Frequencies: (n=12,335) 2002 = 7,023 (57%) Not reported = 5,312 (43%)

State staff noted that approximately 50% of the children that enter foster care exit care within the first 60 days.

The program code is initialized to spaces.

The program code looks for the "legal action" date.

The State's periodic reviews are court reviews. The State has some localities that have Citizen Review Boards (CRB) that conduct the periodic review. The date the CRB conducts a review should be entered and extracted for AFCARS.

The State's program code includes the AFCARS internal consistency check to assess if the child has been in care for nine or more months. If the child has not been in care for nine months or more, but there is a periodic review date, the program code does not extract it. The State should remove this routine from the program code. The State should extract the date of a periodic review even if it occurred before the child would have been in care for nine months. If a periodic review has not been conducted, this element should be blank.

Case file findings: There were 16 out of 47 (34%) records analyzed that were incorrect. Many of these had missing dates that should have been reported to AFCARS. In ten cases, the children had been in care from six months to nine years.

Frequencies: (n=12,335) There are ages up to 28. 7 records show dates with children over the age of 21. One invalid record.

The screen has a field to enter an estimated age.

Frequencies: (n=12,335) 12% American Indian/Alaskan; 2% Asian; 15% black; 1% Hawaiian; 71% white; 6% unable to determine.

This field is initialized to spaces then reset to "0, does not apply." This field should remain initialized to spaces.

The State's code of "888, question not asked" is mapped to "unable to determine." This code should be mapped to blank. Post-site visit analysis: This code is no longer mapped in the program code.

The code "998, other race not listed" is mapped to "unable to determine." The State should remove this option and include actual other races that individuals identify. If an individual refuses, or does not identify themselves with one of the AFCARS race categories, then the State should have a value that represents "refused to provide race."

Remove from program code the condition: "if a - e = 0, then f = 1." If no race is selected in items "a - e," then all items should be left blank.

Frequencies: (n=12,335) Yes = 1,483 (12%); No = 8,954 (73%); Unable to determine = 1,898 (15%)

The field is initialized to spaces then reset to "3, unable to determine." This field should remain initialized to spaces.

State codes "703, unable to determine" and "777, can't find" are mapped to AFCARS value "unable to determine." Code "777" must be mapped to blank.

State maps "none" to "unable to determine." "None" must be mapped to blank.

Case file findings: There were 5 out of 47 (11%) records analyzed that were incorrect. The incorrect cases indicated in AFCARS "unable to determine" and the reviewers were able to determine whether the child was of Hispanic origin or not.

Frequencies: (n=12,335) Yes = 0; No = 12,335

The State has a policy that the child must have a well-child exam within 30 days, which would indicate if further psychological evaluations are needed.

There are several areas in CAMIS in which workers can enter information pertaining to a child's health. These are: the "Person Record," "the Kids Screen," and "Passport." Additionally, workers can enter medical information into their narrative notes, which cannot be extracted.

The program code extracts the disability information for AFCARS from the person record screen. The information for this element is derived based on entries of disability information on the screen. There is a field labeled "disabilities" that the workers use to enter one of the options for disabilities. Workers are not able to enter the AFCARS values of "yes," "no," and "not yet determined."

The program code checks the field for one of the disability codes. If found, then element #10 is set to "yes" and the codes are mapped to the appropriate AFCARS element, #11-15. Otherwise, the response to element #10 is set to "no." This is creating a false "no" in the reported data and makes it appear that the State's children have no medical or emotional problems. The AFCARS definition for "`no" indicates that a qualified professional has conducted a clinical assessment of the child and has determined that the child has no disabilities. "Not Yet Determined'' indicates that a clinical assessment of the child by a qualified professional has not been conducted. The question should be added to the screen in order to accurately capture whether the child has been examined by a physician and diagnosed with a disability or not. Post-site visit analysis: The program code has been modified to initialize to spaces. The changes made to the program code will result in responses of "yes" only, there will be no responses for "no" or "not yet determined." If the program code does not find a disability code entered into the screen, then this element will be blank.

During the site visit the State identified an error in the program code that needs correcting. The program code contains the word "histogram" instead of "read." This is the cause of all the responses of "no." Post-site visit analysis: The program code was changed from "histogram" to "read."

The Federal review team provided the disability resource table to the State team. It is also available on the AFCARS web site. The State should review this table and include any additional medical or psychological conditions that are appropriate.

