Sage Journals: Discover world-class research

Abstract

Background:

Young people with a mental disorder often perform poorly at school and can fail to complete high school. This study aims to compare scholastic performance and high school completion of young people hospitalised with a mental disorder compared to young people not hospitalised for a mental disorder health condition by gender.

Method:

A population-based matched case-comparison cohort study of young people aged ⩽18 years hospitalised for a mental disorder during 2005–2018 in New South Wales, Australia using linked birth, health, education and mortality records. The comparison cohort was matched on age, gender and residential postcode. Generalised linear mixed modelling examined risk of school performance below the national minimum standard and generalised linear regression examined risk of not completing high school for young people with a mental disorder compared to matched peers.

Results:

Young males with a mental disorder had over a 1.7 times higher risk of not achieving the national minimum standard for numeracy (adjusted relative risk: 1.71; 95% confidence interval: [1.35, 2.15]) and reading (adjusted relative risk: 1.99; 95% confidence interval: [1.80, 2.20]) compared to matched peers. Young females with a mental disorder had around 1.5 times higher risk of not achieving the national minimum standard for numeracy (adjusted relative risk: 1.50; 95% confidence interval: [1.14, 1.96]) compared to matched peers. Both young males and females with a disorder had around a three times higher risk of not completing high school compared to peers. Young males with multiple disorders had up to a sixfold increased risk and young females with multiple disorders had up to an eightfold increased risk of not completing high school compared to peers.

Conclusion:

Early recognition and support could improve school performance and educational outcomes for young people who were hospitalised with a mental disorder. This support should be provided in conjunction with access to mental health services and school involvement and assistance.

Keywords

Mental disorders academic performance high school completion

Background

Internationally, mental disorders account for around 16% of the global burden of disease among young people (Kessler et al., 2007). In Australia, around 14% of young people are estimated to experience a mental disorder in 2013–2014 (Lawrence et al., 2016). Mental disorders can have an adverse impact on a young person’s home life, ability to relate to peers, and on their academic performance (Leach and Butterworth, 2012; Suldo et al., 2014). The presence of a mental disorder is usually apparent by age 14 (Kessler et al., 2007); therefore, early recognition, treatment and support are key to improving health and education outcomes.

Poor mental health can be both a cause and effect of academic problems (DeSocio and Hootman, 2004). Inconsistent school attendance can be an early indicator of poor mental health among young people and also the social circumstances contributing to the emergence of a mental disorder (DeSocio and Hootman, 2004). Single or comorbid mental disorders can negatively influence a young person’s educational performance through their effect on a student’s engagement with learning (Breslau et al., 2011; Productivity Commission, 2020). The perception of poor academic competence can in turn result in decreased motivation to engage at school (Breslau et al., 2011). In particular, early-onset depressive disorder in otherwise capable students has been associated with poor performance on school assessments, lower grades and poor peer relationships (Fletcher, 2008).

Some students with a mental disorder struggle to complete high school (Bowman et al., 2017; Leach and Butterworth, 2012; Mental Health Commission, 2021). An estimated one-third of students diagnosed with a mental disorder during adolescence do not enter tertiary college in the United States (Breslau et al., 2008). Poor academic performance can be both a mediator for dropping out of high school (Quiroga et al., 2013) and a predictor of high school dropout (Bowman et al., 2017; Sagatun et al., 2014) for students with a mental disorder.

The consequences of not completing high school can include a lower income, reduced stability in relationships and employment, greater reliance on government financial assistance and poor general health associated with lower socioeconomic status (Breslau et al., 2008, 2011; Kessler et al., 1995; Larson et al., 2017). The likelihood of not completing high school may vary for different types of mental disorders (Breslau et al., 2008) and by gender (Leach and Butterworth, 2012), differences which to date have been underexamined (Deighton et al., 2018). This study aims to compare scholastic performance and high school completion of young people hospitalised for a mental disorder compared to young people not hospitalised for a mental disorder by gender.

Method

A retrospective population-level case-comparison cohort study of young people hospitalised for a mental disorder aged ⩽18 years in New South Wales (NSW), Australia using linked birth, health, education and mortality administrative data collections from 1 January 2005 to 31 December 2018 (Mitchell et al., 2019). Ethical approval and a waiver of consent were obtained from the NSW Population and Health Services Research Ethics Committee (2018HRE0904).

