I have never had trouble with the basic idea and practice of coding. Reducing diagnoses, procedures, visits and so forth to numbers for billing purposes has never proved overly problematic at my practice. What has provided a constant source of amazement – and amusement – is what the powers-that-be have decided to include in and exclude from the coding protocols.

Some time ago I discovered that I am apparently the first doctor in history to diagnose common rhinorrhea, at least as far as the International Classification of Diseases (ICD-9) is concerned. The code book allows for “cerebrospinal fluid rhinorrhea” (349.81); “allergic rhinitis, cause unspecified” (477.9); and “other diseases of nasal cavity and sinuses” (478.1). But in the eyes of the coding gods, plain old runny noses apparently do not exist.

I have found that a number of exceedingly common medical conditions are either not recognized or only obliquely acknowledged by ICD-9. When I first started using electronic health records, I typed “knee pain” in the diagnosis box during a patient visit, fully expecting the corresponding code to appear in the adjacent box. No such code was found. This is because, in the ICD-9 universe, no one suffers from knee pain per se, although they may experience “derangement of meniscus, not elsewhere classified” (717.5); “effusion of joint, lower leg” (719.06); “pain in joint, lower leg” (719.46); or, most generically, “pain in limb” (729.5).

As puzzling as the absences in the ICD-9 code set are, a fair number of the inclusions border on the ludicrous. While I have to resort to artistic license to code for a runny nose, I can be quite direct with my coding for any hapless soul who gets run over by a spaceship: “accident involving spacecraft injuring other person” (E845.9). Of equally dubious value is code E996, “injury due to war operations by nuclear weapons.” Funny how that one never makes the cut for those pocket-sized ICD-9 quick reference guides.

In principle, if a father brought his daughter to see me because she had dyed her hair green and blue, I could code for this “condition”: “variations in hair color” (704.3). But if the father himself complained of scrotal pain, the ambiguous “unspecified disorder of male genital organs” (608.9) or the equally vague “other specified disorders of male genital organs” (608.89) would have to suffice.

I wish the coding sages would ask me to suggest a new code or two. I’ve always thought there should be a code to cover some relatively inexpensive drug or procedure that Medicare or private insurance companies won’t pay for, resulting in them paying for a much more costly intervention down the road. We could call it “medico-fiscal ineptitude, not otherwise specified.” Patients who ask for an antibiotic after I’ve explained to them that they have a virus and that an antibiotic will not help could be assigned a V-code for “microbial incomprehension” along with the “unspecified viral infection” code (079.99).

Clinicians should not feel confused or intimidated by the nuanced absurdity of diagnostic and procedural coding. Like so much of the advice and instruction physicians receive nowadays, it is designed for a world much more orderly and logical than the one in which we practice medicine.