Inpatient CPT Coding Errors Explained: Why Hospitalist Groups Lose 3–5% of Revenue Every Month

Q: Why does a hospitalist group billing thousands of nearly identical encounters every month still manage to lose 3 to 5% of potential revenue to coding errors?

A: Because small, repeated mistakes in how inpatient CPT codes get selected and documented compound quietly across volume, and most of those mistakes trace back to outdated coding habits that haven’t caught up with how the rules actually work today.

Q: What changed recently that makes older coding habits risky?

A: A few significant updates reshaped how inpatient coding works. Observation care and inpatient admission codes, which used to follow separate reporting pathways, now fall under a unified code structure. The standalone observation care E/M code groups were deleted entirely, and the hospital inpatient code groups were updated to absorb observation care services into a single family of codes. Anyone still coding from a pre-2024 cheat sheet is working from a framework that no longer matches reality.

Q: How do you actually select the right code level now?

A: Code level selection comes down to either Medical Decision Making, commonly called MDM, or total time spent on the encounter. One detail that surprises a lot of providers: history and physical examination no longer factor into level selection at all, even though they’re still clinically necessary for patient care. MDM itself depends on the complexity of the problem being addressed, the data reviewed and analyzed, and the risk associated with management decisions.

Q: What about time-based billing? Does that work the way people assume?

A: Not quite, and this is where a lot of coding errors originate. Unlike many other time-based medical billing scenarios, E/M coding doesn’t allow rounding up once you pass the halfway point of a time interval. You have to meet or exceed the full threshold for a code before you’re allowed to bill it.

For initial inpatient services, the breakdown looks like this: 54 minutes doesn’t qualify for the next code level up, but 55 to 74 minutes hits 99222, and 75 to 89 minutes reaches 99223. For prolonged time using add-on code 99418, you have to exceed the base code’s threshold and then complete a full additional 15-minute increment before billing it. So 90 minutes becomes 99223 plus one unit of 99418, while 105 minutes becomes 99223 plus two units.

Subsequent inpatient services follow a similar logic: 35 to 49 minutes lands at 99232, 50 to 64 minutes reaches 99233, and 65 minutes or more allows you to add 99418 on top of 99233.

Q: What’s the single most common mistake providers make with these rules?

A: Treating “close enough” as acceptable. If a code threshold starts at 55 minutes, billing at 54 minutes simply isn’t allowed, no matter how close the documented time comes. That precision requirement catches a lot of providers who are used to more flexible rounding conventions used elsewhere in medical billing.

Q: Beyond timing errors, what other coding risks should inpatient providers watch for?

A: Three patterns show up consistently. Upcoding, billing at a level higher than your documentation actually supports, draws increased scrutiny from payers using automated detection tools and triggers real audit and repayment risk. Downcoding, selecting a lower level than the work actually supports, usually out of uncertainty about the rules, quietly erodes revenue in small increments that add up significantly over time.

The third risk involves incomplete or inaccurate documentation more broadly. Missing support for your MDM level, unclear notes about patient complexity or care decisions, or documentation that simply doesn’t align with the codes billed all create the same downstream problems: denials, delayed reimbursement, increased audit exposure, and sometimes forced downcoding by the payer.

Q: What’s the most effective way to actually fix these issues?

A: Three habits make the biggest difference. First, retire any coding references created before the 2024 changes; using outdated guides is one of the easiest ways to accidentally introduce errors. Second, strengthen documentation regardless of whether you’re coding by MDM or time. Vague notes like “patient stable, continue meds” no longer cut it; documentation needs to clearly reflect problem complexity, data reviewed, and risk involved, or, for time-based coding, the specific activities performed and total time spent on the date of service.

Third, use benchmarking to compare your coding patterns against specialty peers. This surfaces patterns like consistent overcoding or undercoding before a payer audit does, giving your team a chance to course-correct proactively rather than reactively.

Q: Is this complexity worth managing in-house, or does it make sense to bring in outside expertise?

A: That depends on your practice’s size and bandwidth, but it’s worth noting that research on outsourced revenue cycle management consistently shows meaningful cost reductions, often in the 30 to 40% range, alongside improved coding accuracy and faster reimbursement. For practices already stretched thin on administrative capacity, a specialized partner familiar with current inpatient CPT code requirements can close coding gaps far faster than building that expertise internally from scratch.

Q: What’s the bottom line for inpatient groups trying to get this right?

A: Coding accuracy isn’t a one-time fix; it’s an ongoing discipline. Align your codes precisely with the documented work, support every code with documentation detailed enough to withstand scrutiny, and validate your team’s performance regularly through benchmarking rather than waiting for a payer audit to reveal the gaps.