Research reports - New Zealand Health and Ageing Research Group - University of Canterbury - New Zealand

New Zealand Health and Ageing Research Group

Research projects and reports

Scholarships

Over summer 2016/2017 the South Island Health Alliance co-sponsored with UC two summer studentships:

  1. Managing medications as a predictor of aged residential care admission.
    Student recipient: Marina Chen, with Dr Hamish Jamieson.

    The goal of this project was to quantify how much difficulty with medication management is a predictor of aged residential care admission for older persons living in the community without a known cognitive impairment or dementia after adjusting for confounding factors. Confounding factors were, for example, vision, pain and mode of nutrition intake. The national interRAI dataset was used with several exclusions applied and this dataset was linked to mortality and aged residential care admissions data using encrypted NHI numbers. Survival analysis was mainly used, particularly, Competing Risks Regressions which gave unadjusted and adjusted relative risks. Difficulty managing medications and Adherence to Medications was found to be associated with aged residential care admission in older persons after correcting for multiple confounders.
  2. How Social Support Influences Aged Residential Care Admission in People with Dementia in New Zealand.
    Student recipient: Peter Qui, with Dr Hamish Jamieson.

    The aim of this project was to find out to what extent the amount of social support for people with dementia can predict the aged residential care (ARC) admission. InterRAI-Home Care version 9 was used as the assessment. Data were collected between September 2012 and January 2016. Finally, 11,075 elderly people with dementia were selected for the analysis. Competing risks regressions were conducted before and after adjusting for confounders. The result shows that the more often people with dementia have a visit with their family members or friends, the less likely they will enter ARC. However, among all individuals with dementia, those who have had a visit with family members or friends in the last 3 days are most likely to enter ARC. Thus, we believe that regular social interaction appears to be protective of ARC admission, but very frequent social visits appear to indicate increased ARC admission.

In 2015 CDHB funded three scholarships for research students at honours level, or above. These were

  1. Working with NHI and national interRAI (Mortality variables of importance).
    Student recipient: Philip Petrie, with Professor Tim David (UC).

    This project was an analysis of an updated version of the large health data-set obtained from interRAI assessments. The data was comprised of information on 48,000 elderly people in Home Care with updated follow up information on mortality. Our interests were in finding associations and patterns in the data in relation to mortality and survivability. The methods of analysis were classification decision trees such as random forests and boosting. The results gave back variables ranked by importance for predicting mortality. Some of the variables include a previous cancer diagnosis, fatigue, self-reported health, timed four metre walk, change in ADL status and decrease in fluid.
  2. Looking for predictions of mortality within interRAI.
    Student recipient: Matthew Males, with Professor Jennifer Brown (UC).

    The project was a survival analysis of interRAI data seeking predictors of mortality. Analyses focused on CHESS score and selected diagnosed health conditions. Data was linked to known mortality dates, deterministically based on NHI numbers. It was shown that CHESS score provides a useful predictor of mortality for both CDHB and national clients in a home care setting. Results also indicated that those with pre-existing health conditions amplified rates of declining health.
  3. Looking at Parkinson's.
    Student recipient: , with Professor John Dalrymple (UC).

PhD Students

  1. Thesis title: Derivation of Injury Risk Score in Older Adults using interRAI Home Care Data
    Student: Rebecca Abey. Supervisors: Professor Tim Wilkinson (UoC), Dr Hamish Jamieson (UC), and Professor Philip Schluter (UC)
  2. Thesis title: How well do people live across the continuum of aged care in Ryman Healthcare Retirement Villages in Canterbury?
    Student: Robyn Johnston (UC). Supervisor: Associate Professor Ray Kirk (UC).

Previous student projects:

Over summer 2015/2016 the South Island Health Alliance co-sponsored with UC three summer studentships:

  1. Predictors of poor outcomes for dementia.
    Student recipient: Annabelle Bos, with Professor John Dalrymple (UC).

    This project was an analysis of an updated version of the large health data-set obtained from interRAI assessments with a focus on the variable of dementia. The data was comprised of information on 48,000 elderly people in Home Care with updated follow up information on mortality and admittance to rest care facilities. Our interests were in finding associations and patterns in the data that would lead to finding the predictors of admission to rest care for those with a dementia diagnosis.
  2. Predictors of hip fractures amongst older adults who are living at home.
    Student recipient: Losana Vao Latu Latu, with Professor Philip Schluter (UC).

