Vol. 17 Issue 13
Coding Performance & Productivity
How do your coders match up to those in our survey?
Related table 3
There are very few subjects that are as popular as coding performance and productivity. ADVANCE was fortunate to have received 67 responses to its Coding Productivity Survey that ran earlier this year. I am personally pleased to have been asked to summarize the results of the survey for ADVANCE readers. Of the 67 responses, 65 were used for this summary of the responses.
The majority (94 percent) of the responses were submitted by management representatives of whom 38 percent were HIM directors. Facility size included a broad range of licensed beds: seven to 900. Because of this broad range, the results will be displayed in summary format as well as in bed size ranges.
To the right is some basic information from the survey results to provide you with a snapshot of the responses.
The total coding hours do not match the average totals because several of the respondents reported using contract coding support or indicated that their coding team performed other duties.
The Nitty Gritty
This survey had 22 questions. A summary of the results are discussed here. ADVANCE would like to thank everyone who submitted data for this survey.
Using the allocation of hours provided by the respondents, we can establish the average coding time for each of the coding work types. However, the reader should be cautious in their interpretation of these results. Every entity is different and has different expectations for their coding team. This will be apparent when we discuss some of the different activities that consume the coding team's time.
Additionally, the average times are based on allocated authorized hours. While hours may be authorized, they may not have been worked or possibly not available to assign to coding due to other activities in the department. No time study was required to participate in this survey. Actual coding time may be longer or shorter. Table 1 provides an indicator or benchmark from which to assess one's current performance.
One must take into consideration that this represents "authorized" hours rather than "productive" hours and therefore represents an inflated time per record. Using the results for the > 450 bed category and assuming an 85 percent productive time factor (meaning that once one deducts vacations, holidays and other downtime, staff is only productive 85 percent of the time) the production expectation would change as indicated in Table 2.
However, look for the production rate to decline when present on admission (POA) and severity DRGs are implemented as the effort to determine whether the condition was POA and the time to code additional conditions to further demonstrate severity will increase. According to a study conducted at Mayo Clinic (www.hcup-us.ahrq.gov/reports/2006_1.pdf), the POA effort increased coding time by about 2 minutes. While 2 minutes doesn't sound like much, this is approximately an 8 percent increase in time.
The range of minutes per encounter for emergency department (ED) coding was wide. The ED encounter coding did not distinguish between those facili-ties that coded: 1) just diagnosis and procedures, 2) facility levels, 3) professional levels, 4) performed charge entry or 5) any combination of these. However, in the group > 450 beds, there were several facilities that noted that their coders applied professional and facility level codes as well as ambulance coding.
In small hospitals, individuals performing the coding often perform more than one type of coding as well as a variety of other duties in the department totally unrelated to coding. Therefore the speed that accompanies coding one type of record throughout the day and having few interruptions is often not available to those working in the smaller facilities. This is demonstrated in three of the four record type categories.
Conversely, the larger facilities often have more complex cases because they serve as trauma centers and teaching facilities. Having these cases requires more coding effort and time.
As individuals specialize in a certain type of coding, we can see that the coding time declines. The 150 to 450 bed categories seem to support this assumption.
Other Work Types Coded
According to this survey, a variety of other work types are coded by our coding professionals. They include:
Skilled nursing stays
Home health encounters
Outpatient clinic encounters
Additionally, we know that coding may be conducted concurrently and after discharge. Depending on the approach to coding, additional time to code the record may be required to include the travel time to/from the patient care areas.
The duties performed by coders are diverse. Having variety is often preferred by coding professionals. Coding 8 hours a day, while challenging, can also be draining. Having other duties gives the coder some change of venue. The other duties need to be activities that will utilize the skills of the coding professional. While 89 percent of the facility responses indicated that their coders also perform the data abstracting function, some of the other duties that respondents indicated in this survey were:
training registration personnel,
assisting with researching business office queries,
investigating denials that are coding related,
managing medical necessity reports,
assessing CCI edits,
addressing state reporting corrections,
fixing rejection files for patient accounting,
verifying charges on outpatients,
maintaining various registries including: birth certificates and registry, trauma registry, cancer registry, and death reporting and registries,
training physicians and others for documentation improvement purposes,
medical necessity related activities,
clearing billing edits,
resolving patient type and status changes,
tracking unbilled records,
inputting charges for all procedures, ER, injections, infusions, anesthesia, hospitalization and other procedures,
reviewing charts for quality purposes, and
performing various clerical duties including signing out and filing coded records, assembly/analysis,transcription, answering the phone and preparing statistics/census reports.
Again, the smaller the facility, the more often the person performing coding will also have a broad scope of other duties to perform as well. Each of the above listed activities adds minutes to the coding effort.
If these are not clearly differentiated, HIM management will not be able to explain why their staff levels are not consistent with the "best organizations" in comparative productivity/performanceanalyses. Each manager must communicate to his/her administrator the different duties that have been absorbed by the coding professionals in the department to adequately defend staffing levels.
Availability of Technology
It was surprising to see that there are some facilities that do not use encoding technology to assist in the coding function. Of those respondents answering the question about the availability of encoding software, 89 percent utilize it and 11 percent do not.
This response category was interesting to me, and I initially assumed that these facilities were probably not hospitals. Unfortunately, this was not the case. All were hospitals and they ranged in bed size from 50 to 510 beds. The demand on coders' individual skills at these facilities is much greater because they do not have the advantage of the online prompts and references often available in an encoder product.
The topic of compensation is always of interest. We did not capture current rates of pay because this is captured by ADVANCE's 2006 Salary Survey Results, found on the left side under Careers on www.advanceweb.com/him.
The respondents did disclose compensation variations. Overwhelmingly, the majority of respondents (69 percent) did not pay coders differently if they coded inpatient records rather than outpatient records. Of those that did pay differently, 95 percent paid the inpatient coders more. One participant explained the reason why they did not pay a different rate for inpatient vs. outpatient coders was because all coders were completely cross trained.
However, 59 percent did pay a higher rate for those coding professionals who held the certified coding specialist (CCS) credential and 47 percent for the certified professional coder (CPC) credential. Nearly 92 percent of the respondents did not have an incentive plan.
Incentive plans are an alternative approach to improve compensation levels when a facility is in a competitive environment and the facility has not been able to raise salaries sufficiently to compete. Much has been written on the development of incentive plans. Data from this survey can serve as an initial foundation for planning.
Quality expectations vary from organization to organization. The average quality expectation was 95 percent. This expectation was consistent for all categories of bed ranges. Methods of calculating quality differ so widely that comparing quality is almost impossible.
Some organizations base quality on both coding and abstracting quality; others just on coding. Coding quality can be based on a percentage of cases that had a coding error divided by the total number of cases reviewed or the number of wrong codes divided by the total codes. Coding errors could be simply the lack of a fifth digit, the incorrect selection of the principal diagnosis, the wrong code or the omission of a code, to name just several examples. Facilities are encouraged to clearly define their methodology in a departmental compliance procedure, hence ensuring a high standard of coding quality.
Rose T. Dunn is chief operating officer at First Class Solutions, Maryland Heights, MO.