coders with ICD‐11, identify their educational requirements, and assess their willingness to take on new responsibilities if ICD‐11 is implemented.

2. Methods

2.1. Study Design

This cross‐sectional descriptive study was conducted in Iran during 2021–2022.

2.2. Questionnaire and Data Collection

A researcher‐made questionnaire was used as a data collection tool. To identify the items to be included in the questionnaire, we considered available documents related to ICD‐11, including articles and guidelines pertinent to the transition from ICD‐10 to ICD‐11, the implementation of ICD‐11, and training for ICD‐11. Additionally, we reviewed WHO resources, such as the ICD‐11 Reference Guide.

The questionnaire included five parts as follows:

• ‐The first part was the characteristics of the clinical coders, including age, gender, education, last degree, province and city of work, work experience, etc.
• ‐The second part focused on the experiences of clinical coders, including their years of work experience, their involvement in clinical coding—such as coding causes of death—and their familiarity with the classification systems used in Iran.
• ‐The third part discussed familiarity with ICD‐11, including the ICD‐11 browser (blue one), the ICD‐11 browser (orange one), the ICD‐11 Reference Guide, the ICD‐11 coding tool, the cluster coding concept in ICD‐11, postcoordination in ICD‐11, etc. It should be noted that due to the lack of translation of ICD‐11 in Persian language, the English version of ICD‐11 was considered.
• ‐The fourth part was about the readiness to adopt new roles after implementing ICD‐11, including applying the knowledge, rules, and standards of ICD‐11 in the electronic environment for coding diseases or causes of death. It also addressed the skills required for utilizing ICD‐11 codes for reimbursement issues handled through digital systems.
• ‐The fifth part was about the appropriate method for training the clinical coders with ICD‐11, including face‐to‐face training (by mentorship) in hospitals, workshops, online courses, etc.

The content validity of the questionnaire was confirmed by ten experts, all of whom were faculty members in health information management departments and had teaching experience in ICD.

Finally, the content validity ratio (CVR) and the content validity index (CVI) were applied to measure content validity. In addition, test retest method was used to assess the reliability of the questionnaire. Sixteen clinical coders from various hospitals participated in responding to the electronic questionnaire. After 2 weeks passed, they answered the questionnaire again. Pearson's correlation coefficient was used to evaluate the correlation of answers. Based on feedback from the validity and reliability assessment, modifications to the questionnaire were made, resulting in the final version. It is necessary to mention that the clinical coders who were involved in piloting the survey did not complete the survey later.

The questionnaire was designed electronically on Porsline, an Iranian online survey platform. A Likert scoring scale from 0 to 4, from lowest to highest, was used to measure the responses of Sections 2–5, and in Section 5, some yes/no questions were applied.

2.3. Samples

Currently, graduates in medical records and health information technology are responsible for clinical coding in healthcare facilities in Iran. Table 1 shows the status of clinical coding and the use of the WHO Family of Classifications in Iran.

Table 1.

A view of clinical coding in Iran.

Type of healthcare facilities	Subject of clinical coding	Topics of clinical coding	WHO Family of Classifications (WHO‐FIC)
Hospital	Inpatient and outpatient medical records	Primary diagnosis, interim diagnosis, final diagnosis External causes of injury Surgical procedures Causes of deathb	ICD‐10 ICD‐9‐CM for procedures ICD‐O‐3.1a
Primary healthcare centers	Death certificate	Causes of death	ICD‐10
Pathology laboratories	Pathology reports	Pathology diagnosis	ICD‐O‐3.1
State Welfare Organization of Iran	Disability assessment forms	Body function, body structure, activity, and participation	ICF c

^a ICD‐O‐3.1 codes are used only in oncology hospitals or hospitals with an oncology department for the population‐based cancer registry system.

^b In some hospitals, the codes related to the direct and underlying causes of death are recorded by the health information management (medical records) department in the hospital information system.

^c ICF codes are used in the welfare organization to determine the severity of the disability.

