Disability

thanks Tom - I realise it’s a big ask - but I’m equally sure that if there was an easy answer for the huge increase in reported visual impairment in the 2013 post-census disability survey, you would have found it.

Hi, my name is Jo and I am a researcher on health of older people. My concerns coincide with Judith's above - too often results are reported that group all older people (65+) together. There is such a huge difference between those aged say 65-74 and those aged 95+, or even 85+.
I ask that results are more often presented with age categories that split the 65+s - and I understand that may mean surveying more older people to get reliable enough estimates for each age group.
Population ageing in the coming decades means it will become will be increasingly important to understand disability, to estimate how it impacts on living. For example, how do we alter housing design to enable more to stay at home rather than have to move into care? How can we support people during periods of ill-health or other needs, eg as people with acute illness on top of long-term disability transition from hospital back home & recover? Understanding better how widespread are different conditions can inform policies and practices.
I see the census information as informing both the sampling for the Disability Survey and presumably weighting survey results back to represent the population. But I too, along with Jennifer, would like to see the summaries of the original census questions published, together with their equivalents in the Disability Surveys, so we can see ourselves how well they relate.

Hi @lynleyhood, some additional explanation of the changes that occurred across surveys are below.

It also gives some rationale behind the limitations in outputting responses to the sample-identifying disability questions in the census.

The changes that were made to disability questions for the 2013 Census will have had no effect on the size of the disabled population identified in the 2013 Disability Survey.

The disability questions included in the Census helped us to find the sample for the Survey. They were used to ensure that we found enough disabled people for robust survey estimates.

When the survey sample was taken, people were selected from those who indicated disability in Census and also from those who did not. All of the selected people were then asked the survey questions to establish their final disability status. Their Census answers were not used for this so the ordering of the Census questions will not have had any effect on the final population identified.

Some people who didn’t indicate disability in the Census were found to be disabled in the Survey where we asked a much more detailed set of questions. Some people who did indicate disability in the Census were found not to be disabled in the Survey.

Our current view is that the questions used in the 2013 Census did not identify the disabled population very well. A large number of people turned out to be disabled in the Survey but not in the Census. We are also concerned about how well the Census questions worked at helping us find disabled people Survey.

Issues that arose with the 2006 Disability Survey findings are different. The fall in national disability prevalence between 2001 and 2006 was investigated thoroughly at the time. No single major factor was found but a number of possible factors were identified that might have contributed to the 2006 result. This meant that we advised caution in interpreting changes between 2006 and other years. Our statement at the time can be found under the heading Comment on the apparent decline in the disability rate on the webpage: http://www.stats.govt.nz/browse_for_stats/health/disabilities/DisabilitySurvey2006_HOTP06/Technical%20Notes.aspx

The challenges we face in providing quality data about disability are also faced by statistical agencies in other countries. This year we are engaging in a consultation process that will help us to clarify customer needs and their concerns about disability data. This process will inform future changes that we might make to improve disability statistics.

Thanks Digby - I've read the statement on the apparent decline in the disability rate in 2006. As I see it, if "a number of possible factors" contributed to the 2006 result (which seems reasonable to me), surely the same could be said of the result of all the other disability surveys, and, if so, shouldn't the same advice for caution apply to them all? Shouldn't we be equally cautious about comparing the 2001 and 2013 surveys? Okay I realise this notion erodes the usefulness of the surveys - but I guess that relates to the challenge of providing quality data you refer to.
I appreciate the difficulty you face. My concern is that, in the eyes of many decision makers the census content sets the agenda for what matters in NZ. If the disability questions are taken out of the census before anything is in place to replace them, disabled people risk becoming even more marginalised than they are already.

Has it been made clear yet whether a disability survey is in fact to be held in 2018. Th recomendation in the report on projected content was ambivilant

Kia ora, I think smoking behavior is one indicator of population health but I think there are other indicators which need to be understood at the granularity which the Census offers, and which are as or possibly more relevant to todays communities. I believe there is the case for the addition of a physical activity level question which would ascertain the different percentages of people meeting physical activity guidelines in our different communities. This would greatly enhance our ability to design and target intervention. Like smoking, but unlike nutrition, a question around physical activity can be simply asked (although there are issues with over-estimates in self reporting of physical activity.

Hi Tom,

Not sure if my request is particularly applicable to disability but it could be under the Health topic.

Knowing the proportion of individuals that are usually under some form of care, and that type of care, would be very valuable for Civil Defence Emergency Management (CDEM) planning. When planning for public alerting and evacuations knowing what proportions of the population are under the care of another person or institution and also the proportion of the population that are vulnerable (e.g. the elderly) but are still living unsupported also helps us prioritise where our efforts should be focussed both in terms of helping to building community resilience but also in responding to an emergency and recovery.

Thanks
Derek

Bit disappointed to have no answer to my questions? Any help you can give or feedback greatly appreciated.

Apologies for not getting back sooner @chris20

Whether we hold a Disability Survey in 2018 is under review.

Stats NZ and the Office for Disability Issues have established the Disability Data and Evidence Working Group. This group will help us to clarify customer needs and inform future changes that we might make to improve disability statistics. This will include looking at how we measure disability across all our household statistics, including Census.

Click here for more information.

All interesting discussion. particularly the thought that the Disability Surveys may not have measured the population consistently or reliably. That does rather defeat the purpose and I can understand why the case for running them again is debatable.

