Table 2. Results for school absenteeism and suicidal ideation.

References	Suicidal ideation construct	School absenteeism construct	Subgroups	OR (unless otherwise specified)	95% confidence interval	P value	aORa	95% confidence interval	P value
Almansour and Siziya [50]	Past 1 year suicidal ideation	Truancy (not defined)		1.43	1.40–1.47	< 0.001 *	1.26	1.22–1.31	< 0.001 *
Asante et al. [46]	Past 1 year suicidal ideation	Missed class without permission three or more times in the past 30 days		1.52	1.20–1.92	< 0.001	1.40	1.06–1.84	< 0.05
	Past 1 year suicidal plan			1.39	1.11–1.73	< 0.001	1.31	1.01–1.69	< 0.05
Chen et al. [47]	Seriously considered suicide in the past 12 months	Number of days absent from school in the past 30 days (continuous)	Urban	–	–	–	1.1	1.0–1.2	0.046
			Rural	–	–	–	–	–	NSk
Cheng et al. [51]	Seriously considered suicide in the past 12 months	6 or more days absent from school in the past 30 days		Ref	–	–	Ref	–	–
		1–5 days absent from school in the past 30 days		0.7g	0.36–1.36	0.2902	0.66	0.31–1.39	0.2742
		0 days absent from school in the past 30 days		0.45 g	0.28–0.72	0.0011	0.49	0.28–0.84	0.0103
Choquet and Menke [53]	Past 1 year suicidal ideationb	Absent from class in the past 12 months (unclear definition)		2.13 *	1.55–2.93 *	0.0001 *	–	–	–
Davaasambuu et al. [48]	Suicide plan in the past 12 months	Missed class in the past 30 days without permission		1.87	1.59–2.20	< 0.0001	1.53	1.25–1.87	< 0.0001
			Urban	1.98	1.55–2.53	< 0.0001	1.54	1.14–2.07	0.005
			Rural	1.85	1.48–2.31	< 0.0001	1.55	1.19–2.02	0.001
De Man et al. [59]	Suicidal ideation on scale for suicidal ideation (continuous)	Absenteeism in terms of class periods absent (continuous)		r = 0.08h		> 0.001	–		Absenteeism reduces risk
Epstein and Spirito [52]	Considered suicide in the past 12 months	Skipped school because felt unsafe in the past 30 days		–	–	–	–	–	NSk
	Made suicide plan in the past 12 months			–	–	–	–	–	NSk
Fergusson et al. [49]	Suicidal ideation between age 14 and 21 years	Any truancy between age 11 and 15 years		–	–	–	2.09 * i	1.59–2.69 *	< 0.0001 *
Kandel et al. [54]	Suicidal ideation (composite measure of few weeks and 12 months) vs nonec	Cuts class (ever vs never)		1.74 *	1.22–2.49 *	0.0027 *	–	–	–
Lau et al. [45]	Suicidal ideation since earthquake 1 month ago	Absence from school since the earthquake when the school was not closed		2	1.50–2.65 *	< 0.001	1.48	1.05–2.09	< 0.05
Peltzer and Pengpid [57]	Seriously considered suicide in the past 12 months	In the past 30 days, 1 or more days missed class without permission		1.54	1.41–1.69 *	0.0001 *	0.97	0.78–1.2	> 0.05
			Male	–	–	–	0.67	0.48–0.94	< 0.05
			Female	–	–	–	1.19	0.92–1.54	> 0.05
Randall et al. [56]	Seriously considered suicide in the past 12 monthsd	Missed 3 or more days without permission in the past 30 days		–	–	–	0.31j	0.04–2.64	0.261
	Made suicide plan in the past 12 monthsd			–	–	–	0.81j	0.42–1.55	0.492
	Considered suicide or made plan past 12 monthse			2.03 *	1.36–3.03 *	0.0007 *
Sharma et al. [55]	Suicidal ideation in the past 12 months	3 or more days absent in the past 30 days without giving notice to the school		2.07	1.28–3.36	< 0.01	0.64	0.35–1.16	> 0.05
Taliaferro and Muehlenkamp [60]	Suicidal ideation in the past 12 months (and never suicide attempts)	Missed school due to feeling unsafe in the past 30 days	Male	–	–	–	0.70	0.51–0.95	< 0.0033 l
			Female	–	–	–	0.74	0.56–0.97	> 0.0033l
Wilson et al. [58]	Past 12 months, seriously considered attempting suicide or made a planf	Missed 3 or more days in the past 30 days without permission		1.32*	0.84–1.85*	0.1322*	–	–	–

Data is missing where either univariate or multivariate analyses were not reported in the paper and could not be calculated from the data available

^*Results calculated from data available within the paper; OR odds ratio, NS non-significant, significance cutoff not reported in paper; Variables are binary unless otherwise stated; statistically significant results in bold (using P < 0.05 cutoff unless otherwise stated)

^aAdjusted for covariates listed in Table 1

^bDichotomised to never suicidal ideation vs rarely/often and male and female combined

^cDichotomised to suicidal ideation vs some/high and male and female combined

^dMultinomial logistic regression with no ideation vs ideation only vs ideation + plans as outcomes

^eDichotomised to no ideation vs ideation or plan

^fCombined suicidal ideation and planning

^gAdjusted for age and city only

^h r = correlation coefficient, one tailed P value < 0.001 considered as significance cutoff

ⁱConverted from log odds ratio

^jRelative risk ratio

^kResult not reported in multivariate model due to non-significance

^lNS > 0.0033 (Bonferroni correction)