Factors affecting recovery during the first 6 months after hip fracture, using the decision tree model
Pelvic fractures are one of the most common orthopedic problems that can reduce the quality of life in the elderly. In this prospective study, we found that osteoporosis, depression, and socioeconomic status were the most important factors associated with patients’ recovery during the first 6 months after pelvic fracture.
Hip fractures are one of the most common orthopedic problems that can reduce the quality of life in the elderly. Considering that, we aimed to provide a comprehensive assessment of the factors affecting recovery during the first 6 months after hip fracture.
All patients with hip fracture admitted to any of the orthopedic hospitals during July 10, 2011 to July 9, 2012 in Shiraz, Iran were included in this prospective cohort study. Patients’ demographic data and also information regarding their performance and mobility after hip fracture was collected in two interviews at intervals of 6 months. All analyses were done in R software and mostly by party packages and PCAmixdata package. Tree and forest models of conditional inference were used to evaluate the factors affecting the recovery after hip fracture.
Two hundred sixty-six out of 514 patients (51.75%) with hip fracture recovered completely after a 6-month follow-up period. Osteoporosis, new-onset depression after hip fracture, and socioeconomic status (SES) were the most important predictors of patients’ mobility status 6 months after hip fracture. In identifying predictor variables, the conditional inference forest method provided a more appropriate fit for the data than the conditional inference tree.
Awareness of the factors that affect patients’ recovery can be helpful in improving the patients’ health, as well as improving care services, thereby increasing the success of treatment. Osteoporosis, new-onset depression after hip fracture, and SES were the most important factors associated with patients’ recovery. Therefore, focusing on these variables is essential.
KeywordsHip fractures Follow-up studies Decision tree model Iran
We would like to express our special gratitude to Vice-chancellor for Research of Shiraz University of Medical Sciences who financially supported this study. Also, it should be noted that this article has been extracted from the thesis of Fatemeh Jafarzadeh (No: 87/1026).
Compliance with ethical standards
Conflict of interest
- 3.Rostagno C, Buzzi R, Campanacci D, Boccacini A, Cartei A, Virgili G, Belardinelli A, Matarrese D, Ungar A, Rafanelli M (2016) In hospital and 3-month mortality and functional recovery rate in patients treated for hip fracture by a multidisciplinary team. PLoS One 11(7):e0158607CrossRefPubMedGoogle Scholar
- 6.Ioannidis G, Jantzi M, Bucek J, Adachi JD, Giangregorio L, Hirdes J, Pickard L, Papaioannou A (2017) Development and validation of the Fracture Risk Scale (FRS) that predicts fracture over a 1-year time period in institutionalised frail older people living in Canada: an electronic record-linked longitudinal cohort study. BMJ Open 7(9):e016477CrossRefPubMedGoogle Scholar
- 18.Chavent M, Kuentz-Simonet V, Labenne A, Saracco J (2014) Multivariate analysis of mixed data: the PCAmixdata R package. arXiv preprint arXiv:14114911Google Scholar
- 24.Shahnazari B, Keshtkar A, Soltani A, Aghamaleki A, Mansour A, Matin B, Saghafi S, Dini M, Khashayar P, Larijani B (2013) Estimating the avoidable burden of certain modifiable risk factors in osteoporotic hip fracture using Generalized Impact Fraction (GIF) model in Iran. J Diabetes Metab Disord 12(1):10CrossRefPubMedGoogle Scholar