    This study was an analysis of 35,625 older adults who are living in their home. interRAI-Home Care version 9.1 was used and assessments were conducted between 1st July 2012 and 30 June 2014. Deterministic matching of hospital discharge database and mortality database to the interRAI-Home Care database using NHI Encrypted data were undertaken. We were focused on investigating the relationship between falls and hip fractures amongst older(65 or older) adults plus identifying the risk factors of hip fractures.
  3. Big Data and Advanced Care (Mortality & Hospital admission).
    Student recipient: Philip Petrie, with Professor Tim David (UC).

    This project was an analysis of an updated version of the large health data-set obtained from interRAI assessments. The data was comprised of information on 48,000 elderly people in Home Care with updated follow up information on mortality and hospital admission data. Our interests were in finding association and patterns in the data in relation to mortality, CHESS scores and the hospital admission data.

Over summer 2015/2016 University of Otago, Christchurch provided a summer scholarship:

  • Developing a frailty scale from the interRAI.
    Student recipient: Rosie Burn, with Dr Hamish Jamieson (UO) and Professor Jennifer Brown (UC).

    We used data from the interRAI home care assessments to develop an efficient method measure a person's frailty. We derived a 6 point scale which can be used to identify people according to their different needs. This scale has the potential to help guide clinicians, and for health care planning.

Summer 2014/2015 the High Performance Computing unit in the University of Canterbury (UC HPC) sponsored a summer studentship:

  • Statistical Analysis of neurological conditions.
    Student recipient: Philip Petie, with Professor Tim David (UC).

    This project was an analysis of a large health data-set comprised of information on 40,000 elderly people in Home Care, in relation to the neurodegenerative Alzheimers disease. Our interests were in finding associations and patterns in the data. The methods of analysis ranged from simple tests of association to complex classification methods such as random forests and boosting. The results supported evidence of a number of previously believed associations as well as identified relationships previously not considered. These will need to be investigated further. These associations that were supported from our analysis include the relationship between Alzheimers disease and Ethnicity, Cognitive performance, and alcohol consumption. One of the most interesting new relationships found was between Alzheimers disease and Stroke.

Research Reports

Papers

  • Jamieson H.A., Schluter P.J., Pyun J., Arnold T., Scrase R., Abey R., Mor V. and Gray L. (2017). Fecal incontinence is associated with mortality among older adults with complex needs: an observational cohort study. American Journal of Gastroenterology, 112(9):1431-1437. doi: 10.1038/ajg.2017.200.

    Abstract:
    Objectives: Fecal incontinence (FI) is a problem in growing older populations. Validating a suspected association between FI and mortality in community dwelling older adults could lead to improved planning for and management of the increasing complex older population. In a large cohort of New Zealand older adults, we assessed the prevalence of FI, urinary incontinence (UI), combined FI and UI, and their associations with mortality.

    Methods: This study consisted of a retrospective analysis of international standardized geriatric assessment-home care (InterRAI-HC) data from community-dwelling adults aged 65 years or older, who met the criteria required for the InterRAI-HC, having complex needs and being under consideration for residential care. The prevalence of UI and FI was analyzed. Data were adjusted for demography and 25 confounding factors. Mortality was the primary outcome measure.

    Results: The total cohort consisted of 41,932 older adults. Both UI and FI were associated with mortality (P<0.001), and risk of mortality increased with increased frequency of incontinence. In the adjusted model, FI remained significantly related to survival (P<0.001), whereas UI did not (P=0.31). Increased frequency of FI was associated with an increased likelihood of death (hazard ratio 1.28).

    Conclusions: This large national study is the first study to prove a statistically significant relationship between FI and mortality in a large, old and functionally impaired community. These findings will help improve the management of increasingly complex older populations.

     
  • Schluter P.J., Lacey C., Porter R.J. and Jamieson H.A. (2017). An epidemiological profile of bipolar disorder among older adults with complex-needs: a national cross-sectional study. Bipolar Disorders, 19(5):375-385. doi: 10.1111/bdi.12511.

    Abstract:
    Objectives: Research on bipolar disorder (BD) among community-living older adults is scant and often suffers from important methodological limitations. Using a national database, this study presents an epidemiological profile of BD in older community residents within New Zealand.

    Methods: Since 2012, all New Zealand community care recipients have undergone a standardized needs assessment using the Home Care International Residential Assessment Instrument (interRAI-HC). The interRAI-HC elicits information using 236 questions over 20 domains, including BD diagnosis. Those who were assessed between 1 September 2012 and 31 January 2016, who were aged ?65 years, and who provided consent were included. Statistical investigations employed bivariable and multivariable logistic regression models.