The research population consisted of all clinical coders working in Iranian health facilities. Since there were no accurate statistics on the number of clinical coders working in Iran and there was no association or official organization for clinical coders, data collection from the population was done in two ways:

First, the relevant departments within the Ministry of Health and Universities of medical sciences (such as the Center for Health Network Management and the Statistics and Medical Records department in the Vice‐chancellor of Treatment) forwarded the link to the electronic questionnaire to the officials responsible for registering causes of death, as well as to the clinical coders working in the hospitals.

Second, the link to the electronic questionnaire was shared in social networking groups or channels (WhatsApp and Telegram), which typically included clinical coders, health statistics experts, health information technology professionals, health information management specialists, and medical informatics experts.

2.4. Data Analysis

Data were analyzed using descriptive statistics (frequency distribution tables, mean, and standard deviation). We scored our data as follows: Perfectly knowledgeable (4), Very good knowledge (3), somewhat knowledgeable (2), slightly knowledgeable (1), and Unfamiliar (0). Then we calculated the mean score for familiarity with classification systems. We compared the mean score of familiarity with classification systems, age, and working history among participants with or without knowledge about ICD‐11 using Independent Samples t‐test or Mann–Whitney (in the case of non‐normal distribution of data). In addition, we applied Pearson's χ ² to compare familiarity with ICD‐11 among participants' gender, place of work, and education. All analysis was conducted in SPSS Version 24. A significance level of 0.05 was applied and all tests were two‐sided.

2.5. Ethics Approval and Informed Consent Declaration

The Ethics Committee of Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran approved the study with the ethical approval code: IR.AJUMS.REC.1401.102. All participants provided informed consent before completing the questionnaire.

3. Results

A total of 1020 people visited the questionnaire, of which 453 people (44.4%) responded to the questionnaire. Forty‐five respondents had no clinical coding experience, so they were excluded from the study. Four questionnaires were also removed due to incompleteness. Therefore, finally, 404 coders were included in the study.

Table 2 shows the demographic characteristics of the participants. Among them, 352 participants were female (87.1%). The mean (± SD) age of the participants was 37.89 ± 7.02 years, with an age range of 23–60 years. Additionally, the mean work experience among the participants was 12.46 ± 6.93 years, spanning from 1 to 35 years.

Table 2.

Demographic characteristics of participants in the study (n = 404).

Demographic characteristics		Frequency	Percentage
Sex	Male	52	12.9
	Female	352	87.1
Age (year)	20–29	45	11.1
	30–39	185	45.8
	40–49	148	36.6
	50–60	26	6.4
Educational level	Associate degree	39	9.7
	Bachelor	283	70.1
	Masters	77	19.1
	PhD.	5	1.2
Field of study	Medical records	257	63.6
	Health information technology	93	23.0
	Health information management	3	0.7
	Medical informatics	2	0.5
	Other a	49	12.1
Work experiences (year)	< 5	47	11.6
	5–9	69	17.1
	10–14	99	24.5
	15–19	85	21.0
	20–24	55	13.6
	> 24	49	12.1
Workplace	University hospitals (academic/teaching)	168	41.6
	University hospitals (nonteaching)	77	19.1
	Private hospitals	21	5.2
	Government hospitals	41	10.2
	Charity hospitals	6	1.5
	Primary health care (PHC)	73	18.1
	Others	18	4.5
Total		404

^a Such as health services management, public health, midwifery, biostatistics, etc.

Most of the participants (63.6%) were graduates with degrees in medical records. Out of 32 provinces of Iran, clinical coders from 26 provinces participated in the study. The most significant participation was from Tehran and Khuzestan provinces.

3.1. Clinical Coding Experience and Clinical Coders' Familiarity With the WHO Family of Classifications

The average work experience as a clinical coder was 7.6 ± 5.9 years, ranging from 1 to 29 years. Most participants (63.9%) had < 10 years of clinical coding experience.

Fifty‐five coders (13.6%) in the health facilities worked on coding the causes of death recorded on death certificates. As shown in Table 3, most coders reported having adequate skills in using ICD‐10; however, they were unfamiliar with other classifications within the WHO Family of International Classifications.

