State County file, 2002-2005
The sc-MMYY.csv file is a comma separated file without headers. The sc---state county market penetration files-- contains the following fields in the order indicated. . Definitions of these fields are presented beneath. Information on CMS pages with 2006 Rx enrollment is presented on the bottom of the page.
Field Order for 2002-2005
- Part A Aged,
- Part B aged,
- Part AB Aged rate,
- Social Security State county code
Definitions of the fields
Eligibles: The number of original Medicare eligibles---individuals who are either currently or formely entitled or enrolled in either part A or part B original Medicare. We call these the ever enrolled eligibles. The count of the ever enrolled eligibles is typically 2.5%-3% higher than the currently enrolled eligibles. The ever enrolled eligibles also includes those individuals who are enrolled in original Medicare with a future effective date. By law individuals within 4-5 months of their 65th birthday may enroll in original Medicare with an effective date on their birthday. Typically this amounts to 1.5% of the eligible population.
- Enrollees: The number of Medicare Advantage enrollees---individuals who are currently enrolled in a Medicare advantage plan of some type. The types are not listed on this report but include Risk, Cost, Hcpp, Demo (Demonstrations), Private Fee For Service, PPO, PSO, Pace and PPO Demo. Note that Pace plans did not come under the Medicare Advantage program till December 2003.
- State: The eligibles and enrollees are aggregated by State and
- State County Code: The social security 5 character State County code is also given
- Penetration: Penetration is a derived field. It equals the ratio of enrollees over eligibles multiplied by 100. For example if a fictitious county had 1000 enrollees and 10,000 eligibles then its penetration would 1000/10000 *100= 10 (The 10 corresponds to 10%)
- Part A rate: This is the rate computed annually by the Office of the Actuary for the typical Aged individual enrolled in Part A original Medicare. Under the demographic payment system this rate would be multiplied by age-gender-status factors to obtain the payment for any particular individual.
- Part B rate: This is the rate computed annually by the Office of the Actuary for the typical Aged individual enrolled in Part B original Medicare. Under the demographic payment system this rate would be multiplied by age-gender-status factors to obtain the payment for any particular individual.
- Part AB rate: This is a derived field equal to the sum of the part A and part B rates.
Important comments on the data
The privacy laws of the HIPPA act have been interpreted to prohibit publishing aggregate data of 10 or less. Consequently some enrollment fields are blank because of enrollment between 1 and 10. Each state has a fictitious county field, under-11 which aggregates all deleted enrollees. Data users may redistribute the average number of enrollees deleted among the blank cells. For example, if some state had an enrollment count of 50 for the under-11 county and had 25 blank cells then a user could assume that each blank cell had 2 enrollees. This average while not 100% accurate is useful and can help with trend analysis.
The SC file is the only data file listing all counties for which the office of the actuary at CMS computes rates. Consequently if a county has no access it will be listed. In this case the enrollment cell for that county will be blank signifying an absence of MA enrollment.
There is a fictitious county Unusual SCounty with a fictitious state, 99. This county aggregates all eligibles with a state county code outside the US but not in the protectorates (For example people in Asia). It also aggregates all people with state code 99. This record is put in for purposes of completeness and auditing. Most users will probably wish to delete this record. However, eligibles with a state county code in the US but not in the list of State County codes used by the Office of the Actuary and not in the list of the State County codes recognized by the United States Postal service, these records, have been omitted with the assumption that they are miscoded. Typically there are 10-15000 (out of 40-45 million) eligibles with US residence in a county that are not recognizable. The number of eligibles in the Unusual SCounty field is typically 3-400,000.
The sc file has listed Guam since Dec 2003. Since Mar 2004 it has listed Guam (GU), American Samoa (AS), and the Northern Mariana Islands (NI). Prior to these dates the counts for these protectorates were included in the Unusual SCounty record.
The sc file lists enrollees and eligibles by their legal state and county of residence; this is the county used for payment purposes. This affects 1/2 - 1 1/2 % of the enrollees. As an example a person who moved from Baltimore to Los Angeles and enrolled in Pacificare would be paid at the Baltimore rate until the change of address was recorded in the Social Security system.
2006 Rx Enrollment
The CMS has published enrollment files that are similar but not identical to the pre-2006 files. The Related Links Inside CMS and the Related Links Outside CMS sections below contains links to files that present Rx, but not MA-only enrollment, aggregated by state and county. There are also files that contain Rx, but not MA-only enrollment, aggregated by state. The format of these 2006 files is different from the pre-2006 format. For example fips vs. social security state county codes are used. Furthermore, the eligibles are listed in a separate file and not combined with Rx enrollees in the same file.
In naming these files I have used the same formats used in naming the pre-2006 files even though the layout of the file is different.
For example, sc-0306 is a file containing enrollment data by State and County for the period March 2006, while sc-0305 is a file containing enrollment by State and County for the period March 2005. However, as just pointed out these files differ in that the March 2006 file does not contain eligible counts, does not contain MA-only-enrollee counts (but rather Rx enrollee counts), and uses fips codes vs. social security state county codes.
Users should also be aware that the same download url may list the current month enrollment. For example the download url, PartDCountyEnrollmentData, linked to April 2006 data during April 2006 but linked to May 2006 data during May 2006. Currently the url does not work. In other words, not all files are archived. Therefore those who wish to maintain archives should do so personally. I monitor these sites daily and will update file name changes within a day or two.