Data sources

Information on hospital service use was obtained from emergency department (ED) visit and hospital admission data collections. ED visits to public hospitals included information on arrival and departure times, visit type and provisional diagnosis. Hospitalisations included admissions to public and private hospitals, and contained demographic information, diagnoses, separation type (i.e. hospital transfers, death) and clinical procedures. The NSW Registry of Births, Deaths and Marriages provided mortality data for the study period and young people who died during the study period were excluded from analyses (Supplementary Figure 1).

Information on the number and type of ambulatory specialist mental health services at public hospitals was obtained from the ambulatory mental health client contacts database from 1 January 2006 to 30 September 2019. This included mental health day programmes, public psychiatric outpatients and outreach services.

School performance and parental demographic information was obtained from annual National Assessment Plan for Literacy and Numeracy (NAPLAN) assessments conducted in May during 2008 to 2018 for government, Catholic and independent schools. Assessments were conducted on young people in primary school grades 3 (7–9 years of age) and 5 (9–11 years of age), and secondary school grades 7 (11–13 years of age) and 9 (13–15 years of age) and included assessments in both numeracy and reading. Within each assessment, scores represent the same level of achievement over time (Australian Curriculum Assessment and Reporting Authority [ACARA], 2019). Each score is translated into proficiency bands that indicate whether performance was above, at or below the national minimum standard (NMS). Inability to achieve the NMS indicates that a young person will have difficulty making progress in school without assistance (ACARA, 2017).

Information on a young person’s attendance, absence, withdrawal (e.g. philosophical objections to testing or religious beliefs) or exemption due to significant disability was obtained (Supplementary Table 1). Young people exempt from sitting an assessment due to severe disability or language difficulties were rated as scoring below the NMS as per technical guidelines (ACARA, 2014).

A young person was identified as having a language background other than English (LBOTE) if either they or their parent or guardians spoke a language other than English at home (ACARA, 2019). Where there were multiple records of the parents’ level of education, the highest level of education of either parent was identified. Information on high school completion at years 10 (15–16 years of age), 11 (17–18 years of age) and 12 (17–18 years of age) were obtained through the Record of School Achievement, and the Higher School Certificate.

The Centre for Health Record Linkage (CHeReL) identified the population comparison group and linked the birth, health and education records using probabilistic record linkage. Upper and lower probability cut-offs for a link were 0.75 and 0.25 and record groups with probabilities between the cut-offs were clerically reviewed.

Case inclusion criteria

Young people with a year of birth ⩾1997 who were aged ⩽18 years at hospital admission with a principal or additional diagnosis of a mental disorder in up to 50 additional diagnosis classifications (International Classification of Diseases, 10th Revision, Australian Modification [ICD-10-AM]) during 1 January 2005 to 31 December 2018 (Supplementary Table 2). Cases were included if the young person completed both the numerical and reading NAPLAN assessments in a school grade. The number of mental disorders experienced by the young person was categorised as 1 or ⩾2 disorders.

Population-comparison group criteria

A population-based comparison group not hospitalised for a mental disorder from 1 July 2001 to 31 December 2018 was randomly selected from NSW birth records matched 1:1 on age, gender and residential postcode to their counterpart. The comparison selection time frame included a 3.5-year wash out period prior to the case selection time frame to avoid the potential selection of comparisons that may have been hospitalised for a mental disorder prior to the case criteria time frame.

Geographical location and socioeconomic status

Geographical location of residence was assigned using the Australian Statistical Geographical Standard (ASGS) and the young person’s residential postcode based on distance to service centres (Australian Bureau of Statistics, 2013). The ASGS categories were classified as urban (i.e. major cities) and rural (i.e. inner and outer regional, remote and very remote). NAPLAN records provided the remoteness area of the school and was categorised as major city, inner regional, outer regional or remote (Australian Bureau of Statistics, 2013). The young person’s postcode of residence was used to partition socioeconomic disadvantage into quintiles from most (i.e. 1) to least disadvantaged (i.e. 5) (Australian Bureau of Statistics, 2011).