Table 3.

Basic knowledge of the participants on clinical coding (n = 404).

The WHO Family of International Classificationsa	Level of familiarity
	Perfectly	Very	Somewhat (clinical coding ability)	Slightly (general knowledge)‏	Unfamiliar
	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)
ICD‐10	108 (26.7)	226 (55.9)	70 (17.3)	0	0
ICD‐O	6 (1.5)	38 (9.4)	81 (20)	256 (63.4)	23 (5.7)
ICF	2 (0.5)	22 (5.4)	43 (10.6)	159 (39.4)	186 (46)
ICPC	1 (0.2)	13 (3.2)	38 (9.4)	140 (34.7)	212 (52.5)
ICHI	2 (0.5)	9 (2.2)	37 (9.2)	129 (31.9)	227 (56.2)

Abbreviations: ICD, international classification of diseases; ICD‐O, international classification of diseases for oncology; ICF, international classification of functioning, disability, and health; ICHI, international classification of health interventions; ICPC, international classification of primary care.

^a From the WHO Family of International Classifications, the five most significant systems were examined.

3.2. Familiarity With ICD‐11

Out of 404 participants, 118 (29.2%) reported being familiar with ICD‐11. Therefore, approximately three‐quarters of the participants were entirely unfamiliar with ICD‐11. Generally, the level of familiarity was low and had an average score of < 2 out of 4. However, the familiarity with the ICD‐11 coding tool and postcoordination (49.3% and 47.2%) were more than the other items. The lowest level of familiarity was related to the ICD foundation component (15.9%) (Table 4).

Table 4.

Level of familiarity of the participants with the structure and coding instructions of the ICD‐11 (n = 118).

ICD‐11 structure	Level of familiarity
	Perfectly	Very	Somewhat	Slightly	Unfamiliar	Mean (4 score)
	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Mean (%)
Browser (blue)	5 (4.2)	18 (15.3)	33 (28.0)	61 (51.7)	1 (0.8)	1.70 (42.6)
Browser (orange)	4 (3.4)	10 (8.5)	11 (9.3)	30 (25.4)	63 (53.4)	0.83 (20.8)
Reference guide	5 (4.2)	18 (15.3)	36 (30.5)	44 (37.3)	15 (12.7)	1.61 (40.3)
Coding tool	14 (11.9)	19 (16.1)	36 (30.5)	48 (40.7)	1 (0.8)	1.97 (49.4)
ICD foundation component	5 (4.2)	4 (3.4)	10 (8.5)	23 (19.5)	76 (65.4)	0.63 (15.9)
Cluster coding	5 (4.2)	17 (14.4)	38(32.2)	55 (46.6)	3 (2.5)	1.71 (42.8)
Precoordination	5 (4.2)	13 (11.0)	22 (18.6)	45 (38.1)	33 (27.9)	1.25 (31.4)
Postcoordination	5 (4.2)	18 (15.3)	58 (49.2)	33 (27.9)	4 (3.4)	1.89 (47.2)
Stem code	5 (4.2)	19 (16.1)	33 (27.9)	58 (49.2)	3 (2.5)	1.70 (42.6)
Extension code	5 (4.2)	18 (15.3)	34 (28.8)	59 (50.0)	2 (1.7)	1.70 (42.6)
ICD‐10/ICD‐11 mapping tables	4 (3.4)	4 (3.4)	10 (8.5)	30 (25.4)	70 (59.3)	0.66 (16.5)

Abbreviation: ICD, International Classification of Diseases.

There was a statistically significant relationship between education level (p < 0.001), familiarity with other classification systems such as ICF, ICPC, ICD‐O, ICHI (p < 0.001), ICD‐10 (p = 0.016), and ICD‐9‐CM (p = 0.024), and the field of study (p < 0.001). Coders with Ph.D. or master's degrees, graduates in health information management and health information technology, and employees in health facilities and teaching hospitals of metropolitans were more familiar with ICD‐11 than others. Among the participants, the highest level of familiarity was related to participants with a PhD. in health information management. There were no significant relationships between age (p = 0.273), gender (p = 0.298), work experience in the field of coding (p = 0.243), and general work experience (p = 0.11) with the familiarity with ICD‐11 (Table 5).

