Table 1

Subjective assessment of the limitations of the two classification systems for sports medicine diagnoses

Reason for disagreement	ICD-10-AM	OSICS
*Where the correct code had been identified by one coder but was missed by the other two coders.
†Chapter XIII, Diseases of the musculoskeletal system and connective tissue.
‡Chapter XIX, Injury, poisoning and certain other consequences of external causes.
(1) Human error*
(a) Poor organisation of system affected navigation across the documentation	Complexity of musculoskeletal system and organisation based on pathology caused fatigue and boredom, as stated by coders	Poor hierarchical system No redirection guides Codes alphabetised rather than organised anatomically inside subsections
(b) Relationship of codes to diagnosis	Codes in chapter XIII† do not relate to the diagnosis in any recognisable pattern, potentially causing translational errors	Codes relate to the diagnosis (e.g. knee diagnoses start with k) giving some degree of self checking
(c) Lack of formal training in classification system
(i) Erroneous use of “unspecified” codes	Coders chose “unspecified” rather than “not otherwise specified” or “other” codes	Not possible to do as very few codes for “other” or “not otherwise specified” diagnoses
(ii) Omission of part of code	Chapter XIII—4^th letter specifying body location for “other” codes often omitted
(d) Lack of understanding of sports medicine diagnosis	Occurred with expert coder as a result of having no previous experience in the practice of sports medicine	All coders were experienced in the practice of sports medicine
(2) Deficiencies in the codes available
(a) Location of injury
(i) Codes too specific regarding location of injury	e.g. tib. Post. tendon tear—required localising to ankle or foot
(ii) Codes not specific enough for location of injury	e.g. a code for thoracolumbar sprain does not exist, only thoracic or lumbar
(b) Codes do not encompass all details of the diagnosis causing coder to choose:
(i) Pathology v anatomy	e.g. haematoma Achilles does not exist—either injury to Achilles (anatomy) or contusion of ankle (pathology)	e.g. L5/S1 disc degeneration does not exist—either disc injury L5/S1 (anatomy) or disc degeneration (pathology)
(ii) 1 of 2 different pathologies in one injury	Combination injuries, e.g. fracture dislocation, do not have a code in either system
(iii) 1 of 2 structures injured in one body part	e.g. ankle sprains—ICD only allows one injury to be coded. Not possible to code for a sprain of more than one ligament
(c) Post-surgical conditions	Dilemma of coding pre-surgical v post-surgical condition
(3) Deficiencies in diagnosis given
(a) Location e.g. “lateral ligament injury”	Problem—too many “other” codes	Problem—no non-specific codes
(b) Diagnosis, e.g. acceleration injury	Problem—too many “other” codes	Problem—no “other” codes
(c) Terminology, e.g. hamstring impingement syndrome	Problem—too many “other” codes	Problem—no “other” codes
(4) Classification system omits or under-represents diagnoses	Stress fractures, osteitis pubis, osteochondral injuries, specific tendinopathies	Paediatric conditions, e.g. traction apophysitis and generalised osteoarthritis
(5) Chronicity of injury uncertain	Choice of chapter XIII v chapter XIX‡, especially for knee and back
(6) Overuse injuries	Confusion whether overuse injuries can be coded in chapter XIX