    Results: Overall, 71 859 people were eligible; their average age was 82.7 years (range 65-105 years), with 43 802 (61.0%) being female and 798 (1.1%) having a BD diagnosis. Participants' sex, age and ethnic identification were significantly related to BD (all P<.001). Participants with a higher number of comorbidities had greater odds of BD; for those with at least six comorbidities, the adjusted odds ratio (AOR) was 2.32 (95% confidence interval [CI] 1.37-3.92). Almost all considered social and environmental variables were significantly and detrimentally associated with BD, such as living in squalid conditions (2.7% for those with DB vs 1.1% for those without DB; AOR=1.60 [95% CI 1.06-2.42]).

    Conclusions: BD among older adults is not uncommon, and numbers will increase as populations age. Increasingly, health services are moving to home-based integrated models of care. Clinicians and decision-makers need to be aware in their planning and service delivery that significant deficits in environment quality and exposure to stressful living circumstances remain for older adults with BD.

     
  • Schluter P.J., Ward C., Arnold E.P., Scrase R. and Jamieson H.A. (2017). Urinary incontinence, but not faecal incontinence, is a risk factor for admission to aged residential care of older persons in New Zealand. Neurourology and Urodynamics, 36(6):1588-1595. doi: 10.1002/nau.23160.

    Abstract:
    Aims: To determine if urinary incontinence (UI) and fecal incontinence (FI) were independent risk factors for aged resident care (ARC) admissions for older people, after controlling for confounders and applying apposite statistical methods.

    Methods: Since 2012, all community care recipients in New Zealand have undergone a standardized needs assessment using the Home Care International Residential Assessment Instrument (interRAI-HC). The interRAI-HC instrument elicits information on 236 questions over 20 domains, including UI and FI frequency within the last 3 days. Those aged 65+ years with an interRAI-HC assessment between July 1, 2012 and May 31, 2014 were matched to national mortality and ARC databases, and competing-risk regression models applied to those without collection devices or indwelling catheters who were admitted to ARC or alive 30+ days after their interRAI-HC assessment.

    Results: Overall, 32?285 people were eligible, with average age of 82.1 years (range 65, 105 years) of whom 20?627 (63.9%) were female. UI and FI was reported by 36.4% and 12.9% of people, respectively. By June 30, 2014, 5993 (18.6%) had an ARC admission and 5443 (16.9%) had died before any such admission. In the multivariable analysis, the subhazard ratio (SHR) for ARC admission was significant for UI (SHR?=?1.11, 95%CI: 1.05, 1.18) but not for FI (SHR?=?1.07, 95%CI: 0.99, 1.16).

    Conclusions: UI is a common, independent risk factor for ARC admissions. Identifying the extent of incontinence and its impact on ARC admissions is the first vital step in addressing the burgeoning need for better community continence services.

     
  • Crowe, M., Jordan, J., Gillon, D., McCall, C., Frampton, C., & Jamieson, H. (2017). The prevalence of pain and its relationship to falls, fatigue and depression in a cohort of older people living in the community. Journal of Advanced Nursing. [Advance online publication] doi: 10.1111/jan.13328.

    Abstract:
    Aim: To examine the relationship of the pain severity scores with demographic variables (age, gender and ethnicity) and with the frequency of falls, fatigue and depression in a very large New Zealand sample of people over 65 years assessed using the Home Care International Residential Assessment Instrument.

    Background: Pain is reported to be highly prevalent in older people yet it is poorly correlated with tissue damage. There is convincing evidence that it is related to depression and some evidence of its relationship with fatigue and falls.

    Design: This is a cross-sectional study examining a national cohort assessed on referral for a needs assessment for access to publicly funded service provision or support.

    Method: Participants were 45418 adults aged over 65 years referred for a mandatory needs assessment between 2012 - 2014. All variables analysed were drawn from the Home Care International Residential Assessment Instrument. Univariate descriptive statistics were used to characterize the sample in relation to the association of severe pain with age, gender and ethnicity and with the key variables of interest: falls, fatigue and depression. Logistic regression models were used to examine the relationship of the presence or not of severe pain with the other key health-related variables: falls, fatigue and depression respectively, after controlling for the presence of age, gender and ethnicity. Odds ratios are reported to quantify the difference in risk with increasing severity of falls, fatigue and depression.