Table 5.

Factors associated with ICD‐11 familiarity.

Factors		Familiarity with ICD‐11		Test result	p value
		Yes	No
General work history (mean ± SD)		12.6 ± 7.22	13.87 ± 6.8	−1.57a	0.11
Coding work history (mean ± SD)		6.57 ± 5.8	7.19 ± 6.03	14779.5b	0.243
Age (mean ± SD)		37.04 ± 7.64	38.51 ± 6.67	14848.0b	0.273
ICD‐10 familiarity (mean ± SD)		3.06 ± 0.75	2.85 ± 0.72	12822.0b	0.016
ICD‐O familiarity (mean ± SD)		1.69 ± 0.98	1.01 ± 0.89	9180.0b	< 0.0001
ICF familiarity (mean ± SD)		0.97 ± 0.82	0.50 ± 0.74	9947.5b	< 0.0001
ICPC familiarity (mean ± SD)		0.86 ± 0.72	0.47 ± 0.70	10257.0b	< 0.0001
ICHI familiarity (mean ± SD)		0.89 ± 0.82	0.42 ± 0.63	9951.0b	< 0.0001
ICD‐9‐CM familiarity (mean ± SD)		2.79 ± 0.94	2.52 ± 0.95	12493.5b	0.024
Gender n (%)	Male	17 (32.7)	35 (65.3)	1.082c	0.298
	Female	91 (25.9)	261 (74.1)
Field of study n (%)	MR	52 (20.2)	205 (79.8)	20.251c	< 0.0001
	HIT/HIM/MI	43 (43.9)	55 (56.1)
	Other	13 (26.5)	36 (73.5)
Education n (%)	Associate	—	39 (100)	28.77c	< 0.0001
	Bachelor	71 (25.1)	212 (74.9)
	Masters/Ph.D.	37 (45.1)	45 (54.9)
Type of center n (%)	Charity/private	5 (18.5)	22 (81.5)	2.433c	0.657
	Teaching	48 (28.6)	120 (71.4)
	Nonteaching	17 (22.1)	60 (77.9)
	Health facilities	22 (29.7)	52 (70.3)
	Other centers	16 (27.6)	42 (72.4)

^a t‐test.

^b Mann–Whitney U.

^c Pearson's χ ².

3.3. Reasons for Unfamiliarity With ICD‐11

The 286 participants who reported being unfamiliar with ICD‐11 were asked to explain their reasons. Table 6 shows the number of answers given to each possible barrier. Most of the participants believed that the main reasons for unfamiliarity with ICD‐11 included no training courses (67.5%), not knowing about the Iranian implementation plan (48.3%), and not being informed about the release of ICD‐11 (43.4%), respectively.

Table 6.

Reasons why coders were unfamiliar with ICD‐11 (n = 286).

Barriers	Frequency	Percentagea
Assuming that ICD‐11 will not be implemented in Iran soon	57	19.3
Not knowing about the Iranian plan to implement ICD‐11	138	48.3
Until ICD‐11 is implemented, there is plenty of time to learn	36	12.6
Lack of training courses on ICD‐11	193	67.5
Lack of support from the employer (healthcare organizations) for employee training	82	28.7
Lack of Persian resources to learn about ICD‐11	101	35.3
Busy with a family issue	21	7.3
Too busy	90	31.5
Concern about the change (feeling of job risk) caused by implementing ICD 11 (as a new coding system) in the future.	14	4.9
Not being aware of the release of ICD‐11	124	43.4
Others	30	10.5

^a The participants could choose multiple options when responding to the questionnaire.

3.4. The Attitude of the Clinical Coders on Adopting New Roles After the Full Implementation of ICD‐11 in Iran

The participants who were familiar with ICD‐11 were asked to express their readiness to adopt new roles following the implementation of ICD‐11. In all categories, fewer than 25% indicated that they were prepared to assume new roles. The lowest level was associated with membership in the design and development team of software and electronic information systems based on ICD‐11 (Table 7).