Case file findings: There were 24 out of 45 (53%) records analyzed that were incorrect. The reviewers were able to find several diagnosed conditions with "emotionally disturbed" having the highest frequency.

Frequencies: (n=12,335) Applies = 0; Does not apply = 12,335

Initialized to spaces then reset to "zero." These should be left initialized to spaces. Post-site visit analysis: The program code has been modified to initialize to spaces.

Limited number of options for caseworkers to select. State should review the AFCARS disability resource list and decide if additional options should be added to the selection list.

Initialized to spaces and then reset to "zero." These should be left initialized to spaces. Post-site visit analysis: The program code has been modified to initialize to spaces.

The State should review the AFCARS disability resources list to ensure that all possible disabilities are collected and mapped to the AFCARS codes appropriately.

Initialized to spaces and then reset to "zero." These should be left initialized to spaces. Post-site visit analysis: The program code has been modified to initialize to spaces.

The State should review the AFCARS disability resources list to ensure that all possible disabilities are collected and mapped to the AFCARS codes appropriately.

Initialized to spaces and then reset to "zero." These should be left initialized to spaces. Post-site visit analysis: The program code has been modified to initialize to spaces.

The State should review the AFCARS disability resources list to ensure that all possible disabilities are collected and mapped to the AFCARS codes appropriately.

Initialized to spaces and then reset to "zero." These should be left initialized to spaces. Post-site visit analysis: The program code has been modified to initialize to spaces.

The State correctly maps "learning disability" to this element.

The selection list contains the option "other." This needs to be revised to reflect whether it is an other diagnosed medical condition or an other psychological/emotional condition. The State may want to consider removing the option "other" and replacing it with more detailed options.

The State should review the AFCARS disability resource list to ensure that all possible disabilities are collected and mapped to the AFCARS codes appropriately.

Frequencies: (n=12,335) Yes = 210 (2%); No = 12,125; Unable to determine = 0

The State team identified that the frequency for "yes" appeared to be high. State will review these cases to ensure its accuracy.

Screen has a field for "adopted" but there is not a specific question on the screen for the worker to answer. This information is derived from the States' own adoption records. The program code checks for in-State child welfare agency adoptions. Also, the program code looks for an outcome code of "adoption." If it finds one, element #16 is marked "yes." If an outcome of "adoption" is not found, the program marks element #16 as "no." This is resulting in a false "no" response. State shared that this is being modified this summer. A question will be added to the existing screen. Post-site visit analysis: The line checking for a placement outcome code of "adoption" has been commented out.

State asked for clarification on whether this question should include "step-parent" adoptions. The Federal team will check with the policy unit and provide a response to the State. Response: If a child was adopted prior to entering foster care by a step-parent, caseworkers would respond to this question with "yes."

Post-site visit analysis: The State submitted a copy of the screen that is being piloted to collect this data. The "create a placement" screen contains a field "adoption." The choices for the worker to select are: "yes, person is adopted," "no, person not adopted," "unable to determine," or "question not asked." This is mandatory field for workers to complete. Once completed the information is saved in the "person" record and will be permanently stored in the child's file.

If the worker selects "yes,' then an "adoption detail" sub-screen appears. This contains fields for the worker to indicate the type of adoption, i.e., international adoption, private adoption (non-dependent child), DSHS Licensed child placing agency (dependent child), out of state child welfare adoption, or other. The changes to the screen also meet the Inter-Country Adoption Act.

The program code has not been modified to reflect the revised data collection screen.

Initialized to spaces, reset to zero, then reset to "unable to determine" if the "episode outcome" is "adoption." This element should remain initialized to space.

The program code calculates the age at the time of the adoption for this element by subtracting the date of birth from the end of the placement date.

If the age at adoption is not known, the program code defaults to "unable to determine."

The new screen will have a field to enter the age.

Post-site visit analysis: The State submitted a copy of the screen that is being piloted to collect this data. The "create a placement" screen contains an "adoption detail" sub-screen and the worker can enter the child's age at adoption. The program code has not been modified to reflect the changes to the screen.

Initialized to spaces.

This element is analyzed in the program code in three different places. State should modify the program code to analyze and extract this element only once to make the program run more efficiently.

Initialized to spaces.

See the General Requirements findings regarding children placed in foster care for respite services.

The number of removals is calculated by counting the total number of removals from home.

The program code correctly ignores in-home placement from the count. The program code will count all removals that are entered into the system. Workers must enter the date that a "respite" service for in-home family preservation cases begins as a removal date and the date it ends as a discharge date, in addition to entering it as a placement date.

