Objective: The accuracy of clinical coding in COVID-19 is essential for quality of care, disease surveillance, as well as research and reporting. This study aims to describe and categorize consultations between medical coders based on the social media network in Iran on the COVID-19 coding. Method: A clinical coding group in the social network at the national level in Iran was established to follow consultation regarding COVID-19 among coders. We also utilized an online survey, which was designed to extract the problems coders encountered during clinical coding and their opinion on whether these consultations were effective. Herein, we report messages and communication records exchanged among members of this network obtained between 21 February 2020 and 20 November 2020. Finally, we categorised the obtained information and identified the problems for COVID-19 accurate coding in Iran. Results: A total of 1,340 messages in 332 consultations were exchanged amongst 76 coders. We categorised topics of consultations into 11 categories. Most consultations dealt with “suspected or probable” (n = 71), “clinical coding and diagnosis” (n=59) within 332 conversations. In 47% of consultations, the first reply was less than 10 ± 3 min. “Maternal and infant ” and “procedures and drugs ” were the most common subjects with specific answer. Based on the viewpoints of coders, online consultations can reduce the time of clinical coding and increase coding accuracy. Conclusion: The establishment of social networks among medical record coders is an efficient strategy to deal with coding issues during the COVID-19 pandemic and improve the quality of hospital records.