Table 4.

Multilevel logistics regression on factors for avoidable readmissions on all unplanned readmissions

Factor	OR	95.0% C.I.
Same principal diagnosis in both discharge and readmission episode	3.41***	(2.20 - 5.30)
No. of attendance to public outpatient clinics and A&E departments in the past 12 months	0.99	(0.96 - 1.01)
No. of hospitalization in the past 12 months	1.01	(0.96 - 1.07)
Use of sub acute & community services	1.03	(0.61 - 1.75)
Length of stay (in days) of previous linked episode that induce the readmission	0.99	(0.98 - 1.01)
Time (in days) elapsed between discharge from index episode and readmission	0.97**	(0.95 - 0.99)
Fee was not paid by public assistance	1.19	(0.81 - 1.74)
Transferred to rehabilitation hospital	0.79	(0.45 - 1.39)
No. of drugs taken on previous discharge	1.03	(0.97 - 1.09)

Notes:

(i) No. of hospital = 14; No. of patients within each hospital ranged from 19-66

(ii) Intra class correlation = 9.9% implying the correlation of patients within hospital is small

(iii) Regression adjusted for age, gender and principal diagnosis of previous discharge

(iv) ** p-value < 0.01, *** p-value < 0.001, OR refers to Odd Ratio, CI refers to Confidence Interval