Initialized to spaces.

Based on the General Requirements finding that children returned home while in the agency's responsibility for care, placement, or supervision are excluded in the population extraction, this data does not accurately reflect the actual date of discharge from the previous removal episode.

Initialized to spaces.

Workers must enter the date that children are placed in out of home care for respite purposes.

Case file findings: There were 7 out of 47 (15%) records analyzed that were incorrect.

Frequencies: There are fifteen records without a placement date.

Initialized to spaces.

The program code excludes the following types of "placements:" "BA, Birth/Adoptive," "BN, Birth/Adoptive non-custodial," "HO, hospital," "OR, on the run," or missing data. For the report period under review this was correct. However, as of October 1, 2002 the date that placements in a child's own home and runaways began must be reported for this element. In regard to hospitalizations, the State needs to make modifications to collect and report hospital stays that are for non-acute care. If the stay is for a longer period of time, the State needs to report the hospital stay as a placement setting in element #41, the date the placement started in element #23, and increment the placement count by one in element #24 (Child Welfare Policy Manual, Section 1.2.B.7).

Case file findings: There were 9 out of 47 (19%) records analyzed that were incorrect. In two records, the actual dates of placement were later than what was reported in AFCARS. Two records had dates reported in AFCARS that occurred after the report period end date.

Frequencies: There are 26 records that indicate zero placements. There are records with up to 69 placement changes.

The program code correctly excludes from the count of placements: "BA, Birth/Adoptive" "BN, Birth/Adoptive non-custodial," "OR, on the run," and "RE, respite."

The program code also excludes placements in hospitals. This is correct for hospital stays for acute care. If the placement is for a longer period of time, the State needs to report the hospital stay as a placement setting in element #41, the date the placement started in element #23, and increment the placement count by one (Child Welfare Policy Manual, Section 1.2.B.7).

Case file findings: There were 11 out of 45 (24%) records analyzed that were incorrect. There was an almost equal distribution between the number of placements that decreased versus those that increased based on the case file review. Seven records decreased and five increased. The decreases were generally attributed to respite placements being counted in AFCARS when they should not have been included.

Frequencies: (n=12,335) 1 = 1,062; 2 = 10,197; Not Yet Determined = 1,076

The State's program code contains values that are no longer used, but are in older cases. Therefore, the codes must be included until these children age out of foster care or are discharged from their current removal episode.

Initialized to spaces then reset to "3, not yet determined." Must remain initialized to spaces.

The State's code "PC, protective/police" is an option on the input screen but is not mapped into AFCARS. This code needs to be separated into two different options and mapped separately to AFCARS. "Protective custody" should be mapped to "court ordered." "Police authority" should be mapped to "not yet determined."

The State staff shared that the legal and placement component of CAMIS has several problems. For instance, it is not tied to a specific legal order that gave the agency authority to place the child. The State is currently building a windows based interface using Graphical User Interface (GUI).

Case file findings: There were 8 out of 46 (17%) records analyzed that were incorrect. There were two records indicating "not yet determined" in the AFCARS file. In one case the child had been in care 24 days but it was under a voluntary placement agreement. In the other record, the child had been in care since November 1998. The remaining six cases that were incorrect were due to children entering foster care under a voluntary placement agreement and then the agency needing to obtain a court order after the 180 day timeframe to obtain custody of the child. The AFCARS reports reflected a "court order" for manner of removal, not a "voluntary placement agreement."

For FC elements #26 - 40, the State's input screen has space for only four conditions, and the worker is required to enter only one condition.

The system must support the recording of all possible combinations of circumstances associated with removal.

This information must be "frozen" to reflect only the information known at the time of removal.

Case file findings: For elements #26 - 40, the AFCARS data were incorrect because not all applicable conditions were selected by the workers.

Post-site visit analysis: The State submitted a copy of the screen that is being piloted to collect this data. The State modified the screen to allow caseworkers to select all the reasons that apply and the reasons associated with removal will be frozen 30 days after the latest removal date or the end of a removal episode, depending which occurs first. The field for "circumstances associated with removal" is now a mandatory field when a new removal from home occurs.

Frequencies: (n=12,335) Does not apply = 11,360; Applies = 975

Case file findings: There were 10 out of 47 (21%) records analyzed that were incorrect.