Table 7.

Readiness to accept new roles after the full implementation of ICD‐11 in Iran (n = 118).

The new roles after the full implementation of ICD‐11 in Iran	Level of readiness
	I lack this skill	I have basic skill	I have great skill
	Frequency (%)	Frequency (%)	Frequency (%)
Applying the knowledge, rules, and standards of ICD‐11 electronically for coding diseases or causes of death in healthcare facilities	29 (24.6)	63 (53.4)	26 (22.0)
Using ICD‐11 codes in payment systems electronically	21 (17.8)	53 (44.9)	44 (37.3)
Evaluating the quality of assigned codes with ICD‐11 in electronic systems	23 (19.5)	61 (51.7)	34 (28.8)
Evaluating of rules and guidelines for the use of ICD‐11	24 (20.3)	58 (49.1)	36 (30.5)
Managing the quality assessment of data coded with ICD‐11 and making necessary correction	25 (21.2)	63 (53.4)	30 (25.4)
Implementing and evaluating ICD‐11 coding software or computer coding systems with ICD‐11	18 (15.3)	48 (40.7)	52 (44.1)
Using quality assessment guides and coding quality audits for ICD‐11	26 (22.0)	53 (44.9)	39 (33.1)
Analyzing health information and reporting to the Ministry of Health/WHO based on ICD‐11	28 (23.7)	56 (47.5)	34 (28.8)
Compiling and designing ICD‐11 educational content for participating in ICD‐11 training	25 (21.2)	53 (44.9)	40 (33.9)
Cooperating in designing electronic health information systems based on ICD‐11	14 (11.9)	50 (42.3)	54 (45.8)

Abbreviation: ICD, International Classification of Diseases.

3.5. The Attitude of the Participants About the Appropriate Method to Teach ICD‐11 to Clinical Coders in Iran

Participants preferred face‐to‐face training, such as workshops and mentoring in hospitals, over online training (Table 8). Additionally, in response to the open question regarding suggested training methods, the following points were emphasized:

• Using online (virtual) training as a supplement to face‐to‐face training.

• Continuity of the training process even after the implementation.

• Teaching how to document diagnosis according to ICD‐11 principles to physicians and coders simultaneously.

• The necessity of nationwide training (providing an educational program in all Iranian universities of medical sciences).

Table 8.

The appropriate ways to teach ICD‐11 to coders in Iran based on the participants' attitude (n = 118).

Teaching methods	Agree	Somewhat agree	Neither agree nor disagree	Somewhat disagree	Disagree	Mean (4 score)
	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Frequency (%)	Mean (%)
Face‐to‐face training in the form of mentoring in healthcare facilities (such as hospitals)	66 (55.9)	25 (21.2)	22 (18.6)	4 (3.4)	1 (0.8)	3.28 (82)
Face‐to‐face training in the form of a workshop	73 (61.9)	29 (24.6)	12 (10.2)	4 (3.4)	0	3.44 (86)
Online (virtual) training	23 (19.5)	28 (23.7)	35 (29.7)	24 (20.3)	8 (6.8)	2.28 (57)
Offline (virtual) training	27 (22.9)	23 (19.5)	31 (30.5)	27 (22.9)	10 (8.5)	2.25 (56.3)

The participants believed that the most suitable method for selecting cases to teach coding was using real medical records (63.5%). Using unique diagnoses (14.4%), prewritten scenarios (13.6%), and routine diagnoses (13.4%) were the others, respectively.

3.6. The Attitude of the Participants on ICD‐11 Educational Content

Table 9 shows that the coders familiar with ICD‐11 believed that skill in using ICD‐11 tools was a higher priority than other educational topics.

Table 9.

Suggested content for ICD‐11 training in Iran (n = 118).

