Limits...
Predictors for Nursing Home Admission and Death among Community-Dwelling People 70 Years and Older Who Receive Domiciliary Care.

Wergeland JN, Selbæk G, Bergh S, Soederhamn U, Kirkevold Ø - Dement Geriatr Cogn Dis Extra (2015)

Bottom Line: Through bivariate and multivariate Cox proportional hazards regression models, associations between a covariate and the outcomes were analyzed.The Neuropsychiatric Inventory sub-syndrome psychosis, poorer functional impairment and age were associated with NHA.Female sex, age, worse medical health and functional impairment were associated with death.

View Article: PubMed Central - PubMed

Affiliation: Norwegian Center for Research, Education and Service Development, Vestfold Hospital Trust, Tønsberg, Norway ; Faculty of Health and Sport Sciences, University of Agder, Kristiansand, Norway.

ABSTRACT

Aim: The aim of this study was to analyze which variables predicted nursing home admission (NHA) and death.

Methods: 1,001 recipients of domiciliary care were assessed three times in a 3-year period. Through bivariate and multivariate Cox proportional hazards regression models, associations between a covariate and the outcomes were analyzed.

Results: Participants with dementia had a higher risk of NHA (odds ratio 3.88, 95% confidence interval 2.92-5.16) compared to participants without dementia. The Neuropsychiatric Inventory sub-syndrome psychosis, poorer functional impairment and age were associated with NHA. Female sex, age, worse medical health and functional impairment were associated with death.

Conclusion: Support to the caregiver and education on how to meet and cope with behavioral disturbance, depressive mood and sub-syndrome psychosis will probably enable the family to better adapt to late life changes.

No MeSH data available.


Related in: MedlinePlus

Time until NHA and death of the whole sample and of participants with and without dementia.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4608662&req=5

Figure 2: Time until NHA and death of the whole sample and of participants with and without dementia.

Mentions: To ensure that the effects of the explanatory variables were independent of time from baseline, we performed two different tests. For the continuous explanatory variables (age, GMHR, PSMS, NPI and NPI sub-syndromes), we estimated the Cox model and saved the partial residuals. Then we tested the correlation between the partial residuals and the time variable. If these correlations were <0.2, the explanatory variable was judged to be independent of time. All the tested continuous variables were independent of time. For dichotomous variables (sex, living alone, dementia and any use of psychotropic drugs), we plotted the survival curves with a Kaplan-Meier plot for each variable split on the two possible scores (fig. 2) and checked for crossing lines. Variables with non-crossing lines were assumed to be independent of time [32]. This was the case for all dichotomous variables. All significance tests were two-tailed and the level of significance was set at <0.05.


Predictors for Nursing Home Admission and Death among Community-Dwelling People 70 Years and Older Who Receive Domiciliary Care.

Wergeland JN, Selbæk G, Bergh S, Soederhamn U, Kirkevold Ø - Dement Geriatr Cogn Dis Extra (2015)

Time until NHA and death of the whole sample and of participants with and without dementia.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4608662&req=5

Figure 2: Time until NHA and death of the whole sample and of participants with and without dementia.
Mentions: To ensure that the effects of the explanatory variables were independent of time from baseline, we performed two different tests. For the continuous explanatory variables (age, GMHR, PSMS, NPI and NPI sub-syndromes), we estimated the Cox model and saved the partial residuals. Then we tested the correlation between the partial residuals and the time variable. If these correlations were <0.2, the explanatory variable was judged to be independent of time. All the tested continuous variables were independent of time. For dichotomous variables (sex, living alone, dementia and any use of psychotropic drugs), we plotted the survival curves with a Kaplan-Meier plot for each variable split on the two possible scores (fig. 2) and checked for crossing lines. Variables with non-crossing lines were assumed to be independent of time [32]. This was the case for all dichotomous variables. All significance tests were two-tailed and the level of significance was set at <0.05.

Bottom Line: Through bivariate and multivariate Cox proportional hazards regression models, associations between a covariate and the outcomes were analyzed.The Neuropsychiatric Inventory sub-syndrome psychosis, poorer functional impairment and age were associated with NHA.Female sex, age, worse medical health and functional impairment were associated with death.

View Article: PubMed Central - PubMed

Affiliation: Norwegian Center for Research, Education and Service Development, Vestfold Hospital Trust, Tønsberg, Norway ; Faculty of Health and Sport Sciences, University of Agder, Kristiansand, Norway.

ABSTRACT

Aim: The aim of this study was to analyze which variables predicted nursing home admission (NHA) and death.

Methods: 1,001 recipients of domiciliary care were assessed three times in a 3-year period. Through bivariate and multivariate Cox proportional hazards regression models, associations between a covariate and the outcomes were analyzed.

Results: Participants with dementia had a higher risk of NHA (odds ratio 3.88, 95% confidence interval 2.92-5.16) compared to participants without dementia. The Neuropsychiatric Inventory sub-syndrome psychosis, poorer functional impairment and age were associated with NHA. Female sex, age, worse medical health and functional impairment were associated with death.

Conclusion: Support to the caregiver and education on how to meet and cope with behavioral disturbance, depressive mood and sub-syndrome psychosis will probably enable the family to better adapt to late life changes.

No MeSH data available.


Related in: MedlinePlus