Frequencies: (n=12,335) Does not apply = 9,583; Applies = 2,752

Case file findings: There were 8 out of 47 (17%) records analyzed that were incorrect.

Frequencies: (n=12,335) Does not apply = 10,932; Applies = 1,403

Case file findings: There were 8 out of 47 (17%) records analyzed that were incorrect.

Frequencies: (n=12,335) Does not apply = 12,335; Applies = 0

State does not do relinquishments. Instead, the family/individual is referred to a private adoption agency. If the State changes its practice, the program code must be revised.

Frequencies: (n=12,335) Pre-Adopt Home = 136; FFH-Relative = 3,759; FFH-non-relative = 6,792; Group Home = 581; Institution = 387; Sup. IL = 16; Runaway = 206; THV = 1; Not reported = 277

Initialized to spaces.

The program code maps "RE, respite" to "foster family home, non-relative." The program code must check whether the family is also a relative of the child. For those children that in a foster home as a respite placement from their own home, the State should not code these as "respites" so that it will not be confused with respite placements for children in foster care. The State needs to develop another code to use for these placements.

If the placement setting is a hospital, the program code ignores it. This needs to be modified.

Placements in a hospital for acute care purposes are not considered placements. However, if the placement is for a longer period of time, the State needs to report the hospital stay as a placement setting, the date the placement started in element #23, and increment the placement count in element #19 by one (Child Welfare Policy Manual, Section 1.2.B.7).

If the placement setting is "BA, birth/adopt" or "BN, birth/adopt (non-custodial)," then the program code checks for a flag set to "yes." If the flag is found, then it is mapped to "trial home visit."

If an institutional setting is found, the program code maps element #49 to "not applicable," fills elements #50 and #51 with "0000" and sets elements #52 - 55 to zero.

At the end of the routine, if a "3" is in the work field, and #41 is blank, then it sets #41 to "3, FFH, non-relative." This includes all respites. State must remove this line of program code.

State has new facility codes that are not mapped in the AFCARS program code. State must add these codes to their program code. The codes are: "CS = Secured Crisis Residential Center," "CF = Family Crisis residential center" and "CG = Group crisis residential center." Post-site visit analysis: The State added the codes "CS = Secured Crisis Residential Center," "CF = Family Crisis residential center," and "CG = Group crisis residential center." "CS" and "CG" are mapped to "group home" and "CF" is mapped to "foster family home, non-relative." Also the State changed the mapping of "RA, regional assessment center" from "institution" to "group home."

Frequencies: (n=12,335) Yes = 648; No = 11,672 Invalid values reported (0) = 15

This element is initialized to spaces then reset to zero. Must remain initialized to spaces.

If "WA" is in the address, then sets element #42 to "no." Otherwise, maps to "yes."

Frequencies: (n=12,335) Reunify = 8,221; Live w/other relative = 221; Adoption = 2,365; LTFC = 421; Emancipation = 115; Guardianship = 683; Not yet established = 309

State requires a goal within 30 days, and no later than 60 days. If unknown is selected by the worker after 60 days there is a reminder to the worker to update the information.

Initialized to spaces.

"None" is mapped to "7, not yet established." Blanks and "unknown" are also mapped to "not yet determined." Modify program code to remove "none" from program code and map to blank. Map "unknown" to "not yet established." The program code must be set to blank for "not yet established" after 60 days. Post-site visit analysis: The program code was modified to map missing data to blanks. "U, not yet determined," is mapped to AFCARS value "7, case plan goal not yet established."

Case file findings: There were 11 out of 46 (24%) records analyzed that were incorrect. The incorrect records reflected case plan goals that had not been updated to reflect the most recent case plan goal. The majority of the cases were reported to AFCARS with a goal of "reunify."

Frequencies: (n=12,335) Married couple = 2,208; Unmarried couple = 1,655; SF = 6,155; SM = 559; Unable to determine = 1, 758 (14%)

Initialized to spaces.

Looks for information on the home from which the child was removed.

If no value is found, program code is set to "unable to determine." Must be set to blank. Post-site visit analysis: The program code was modified to map missing data to blank.

Freeze marital status at the time of removal.

Individuals that are "separated" must be mapped to "married."

Case file findings: There were 11 out of 46 (24%) records analyzed that were incorrect. In eight of the incorrect records, the AFCARS data indicates "unable to determine," but the reviewers found the information in the case file.

Frequencies: (n=12,335) Invalid data (0000) = 16; Not reported = 1730 (Totals 1746) Some records have birth years of 1901, 1903, 1910 1917, 1919, 1921, 1992 - 2002.

