| Gerber's E-Myth theory applies well to billing: most | | | | categories: encoders, revenue-cycle management |
| billing services fail because the founders are | | | | tools, and artificial intelligence systems: |
| "technicians" who are inspired to start a business | | | | |
| without knowledge of how successful businesses | | | | 1. Encoders help coding personnel improve |
| run. Typically billing "technicians," who are skilled at | | | | consistency. An encoder is a database-driven |
| billing and may enjoy coding, start their own billing | | | | software, which prints a list of codes in return to |
| operation and continue doing the work they are | | | | a procedure name or problem description. Typical |
| skilled at. However without access to solid | | | | encoders include ICD-9, CPT, and HCPCS books. |
| technology and industrial-grade processes, these | | | | Additionally, encoders provide access to |
| "technicians" soon find themselves unable to scale | | | | transmittals and bulletins from Medicare, advice |
| up. Rather than working "on" the business, they | | | | from the AMA's CPT Assistant, CCI, and details |
| work "in" the business, merely owning a job | | | | from national coverage decisions (NCDs). The |
| instead of a business. | | | | database must be continuously updated, which |
| Billing is especially hard because of coding | | | | can be accomplished with ongoing Internet |
| complexity and payer adversity. The shear | | | | research. |
| number of codes and rules create an | | | | 2. Revenue-Cycle Management tools scrub claims |
| environment, where a coder is unable to perform | | | | and manage the claim payment cycle. Scrubbing |
| consistently. It includes more than 8,500 | | | | starts with missing field identification and testing |
| procedure codes and modifiers, over 16,000 | | | | of medical necessity rules, typically including CCI, |
| diagnosis codes, and millions of rules for medical | | | | LMRP, and, most importantly, local payer rules. |
| necessity, correct coding initiative (CCI), local | | | | The knowledge base, consisting of rules like "For |
| medical review policy (LMRP) and bundling. Even | | | | payer X and CPT code Y, do Z," must be |
| highly trained coders have difficulty maintaining | | | | continuously updated. However, in contrast to |
| coding consistency. Their CPT choices are | | | | encoders, up-to-date maintenance of the |
| inconsistent fifteen percent of the time, while their | | | | knowledge base cannot be accomplished by |
| ICD-9 codes disagree with their own earlier | | | | research alone. Since some payers do not publish |
| choices almost half the time (Perspectives in | | | | their payment policies, the rule base must be built |
| Health Information Management, Fall 2006). | | | | through a live trial-and-error process. In this |
| Billing complexity also generates opportunities for | | | | process, a delay or a denial triggers a dialog with |
| providers to commit fraud and for payers to | | | | the payer, the discovery of a missing rule, and |
| benefit at the provider's expense. An in-house | | | | software changes to encode the new rule. This |
| billing operation and a naive outsourced billing office | | | | procedure is much more time consuming and |
| owner are both helpless against insurance | | | | expensive for both the vendor and the provider |
| companies armed with a powerful three-pronged | | | | that happens to experience the inconsistency. |
| system to keep providers' money: solid business | | | | 3. Artificial Intelligence systems typically use |
| strategy, well-documented and professionally | | | | natural language processing algorithms to make |
| managed processes, and leading-edge technology.t | | | | human-computer interaction more intuitive and |
| A scalable billing service requires disciplined | | | | easy. Such tools accept an electronic medical |
| performance measurement, process consistency, | | | | record of the encounter and return a coded |
| and industrial-grade technology. Provider-side billing | | | | superbill, which include diagnosis and procedures |
| technology can be roughly divided into three | | | | codes. |