Identification of chronic health conditions

Common comorbidities for young people were identified from prevalence studies (Edwards et al., 2012; Miller et al., 2013; Mitchell et al., 2017) and were those that would reasonably be expected to last 12 months or result in the need for ongoing healthcare (Miller et al., 2013). For this study, a chronic health condition was identified using a 3-year lookback period (to 1 January 2002) and diagnosis classifications from ICD-10-AM (Supplementary Table 3).

Data organisation and analysis

Data analysis was conducted using SAS 9.4 (SAS Institute, Cary, NC). All hospital episodes of care related to the same event were linked to form a period of care. Chi-square tests of independence and Wilcoxon Mann–Whitney tests were used to examine characteristics of young people hospitalised for a mental disorder and their counterpart. The number of ED visits, hospital admissions and hospital length of stay (LOS) before and during and also after the index admission were identified for both the young person with a mental disorder and their matched peer for each NAPLAN assessment. The calculation of hospital LOS was cumulative and included transfers between hospitals.

To examine school performance at each grade, Poisson regression models with robust variance estimators (Zou, 2004) using PROC GENMOD were used to assess the difference in proportions of performances below the NMS for each assessment domain for the school grades 3, 5, 7 and 9 for young people hospitalised for a mental disorder and their matched peers. Adjusted relative risks (ARRs) and 95% confidence intervals (CIs) were calculated. For each domain, sequential forward selection was used to sequentially add covariates to the model (Mitchell et al., 2019) and significance was assessed using p-values (p < 0.05) to examine the overall effect in the model. The final models included mental disorder status, gender, comorbidity status (Y/N), socioeconomic status, highest level of education for any parent/guardian (i.e. ⩽Year 12; certificate I-IV, trade or diploma; bachelor or higher degree or other), highest level of any parent/guardian occupation (i.e. senior manager/qualified professional; business manager/associate professional; trades/clerks/skilled office/sales and service; machine operators/hospitality/assistants/labourers; or not in paid work in last 12 months) and a log of hospital LOS. As comparison group members could have nil hospital LOS, a constant value was added to LOS before transformation (Woodward, 2005) (Supplementary Table 4).

Generalised linear mixed modelling (GLMM) was conducted to perform multi-level modelling of school performance below the NMS for the numeracy and reading assessment domains for young people and their counterpart who had completed multiple grades of schooling. For each domain, PROC GLIMMIX was used with a binary distribution, log link function and Kenward and Roger denominator degrees of freedom. The residual option of the random statement was used to model R-side covariance and data were analysed to account for within student correlation in the longitudinal data and repeated measurements using an autoregressive covariance structure. ARRs and 95% CIs were generated. The final models for all persons included mental health status, NAPLAN grade (i.e. 3, 5, 7, 9), gender, comorbidity status (Y/N), LBOTE, socioeconomic status, highest level of education for any parent/guardian (i.e. bachelor or higher degree or other), log of hospital LOS and a school sector (i.e. government, Catholic, independent). Each model was then separately run by gender using the same covariates, excluding gender.

Factors associated with high school completion at either year 10, 11 or 12 for young people hospitalised for a mental disorder compared to their counterparts was examined using generalised linear regression using PROC GENMOD. For each grade, models were fitted using generalised estimating equations (GEEs) with binomial distribution and a log function. ARR and 95% CIs were calculated. The final models for all persons included mental health status, gender, comorbidity status (Y/N), LBOTE, socioeconomic status, highest level of education for any parent/guardian and geographic location of residence. Where sample sizes allowed, modelling was conducted separately by gender for all mental disorders and the number of mental disorders using the same covariates, excluding gender.

Results

During 2015–2018, there were 7069 young people hospitalised with a mental disorder for whom a matched comparison could be identified and who completed the NAPLAN assessments in Grade 3; 6203 in Grade 5; 5009 in Grade 7; and 3374 in Grade 9. There were 8575 young people with a mental disorder hospitalised with a matched peer who could have completed year 10; 7497 year 11; and 6063 year 12 of high school.

Across grades 3 to 9, there was a higher proportion of young females (57.9–63.1%) compared to males, and a higher proportion of young people in urban locations (70.7–73.2%) compared to rural locations hospitalised for a mental disorder. Most young people (⩾83%) were from English-speaking households and almost all (⩾96%) had no other health condition identified. Young people with a mental disorder had a higher proportion of parents completing year 11 high school or equivalent and were not in paid employment in the last 12 months than their matched peers for each grade. Young people hospitalised for a mental disorder had higher mean ED visits, hospital admissions, hospital LOS, and pre- and post-ambulatory mental health service contacts than their matched counterparts in each grade (Table 1).