Initialized to "0000."

State believes that there may be a problem with where the information is extracted from based on the case file review.

Case file findings: There were 10 out of 46 (22%) records analyzed that were incorrect.

Frequencies: (n=12,335) Reported = 3,750 Not reported = 8,585; Invalid data (0000) = 29

State believes that there may be a problem with where the information is extracted from based on the case file review.

Case file findings: There were 6 out of 46 (13%) records analyzed that were incorrect.

Frequencies: (n=12,335) Not Applicable = 1,468; Married = 5,937; Unmarried Couple = 419; Single Female = 3459; Single Male = 492; Not reported = 560

Initialized to spaces.

If no value found, sets to "not applicable." Must be set to blank. Post-site visit analysis: The program code was modified to map missing data to blank.

"Separated" must be mapped to "married."

Case file findings: There were 6 out of 40 (15%) records analyzed that were incorrect.

Frequencies: (n=12,335) Not reported = 1468; Reported = 10,867 matches records in #41; Invalid data (0000) = 1,942 - Some records going back to 1901.

Initialized to "0000." Should be initialized to blank. Post-site visit analysis: The program code was modified to map missing data to blank.

If an institutional setting is found in #41, then this is set to "00." Remove this line. Post-site visit analysis: The program code was modified by commenting out this line.

Case file findings: There were 6 out of 26 (23%) records analyzed that were incorrect.

Frequencies: (n=12,335) Not reported = 1,468; Invalid data = 4,577

Initialized to "0000." Should be initialized to blank. Post-site visit analysis: The program code was modified to map missing data to blank.

If an institutional setting is found in #41, then this element is set to "0000." This program code should be removed. Post-site visit analysis: The program code was modified by commenting out this line.

Case file findings: There were 6 out of 26 (23%) records analyzed that were incorrect.

Frequencies: (n=12,335) Unable to determine = 993 (8%); Two or more races: Applies = 164 (1%)

This field is initialized to spaces then is reset to "0, does not apply." This field should remain initialized to spaces.

The State's code of "888, question not asked" is mapped to "unable to determine." This code should be mapped to blank. Post-site visit analysis: This code is no longer mapped in the program code.

The code "998, other race not listed" is mapped to "unable to determine." The State should remove this option and include actual other races that individuals identify. If an individual refuses, or does not identify themselves with one of the AFCARS race categories, then the State should have a value that represents "refused to provide race."

Remove from program code the condition: "if a - e = 0, then f = 1." If no race is selected in "a - e," then "f" should be blank. Post-site visit analysis: The State removed the condition: "if a - e = 0, then f = 1" from program code.

Case file findings: There were 6 out of 25 (24%) records analyzed that were incorrect.

Frequencies: (n=12,335) Not applicable = 1,468; Yes = 553; No = 6,251; Unable to determine = 2,507 (20%) Not reported = 1,556

If an institutional setting is found in #41, then this element is set to "0." The Federal team recommends leaving this blank. Post-site visit analysis: The program code was modified by commenting out this line.

Initialized to spaces then reset to "3." This field should remain initialized to spaces. Post-site visit analysis: This line is commented out in the program code.

State codes "703, unable to determine" and "777, can't find" are mapped to AFCARS value "unable to determine." Code "777" must be mapped to blank.

State maps "none" to "unable to determine." "None" must be mapped to blank. Post-site visit analysis: "None" is mapped to blanks.

Case file findings: There were 5 out of 25 (20%) records analyzed that were incorrect.

Frequencies: (n=12,335) Unable to determine = 596; Two or more races = 81

This field is initialized to spaces then is reset to "0, does not apply." This field should remain initialized to spaces.

The State's code of "888, question not asked" is mapped to "unable to determine." This code should be mapped to blank. Post-site visit analysis: This code is no longer mapped in the program code.

The code "998, other race not listed" is mapped to "unable to determine." The State should remove this option and include actual other races that individuals identify. If an individual refuses, or does not identify themselves with one of the AFCARS race categories, then the State should have a value that represents "refused to provide race."

Remove from program code the condition: "if a - e = 0, then f = 1." If no race is selected in "a - e," then "f" should be blank.

Case file findings: There were 5 out of 25 (20%) records analyzed that were incorrect.

Post-site visit analysis: The State removed the condition: "if a - e = 0, then f = 1" from program code. If there are no foster parents this information will be left blank.

