Aged Care Assessment Program National Data Repository

Chapter 5: Potential for improving MDS v2

Up to Aged Care Assessment Teams

prev page| TOC |next page

Chapter 5: Potential for improving MDS v2

This second Annual Report by the National Data Repository provides an opportunity to review the content of the MDS v2 and to comment on how well the data are being collected. There are three main issues: what is missing in the MDS v2 framework, especially in comparison to MDS v1: missing data and internal consistency.

Section 11: What is missing in the MDS v2 framework?

11.1 Continence disability

In MDS v1, the level of disability of ACAT clients was indicated in three key areas: mobility, continence and orientation. The presence or absence of each disability was recorded at the end of the assessment process when a long-term care plan was being developed for the client. An individual client could be recorded dependent in one or more of these areas. The percentages recorded as disabled in 2002-2003 were 36.8% for mobility, 37.2% for continence, and 29.2% for orientation (Lincoln Centre for Ageing and Community Care Research, 2004).

The items intended to replace disability in the MDS v2 were:

• For Mobility disability, the Activity limitation of movement (i.e., transfers) and moving around and the health conditions;
• For Orientation disability, the Health conditions dementia, confusion and memory loss; and
• For Continence disability, the Health conditions urinary and faecal incontinence.

Whether MDS v2 items are equivalent to MDS v1 indications can be measured by comparing the prevalence of these conditions within the data set.

Table 47: Comparison of MDS v1 and MDS v2 indications of disability

MDS v1		MDS v2 equivalent
% (2002-2003)		% 2003-2004	% 2003-2004	% 2004-2005	% 2005-2006
Mobility disability	36.8	Needs help with Movement or Moving around	51.8	53.7	49.6
Continence disability	37.2	Health condition Bowel or urinary incontinence (1402, 1707, 1708)	5.7	4.1	7.4
Orientation disability	29.2	Health condition Dementia (0500-0532) or 1716 (Disorientation)	28.9	28.5	29.6

Top of PageNote: MDS v2 records, Complete assessments only.

It is clear that Orientation disability is well matched by the Health conditions dementia and disorientation, but Mobility disability and Continence disability have no real equivalent in MDS v2. Needing help with Movement or Moving around is far more inclusive than Mobility disability was, which is puzzling but not disastrous.

More concerning, the proportion of clients with a diagnosis or symptom of incontinence in MDS v2 is far lower than would be expected. Using Health condition to record incontinence in MDS v2 has not been successful.

It is important that the issue of recording incontinence adequately in the MDS be resolved, since incontinence is a critical factor in precipitating a need for residential care (Lincoln Centre for Ageing and Community Care Research, 2004).

11.2 On-going involvement

On-going involvement was indicated in MDS v1 by an item that reflected the broader role assigned to assessment teams in the Commonwealth Government Guidelines for Assessment Services: in particular the requirement to “facilitate the provision of appropriate services to meet assessed needs and follow up clients to monitor their satisfaction with support/services provided”. ACAP staff recorded:
monitoring that they intended to do, either directly or through another agency to ascertain whether or not the care plan had been implemented;
active assistance, which includes brokerage between the client and community services, residential care facilities; and
direct service provision by the team; for example, counselling or equipment demonstrations.

Nationally in 2002-2003, 46.5% of clients received no further support and 27.4% received monitoring only. Active assistance by the ACAT was to be provided for 21.7% of clients.

On-going involvement was not included since the 2003-2004 Annual Report was written, plans have been developed to record on-going involvement with clients using two new data items to be included in the MDS: Care coordination in MDS v2. There were concerns about how accurately the item was completed in MDS v1 and difficulty in coming to an agreement on how such an item should be coded. Further, for an item that could only record intentions once an assessment had been completed, it could never accurately indicate what actually happened to clients after the assessment.

In 2005-2006, Care coordination received and Care coordination closure date were introduced to the MDS. Care coordination is the term used describe those activities of ACATs associated with organising and monitoring services for clients that should be reported separately from assessment. It is defined as activity which assists the client to implement the care plan, and is coded by level. Level 1 care coordination is follow-up and monitoring, while Level 2 is intensive coordination. Clients can be provided with either, both, or none.