Table 1.

Demographic and healthcare use characteristics of young people hospitalised for a mental health condition and their matched comparison by school grade, linked health and school performance data NSW, 2005–2018.

Characteristics	Grade 3^a				Grade 5^b				Grade 7^c				Grade 9^d
	Case (n = 7069)		Comparison (n = 7069)		Case (n = 6203)		Comparison (n = 6203)		Case (n = 5009)		Comparison (n = 5009)		Case (n = 3374)		Comparison (n = 3374)
	n	%	n	%	n	%	n	%	n	%	n	%	n	%	n	%
Gender
Male	2977	42.1	2977	42.1	2478	40.0	2478	40.0	1874	37.4	1874	37.4	1246	36.9	1246	36.9
Female	4092	57.9	4092	57.9	3725	60.1	3725	60.1	3135	62.6	3135	62.6	2218	63.1	2218	63.1
Location of residence^e
Urban	4999	70.7	4999	70.7	4424	71.3	4424	71.3	3603	71.9	3603	71.9	2470	73.2	2470	73.2
Rural	2069	29.3	2069	29.3	1778	28.7	1778	28.7	1405	28.1	1405	28.1	903	26.8	903	26.8
Socioeconomic status^f
Most disadvantaged	1306	18.5	1306	18.5	1154	18.6	1154	18.6	905	18.1	905	18.1	588	17.4	588	17.4
2	1610	22.8	1610	22.8	1424	23.0	1424	23.0	1125	22.5	1125	22.5	726	21.5	726	21.5
3	1582	22.4	1582	22.4	1369	22.1	1369	22.1	1101	22.0	1101	22.0	727	21.6	727	21.6
4	810	11.5	810	11.5	684	11.0	684	11.0	534	10.7	534	10.7	378	11.2	378	11.2
Least disadvantaged	1758	24.9	1758	24.9	1569	25.3	1569	25.3	1342	26.8	1342	26.8	953	28.3	953	28.3
LBOTE^g
Non-LBOTE	6018	85.1	5856	82.8	5277	85.1	5147	83.0	4254	84.9	4148	82.8	2834	84.0	2768	82.0
LBOTE	1031	14.6	1198	17.0	916	14.8	1048	16.9	750	15.0	855	17.1	538	16.0	601	17.8
Not known	20	0.3	15	0.2	10	0.2	8	0.1	5	0.1	6	0.1	<5	–	5	0.2
Health condition
0	6769	95.8	6967	98.6	5936	95.7	6107	98.5	4805	95.9	4922	98.3	3248	96.3	3309	98.1
⩾1	300	4.2	102	1.4	267	4.3	96	1.6	204	4.1	87	1.7	126	3.8	65	1.9
Parent highest level of education
Year 11 or equivalent	788	11.2	571	8.1	715	11.5	511	8.2	544	10.9	399	8.0	351	10.4	251	7.4
Year 12 or equivalent	307	4.3	321	4.5	290	4.7	310	5.0	247	4.9	243	4.9	164	4.9	170	5.0
Certificate I-IV or trade	2119	30.0	2215	31.3	1862	30.0	1975	31.8	1480	29.6	1556	31.1	958	28.4	1045	31.0
Advanced diploma/diploma	1144	16.2	1203	17.0	1016	16.4	1053	17.0	823	16.4	879	17.6	547	16.2	583	17.3
Bachelor degree or higher	2516	35.6	2583	36.5	2235	36.0	2267	36.6	1853	37.0	1869	37.3	1313	38.9	1278	37.9
Not stated/not known	195	2.8	176	2.5	85	1.4	87	1.4	62	1.2	63	1.3	41	1.2	47	1.4
Parent highest occupation
Senior manager/qualified professional	2148	30.4	2267	32.1	1922	31.0	2003	32.3	1638	32.7	1671	33.4	1141	33.8	1144	33.9
Business management/ associate professional	1628	23.0	1872	26.5	1461	23.6	1655	26.7	1184	23.6	1364	27.2	827	24.5	935	27.7
Trades, clerks, skilled office, sales and service	1386	19.6	1490	21.1	1203	19.4	1309	21.1	951	19.0	1047	20.9	634	18.8	706	20.9
Machine operators, hospitality, assistants, labourers	921	13.0	867	12.3	793	12.8	772	12.5	636	12.7	607	12.1	383	11.4	386	11.4
Not in paid work in last 12 months	745	10.5	373	5.3	641	10.3	308	5.0	460	9.2	211	4.2	283	8.4	120	3.6
Not known	241	3.4	200	2.8	183	3.0	156	2.5	140	2.8	109	2.2	106	3.1	83	2.5
Health care use	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
ED visits	1.6	4.9	0.7	2.5	1.9	5.6	0.8	2.5	2.1	6.1	0.9	2.5	3.2	7.0	1.2	2.8
Hospital admissions	1.2	4.9	0.2	1.0	1.3	5.3	0.2	1.1	1.4	5.5	0.2	1.2	1.8	6.0	0.3	2.3
Hospital length of stay	3.9	21.4	0.7	3.8	4.2	23.3	0.7	3.6	5.6	32.7	0.7	3.3	9.6	40.5	0.8	4.5
Ambulatory mental health service contacts
Prior to index admission	2.9	5.5	0.2	1.3	3.0	5.5	0.2	1.4	3.1	5.7	0.2	1.5	2.9	5.2	0.2	1.4
Post index admission	5.3	7.4	0.2	2.1	5.6	7.6	0.2	2.2	5.8	7.4	0.2	2.2	5.8	7.2	0.2	1.5

