2018-19 Section 7-Neoplastic and Inflammatory Diseases of the Head and Neck eBook

TABLE III. Multivariate Logistic Regression With Unadjusted and Adjusted Rate of Morbidity and Mortality.

Unadjusted OR (95% CI)

Adjusted OR (95% CI)

Adult abscess specific morbidity and mortality Timing of surgical drainage after admission 0 days

Ref

Ref

1–2 days

1.30 (0.64–2.62)

1.27 (0.59–2.74)

3–7 days

2.56 (1.19–5.53) 1.64 (0.90–2.97)

2.38 (1.01–5.59) 2.12 (1.08–4.15)

Gender, female

ASA class 3, 4, or 5

1.78 (0.96–3.30)

1.51 (0.76–2.96)

Preoperative sepsis or septic shock

4.43 (2.36–8.32) 1.09 (1.04–1.14)

3.32 (1.65–6.67) 1.07 (1.01–1.11)

White blood cell count (1,000 cell/ l L)

Pediatric abscess specific morbidity and mortality Timing of surgical drainage after admission 0 days

Ref

Ref

1–2 days

0.74 (0.34–1.62)

0.59 (0.25–1.43)

3–7 days

0.89 (0.29–2.76) 1.92 (0.92–3.98)

1.00 (0.30–3.37) 1.88 (0.84–4.21)

Gender, female

Congenital malformation

3.62 (1.30–10.04)

3.68 (1.16–11.67)

ASA class 3 or 4

6.66 (2.54–17.42)

5.64 (2.45–13.00)

Preoperative sepsis or septic shock

2.13 (1.00–4.52)

1.91 (0.79–4.59)

White blood cell count (1,000 cell/ l L)

1.01 (0.96–1.06)

0.99 (0.93–1.04)

Adjusted OR shows the final multivariate logistic regression model. Unadjusted OR shows bivariate logistic regression. ASA 5 American Society of Anesthesiology; CI 5 confidence interval; OR 5 odds ratio; Ref 5 reference value.

After this adjustment, C indices varied from 0.75 to 0.76 in adults and from 0.72 to 0.75 in pediatrics, with a significant trend in increased M&M seen with delay in surgical drainage in adults but not in children. This result supports the robustness of our findings and suggests that they were not significantly influenced by our categorizations.

DISCUSSION This study examines the timing of surgical drainage of deep neck abscess in a prospective, multicenter, risk- adjusted cohort in both adult and pediatric populations. We hypothesized that delaying surgical drainage of deep neck infections for a trial of medical therapy would be associated with increased abscess-specific M&M. We

TABLE IV. Sensitivity Analysis to Test Robustness of Association With Varying Definitions of Time From Admission to Surgery.

Timing of Surgery Groups

Groups

Adjusted OR (95% CI)

C Statistic

H-L Test

Adult Primary analysis: day 0, day 1–2, day > 3

Day 0

Ref

0.75

0.25

Day 1–2

1.27 (0.60–2.69)

Day 3–7

2.55 (1.11–5.87

Alternate analysis: day 0–1, day 2–3, day 4–7

Day 0–1 Day 2–3

Ref

0.75

0.48

0.98 (0.40–2.40)

Day 4–7

4.31 (1.66–11.20)

1 1 day

Alternate analysis: each day continuous variable

1.25 (1.03–1.52)

0.76

0.34

Pediatric Primary analysis: day 0, day 1–2, day > 3

Day 0

Ref

0.75

0.74

Day 1–2 Day 3–7

0.59 (0.25–1.43) 1.00 (0.30–3.37)

Alternate analysis: day 0–1, day 2–3, day 4–7

Day 0–1

Ref

0.72

0.34

Day 2–3 Day 4–7

0.93 (0.44–1.98) 1.86 (0.83–4.16)

1 1 day

Alternate analysis: each day continuous variable

0.92 (0.64–1.32)

0.73

0.16

CI 5 confidence interval; H-L test 5 Hosmer-Lemeshow test for goodness of fit; OR 5 odds ratio; Ref 5 reference.

Laryngoscope 126: August 2016

Cramer et al.: Delay in Drainage of Deep Neck Abscess

215