ORIGINAL ARTICLE   Free access

European Journal of Physical and Rehabilitation Medicine 2019 December;55(6):735-42

DOI: 10.23736/S1973-9087.19.05840-4

Copyright © 2019 EDIZIONI MINERVA MEDICA

language: English

Risk factors for long-term care after hemiplegia from cancer-related brain surgery: a pilot study for new prediction model

Arturo ZANCAN ¹ ✉, Alessandra RODIGARI ², Francesca GIGLI BERZOLARI ³, Paola BORRELLI ³

¹ Subacute Care Unit, Istituti Clinici Scientifici Maugeri Spa, Società Benefit and IRCCS, Pavia, Italy; ² Unit of Rehabilitation and Functional Recovery, Istituti Clinici Scientifici Maugeri Spa, Società Benefit and IRCCS, Pavia, Italy; ³ Unit of Biostatistics and Clinical Epidemiology, Department of Public Health Experimental and Forensic Medicine, University of Pavia, Pavia, Italy

BACKGROUND: Hemiplegia is a worldwide-represented neurological condition leading to long-term disability. The most common cause of hemiplegia is stroke; 25% to 50% of stroke survivors require some assistance after hospital discharge; approximately half of them become dependent, while only 14% achieve full recovery of activities of daily living (ADL). Cancer and cancer-related surgery are other causes of hemiplegia and rehabilitation in cancer patients has been recognized as important factor in order to help maintaining quality of life as long as possible. Many studies have been done in order to assess a reliable prediction about outcome of hemiplegia from stroke, but outcome prediction for cancer-related hemiplegia still remains a challenge and no clinical prediction tool has been developed being better than physician’s informal prediction.
AIM: Aim of this pilot study was: a) to detect risk factors associated with need for long-term care for patients suffering from hemiplegia due to cancer-related brain surgery, b) to build an algorithm-based model from detected risk factors in order to predict the need for long-term care after rehabilitation 3) to assess the feasibility of a subsequent study on a larger sample of subjects, in order to validate of the model. The subsequent study will be considered feasible if the model developed by the pilot study will be able to correctly predict more than 85% of patients needing or not long-term care after rehabilitation.
DESIGN: Observational retrospective study.
SETTING: Neurorehabilitation Unit.
POPULATION: Inpatients affected by hemiplegia due to cancer-related surgery.
METHODS: The observational retrospective study involved 158 subjects affected by hemiplegia due to cancer-related brain surgery. All of the subjects underwent rehabilitation therapy, while radio /chemotherapy was administered if needed. Stroke prognostic factors and other clinical variables were recorded for all subjects. The endpoint variable was Functional Independence Measure (FIM®) Score at discharge after rehabilitation. Variables were then associated with patient’s FIM Score <61 at discharge after rehabilitation, as predictor of long-term care at home.
RESULTS: After statistical evaluation, age, comorbidity Charlson Index >3, complete absence of motricity of the affected limbs, hypoesthesia, trunk-control deficit, dysphagia, language disorder, urinary or fecal incontinence were found to be risk factors for FIM Score <61 at discharge. From detected factors an algorithm-based model was built in order to estimate patient’s overall probability to need or not an intensive long-term care after rehabilitation.
CONCLUSIONS: The model developed by the pilot study allowed correct positive or negative prediction for long-term care need after rehabilitation for 90.6% of the patients suffering from cancer-related hemiplegia. A subsequent study on a larger sample of subjects resulted therefore feasible because overall correct prediction was higher than 85%.
CLINICAL REHABILITATION IMPACT: Risk factors for intensive care at home (complete absence of motricity of affected limbs, trunk control deficit, fecal incontinence, dysphagia and comorbidity Charlson Index >3) can be useful to evaluate patients suffering from hemiplegia due to cancer-related brain surgery, at admission into Rehabilitation Unit. The algorithm-based model seems to be a promising tool to estimate the probability of intensive home care for that type of hemiplegic patients.

KEY WORDS: Hemiplegia; Neoplasms; Brain; Surgery; Long-term care