Table 4.

Predictors of prevalence of correctly completed death certificates: Logistic regression analysis, univariate analysis

Death certificate characteristics	Unadjusted OR (95% CI)	p value	GP characteristics and associated parameters.	Unadjusted OR (95% CI)	p value
Elderly (age ≥ 60 yr.)	1.2 (0.8–1.8)	0.2	Age	0.8 (0.5–1.3)	0.3
Female	1.1 (0.8–1.6)	0.5	Female	0.6 (0.3–1.3)	0.1
Hospital Sizes			Medical School Sizes
▪ 400–700 beds	Reference		▪ > 200 graduates/ year	Reference
▪ 701–1000 beds	1.1 (0.7–1.7)	0.6	▪ 101–200 graduates/ year	0.9 (0.4–1.9)	0.7
▪ More than 1000	1.4 (0.9–2.0)	0.1	▪ 51–100 graduates/ year	0.5 (0.2–1.7)	0.2
COD disease groups			▪ ≤ 50 graduates/ year	1.6 (0.5–5.1)	0.3
▪ Cardiovascular disease	2.3 (1.6–3.5)	< 0.001	Future plans for specialist training
▪ Infectious disease	0.4 (0.2–0.7)	0.001	▪ Specialist training plans	1.3 (0.5–3.1)	0.6
▪ Cancer group	3.7 (2.2–6.0)	< 0.001	GPA Group
▪ Gastrointestinal disease	0.9 (0.6–1.5)	0.7	▪ GPA ≥ 3.5	3.5 (1.5–7.9)	0.003
▪ Pulmonary disease	1.2 (0.7–2.2)	0.4	Time spent during medical curriculum
▪ Endocrine disease	0.1 (0.04–0.4)	< 0.001	▪ Time spent ≥2 h	1.0 (0.4–2.3)	0.9
▪ External cause	0.6 (0.3–1.2)	0.1	COD coding experience
▪ Nephrology disease	0.7 (0.2–1.8)	0.4	▪ Have experience in completing DCs by themselves during medical school	1.8 (0.9–3.8)	0.1
▪ Neurologic disease	0.5 (0.1–1.7)	0.2	▪ Have experience completing DCs by themselves >5cases/month after graduation	1.2 (0.6–2.5)	0.6
▪ Musculoskeletal and rheumatologic disease	0.2 (0.02–1.5)	0.07	▪ Established COD coding training program in work hospital	0.9 (0.4–1.9)	0.8
▪ Unknown cause of death/ Other	0.4 (0.2–1.1)	0.05	Understands that DCs are used in public health research and policy	1.5 (0.7–2.9)	0.2

Abbreviations: DCs Death certificates, OR Odds ratio, CI Confidence interval, COD Cause of death, DC Death certificate, DCs Death certificates, GP General practitioner, GPA Grade point average, yr. year