LBOTE: language background other than English; ED: emergency department.

Grade 3 chi-square test: LBOTE p < 0.001; Health conditions p < 0.0001; Parent highest level of education p < 0.0001; Parent highest occupation p < 0.0001; and Wilcoxon Mann–Whitney tests: ED visits p < 0.0001; Hospital admissions p < 0.0001; Hospital length of stay p < 0.0001; and Mental health contacts prior p < 0.0001 and post p < 0.0001 index admission.

Grade 5 chi-square test: LBOTE p < 0.01; Health conditions p < 0.0001; Parent highest level of education p < 0.0001; Parent highest occupation p < 0.0001 and Wilcoxon Mann–Whitney tests: ED visits p < 0.0001; Hospital admissions p < 0.0001; Hospital length of stay p < 0.0001; and Mental health contacts prior p < 0.0001 and post p < 0.0001 index admission.

Grade 7 chi-square test: LBOTE p < 0.05; Health conditions p < 0.0001; Parent highest level of education p < 0.0001; Parent highest occupation p < 0.0001; and Wilcoxon Mann–Whitney tests: ED visits p < 0.0001; Hospital admissions p < 0.0001; Hospital length of stay p < 0.0001; and Mental health contacts prior p < 0.0001 and post p < 0.0001 index admission.

Grade 9 chi-square test: LBOTE p < 0.0001; Health conditions p < 0.0001; Parent highest level of education p < 0.001; Parent highest occupation p < 0.0001; and Wilcoxon Mann–Whitney tests: ED visits p < 0.0001; Hospital admissions p < 0.0001; Hospital length of stay p < 0.0001; and Mental health contacts prior p < 0.0001 and post p < 0.0001 index admission.

Excludes not known location of residence for one case and one comparison in each grade.

Excludes not known socioeconomic status for three cases and three comparisons in grades 3 and 5 and two cases and two comparisons in grades 7 and 9.

Language background other than English and excludes two not known LBOTE for cases and five comparisons in grade 9.

While all young people predominantly attended government schools in major cities, the type of school attended and its geographic location differed for young people with a mental disorder compared to matched peers for each grade. There was a higher proportion of young people with a mental disorder attending government schools and attending schools in major cities compared to matched peers. At least double the proportion of young people hospitalised with a mental disorder did not achieve the NMS for their school assessments across grades 3–9 compared to their matched counterparts (Table 2).

Table 2.

School and NAPLAN assessment characteristics of young people hospitalised for a mental health condition and their matched comparison by school grade, linked health and school performance data NSW, 2005–2018.