Frequencies: (n=12,335) Not applicable = 1,468; Yes = 342; No = 4,269; Unable to determine = 1,752 (14%); not reported = 4504

Initialized to spaces then reset to "3, unable to determine." This field should remain initialized to spaces. Post-site visit analysis: The line resetting this element to "unable to determine" is commented out.

State codes "703, unable to determine" and "777, can't find" are mapped to AFCARS value "unable to determine." Code "777" must be mapped to blank.

State maps "none" to "unable to determine." "None" must be mapped to blank. Post-site visit analysis: "None" is mapped to blank.

Case file findings: There were 5 out of 25 (20%) records analyzed that were incorrect.

Frequencies: (n=12,335) 2002 = 3,570; Not reported = 8,765

Workers must enter the date that children are discharged from the agency's responsibility for care, placement or supervision. See the General Requirements findings for children in their own home while under the agency's responsibility for care, placement or supervision and children placed in out-of-home care for respite purposes.

There is a discrepancy between the numbers reported in element #56 and #58.

Frequencies: (n=12,335) Not applicable = 8,769; Reunification = 2,311; Live w/other relative = 0; Adoption = 583; Emancipation = 195; Guardianship = 303; Transfer = 169; Runaway = 0; Death of child = 5

Workers must enter the date that children are discharged from the agency's responsibility for care, placement or supervision. See the General Requirements findings for children in their own home while under the agency's responsibility for care, placement or supervision and children placed in out-of-home care for respite purposes.

There is a discrepancy between the numbers reported in element #56 and #58.

Initialized to zero. Must be initialized to blank.

Post-site visit analysis: The program code has been modified to initialize to spaces.

Initialized to "0." Must be initialized to blank.

Post-site visit analysis: The program code has been modified to initialize to spaces.

State does not claim title IV-A for foster care maintenance payments. The response to this element should always be "does not apply."

The program code does include mapping that if it finds a person income code of "AFDC" this element will be set to "applies."

Initialized to "0." Post-site visit analysis: The program code has been modified to initialize to spaces.

There is no program code to look for title IV-D and there is no interface to the Child Support Enforcement system. This is currently being addressed in the State's Statewide Automated Child Welfare Information System (SACWIS) Assessment Review action plan.

Frequencies: (n = 12,335) Does not apply = 10,057; Applies = 2,278

Initialized to zero. Must be initialized to blank. Post-site visit analysis: The program code has been modified to initialize to spaces.

Currently, CAMIS is not interfaced with the Medicaid system (ACES). The work to interface with ACES is in progress. The screen that the Federal funding unit uses has a field for Medicaid eligible. The AFCARS program code checks this file, but without the interface this information is not being automatically populated. The Federal funding unit may enter information manually. The State team believes that the number of records reported as "applies" is an undercount and reflects data that the Federal funding unit staff has entered into CAMIS.

Initialized to "0."

There is no program code written to extract this data.

Post-site visit analysis: The State's revised program code (dated July 29, 2003) reflects the addition of this element. However, based on the Federal team's review of the program code and discussions with the State, there needs to be additional modifications. The State is to report the amount of the monthly foster care payment (regardless of source, i.e., Federal, State, county, municipality, tribal, and private payments). If title IV-E is paid on behalf of the child, the amount indicated should be the total computable amount. If the payment made on behalf of the child is not the same each month, indicate the amount of the last full monthly payment made during the reporting period. If no monthly payment has been made during the period, enter all zeros. This amount should include all payments made on behalf of the child during the month. If no payment was made, use zeros.

Frequencies: (n=530) Yes = 530

Initialized to "yes." The agency is only reporting adoptions of children from its foster care system or from private agency adoptions where an adoption agreement is in place.

There is a question on the "Adopted" screen with responses of "yes" or "no." However, this information was found to be inaccurate because workers were responding "no" for private agencies under contract with the State agency. Therefore, the program code does not extract information from this field.

Frequencies: (n= 530) American Indian/Alaska Native = 46; Asian = 6; Black/African American = 82; Native Hawaiian/Pacific Islander = 4; White = 403; Unable to determine = 22; Two or more Races = 32

This field is initialized to spaces, but then is reset to "0, does not apply." This field should remain initialized to spaces.

The State's code of "888, question not asked" is mapped to "unable to determine." This code should be mapped to blank.