Data on care coordination activity were downloaded to the NDR from some jurisdictions in 2005-2006, and these figures are provided in the tables in Appendix A and explored below.

Three jurisdictions (Victoria, Western Australia and Tasmania) were downloading care coordination data by Quarter 4, 2005-2006. There were difficulties in Victoria and in Western Australia in fully complying with the guidelines for recording and downloading the data, however, so the following exploration of the data should be treated with some caution.

Extent of care coordination is explored in the following table. Care coordination was determined to have occurred if Level 1 and/or Level 2 were recorded. Cases were selected if there were valid values for Care coordination level (i.e., includes None but not Missing).

Top of Page

Table 48: Extent of care coordination activity, by jurisdiction, Q4 2005-2006

	Level 1 and/or Level 2 N	Level 1 and/or Level 2 %
Vic	1,389	11.5
WA	1,375	38.8
Tas	69	5.3
All included jurisdictions	2,833	16.7

Overall, about 17% of clients were recorded as having received some care coordination from the ACAP. Western Australian teams provided care coordination to a much higher proportion of clients than this average, while Victorian and Tasmanian teams provided care coordination to a lower proportion. The following table provides more information on the levels of care coordination provided.

Table 49: Level of care coordination activity, by jurisdiction, Q4 2005-2006

	Level 1	Level 2	Level 1 & 2	None	N
Vic	8.3	1.9	1.3	88.5	12,083
WA	27.6	5.5	5.6	61.2	3,541
Tas	3.7	1.1	0.5	94.7	1,308
All included jurisdictions	12.0	2.6	2.2	83.3	16,6932

In all jurisdictions, the majority of clients received no care coordination. Western Australia was the only jurisdiction where a substantial proportion of clients was recorded as receiving care coordination. Most of this was Level 1 care coordination. In Victoria and Western Australia, Level 1 care coordination was about 3 times more common than Level 2 care coordination; in Tasmania, this ratio was more like 2.5 times.

Duration of the care coordination episode can also be reported. At this stage, estimates of the duration of care coordination are likely to be under-estimates.

Table 50: Duration of care coordination episode, by jurisdiction, Q4 2005-2006

	mean	median	90th percentile	N
Vic	15.6	6	45	1,477
WA	16.9	10	43	1,609
Tas	28.6	20	70	79
All included jurisdictions	16.6	6	45	3,165

The average duration of care coordination was much longer in Tasmania than in the other two jurisdictions.

Of interest is the question of when care coordination is more likely to occur. The following series of tables addresses this question.

Top of Page

Table 51: Intensity of care coordination by characteristics of the assessment

	Level 1	Level 2	Level 1 & 2	None	N
Complete/Incomplete assessment
Incomplete	4.1	0.9	0.5	94.6	1,998
Complete	13.1	2.8	2.4	81.7	14,927
Location of assessment
Hospital	6.9	1.4	2.1	89.5	1,321
Other inpatient setting	7.3	2.0	1.9	88.9	963
Residential aged care service	11.2	3.0	2.0	83.8	10,228
Other	15.5	3.3	2.5	78.7	1,514
Not stated	2.7	.4	0.2	96.7	16,932
Total	12.0	2.6	2.2	83.3	1,321

Care coordination was more likely to be recorded for complete assessments than for incomplete assessments and for assessments that took place in community settings rather than in hospitals. Surprisingly, assessment activity in residential care settings also resulted in a relatively high degree of care coordination activity. Such clients include those in respite as well as clients living in a residential care setting at assessment.