Characteristics	Grade 3^a				Grade 5^b				Grade 7^c				Grade 9^d
	Case (n = 7069)		Comparison (n = 7069)		Case (n = 6203)		Comparison (n = 6203)		Case (n = 5009)		Comparison (n = 5009)		Case (n = 3374)		Comparison (n = 3374)
	n	%	n	%	n	%	n	%	n	%	n	%	n	%	n	%
School sector
Government	5288	74.8	4907	69.4	4613	74.4	4298	69.3	3688	73.6	3425	68.4	2464	73.0	2249	66.7
Catholic	1163	16.5	1496	21.2	1064	17.2	1312	21.2	878	17.5	1093	21.8	615	18.2	779	23.1
Independent	616	8.7	666	9.4	526	8.5	593	9.6	443	8.8	491	9.8	295	8.7	346	10.3
Remoteness area of school
Major city	4979	70.4	4795	67.8	4394	70.8	4255	68.6	3617	72.2	3514	70.2	2481	73.5	2413	71.5
Inner regional	1641	23.2	1851	26.2	1431	23.1	1575	25.4	1124	22.4	1244	24.8	727	21.6	793	23.5
Outer regional/remote	447	6.3	423	6.0	378	6.1	373	6.0	267	5.3	250	5.0	164	4.9	167	5.0
NAPLAN assessment
Numeracy (Below NMS)	601	8.5	269	3.8	589	9.5	271	4.4	340	6.8	147	2.9	242	7.2	95	2.8
Reading (Below NMS)	598	8.5	267	3.8	644	10.4	359	3.8	399	8.0	185	3.7	342	10.1	170	5.0

NAPLAN: National Assessment Plan for Literacy and Numeracy; NMS: national minimum standard.

Grade 3 chi-square test: School sector (excluding two home schooled) p < 0.0001; Remoteness area of school (excluding two missing) p < 0.001; All NAPLAN assessments p < 0.0001.

Grade 5 chi-square test: School sector p < 0.0001; Remoteness area of school p < 0.01. All NAPLAN assessments p < 0.0001.

Grade 7 chi-square test: School sector p < 0.0001; Remoteness area of school (excluding two missing) p < 0.0001. All NAPLAN assessments p < 0.0001.

Grade 9 chi-square test: School sector p < 0.0001; Remoteness area of school (excluding three missing) p = 0.1. All NAPLAN assessments p < 0.0001.

For young people who were hospitalised with a mental disorder and completed school assessments, the most common diagnoses for males between school year 3 to year 9 were conduct (27.8–41.9%) and anxiety (29.3–31.0%) disorders. For females, the most common diagnoses were anxiety (38.2–39.1%) and mood (23.0–27.2%) disorders. Young females (11.8–14.0%) had a higher proportion of eating disorders than young males (2.0–2.3%) in all grades (Table 3). In the comparison of high school completions from year 10 to year 12, anxiety (26.9–27.7%), conduct (21.5–23.1%) and substance (22.7–24.2%) disorders were the most common diagnoses among young males. Anxiety (37.1–37.8%) and mood (30.2–30.3%) disorders were the most common diagnoses for young females in this schooling period (Table 4).

Table 3.

Characteristics of young people hospitalised for a mental disorder by school grade, linked health and school performance data NSW, 2005–2018.

Characteristic	NAPLAN Grade 3				NAPLAN Grade 5				NAPLAN Grade 7				NAPLAN Grade 9
	Male (n = 2977)		Female (n = 4092)		Male (n = 2478)		Female (n = 3725)		Male (n = 1874)		Female (n = 3135)		Male (n = 1246)		Female (n = 2128)
	n	%	n	%	n	%	n	%	n	%	n	%	n	%
Mental disorder
Substance disorders	363	12.2	341	8.3	334	13.5	312	8.4	287	15.3	268	8.5	232	18.6	192	9.0
Psychotic disorders	81	2.7	61	1.5	69	2.8	57	1.5	53	2.8	52	1.7	45	3.6	38	1.8
Mood disorders	351	11.8	940	23.0	332	13.4	920	24.7	307	16.4	830	26.5	231	18.5	578	27.2
Depression	290	9.7	744	18.2	262	10.6	700	18.8	241	12.9	628	20.0	172	13.8	446	21.0
Anxiety disorders	873	29.3	1563	38.2	754	30.4	1425	38.3	581	31.0	1218	38.9	366	29.4	832	39.1
Eating disorders	61	2.0	484	11.8	56	2.3	469	12.6	42	2.2	423	13.5	25	2.0	297	14.0
Conduct disorders	1248	41.9	703	17.2	933	37.7	542	14.6	604	32.2	344	11.0	347	27.8	191	9.0
Co-occurring disorders
1 disorder	2727	91.6	3541	86.5	2248	90.7	3196	85.8	1678	89.5	2676	85.4	1109	89.0	1814	85.2
⩾2 disorders	250	8.4	551	13.5	230	9.3	529	14.2	196	10.5	459	14.6	137	11.0	314	14.8

