DQ – Disabling Condition Not Functioning Correctly Due to Specification Conflict

[csmith-caseworthy]

The specs for setting tlsa_Enrollment.DisabilityStatus in 3.4, never indicates a value of 99 but it has NULL to be set for any value other than 0 or 1. Both 10.12 & 11.7 spec states to look for DisabilityStatus value of 99 and is programmed as such, not NULL. This is leaving DQ counts for Disabling Condition at 0 when there are DQ issues.

3.4. HMIS Client Enrollments (tlsa_Enrollment)

DisabilityStatus

Because it is relevant and used repeatedly in subsequent steps both for demographic reporting and for identification of people and households who are part of specific populations of interest (e.g, Households with a Disabled Adult or Head of Household) , a preliminary enrollment-level value is included in tlsa_Enrollment.

Enrollment DisablingCondition Value	DisabilityStatus
0	0
1	1
(any other)	NULL

10.12. DQ – Disabling Condition

This is a subset of ClientEntry (the Value for ReportRow 910). The Value for ReportRow 913 is a count of distinct EnrollmentIDs in tlsa_Enrollment for ProjectIDs in lsa_Project where where AIR = 1 or (LSAScope <> 3 and Active = 1) and:

• DisabilityStatus = 99.

insert into lsa_Calculated (Value, Cohort, Universe, HHType , Population, SystemPath, ReportRow, ProjectID, ReportID, Step) select count (distinct n.EnrollmentID), 1, 10, 0, 0, -1, 913, n.ProjectID, rpt.ReportID, '10.12' from lsa_Report rpt inner join tlsa_Enrollment n on n.EntryDate <= rpt.ReportEnd where (n.AIR = 1 or (rpt.LSAScope <> 3 and n.Active = 1)) and n.DisabilityStatus = 99 group by n.ProjectID, rpt.ReportID

11.7. Data Quality: Enrollment Data

DisablingCond This is a subset of ClientEntry. A count of distinct EnrollmentIDs in tlsa_Enrollment where where AIR = 1 or (Active = 1 and LSAScope <> 3) and:

• DisabilityStatus = 99.

update rpt set rpt.DisablingCond = (select count(distinct n.EnrollmentID) from tlsa_Enrollment n where (n.AIR = 1 or (n.Active = 1 and rpt.LSAScope <> 3)) and n.DisabilityStatus = 99 ) from lsa_Report rpt

[candiceh214]

We are getting CoCs asking about this as well. Any update? Seems that CoCs were told that CDK and CPNTA would not be treated as missing but we only have 0, 1, and 99 in the specs.

[MollyMcEvilley]

The current status on this is that the threshold has been raised from 5% to 10%, which brings down the number of CoCs from "more than half" to 184 -- which is still very high. RJ is communicating with the HDX analysis team re: next steps. One possibility is delaying implementation of this flag until next year with an update to the specs to distinguish between CDK/CPTNA and Data Not Collected for LSAPerson.DisabilityStatus.

I'm leaving this open for now as a reminder and will update as more info becomes available.