Home > Research > Publications & Outputs > Identifying the long-term care beneficiaries di...

Electronic data

  • ACER 04_Manuscript_reviewed

    Rights statement: The final publication is available at Springer via http://dx.doi.org/10.1007/s40520-019-01418-w

    Accepted author manuscript, 169 KB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License


Text available via DOI:

View graph of relations

Identifying the long-term care beneficiaries differences between risk factors of nursing homes and community-based services admissions

Research output: Contribution to Journal/MagazineJournal articlepeer-review

<mark>Journal publication date</mark>1/10/2020
<mark>Journal</mark>Aging Clinical and Experimental Research
Number of pages12
Pages (from-to)2099–2110
Publication StatusPublished
Early online date28/11/19
<mark>Original language</mark>English


The Portuguese long-term care sector is classified into home and community-based services (HCBS) and three nursing home (NH) units: convalescence, medium term and rehabilitation, and long term and maintenance.

To identify the main factors of admission into each care setting and explore to what extent these populations are different. 14,140 patients from NH and 6844 from HCBS were included from all over the country.

A logistic regression was estimated to identify determinants of admission into NH care, using sociodemographic characteristics, medical conditions and dependence levels at admission as independent variables, and region of care, referral entity and placement process as control variables. Then, ordered logistic regression was used to identify the contribution of the above factors in each specific NH unit.

Being female, not being married, not having family/neighbour support, being literate, having mental illness, being cognitively or physically impaired are the main predictors of being admitted into a NH. Within the NH units, placements of the large majority of patients were accurately predicted, based on the available variables. However, for around half of the patients referred to long-term care units, the model expected placements into medium-term units, while for those admitted into short-stay units, the model returned that 29% could have benefited from being admitted into a medium-term care unit.

Discussion and conclusions
Patients’ accurate placement is a highly complex and challenging process, demanding more variables than the ones available for the model here presented. Our work confirms the need to collect other type of variables to improve the placement decision process.