Skip to main content

Table 3 Factors associated with the total length of hospital stay (LOS)

From: Hospital stay in patients admitted for acute bipolar manic episodes prescribed quetiapine immediate or extended release: a retrospective non-interventional cohort study (HOME)

 

Estimate (95% CI)

t-statistic

p-value

Site characteristics

   

Public funding vs private

1.442 (1.259, 1.651)

5.28

< 0.0001

Less than 50 beds vs more than 600 beds

1.015 (0.851, 1.212)

0.17

0.8651

50 to 200 beds vs more than 600 beds

0.936 (0.846, 1.035)

−1.29

0.1966

201 to 600 beds vs more than 600 beds

1.011 (0.923, 1.107)

0.24

0.8101

Patient demographics

   

Lives accompanied vs alone

0.889 (0.812, 0.974)

−2.54

0.0112

Any present alcohol/drug abuse or dependence

0.883 (0.805, 0.968)

−2.66

0.0079

Medical history

   

Any psychiatric medical history

0.868 (0.770, 0.978)

−2.33

0.0202

Any non-psychiatric medical history

1.064 (0.980, 1.156)

1.48

0.1392

Any BD hospital admission in prior 12 months

1.179 (1.071, 1.297)

3.37

0.0008

No manic event in prior 12 months

1.188 (1.076, 1.312)

3.41

0.0007

Severity of manic event

   

Moderate vs mild

0.931 (0.763, 1.136)

−0.71

0.4799

Severe vs mild

1.250 (1.031, 1.514)

2.27

0.0231

Department of admission

   

ICU vs ER

0.991(0.782, 1.255)

−0.08

0.9397

General ward vs ER

1.236 (0.985, 1.551)

1.84

0.0667

Psychiatric ward vs ER

1.180 (1.035, 1.345)

2.47

0.0137

Treatment

   

Maximum no. of new medications administered

1.154 (1.122, 1.187)

10.00

< 0.0001

 

0.969 (0.903, 1.040)

−0.87

0.3861

Did not receive lithium

1.211(1.107, 1.325)

4.18

< 0.0001

Did not receive anxiolytics, sedatives/hypnotics

1.222 (1.119, 1.334)

4.47

< 0.0001

Received antidepressants

1.144 (0.971, 1.348)

1.61

0.1069

% quetiapine use

0.999 (0.996, 1.001)

−1.29

0.1970

  1. BD, bipolar disorder; CI, confidence interval; ER, emergency room; ICU, intensive care unit.
  2. Factors associated with a LOS are shown in bold. Estimates with 95% CI were calculated by zero-truncated negative binomial regression multivariable analysis. Only clinical factors included in the final model are shown. Countries were included in the model but are not shown. Clinical factors with a p-value of < 0.05 were considered significant.