NAPLAN: National Assessment Plan for Literacy and Numeracy.

Table 4.

Characteristics of young people hospitalised for a mental disorder by school completion year, linked health and school performance data NSW, 2005–2018.

Characteristic	Year 10				Year 11				Year 12
	Male (n = 3233)		Female (n = 5342)		Male (n = 2766)		Female (n = 4731)		Male (n = 2200)		Female (n = 3863)
	n	%	n	%	n	%	n	%	n	%	n	%
Mental disorder
Substance disorders	735	22.7	570	10.7	639	23.1	506	10.7	532	24.2	424	11.0
Psychotic disorders	152	4.7	122	2.3	131	4.7	113	2.4	104	4.7	92	2.4
Mood disorders	644	19.9	1613	30.2	564	20.4	1431	30.2	456	20.7	1172	30.3
Depression	438	13.5	1193	22.3	376	13.6	1047	22.1	295	13.4	853	22.1
Anxiety disorders	895	27.7	2017	37.8	757	27.4	1774	37.5	592	26.9	1435	37.1
Eating disorders	59	1.8	641	12.0	52	1.9	579	12.2	44	2.0	482	12.5
Conduct disorders	748	23.1	379	7.1	623	22.5	328	6.9	472	21.5	258	6.7
Co-occurring disorders
1 disorder	2739	84.7	4363	81.7	2341	84.6	3867	81.7	1864	84.7	3140	81.3
⩾2 disorders	494	15.3	979	18.3	425	15.4	864	18.3	336	15.3	723	18.7

Young males hospitalised for a mental disorder had over a 1.7 times higher risk of not achieving the NMS on school assessments for numeracy (ARR: 1.71; 95% CI: [1.35, 2.15]) and reading (ARR: 1.99; 95% CI: [1.80, 2.20]) compared to matched counterparts. For psychotic and conduct disorders, young males had around twice the risk, and for anxiety males had over 1.5 times higher risk of not achieving the NMS in school assessments compared to matched peers. Young females with a mental disorder had around 1.5 times higher risk of not achieving the NMS for numeracy (ARR: 1.50; 95% CI: [1.14, 1.96]) compared to matched peers. For young females, there was around twice the risk of not achieving the NMS for conduct disorders for each assessment compared to matched peers. Both young males and females with depression had around a 1.5 times higher risk of not achieving the NMS for numeracy (ARR: 1.56; 95% CI: [1.22, 1.99]) and reading (ARR: 1.45; 95% CI: [1.04, 2.02]), respectively, compared to matched counterparts. There was no increased risk of not achieving the NMS for mood disorders for males and females or for eating disorders for females compared to matched peers. Both male and female students with ⩾2 disorders had twice the risk of not achieving the NMS for numeracy (ARR: 2.35; 95% CI: [1.28, 4.32] and ARR: 2.39; 95% CI: [1.86, 3.08], respectively) compared to matched peers (Table 5).

Table 5.

Multilevel model of mental disorders associated with a below NMS NAPLAN assessment for young people with an index hospitalisation for during 2005–2018 compared to a matched comparison by gender and assessment type, linked health and school performance data NSW.