The contents	Content details	Frequency (%)a
Familiarity with ICD‐11	The structure of ICD‐11 and the introduction of its chapters	98 (83.1)
	Differences between ICD‐11 with ICD‐10	86 (72.9)
	Instructions for using ICD‐11	101 (85.6)
	Stem code	49 (41.5)
	Extension code	71 (60.2)
	Cluster coding	49 (41.5)
	Precoordination coding	48 (40.7)
	Postcoordination coding	49 (41.5)
	Mean (Percent)	68.9 (58.4)
Skill in using ICD‐11 tools	ICD‐11 browser	97 (82.2)
	ICD‐11 coding tool	98 (83.1)
	ICD‐11 Reference Guide	99 (83.9)
	ICD‐10/ICD‐11 mapping tables	68 (57.6)
	Mean (Percent)	90.5 (76.7)
Advanced training	Application of ICD‐11 to medical reimbursement cases	70 (59.3)
	Application of ICD–11 for quality and patient safety	73 (61.9)
	Familiarity with mortality coding in ICD‐11	95 (80.5)
	Evaluation of the quality of ICD‐11 coded data	68 (57.6)
	The method of reporting coded information based on ICD‐11 to the Ministry of Health/WHO	65 (55.1)
	Design and development of electronic health information systems compatible with ICD‐11	79 (66.9)
	Principles of documentation and diagnosis recording according to ICD‐11	66 (55.9)
	Application of WHO Disability Assessment Schedule 2.0 (WHODAS 2.0)	34 (28.8)
	Mean (Percent)	68.75 (58.3)

^a The participants could choose multiple options when responding to the questionnaire.

4. Discussion

The results showed that 29% of the participants were familiar with the ICD‐11. Additionally, the overall score reflecting familiarity with ICD‐11 features was low, with an average score of < 2 out of 4. Based on these findings, it can be concluded that, overall, familiarity with ICD‐11 was low at the time of the study. To the best of our knowledge, there are few studies regarding the educational needs of coders concerning ICD‐11.

Some studies show that the coding accuracy with ICD‐11 is lower than that of ICD‐10 [35, 36, 37]. One reason for this discrepancy is the coders' insufficient knowledge of ICD‐11. A study in Japan has shown that the coder had more problems with ICD‐11 than with ICD‐10, primarily because of the ambiguity of the codes in ICD‐11 [35] and the lack of ICD‐11 training [38]. In addition, a study in South Korea showed that implementing ICD‐11 led to less coding accuracy than ICD‐10. Therefore, it is essential to improve the knowledge and skills of coders, especially in choosing the stem codes and postcoordination process [36].

The results indicated that familiarity with ICD‐11 was not associated with age, gender, ICD‐10 coding skills, or work experience, as these factors were not statistically significant. But individuals who graduated in fields other than health information management and health information technology, those working in small cities, and those who are not familiar with the WHO Family of International Classifications had significantly less familiarity with ICD‐11 compared to other clinical coders. In addition, 63.6% of the participants graduated with degrees in medical records, and their knowledge was lower than those who graduated in health information management and health information technology. It should be noted that the field of health information technology has replaced the field of medical records in Iran since 2010. In response to this change, the curriculum for this field was revised, and a new course was introduced to help students become familiar with specialized classification systems, such as ICF and ICD‐O. Considering the graduation period of these individuals and their limited familiarity with specialized classification systems, more effort is needed in comprehensive planning to effectively teach them the ICD‐11. Furthermore, the findings state that lack of training courses, information about Iran's plan to implement ICD‐11, and knowledge about releasing ICD‐11 were the main reasons for coders' unfamiliarity with ICD‐11.

The participants believed that the most effective way to teach ICD‐11 was through in‐person training, utilizing mentoring and real medical records. A research group in Canada has published a report on the structure and content of ICD‐11 training. They explained their educational materials covered class presentation slide sets, line coding, case scenarios, and a knowledge assessment quiz. The experience of implementing this educational program showed that the coders were satisfied with the duration of the training, the training steps, and the number of face‐to‐face sessions. They also believed that peer‐to‐peer discussions were more helpful than slide presentations. Moreover, according to the coders' attitude, access to experienced coders during coding, scenario coding, and coding a complete medical record, interaction with other coders, and class sessions have been very useful. Canadian researchers believed that training in small groups of coders enhanced interaction, and they found that 20 h of training (consisting of 10 sessions of 2 h each) was sufficient for ICD‐11 training [11]. The Canadian coders' attitude regarding in‐person training and access to experienced mentors are similar to ours. However, there was a difference regarding training coders using standard diagnoses and scenarios.