Table 52: Intensity of care coordination by Accommodation setting—usual and Recommended long-term care setting

	Level 1	Level 2	Level 1 & 2	None	N
Accommodation setting—usual
Living in Community	13.0	2.8	2.4	81.8	13,301
Living in Residential care	9.1	1.9	1.7	87.3	1,270
Living in Other/Missing	4.4	.5	.4	94.7	1,209
Total	12.0	2.6	2.1	83.2	15,780
Recommended long-term care setting
Recommended to community	15.0	3.0	2.3	79.7	8,918
Recommended to residential care	10.3	2.4	2.5	84.8	5,987
Total	13.1	2.8	2.4	81.7	14,905

Assessments were more likely to be followed by care coordination activity if Clients living in community settings rather than in residential care. However, a surprisingly high proportion of people living in residential care at assessment were also provided with care coordination.

Care coordination was more likely to be recorded for clients recommended to the community than for those recommended to residential care. However, a surprisingly high proportion of clients recommended to residential care also received care coordination.

In examining the impact of client characteristics on care coordination activity subsequent to the assessment, it is worth selecting clients who live in the community at assessment (since these characteristics are likely to be confounded with Accommodation-setting—usual).

Top of Page

Table 53: Intensity of care coordination by client characteristics, clients living in the community at assessment

	Level 1	Level 2	Level 1 & 2	None	N
Carer availability and co-residency
Co-resident carer	13.5	3.1	2.3	81.1	5,554
Non-resident carer	13.2	2.7	2.4	81.6	5,066
No carer	11.7	2.7	2.3	83.3	2,504
Total	13.1	2.9	2.3	81.7	13,124
Dementia diagnosis
No dementia diagnosis	12.8	2.6	2.1	82.4	10,163
Dementia diagnosis	13.6	3.5	3.1	79.8	3,138
Total	13.0	2.8	2.4	81.8	13,301
Dependency
Low dependency	11.8	3.0	1.6	83.6	3,073
Medium dependency	14.7	3.0	2.8	79.5	5,829
High dependency	12.6	2.8	2.4	82.2	3,932
Total	13.4	2.9	2.4	81.3	12,834

There was little difference in care coordination activity by carer availability. This is surprising, as it was expected that clients without carers would be more likely to require care coordination than clients with carers. However, clients with no carers were slightly less likely to be provided with care coordination that clients with carers.

Clients with dementia were somewhat more likely to be given care coordination than clients without. However, this difference was slight. Similarly, Care coordination was not closely related to client dependency. Clients in the medium range were slightly more likely than clients at either extreme to be given care coordination.

Conclusion

The care coordination items have the potential to yield useful information on the care coordination activity of ACATs. However, only 3 of the 8 jurisdictions downloaded data on care coordination activity in Quarter 4 2005-2006, and some of this data looked unreliable. There is still some way to go before comprehensive reports on care coordination activity can be provided.

Given the data that has been provided to the NDR, a minority of clients (probably about 16%) receives care coordination in the ACAP. Most of this care coordination was low intensity (Level 1) and of relatively short duration (median 8 days).

We look forward to being able to report more extensively on these new data items in the 2006-2007 Annual Report.

11.3 Other matters

In the two previous reports, brief attention was given to recording reassessment and to draft items (e.g., Language spoken). There has been no change in recording either of these, except at the jurisdictional level, and they are not included in the National MDS.

Summary and Discussion: Potential for improving the MDS

Disability

Orientation disability is fairly well matched by the Health conditions dementia and disorientation, but Mobility disability and Continence disability have no real equivalent in MDS v2. Needing help with Movement or Moving around is far more inclusive than Mobility disability was.

In contrast, the proportion of clients with a diagnosis or symptom of incontinence in MDS v2 is far lower than would be expected. Feedback from some teams indicates that some staff members believed that incontinence recorded under ACCR items in Part 5 of the form would be reflected in the MDS, which is not true. Using Health condition to record incontinence in MDS v2 has not been successful.

Care coordination

Care coordination received and Care coordination closure date have been included in the ACAP MDS to replace the MDS v1 item Ongoing involvement. These two new items were downloaded by only 3 jurisdictions in Quarter 4 2005-2006. They have the potential to yield useful information.

