The International Classification of Diseases, Ninth Revision, Tenth Revision, and Eleventh Revision, are the systems used to assign diagnoses of diseases, conditions, and subconditions for the sake of insurance billing purposes, cause of death on death certificates and similar record-keeping needs. Think of ICD-9 and ICD-10—the two currently in active use in the US—as the Dewey Decimal System of health conditions in the US. These codes can go beyond the name of a particular condition by also establishing the cause of a condition, whether biological or external (such as codes explicitly associated with child abuse). Although the World Health Organization currently accepts data using ICD-10 and ICD-11, the U.S. has said it will not be able to transition to using ICD-11 until 2023 at the earliest.
Deeper dive
These codes are essential for billing purposes, but they are also frequently used in medical research as a marker for determining a person’s diagnosis or comorbidities. These can also be the source of confusion when it comes to how doctors use them. For example, some electronic health records only allow doctors to input one or two codes, which means highly relevant medical information about a condition could be left out, such as in the case of a mother giving birth. The code(s) used might refer only to the birth and/or one complication without including relevant comorbidities or other complications. In fact, the codes doctors do (or don’t) use during childbirth or that coroners do (or don’t) use on death certificates has been the source of much of the confusion and uncertainty related to official maternal mortality statistics. Any studies using ICD codes to establish diagnoses and/or comorbidities should be scrutinized in terms of how using these codes might limit the study’s findings or conclusions.