The participants believed that educational subjects such as the structure and instructions for using ICD‐11, the use of reference guide, coding tool, browser, and mortality coding were more necessary than other subjects. In contrast, Canadian coders considered subjects such as postcoordination coding, coding injuries and harms, complications, adverse reactions, and extension codes necessary for training [11]. The results of these studies were somewhat different in the priority of educational needs. The difference may be due to the timing of distributing the questionnaires. In the Canadian study, the coders filled out the questionnaire after completing the training course, but in our research, no training course was held for the participants. It should be noted that ICD‐11 has many differences from ICD‐10 [8]. Therefore, the ICD‐11 training program should be designed with an emphasis on all new aspects.

Coding with ICD‐11 is different from ICD‐10, and this new classification supports electronic coding. Therefore, coders should be ready to accept new roles (in addition to coding medical records). However, the results of this study indicated that fewer than 25% of the participants were prepared to adopt these new roles. Therefore, this issue should be considered in designing and implementing ICD‐11 educational content.

5. Strengths and Limitations

As far as we know, this study was the first study at a broad and national level regarding the educational needs of coders for using ICD‐11. This is the strength of the study. However, this study had some limitations. Due to the lack of access to clinical coders, an online survey was used to distribute the questionnaires. As a result, the distribution of the participants from different provinces of the country was not the same. In addition, due to the unclear number of coders in the country, an accurate estimate of the total number of clinical coders was not available. It is important to note that this study focuses on the period of 2021–2022. Since then, several workshops have been held locally in Iran to familiarize health staffs with ICD‐11. The Iranian Ministry of Health has launched a training program for coders responsible for death registration. Therefore, repeating the study may yield different results.

6. Conclusion

The results showed that only 29% of the clinical coders declared that they were familiar with ICD‐11, and among them, the level of familiarity with most of the topics inquired in this study was low. In addition, this study did not confirm the statistically significant relationship between the level of familiarity with ICD‐11 and age, gender, ICD‐10 coding skill, and work experience. However, older graduates, those working in small towns, and those unfamiliar with the WHO Family of International Classification Systems were far less familiar with ICD‐11. The lack of training courses, lack of information about the national plan for implementing ICD‐11, and lack of information about the release were the main reasons for not being familiar with ICD‐11. Additionally, the best method for training ICD‐11 is face‐to‐face training in the form of mentoring in healthcare facilities (such as hospitals) and face‐to‐face training in the form of a workshop using complex and actual medical records. The educational preferences of the coders included ICD‐11 structure, instructions for its use, use of reference guide, coding tool, browser, and ICD‐11 coding rules.

7. Recommendations

It is suggested to develop comprehensive content for training programs according to the target groups' needs. These programs should include the content of the course, educational resources, duration of the training, how to train, and how to certify and evaluate the training process. In addition, these programs should be developed in cooperation with the board of health information management, the ICD‐11 pilot team, and the educational departments of health information management and technology at universities. The clinical coding courses in curriculums should be revised to include ICD‐11. As suggested by the Canadian research team, countries' educational content can be shared internationally.

Author Contributions

Javad Zarei: conceptualization, methodology, data curation, formal analysis, writing – original draft, writing – review and editing, funding acquisition, investigation, validation. AliMohammad Hadianfard: conceptualization, methodology, writing – review and editing, writing – original draft. Abbas Sheikhtaheri: conceptualization, methodology, supervision, formal analysis, writing – original draft, writing – review and editing, investigation, validation.

Conflicts of Interest

The authors declare no conflicts of interest.

Transparency Statement

The lead authors, Javad Zarei and Abbas Sheikhtaheri, affirm that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.

Data Availability Statement

The authors confirm that the data supporting the findings of this study are available within the article.