Section 12: Data quality in MDS v2

12.1 Validation reports

One way of assessing data quality is to examine the proportion of records that pass validation and the number of errors and warnings as a ratio of the total number of records. The following table summarises the validation reports by jurisdiction for Q4 2003-2004, 2004-2005, and 2005-2006. (The Index for validation warnings includes only type 1 validation warnings; that is, those that do not reflect care plans interim to entering residential care.) It is desirable for the percentage of validated records to increase but for the Indices of errors and warnings to decrease over time. However, increases in the Index for validation warnings are caused partly by increasing the number of variables in the data set and of the number of validation criteria.

Top of Page

Table 54: Comparison of validation reports by jurisdiction for Q4 in 2003-2004 and 2004-2005

		Q4 2003-2004	Q4 2004-2005	Q4 2005-2006
New South Wales	% of records validated	67.8	58.0	56.8
	Index for errors	2.3	0.8	0.6
	Index for validation warnings	92.6	98.4	100.8
Victoria	% of records validated	70.5	89.4	94.9
	Index for errors	6.5	0.2	0.4
	Index for validation warnings	24.8	12.2	4.9
Queensland	% of records validated	NA	NA	73.8
	Index for errors	NA	NA	0.0
	Index for validation warnings	NA	NA	0.0
South Australia	% of records validated	14.5	23.4	35.7
	Index for errors	14.6	18.9	4.5
	Index for validation warnings	194.8	123.9	98.5
Western Australia	% of records validated	17.5	7.5	72.8
	Index for errors	0.0	0.0	0.0
	Index for validation warnings	345.1	519.3	58.9
Tasmania	% of records validated	96.7	97.4	99.3
	Index for errors	17.8	16.4	0.0
	Index for validation warnings	39.3	42.7	0.6
Northern Territory	% of records validated	50.3	54.5	61.7
	Index for errors	17.8	16.4	10.6
	Index for validation warnings	39.3	42.7	10.6
ACT	% of records validated	50.5	91.0	52.1
	Index for errors	56.8	0.3	0.0
	Index for validation warnings	4.4	75.8	76.8
		Q4 2003-2004	Q4 2004-2005	Q4 2005-2006
New South Wales	% of records validated	67.8	58.0	56.8
	Index for errors	2.3	0.8	0.6

The data presented in this table show considerable improvement for data from most jurisdictions. However, New South Wales is an exception to this trend. Data from the ACT have improved, but ironically the inclusion of more variables in the MDS has led to a larger rate for validation warnings. Queensland’s data are generally of a high quality. The data from Tasmania are excellent.

Top of Page

12.2 Missing data

Rates of missing data are compared below from final annual data sets in the years 2004-2005 and 2005-2006. An acceptable level of missing data in data sets such as this is commonly set at 1.0.
There was no missing data on Sex (only 31 cases in the entire data set).

Table 55: Missing data on items required for all records (MDS v1 and v2)

	2004-2005		2004-2005
	Age (from date of birth) %	Assessment end date %	Age (from date of birth) %	Assessment end date %
NSW	0.1	0.0	NSW	0.1
Vic	0.0	0.0	Vic	0.0
Qld	0.0	0.0	Qld	0.0
SA	0.0	0.0	SA	0.0
WA	1.5	0.0	WA	1.5
Tas	0.0	0.0	Tas	0.0
NT	0.0	0.0	NT	0.0
ACT	0.0	0.0	ACT	0.0
Australia	0.2	0.0	Aus	0.2

Missing end-date implies a date not in the quarter
Missing age implies Date of birth missing or invalid range (less than 5 or more than 110)

Table 56: Missing data on items that should be present for all Level 3 records (has first face-to-face contact date; MDS v1 and MDS v2)

	2004-2005		2004-2005
	Postcode %	Location of assessment %	Postcode %	Location of assessment %
NSW	0.9	1.4	NSW	0.9
Vic	2.7	0.5	Vic	2.7
Qld	8.8	0.5	Qld	8.8
SA	0.0	5.9	SA	0.0
WA	0.0	0.4	WA	0.0
Tas	0.1	0.1	Tas	0.1
NT	0.0	0.1	NT	0.0
ACT	1.0	0.0	ACT	1.0
Australia	2.5	1.1	Aus	2.5

