Predicting clinical instability of older patients in post-acute care units: A nationwide cohort study

Wei Ju Lee, Ming Yueh Chou, Li Ning Peng, Chih Kuang Liang, Li Kuo Liu, Chien Liang Liu, Liang Kung Chen*, Yung Hung Wu, Ching Yi iu, Weng Pin Yu, Yao Hsien Tseng, Ta Wei Luo, Chu Sheng Lin

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

12 Scopus citations


Aim: Although patients admitted to post-acute care (PAC) units are usually clinically stable, unexpected medical conditions requiring acute ward readmissions still occur and can jeopardize the clinical effectiveness of PAC services. The main purpose of the present study was to evaluate predictive factors for clinical instability of patients in PAC units to improve the quality of PAC services. Methods: This was a nationwide multicenter cohort study that recruited patients from five PAC units in Taiwan between July 2007 and June 2009. All patients received the comprehensive geriatric assessment (CGA) within 72h of PAC unit admissions. Conditions requiring acute ward re-admissions from PAC units were defined as clinical instability. Causes of clinical instability for all patients and data of CGA were collected for analysis. Results: Of 918 enrolled participants, 119 (12.9%) experienced acute ward readmissions, including 106 (89.1%) admissions related to medical conditions and 13 (10.9%) for surgical causes. Common conditions included diseases of the respiratory system (n=32, 26.9%), genitourinary system (n=24, 20.2%) and digestive system (n=14, 11.8%). Surgical conditions, mainly fractures and dislocation of upper limbs, were significantly more likely to occur later (P=0.05) in the PAC unit admissions than medical conditions. Compared with the non-readmission group, the readmission group was leaner (mean body mass index 21.1±2.8 vs 22.0±3.8kg/m2, P=0.007), having poorer functional status (mean Barthel Index 41.0±19.4 vs 45.4±20.3, P=0.02; mean IADL: 1.3±1.6 vs 1.7±1.8, P=0.016), poorer cognitive function (mean Mini-Mental State Examination: 16.8±6.4 vs 18.3±6.4, P=0.022), poorer ambulation (mean Timed Up & Go test 32.7±18.7 vs 26.6±11.7s, P=0.039) and poorer nutritional status (mean Mini-Nutrition Assessment 13.3±5.7 vs 15.4±5.8, P<0.001), but similar in depression status (mean Geriatric Depression Score 3.7±3.3 vs 3.4±2.8, P=0.247). In multivariate logistical regression model, lower Mini-Mental State Examination score was the only independent predictor for clinical instability (odds ratio 3.8, 95% confidence interval 1.348-10.870, P=0.012). Conclusion: Approximately 13% of PAC patients might experience acute ward readmissions, and nearly 90% of them are caused by medical conditions. Poor cognitive function is a significant predictive factor for clinical instability in PAC, which deserves more clinical attention for all PAC patients.

Original languageEnglish
Pages (from-to)267-272
Number of pages6
JournalGeriatrics and Gerontology International
Issue number2
StatePublished - 2014


  • Comprehensive geriatric assessment
  • Elderly
  • Post-acute care
  • Readmission


Dive into the research topics of 'Predicting clinical instability of older patients in post-acute care units: A nationwide cohort study'. Together they form a unique fingerprint.

Cite this