October 20, 2014

Anesthesia time “starts when the anesthesia practitioner begins to prepare the patient for anesthesia services in the operating room or an equivalent area,” according to the Medicare regulations and the ASA Relative Value Guide^®, or “begins to prepare the patient for the induction of anesthesia” as stated in the CPT^® Anesthesia Guidelines.  What do those words mean—and can there be any remaining controversies after all these years?

To “begin to prepare the patient” involves “doing something to the patient” or placing “hands on,” in the memorable words of a former chair of the ASA Committee on Economics.  Thus anesthesia time starts with an intervention performed after completion of the preoperative assessment, whether that assessment is done in the holding area or in the operating room itself.  The start of anesthesia time occurs before anesthesia is induced.  Our concern is with the period of time from the first act of physical preparation, such as the placement of monitors or the administration of pre-anesthesia sedation, until induction.

More particularly, our concern is with recording the minute when anesthesia time begins for billing purposes.  Quite a few anesthesiologists only start running the clock when the patient is in the room (PIR).  That may happen several minutes after the anesthesiologist or CRNA places an i.v. line, gives an anxiolytic and begins monitoring the patient while they are still in the holding area, i.e., actual “Anesthesia Start Time” (AST).  One reason to delay recording the AST is to avoid issues of concurrency.  Another is that the physical layout of the OR suite and the preoperative areas as well as workflow may delay the patient’s arrival in the room by five or ten minutes or even more.  At one major health system in the Midwest, an analysis of AST and PIR times showed that an average of eight minutes actually elapsed between AST and PIR times in the roughly 17 percent of cases in which the recorded AST and PIR times were less than a minute apart.  The layout of the OR suite would have made it difficult for AST and PIR to occur in fact within a minute of each other, however, and further inquiry revealed that a number of providers would simply wait until they were in the room to document the two events simultaneously, even if they had been monitoring the patient continuously, because it was not convenient in the preoperative area.

This matters because, as we all know, time is money in anesthesiology practice.  Leaving eight minutes of anesthesia time off the claim in thousands of cases can add up to appreciable sums.  To illustrate, let us assume that there is an eight-minute lag in 1,500 Medicare cases, or 15 percent of a hypothetical practice’s annual volume of 10,000 Medicare cases.  The national average anesthesia conversion factor is currently $22.62, and there are 15 minutes in a Medicare time unit, so the value of one minute of billable anesthesia time is $1.508 ($22.62/15).  The practice would be underbilling by $18,096.00  ($1.508 x 8 minutes x 1,500 cases).

For commercial cases, the missed revenues would be greater.  The national average commercial conversion factor is about $70 (Stead SW, Merrick SK. ASA 2014 Survey Results for Commercial Fees Paid for Anesthesia Services.  ASA Newsl. 2014:78(10);63-69).  Assuming a total commercial volume of 20,000 cases, as well as the 15-percent underreporting rate in the previous example, the practice would be losing eight minutes per case in each of 3,000 cases.  Let us assume further that the payer contracts in one half of those 3,000 cases followed the Medicare method of dividing the total number of minutes in a case by 15 to obtain the number of units to one decimal place.  The dollar value of each minute of these 1,500 cases would be $4.67 ($70/15), so the practice would have underbilled by a total of $56,040 ($4.67 x 8 minutes x 1,500 cases).  That leaves the other 1,500 cases, for which we assume that the payer contract provides for rounding up to the next whole time unit if the remainder (after dividing the total minutes by 15) is between 8 and 14 minutes.  If one-half of these 1,500 cases could thus be rounded up to the next whole unit, we would have a further underbilled amount of $52,500 (750 cases x 1.0 extra unit x $70).  Obviously the Medicare and commercial conversion factors, as well as the rounding rules, will vary between payers and locales, so the numbers here are not necessarily representative.

In our hypothetical practice that performs 10,000 Medicare cases and 20,000 commercial cases in a year, the total value of the time not billed would thus be $126,636.  If there are no concurrency concerns dictating otherwise, why leave this money on the table?

An informal poll of a number of other academic anesthesiology departments revealed that the problem was widespread.  The main reason for failure to start the clock at the point of AST was that it simply was not convenient (or even technically feasible, in some cases) to log into the Anesthesia Information Management System (AIMS) and document AST in the preoperative holding area. Instead, providers waited until they were in the operating room and clicked both the AST and PIR buttons simultaneously when logging in to the AIMS.  In this manner they also hoped to avoid any potential auditor’s questions about starting time too early—or about concurrency, in cases where times might appear to overlap between more than the permitted number of concurrent cases.

The same “flight to conservatism” in reporting pre-PIR time is common in private practice.  Many anesthesiologists prefer not to chance disputes with Medicare over when they actually began to prepare the patient for anesthesia services or over the number of concurrent cases and systematically start the clock when both they and the patient are in the OR.  This would appear to be true even when there is no AIMS—and there is no AIMS in most holding areas—and no need to document the event that should start the clock by entering information into a computer in real time; the paper record offers greater flexibility in this respect.  Starting to count anesthesia time in the OR, even if the anesthesiologist has begun to prepare the patient in the holding area, is a reasonable option for practices that do not want to allow any possibility of exceeding medical direction concurrency limits, we would remind readers.  The additional minutes that could be billed and paid may not be worth it. Analysis of how your own OR functions is key.

In OR suites where the reason for delaying the start of anesthesia time is the perceived hassle of documenting AST in the AIMS while preparing and monitoring the patient in the holding area, or where the AST button is in the intraoperative AIMS module, there is a technology solution:  provide clinical decision support and create a targeted alert that will prompt the anesthesiologists for a correction, during the case, when the interval between AST and PIR is too short (or, for that matter, too long).  One product that provides this feature is the AlertWatch™ system, developed at the University of Michigan for use by clinicians for secondary monitoring of patients within operating rooms.  AlertWatch accomplishes this by aggregating data from networked physiologic monitors, AIMS and patient medical records and displaying them at a central location.  (Disclosure:  I am a member of the board of directors of AlertWatch, Inc.  ABC and AlertWatch have a joint marketing agreement.)

In two abstracts presented at the January 2014 Society for Technology in Anesthesia conference, preliminary data showed improved intraoperative hemodynamic management and perioperative glycemic control.  In September, AlertWatch was chosen as a semi-finalist in the 2014 Accelerate Michigan Innovation Competition. The annual event brings over 50 Michigan startups together to compete for a multitude of prizes; this year’s winners will be announced in November.

The system is beginning to reveal how it can help anesthesiologists correct documentation errors while the case is still underway and providers’ recollection fresh, avoiding the problems of having billing and coding staff track down the clinicians perhaps days after the case.  Documentation alerts that are being added include “attending physician not signed in” and “pre-operative diagnosis missing.”  Other vendors offer similar compliance and documentation decision support technologies designed to alert to various potential reporting problems, including concurrency issues.  Such technologies can be deployed to help prevent both overbilling and underbilling, as in the case of recording anesthesia start time.

With best wishes,

Tony Mira
President and CEO