Includes all records with a First face-to-face contact date

Top of Page

Table 57: Missing data on items required for all complete Assessments (MDS v1 and v2)

	2004-2005		2005-2006
	First face-to-face contact date %	Accommodation setting—usual %	First face-to-face contact date %	Accommodation setting—usual %
NSW	0.5	2.2	0.5	1.7
Vic	0.1	0.1	0.1	0.1
Qld	0.0	1.8	0.0	0.1
SA	28.2	2.6	16.6	1.7
WA	10.8	12.8	0.0	0.0
Tas	0.0	0.0	0.0	0.0
NT	0.0	0.5	0.0	0.3
ACT	0.0	42.9	0.0	46.8
Australia	3.4	3.2	1.7	1.5

Includes invalid values.
Table includes all complete assessments.

Table 58: Missing data on items required for all MDS v2 records

	2004-2005			2005-2006
	Letters of name %	Priority category %	Reason for ending assessment %	Letters of name %	Priority category %	Reason for ending assessment %
NSW	4.0	9.2	2.2	0.0	9.6	0.4
Vic	0.0	4.7	0.1	0.0	4.2	0.1
Qld				0.0	12.6	0.3
SA	0.0	11.5	0.0	0.0	13.0	0.0
WA	0.0	29.2	4.6	0.0	0.0	0.0
Tas	0.0	0.4	0.0	0.0	0.2	0.0
NT	0.0	0.4	0.1	0.0	0.0	0.5
ACT	0.0	1.0	0.2	0.0	0.6	0.4
Australia	1.5	10.0	1.5	0.0	7.5	0.2

Top of Page

Table 59: Missing data on items required for all Complete MDS v2 assessments

	2004-2005				2005-2006
	First inter-vention date %	Activity Limit-ation %	Assessor Pro-fession %	First Health condition %	First inter-vention date %	Activity Limit-ation %	Assessor Pro-fession %	First Health condition %
NSW	0.1	1.9	3.8	2.1	0.0	1.7	3.2	1.1
Vic	0.0	0.8	1.7	0.2	0.0	0.4	0.6	0.1
Qld					0.0	0.5	6.5	0.5
SA	27.9	0.2	1.6	2.2	15.9	0.1	1.6	2.1
WA	10.8	12.9	13.0	6.8	0.0	0.3	0.0	0.0
Tas	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
NT	3.1	0.0	0.0	1.0	0.0	0.0	0.2	0.3
ACT	0.1	1.7	2.4	0.0	0.0	0.3	2.7	0.0
Australia	3.9	2.5	3.8	1.9	1.5	0.9	2.3	0.7

Activity is missing if coded NA, UTD, or NS: Assessor profession is missing if coded NS: Health condition missing if HC1 not stated.
Table is based on complete MDS v2 assessments.

Table 60: Missing data on items required for all complete community-dwelling MDS v2 assessments

	2004-2005			2005-2006
	Living arrangement %	Carer availability %	Respite care use %	Living arrangement %	Carer availability %	Respite care use %
NSW	2.0	4.2	8.7	1.8	2.1	3.4
Vic	1.1	2.0	5.3	0.1	0.1	1.5
Qld				0.2	0.3	3.4
SA	6.9	10.5	20.5	2.9	2.8	3.3
WA	0.1	0.1	1.2	0.1	0.2	0.9
Tas	0.2	0.5	0.7	0.0	0.0	0.0
NT	0.9	3.8	81.6	0.0	0.2	68.6
ACT	9.1	4.5	0.1	9.5	1.5	0.0
Australia	2.0	3.5	7.9	1.1	1.1	2.8

Missing includes NA and NS.
Table is based on complete MDS v2 assessments.

