Table 4 The effect of trazodone on daytime impairment symptoms by domain and ontology

From: Medical ontology learning framework to investigate daytime impairment in insomnia disorder and treatment effects

Endpoint Ontology

Number of daytime impairment symptoms [events per 100 patient years]a

Percentage increase in endpoint after treatment ± SD (P-value)b

Before treatment

During treatment

Average treatment effect [95% CI]

All domains

ICD-clin

12.6

19.4

6.8 [−0.0, 13.5]

53.5% ± 134.7% (P = 0.014)

ICD-clin-DiSMOL

69.6

91.8

16.4 [−1.3, 34.1]

32.0% ± 351.9% (P < 0.001)

DiSMOL

137.7

150.1

12.4 [−7.4, 32.1]

9.0% ± 392.7% (P = 0.094 [ns])

Cognition domain

ICD-clin

12.6

19.4

6.8 [−0.0, 13.5]

53.5% ± 134.7% (P = 0.014)

ICD-clin-DiSMOL

34.9

43.4

6.4 [−3.6, 16.3]

24.3% ± 198.5% (P = 0.037)

DiSMOL

95.1

99.4

4.2 [−10.2, 18.6]

4.4% ± 286.1% (P = 0.465 [ns])

Emotional domain

ICD-clin

12.3

18.4

6.1 [−0.6, 12.8]

49.9% ± 134.0% (P = 0.018)

ICD-clin-DiSMOL

43.0

58.0

9.2 [−5.6, 24.1]

35.1% ± 295.4% (P = 0.002)

DiSMOL

87.1

98.3

11.2 [−4.8, 27.2]

12.9% ± 317.6% (P = 0.071 [ns])

Physical domain

ICD-clin

12.5

18.8

6.3 [−0.5, 13.0]

50.1% ± 134.4% (P = 0.018)

ICD-clin-DiSMOL

19.1

24.4

5.3 [−2.6, 13.2]

27.8% ± 157.8% (P = 0.072 [ns])

DiSMOL

110.8

118.6

7.7 [−9.1, 24.6]

7.0% ± 334.5% (P = 0.237 [ns])

  1. Trazodone did not demonstrate statistically significant changes across any domain (P-values ≥ 0.071) when using the DiSMOL method. However, when assessed using ICD-clin and ICD-clin-DiSMOL, trazodone was associated with a statistically significant increase in daytime impairment symptoms across the all domains, cognition domain, emotional domain, and physical domain categories. Notably, the observed increase in daytime impairment may be overestimated when relying on ICD-clin and ICD-clin-DiSMOL.
  2. CI confidence interval, DiSMOL Disease-Specific Medical Ontology Learning, ICD International Classification of Diseases, ns not significant, SD standard deviation.
  3. aThe follow up period was from 182 days before treatment to approximately 3 months (mean) during treatment.
  4. bAn increase shows worsening of daytime functioning (more daytime impairment symptoms) after treatment compared with before treatment, while a decrease shows an improvement in daytime functioning (fewer daytime impairment symptoms).