In older people, arguably the single biggest predictor of all sorts of health-related outcomes is self-perceived health. This is very easily measured, eg a 5-point scale excellent-very good-good-poor-very poor. It is not a measure of disability, but I wonder whether it may be useful for sample selection for the more detailed Disability Survey - in conjunction with measures of function if some can be found that do reliably indicate disability? A simple measure of health in itself, at a small-area level, could indicate areas of unmet or poorly-met need, for use in service planning or health promotion, and be able to describe groups by age and living situation (in care, living alone etc) too.

For us researching health of older people, variations in health service use by location may indicate that there is differential provision or access to services. But it may be that actual health needs differ and these accounts for differences in utilisation. We can make comparisons e.g. by analysing DHB-specific hospitalisation rates for particular conditions or the proportion in long-term residential care, but at present we have no idea what the health of the population is, except through the Health and Disability surveys which cannot give detail by location.
So my suggestion is that an additional question, self-rated health, be included in the census. I am not suggesting that we revert to a medical model of health alone, but that that model still offers information that is useful.

I know, Tom, that this thread is not about health as such, but is about disability. I don't see a group for health so I've popped it in here! Without any evidence to support my argument, I do think it would be helpful for sample selection as well as function/disability/limitations themselves.

As an aside (and rather tentatively) I wonder whether the question about function/disability/limitations need to include a comparator, eg 'Compared to most other people your age, do you ...'? That seems like it could be divisive or alienating, I know, but it may improve the identification of those living with great challenges. If you discern that I am not up-to-date with latest appropriate terms, then well done, you would be correct, I'm just putting the idea out there. If a main purpose of the questions is to help in sample selection by identifying those most challenged, then somehow the questions have to reflect that.

And please, when results are produced, please could age breakdowns be used that describe meaningful groups. "65+" is not sufficient & is too diverse a group to be useful. For example, we would want to divide the 65+s into perhaps 65-74, 75-84, 85-94 and 95+ where counts permit, or to 85+ if they do not. Users can always add some together for their own purposes, but cannot split the groups. And with population ageing, age breaks will increasingly be needed to describe trends over time.

re other comments above - could some of the variation between the different disability survey results sometimes be the inclusion/non-inclusion of people living in long-term care facilities?

This information is also important for children. The 'disability' question in Census is important as a guide and to provide detail about geographical areas or population groups which will only be identified in a whole of population sample. I think the question should be retained whether or not there is a disability-specific household survey planned.

Hi @alexiagarbutt
Thank you for your feedback on the disability statistics. We have noted your interest in data about former refugees with disabilities.

The definition of disability used in the 2013 Disability Survey includes people with psychological/psychiatric impairments, such as long term issues with anxiety and depression. The Disability Survey data contains information on country of birth and years since arrival in New Zealand, but we do not collect information that would allow us to identify former refugees. Please see the disability page for more information about the Disability Survey.

Information on mental health, country of birth and years since arrival are also collected in other surveys including the NZ General Social Survey and the NZ Health Survey, but these surveys may not fully meet your needs either as they do not identify former refugees.

Hi everyone,
Just a quick reminder that the consultation period is quickly coming to an end. There’s been some great discussion in this thread so far and I just wanted to encourage all of you to make a formal submission here before our engagement and consultation closes. All the discussions on loomio will be formally assessed but sending in a formal submission will add weight to whatever your position is. All formal submissions will be assessed against the Content Determination Framework.

Can you do a formal submission if you are an individual?

Hi @jennifermcdonald,

Yes, anyone can make a formal submission, regardless of whether you represent an organisation or not.

No comments in response to an overall measure of self-rated health??

That would make the census soooo much more valuable - for researchers interested in housing, transport, employment, demographics, refugees, immigration, local bodies, education, SES...
A simple 5-point scale would be all that is needed.

Kia Ora everybody

Jo, this is an excellent idea! In 2011, for example, the ONS asked a 5-point scale question for "General Health" in addition to a question on limiting illnesses/disability (Yes a little, Yes a lot, no),

There is a mountain of evidence from the UK that Self-rated health is associated with many determinants. It would be invaluable for research and policy.

A question on health - for which results are publically available - would be extremely valuable and applicable to many situations!

I also support inclusion of a self-rated health question in the census. This would be very helpful. Research has shown that children as young a s five can
participate in self-rating their health – so this would be a very useful universal question.

Mavis

Thanks @Tom for the reminder that the deadline is looming! Since the question in the 2013 Census about disability is not available for analysis, would it be possible to open a new discussion thread about the value (or not) of a general health question?

Having a question in the census that is not able to be disseminated at any level seems an important impetus for alternative measures of health/disability of the whole population.

Cheers
Dan

Hi @danieljohnexeter, @joannabroad and @mavisduncanson you might be interested in the new thread that I just created specifically on the idea of having a general health indicator in the 2018 Census

Thanks Tom, I will certainly make a submission as a near graduate student of disability studies

Hi

Have been following the conversation re health with interest as the issues are very similar to those of the disability community. I feel that Q16, or one similar, should be retained. I understand that there are issues with over and undercounting because of incorrect self identification of disability, however currently territorial authorities have real difficulty in planning for disabled people as there isn't local data available about us. This has many very real and immediate implications for our ability to participate in and move around our communities. Given the degree of social reform that is being signalled by central government I feel that there should be a serious attempt to correct this gap,