Top of Page

Table 61: Missing data on items required for all complete community-dwelling MDS v2 assessments with any activity limitation

	2004-2005		2005-2006
	Current assistance with activities %	Source of current assistance with activities %	Current assistance with activities %	Source of current assistance with activities %
NSW	6.3	9.7	5.3	3.8
Vic	3.8	1.8	2.2	0.3
Qld			0.7	5.8
SA	1.1	1.4	0.2	0.0
WA	0.6	15.0	0.5	0.0
Tas	0.2	0.7	0.2	0.0
NT	1.8	4.4	2.1	2.4
ACT	0.0	77.9	0.0	78.4
Australia	4.1	7.4	2.7	3.0

Current assistance is selected for Complete, community-dwelling, and has an activity limitation. Missing includes NA, UTD and NS.
Source of current assistance is selected for Complete, community-dwelling, has an activity limitation, and has assistance on any of the items. Missing is coded NA or Not Stated/ Inadequately described for any of the items where it has been indicated that help is provided.
Table is based on complete MDS v2 assessments of clients living in the community at assessment.

Table 62: Missing data on items required for all Complete MDS v2 assessments recommended to community

	2004-2005		2005-2006
	Recommended Government services %	Recommended respite %	Recommended Government services %	Recommended respite %
NSW	6.1	8.1	3.6	4.7
Vic	3.1	4.0	1.8	2.1
Qld			2.4	2.1
SA	0.6	3.5	0.4	2.4
WA	19.3	5.8	3.3	6.7
Tas	0.6	0.8	0.5	0.5
NT	1.8	14.5	3.4	13.8
ACT	1.9	2.0	2.6	16.5
Australia	5.6	5.8	2.6	3.9

Missing if coded NA, UTD, or NS.
Table is based on complete MDS v2 assessments of clients recommended to the community.

Top of Page

Table 63: Missing data on items required for most Complete MDS v2 Assessments

	2004-2005		2005-2006
	Indigenous status %	DVA status %	Indigenous status %	DVA status %
NSW	5.3	13.7	3.4	8.0
Vic	1.8	8.6	1.5	2.3
Qld			5.6	14.8
SA	3.7	8.9	3.3	7.2
WA	1.6	4.9	0.2	0.5
Tas	0.1	0.2	0.2	0.1
NT	1.9	1.1	1.0	1.4
ACT	5.2	100.0	8.4	100.0
Australia	3.4	11.8	2.8	7.7

Table is based on Complete MDS v2 assessments.

Table 64: Missing data on items required for most complete Assessments (MDS v1 and v2)

	2004-2005	2005-2006
	Country of birth %	Country of birth %
NSW	4.2	2.9
Vic	3.4	3.6
Qld	2.9	2.1
SA	2.8	2.4
WA	0.7	0.2
Tas	0.0	0.0
NT	5.5	4.6
ACT	1.2	1.1
Australia	3.2	2.6

Table is based on Complete MDS v2 assessments.

Table 65: Missing data on Delegation dates

	2004-2005	2005-2006
	Missing/invalid delegation dates %	Missing/invalid delegation dates %
NSW	0.3	0.0
Vic	0.2	0.0
Qld		0.0
SA		0.0
WA	0.0	0.0
Tas	0.0	0.0
NT	8.1	3.0
ACT	7.8	8.7
Australia	0.3	0.1

South Australia has no information on Approvals and few records have Delegation dates. Hence it is not possible to determine the level of missing data on Delegation dates.

Delegation date is missing if the record contains Approvals but there is no Delegation date. Delegation date is invalid if it falls before the Assessment end date.

Top of Page

Table 66: Missing data on items required for Complete MDS v2 with carers, living in the community at assessment

	2004-2005		2005-2006
	Carer co-residence %	Carer relationship %	Carer co-residence %	Carer relationship %
NSW	0.6	2.6	0.7	4.0
Vic	0.3	2.1	0.2	0.7
Qld			0.6	0.4
SA	0.7	0.5	1.8	0.3
WA	0.0	0.1	0.1	0.2
Tas	0.1	0.1	0.1	0.2
NT	0.7	2.6	2.0	2.2
ACT	0.2	0.4	1.1	0.5
Australia	0.4	1.8	0.6	1.8

Overall, levels of missing data and invalid values have fallen over the past year. Some jurisdictions have particular problems with some items, whereas other items are problems in most jurisdictions (e.g., DVA status and Recommended respite). Problem items by jurisdiction are listed below, in Table 65. Some of these issues will become less problematic as business rules are progressively incorporated into data collection software.