The code "998, other race not listed" is mapped to "unable to determine." The State should remove this option and include actual other races that individuals are identifying with. If an individual refuses to identify, or does not identify with one of the AFCARS race categories, then the State should have a value that represents "refused to provide race."

Remove from program code the condition: "if a - e = 0, then f = 1." If no race is selected in "a - e," then "f" should be blank.

Post-site visit analysis: The State removed the initialization to zeroes.

Frequencies: (n= 530) Yes = 71; No = 378; Unable to determine = 81

This field is initialized to spaces then reset to "3, unable to determine." This field should remain initialized to spaces.

State codes "703/unable to determine" and "777/can't find" are mapped to AFCARS value "unable to determine." Code "777" must be mapped to blank.

State maps "none" to "unable to determine." "None" must be mapped to blank.

Frequencies: (n= 530) Yes = 370; No = 160

This element is derived from adoption element #10.

Initialized to "no."

See notes in adoption element #10.

Case file findings: There were 3 out of 18 (17%) records analyzed that were incorrect. In the three cases that did not match, the AFCARS data indicates "no" and the reviewer indicated the response is "yes."

Frequencies: (n=530) Not applicable = 160; Race = 19; Age = 113; Sibling group = 6; Medical/Mental Condition = 232; other = 0

Initialized to "zero." This needs to be initialized to blank.

The program code selects a primary basis for AFCARS reporting. The program code extracts the first condition that is selected by the worker. The State needs to redesign the collection of this information in order to specifically designate a primary basis for special needs.

"Other medical" is not listed as an option on the screen.

State needs to ensure that "at risk" conditions are mapped to "other."

Case file findings: There were 3 out of 18 (17%) records analyzed that were incorrect.

Frequencies: (n= 530) Does not apply = 526; Applies = 4

Initialized to "zero." This needs to be initialized to blank.

Post-site visit analysis: The program code was modified to initialize to spaces.

Frequencies: (n= 530) Does not apply = 530; Applies = 0

Initialized to "zero." This needs to be initialized to blank.

There is an error in the program code and this information is not getting extracted.

Post-site visit analysis: The program code was modified to initialize to spaces.

Frequencies: (n= 530) Does not apply = 497; Applies = 33

Initialized to "zero." This needs to be initialized to blank.

Post-site visit analysis: The program code was modified to initialize to spaces.

Frequencies: (n= 530) Does not apply = 331; Applies = 199

Initialized to "zero." This needs to be initialized to blank.

Post-site visit analysis: The program code was modified to initialize to spaces.

Frequencies: (n= 530) Does not apply = 530; Applies = 0

Initialized to "zero." This needs to be initialized to blank.

This is never assessed in the program code and is not an option on the input screen. Program code needs to be added to extract and map appropriately other diagnosed conditions based on the new screen.

The program code includes a routine that if the response to this element is "medical conditions or mental, physical or emotional disabilities," and there are zeroes in elements 11-14, then the response to element #15 is "applies."

Post-site visit analysis: The program code was modified to initialize to spaces.

Frequencies: (n= 530) Not reported = 232

Upon further analysis during the Post-site visit analysis it was determined the rating factor given during the on-site visit was incorrect. However, based on the case file analysis it was determined that there needs to be further attention given to the accuracy of this data and workers need to enter the information into the system.

Case file findings: There were 11 out of 22 (50%) records analyzed that were incorrect. Reviewers noted that this information could be found on the child's birth certificate.

Frequencies: (n=530) Unable to determine = 530

Initialized to "3, unable to determine."

Currently, the State does not collect this information and there is no program code to extract this information. This is being addressed with the new system design changes.

Case file findings: There were 9 out of 9 (100%) records analyzed that were incorrect. In the records analyzed, the reviewers were able to find the information in the case file.

Post-site visit analysis: The State submitted a screen print with this question added to it. This approach is adequate and meets the reporting requirement as a short term solution. However, the question is on an adoption screen and should be on a screen that collects information closer to the time the child is removed from his/her home. The State should add this to the changes in the "placement/legal" screen.

State may consider removing the program code that is no longer needed regarding "legally free" status.

Based on the case file findings, the State should assess the cause for the inaccurate data and make appropriate corrections.

Case file findings: There were 3 out of 24 (13%) records analyzed that were incorrect.

Post-site visit analysis: The State removed the program code regarding "legally free" status.

State may consider removing the program code that is no longer needed regarding "legally free" status.

Based on the case file findings, the State should assess the cause for the inaccurate data and make appropriate corrections.