    Results: The mean age of the cohort was 82.48 years (SD 7.48) and 48.5% of the sample reported experiencing daily pain. The rates of severe daily pain were 12-18% with those in the 65-74 years group having the highest reported rate of severe pain (18%) and a gradual decrease in severe pain with age. After co-varying for age, ethnicity and gender, the presence of severe daily pain was predicted by increasing frequency of falls and fatigue and having current depression. While two-thirds reported their pain control was adequate, a third reported that their pain control was not adequate despite adhering to a therapeutic regime.

    Conclusion: The study highlights how pervasive pain is in this group and its correlation with falls, fatigue and depression. While the effect sizes were relatively small the results may be clinically significant and it may be clinically important that the presence of pain in older people could indicate the need for further nursing assessment in relation to falls, fatigue and depression

     
  • Schluter P.J., Ward C., Arnold E.P., Scrase R. and Jamieson H.A. (2016). Urinary incontinence, but not faecal incontinence, is a risk factor for admission to aged residential care of older persons in New Zealand. Neurourology and Urodynamics. [Epub ahead of print] doi: 10.1002/nau.23160.

    Abstract:
    Aims: To determine if urinary incontinence (UI) and faecal incontinence (FI) were independent risk factors for aged resident care (ARC) admissions for older people, after controlling for confounders and applying apposite statistical methods.

    Methods: Since 2012, all community care recipients in New Zealand have undergone a standardised needs assessment using the Home Care International Residential Assessment Instrument (interRAI-HC). The interRAI-HC instrument elicits information on 236 questions over 20 domains, including UI and FI frequency within the last 3 days. Those aged 65+ years with an interRAI-HC assessment between 1 July 2012 and 31 May 2014 were matched to national mortality and ARC databases, and competing risk regression models applied to those without collection devices or indwelling catheters who were admitted to ARC or alive 30+ days after their interRAI-HC assessment.

    Results: Overall, 32,285 people were eligible, with average age of 82.1 years (range 65, 105 years) of whom 20,627 (63.9%) were female. UI and FI was reported by 36.4% and 12.9% of people, respectively. By 30 June 2014, 5,993 (18.6%) had an ARC admission and 5,443 (16.9%) had died before any such admission. In the multivariable analysis, the subhazard ratio (SHR) for ARC admission was significant for UI (SHR=1.11, 95% CI: 1.05, 1.18) but not for FI (SHR=1.07, 95% CI: 0.99, 1.16).

    Conclusion: UI is a common, independent risk factor for ARC admissions. Identifying the extent of incontinence and its impact on ARC admissions is the first vital step in addressing the burgeoning need for better community continence services.

     
  • Schluter P.J., Ahuriri-Driscoll A., Anderson T.J., Beere P., Brown J., Dalrymple-Alford J., David T., Davidson A., Gillon D.A., Hirdes J., Keeling S., Kingham S., Lacey C., Menclova A.K., Millar N., Mor V. and Jamieson H.A. (2016). Comprehensive clinical assessment of home-based older persons within New Zealand: an epidemiological profile of a national cross-section. Australian and New Zealand Journal of Public Health, 40(4):349-355. doi: 10.1111/1753-6405.12525.

    Abstract:
    Objective: Since 2012, all community care recipients in New Zealand have undergone a standardized needs assessment using the Home Care International Residential Assessment Instrument (interRAI-HC). This study describes the national interRAI-HC population, assesses its data quality, and evaluates its ability to be matched.

    Methods: The interRAI-HC instrument elicits information on 236 questions over 20 domains; conducted by 1,800+ trained health professionals. Assessments undertaken between 1 July 2012 and 30 June 2014 were reported here. Stratified by age, demographic characteristics were compared to census estimates, and selected health profiles described. Deterministic matching to the Ministry of Health’s mortality database was undertaken.

    Results: Overall, 51,232 interRAI-HC assessments were conducted, with 47,714 (93.1%) consented for research from 47,236 unique individuals; including 2,675 Maori and 1,609 Pacific people. Apart from height and weight, data validity and reliability were high. A 99.8% match to mortality data was achieved.

    Conclusions: The interRAI-HC research database is large, ethnically diverse, with high consent rates. Its generally good psychometric properties and ability to be matched enhances its research utility. Implications: This national database provides a remarkable opportunity for researchers to better understand older persons’ health and health care so as to better sustain older people in their own homes.

     

Other Reports

Conference Presentations

  • New Zealand Health and Ageing Research Group
    University of Canterbury
    Private Bag 4800, Christchurch
    New Zealand
  • Phone +64 3 364 2696
    Jennifer.Brown@canterbury.ac.nz
  • Follow us
    FacebookYoutubetwitterLinked In