Background: Homeless patients have a higher incidence of medical and psychiatric diagnoses, which often necessitate hospitalization. Studies show that homeless persons have increased lengths of stay with an average excess of 4.1 days, accounting for an average excess cost of $2414 per admission. Additionally, homeless persons have high rates of 30-day inpatient readmissions, as high as 50% in one study and 70% when ED visits were included.  Several factors contribute to high readmission rates, including cognitive impairments, concomitant mental illness and substance abuse, limited support and difficulty following specialized diets and mobility precautions given the lack of stable funds and housing. The objective of this study was to understand and further characterize the severity of which cognitive impairments affect homeless patients who are recently hospitalized.

Methods:

The Montreal Cognitive Assessment (MoCA) was developed in 2005, and is a clinician-administered cognitive screener used in the detection of mild cognitive impairment and dementia. It has since been proven to be more sensitive than other commonly used screeners including the Mini Mental Status Exam (MMSE).  Eight cognitive domains are evaluated including visuospatial/executive, naming, memory, attention, language, abstraction, delayed recall and orientation for a maximum score of 30 points. A score of 26 points and above is considered within normal range, and this score is used as the lower limit of normal range.

The MoCA was given to 145 patients admitted to the medical respite program, which is a program that gives homeless individuals a guaranteed bed in a shelter, to recuperate upon discharge from an acute hospital stay. The test was administered and scored by a licensed psychologist or psychology student supervised by a licensed psychologist. Descriptive statistics, including mean and standard deviations were calculated for individual item scores.

Results:

For the 145 patients sampled, the average MoCA score was 20.39, with a standard deviation of 4.80. Additionally, 84.7% of the patients fell below a score of 26 points, the lower limit of normal. There was no statistic significant difference between male and female, however there was a negative correlation between age and overall score. A score of 18 points or less is considered to be the cutoff for dementia, and 22.2% of patients sampled scored below 18 points. The lowest scores were on the language fluency (39% correct), delayed memory (average score of 1.42/5), language abstraction (average of 0.74/2), language repetition (average of 0.97/2), cube copy (49% correct), alternating trails (51% correct), and clock drawing hands (55% correct) items.

Conclusions:

These findings provide evidence for a high rate of cognitive impairment in a medically-ill homeless population recently hospitalized and its severity. As transitions of care programs are being developed to improve readmission rates and discharge planning, recognizing the high prevalence and severity of cognitive impairments will have implications on program development and discharge planning.