Case file findings: There were 3 out of 23 (13%) records analyzed that were incorrect.

Post-site visit analysis: The State removed the program code regarding "legally free" status.

Line 5110 in the program code needs to be moved to an earlier location in the program code. For children that have two finalization dates, due to a dissolved adoption, the program code currently picks up the older date and not the new legalization date. If the code is moved up in the logic, it will work correctly.

Post-site visit analysis: This line has been moved to line 1750.

Frequencies: (n= 530) Married Couple = 371; Unmarried couple = 17; Single female = 129; Single male = 9; Not reported = 4

Individuals that are "separated" must be mapped to "married."

State should map same sex couples to unmarried couple.

1st loop looks in the adoption support record. If one is not found then it goes to the placement setting, assuming that it is the adoption family. Missing data are probably kids in group homes. Needs to be corrected.

Frequencies: (n=530) American Indian/Alaska Native = 41; Asian = 6; Black/African American = 58; Native Hawaiian/Pacific Islander = 3; White = 403; Unable to determine = 33; Two or more races = 26

This field is initialized to zeroes. This field should be initialized to spaces. The State's code of "888, question not asked" is mapped to "unable to determine." This code should be mapped to blank.

The code "998, other race not listed" is mapped to "unable to determine." The State should remove this option and include actual other races that individuals identify. If an individual refuses to identify, or does not identify with one of the AFCARS race categories, then the State should have a value that represents "refused to provide race."

Remove from program code the condition: "if a - e = 0, then f = 1." If no race is selected in "a - e," then "f" should be blank.

The program code checks for an "f" for female and maps the data accordingly. All other information is mapped to the elements for the father's information.

There should never be missing data for this element.

Frequencies: (n=530) Not applicable = 13; Yes = 71; No = 366; Unable to determine = 80

Initialized to zero then reset to "2, no." This field should be initialized to spaces.

State codes "703, unable to determine" and "777, can't find" are mapped to AFCARS value "unable to determine." Code "777" must be mapped to blank.

State maps "none" to "unable to determine." "None" must be mapped to blank.

The program code checks for an "f" for female and maps the data accordingly. All other information is mapped to the elements for the father's information.

Frequencies: (n=530) American Indian/Alaska Native = 32; Asian = 6; Black/African American = 39; Native Hawaiian/Pacific Islander = 3; White = 336; Unable to determine = 35; Two or more races = 21

This field is initialized to zero. This field should be initialized to spaces.

The State's code of "888, question not asked" is mapped to "unable to determine." This code should be mapped to blank.

The code "998, other race not listed" is mapped to "unable to determine." The State should remove this option and include actual other races that individuals identify. If an individual refuses, or does not identify themselves with one of the AFCARS race categories, then the State should have a value that represents "refused to provide race."

Remove from program code the condition: "if a - e = 0, then f = 1." If no race is selected in a - e, then f should be blank.

The program code checks for an "f" for female and maps the data accordingly. All other information is mapped to the elements for the father's information. This will not pick up data on same sex couples that adopt children.

There should never be missing data for this element.

Frequencies: (n=530) Not applicable = 100; Yes = 59; No = 306; Unable to determine = 65

Initialized to zero then reset to "3." This field should remain initialized to spaces.

State codes "703, unable to determine" and "777, can't find" are mapped to AFCARS value "unable to determine." Code "777" must be mapped to blank.

State maps "none" to "unable to determine." "None" must be mapped to blank.

The program code checks for an "f" for female and maps the data accordingly. All other information is mapped to the elements for the father's information. This will not pick up data on same sex couples that adopt children.

There should never be missing data for this element.

Line 4080 in the program code checks for birth mom or dad. State does not know why this is in the code. This needs to be removed. This may explain why there is a zero response for relative adoption.

State needs to review and revise the program code for these elements and make modifications.

Post-site visit analysis: The State submitted revised screen prints with this information added as a field on the screen that is mandatory. The screen allows for a single selection only and must be modified to allow worker to enter all that apply.

Frequencies: (n=530) Applies = 203; does not apply = 327

Initialized to "zero." Post-site visit analysis: The State removed the initialization to zero.

Derived from 29 - 31, if all are zero, then element #32 is "applies" (8310). This should not be derived. State needs to ensure that information for elements #29 - 32 are each individually collected.

Case file findings: There were 4 out of 21 (19%) records analyzed that were incorrect. See the note in adoption element #30.

Frequencies: (n=530) Yes = 520; No = 10