Skip to main content

Table 5 Predictors of log of inpatient hospital charges among patients with gout who were admitted to the hospital after presenting to the Emergency Department (ED) with gout using linear regression

From: Gout-related inpatient utilization: a study of predictors of outcomes and time trends

 

Univariate

Multivariable-adjusted

 

B-estimate (95 % CI)

P value

B-estimate (95 % CI)

P value

Age

 <50

Ref

 

Ref

 

 50– <65

0.11 (0.03, 0.19)

0.0055

0.10 (0.02, 0.18)

0.0129

 65– <80

0.13 (0.06, 0.21)

0.0009

0.06 (−0.03, 0.16)

0.1681

 ≥80

0.13 (0.05, 0.21)

0.0024

0.04 (−0.07, 0.14)

0.4682

Gender

 Female

Ref

 

Ref

 

 Male

−0.02 (−0.07, 0.03)

0.4747

−0.02 (−0.07, 0.03)

0.4667

Median household income

 1st quartile (<$38,999)

Ref

 

Ref

 

 2nd quartile (39,000 to 47,999)

0.02 (−0.09, 0.12)

0.7360

−0.02 (−0.12, 0.08)

0.7445

 3rd quartile (48,000 to 62,999)

0.09 (−0.03, 0.20)

0.1558

0.01 (−0.11, 0.13)

0.8524

 4th quartile ($63,000 or more)

0.10 (−0.05, 0.26)

0.1891

0.00 (−0.17, 0.16)

0.9868

Primary payer

 Medicare

Ref

 

Ref

 

 Medicaid

−0.04 (−0.15, 0.06)

0.4274

−0.05 (−0.16, 0.06)

0.3690

 Private insurance

−0.11 (−0.18, −0.03)

0.0090

−0.07 (−0.16, 0.02)

0.1432

 Self-pay/no charge

−0.11 (−0.21, 0.00)

0.0851

−0.05 (−0.17, 0.07)

0.4134

 Other

−0.15 (−0.3, 0.02)

0.0814

−0.25 (−0.42, −0.08)

0.0040

Patient location (residence)

 Micropolitan/not metro

Ref

 

Ref

 

 Metro (large or small)

0.23 (0.12, 0.35)

0.0001

0.22 (0.11, 0.33)

0.0001

Hospital region

 Northeast

Ref

 

Ref

 

 Midwest

−0.13 (−0.32, 0.06)

0.1889

−0.13 (−0.33, 0.07)

0.2064

 South

−0.13 (−0.33, 0.08)

0.2283

−0.11 (−0.33, 0.11)

0.3367

 West

0.40 (0.18, 0.61)

0.0003

0.45 (0.24, 0.66)

<0.0001

Teaching status of hospital

 Metropolitan non-teaching or non-metro

Ref

 

Ref

 

 Metropolitan teaching

0.01 (−0.12, 0.15)

0.8413

0.00 (−0.13, 0.13)

0.9667

Comorbidities

 CHD (ref: no)

0.10 (0.03, 0.16)

0.0053

0.04 (−0.03, 0.11)

0.2616

 Hyperlipidemia (ref: no)

−0.02 (−0.07, 0.03)

0.4388

−0.05 (−0.10, 0.00)

0.0618

 Renal failure (ref: no)

0.15 (0.09, 0.22)

<0.0001

0.12 (0.07, 0.18)

<0.0001

 Heart failure (ref: no)

0.16 (0.10, 0.23)

<0.0001

0.13 (0.06, 0.19)

0.0001

 Hypertension (ref: no)

−0.00 (−0.07, 0.07)

0.9331

−0.05 (−0.12, 0.03)

0.2028

 Diabetes (ref: no)

0.08 (0.02, 0.13)

0.0058

0.04 (−0.02, 0.10)

0.1590

 COPD (ref: no)

0.10 (0.02, 0.18)

0.0167

0.06 (−0.02, 0.14)

0.1744

 OA (ref: no)

0.01 (−0.07, 0.09)

0.7755

0.02 (−0.05, 0.10)

0.5026

  1. Significant odds ratios are in bold
  2. CI confidence interval, Ref reference category, CHD coronary heart disease, COPD chronic obstructive pulmonary disease, OA osteoarthritis
  3. For interpretation in numeric terms for hospital charges, the coefficients from this regression with log (hospital charges) as an outcome should be transformed as ex, where x = beta-estimate. For example, compared to age <50, hospital charges for ages 50 to <65 was 1.10 times higher (beta coefficient = 0.10; e0.10 = 1.10). On the other hand, compared to Medicare, “other” insurance was associated with 0.78 times the duration of hospital stay (beta coefficient =− 0.25; e−0.25 = 0.78). Positive beta coefficients in this regression with log outcome indicate higher charges and negative beta coefficients indicate lower hospital charges. Patients residing in metropolitan area had hospital charges 1.25 times (reference, non-metro) and in the West 1.57 times hospital charges (reference, Northeast); patients with renal failure had 1.12 times charges and with heart failure, 1.24 times hospital charges
Back to article page

Arthritis Research & Therapy

ISSN: 1478-6362

Contact us