Table 67: Items needing attention in each jurisdiction

Jurisdiction	Items with more than 10% missing or invalid data	Items with more than 1% missing data
NSW	None	Accommodation setting—usual; Priority category; Activity limitation; Assessor profession; First health condition; Living arrangement; Carer availability; Respite care use; Current assistance with activities; Source of current assistance; Recommended government services; Recommended respite; Indigenous status; DVA status; Country of birth; Carer relationship
Vic	None	Priority category; Respite care use; Current assistance with activities; Recommended government services; Recommended respite; Indigenous status; DVA status; Country of birth
Qld	Priority category; DVA status	Postcode; Assessor profession; Respite care use; Source of current assistance with activities; Recommended government services; Recommended respite; Indigenous status; Country of birth
SA	First face-to-face contact date; Priority category; First intervention date	Location of assessment; Accommodation setting—usual; Assessor profession; First health condition; Living arrangement; Carer availability; Respite care use; Recommended respite; Indigenous status; DVA status; Country of birth; Carer co residence
WA	None	Recommended government services; Recommended respite
Tas	None	None
NT	Respite care use; Recommended respite	Current assistance with activities; Source of current assistance; Recommended government services; Indigenous status; DVA status; Country of birth; Delegation date; Carer co residence; Carer relationship
ACT	Accommodation setting—usual; Assessor profession; Source of current assistance; Recommended respite; DVA status	Living arrangement; Carer availability; Recommended government services; Indigenous status; Country of birth; Delegation date; Carer co residence

Top of Page

12.3 Data inconsistencies and coding issues

13.3.1 Date sequencing

All jurisdictions except Tasmania and Western Australia have had some difficulties with date sequencing. Most of these errors are Delegation dates that precede the assessment end date. The business rule that Delegation date must occur on or after the End of Assessment date was introduced into the MDS during the second half of the 2003-2004 financial year. As with other errors, we expect the number of occurrences to decrease with time. The following table compares jurisdictions on invalid values for date sequences (MDS v2 data only). The percentages in this table include both values where the date precedes the previous point (e.g., First face-to-face contact date is before the First intervention date) and those where the interval is more than 364 days.

Table 68: Difficulties with date sequencing by jurisdiction (invalid %), 2005-2006

	NSW	Vic	Qld	SA	WA	Tas	NT	ACT	Total
R to FI date	0.2	0.0	0.0	0.4	0.0	0.0	2.8	0.1	0.1
FI date to FTF Date	0.0	0.0	0.0	0.0	0.0	0.0	1.7	0.0	0.0
FTF date to End date	0.3	0.0	0.0	0.4	0.0	0.0	1.3	0.0	0.1
End date to Delegation date	0.0	0.1	0.1	9.1	0.0	0.0	2.7	0.0	0.6

Note: MDS v2 records, not including missing values.

There has been an overall improvement in the quality of data in each of the two preceding years. The jurisdictions with the worst date sequencing problems were South Australia and the Northern Territory. Western Australia and Tasmania recorded no date sequencing errors.

Summary and Discussion: Data quality in MDS v2

On the whole, the quality of the ACAP MDS has improved in each year since 2003-2004. Levels of missing data improved on many data items. Tasmania’s data were consistently of a very high standard. There was room for improvement in all other jurisdictions.

Data sequencing was a large problem only in South Australia and the Northern Territory. However, for all jurisdictions except Western Australia and Tasmania, the date validation criteria had to be turned off for every data set submitted to the NDR in 2005-2006. The date sequence that caused the most problems was Delegation date preceding the Assessment end date.

It is expected that with software developments currently being put into place to minimise errors, the quality of the data in MDS v2 will improve during 2006-2007.

Top of Page

prev page| TOC |next page