	Numeracy		Reading
	ARR^a	95% CI	ARR^a	95% CI
All persons
Any disorder^a	1.62*	[1.35, 1.94]	1.53*	[1.30, 1.81]
Substance disorders^b	1.37***	[1.02, 1.85]	1.35**	[1.07, 1.71]
Psychotic disorders^c	3.15*	[1.90, 5.23]	2.47**	[1.54, 3.95]
Mood disorders^d	1.12	[0.89, 1.40]	1.13	[0.93, 1.38]
Depression^e	1.56**	[1.22, 1.99]	1.29***	[1.04, 1.62]
Anxiety disorders^f	2.03*	[1.74, 2.36]	1.72*	[1.49, 1.98]
Conduct disorders^g	2.01*	[1.58, 2.55]	2.15*	[1.69, 2.72]
Number of disorders
1 disorder^h	1.56*	[1.29, 1.88]	1.53*	[1.29, 1.82]
⩾2 disordersⁱ	2.39*	[1.86, 3.08]	1.41	[0.87, 2.28]
Male
Any disorder^a	1.71*	[1.35, 2.15]	1.99*	[1.80, 2.20]
Substance disorders^b	1.23	[0.86, 1.77]	1.22	[0.93, 1.60]
Psychotic disorders^c	2.59**	[1.38, 4.84]	2.61**	[1.44, 4.72]
Mood disorders^d	0.92	[0.60, 1.40]	0.99	[0.71, 1.39]
Depression^e	1.39	[0.94, 2.06]	1.45**	[1.04, 2.02]
Anxiety disorders^f	1.87*	[1.49, 2.35]	1.79*	[1.46, 2.19]
Eating disorders	NA	NA	NA	NA
Conduct disorders^g	2.02*	[1.55, 2.65]	2.16*	[1.65, 2.83]
Number of disorders
1 disorder^h	1.56*	[1.29, 1.88]	1.53*	[1.29, 1.82]
⩾2 disordersⁱ	2.35**	[1.28, 4.32]	1.40	[0.87, 2.27]
Female
Any disorder^a	1.50**	[1.14, 1.96]	1.16	[0.91, 1.49]
Substance disorders^b	1.87	[0.51, 6.87]	0.85	[0.32, 2.24]
Psychotic disorders^c	4.25**	[1.85, 9.79]	2.39***	[1.06, 5.44]
Mood disorders^d	1.20	[0.92, 1.56]	1.22	[0.95, 1.57]
Depression^e	1.70**	[1.24, 2.34]	1.21	[0.91, 1.61]
Anxiety disorders^f	2.17*	[1.76, 2.67]	1.72*	[1.41, 2.09]
Eating disorders^j	0.77	[0.49, 1.21]	0.68	[0.43, 1.08]
Conduct disorders^g	2.12**	[1.29, 3.49]	2.07**	[1.27, 3.39]
Number of disorders
1 disorder^h	1.56*	[1.29, 1.88]	1.53*	[1.29, 1.82]
⩾2 disordersⁱ	2.39*	[1.86, 3.08]	1.40	[0.87, 2.27]

ARR: adjusted relative risk; CI: confidence interval; NAPLAN: National Assessment Plan for Literacy and Numeracy; NMS: national minimum standard; LBOTE: language background other than English; LOS: length of stay; NA: not applicable as cell sizes too small.

Adjusted for disorder present, gender, health condition, LBOTE, socioeconomic status, NAPLAN grade, parental education, school sector and hospital LOS. Excludes 20 young people with missing socioeconomic status, 7 with missing LBOTE and 2 who were home schooled.

All persons and males numeracy excludes LBOTE; males reading excludes LBOTE and hospital LOS; females reading excludes LBOTE.

All persons numeracy and reading exclude LBOTE and hospital LOS; males numeracy excludes LBOTE and hospital LOS; females reading excludes LBOTE and hospital LOS.

All persons numeracy and reading exclude health condition, LBOTE, hospital LOS and school sector; males and females numeracy and reading exclude health condition, LBOTE, socioeconomic status, hospital LOS and school sector.

All persons, males and females numeracy and reading exclude health condition, LBOTE and hospital LOS.

All persons numeracy excludes health condition and hospital LOS and reading excludes health condition; males numeracy and reading exclude health condition, LBOTE and hospital LOS; females numeracy excludes health condition and hospital LOS and reading excludes health condition, LBOTE and hospital LOS.

All persons numeracy excludes health condition and LBOTE and reading excludes health condition, LBOTE and socioeconomic status; males numeracy excludes health condition and reading excludes health condition and LBOTE; females numeracy and reading exclude health condition.

All persons and females numeracy excludes health condition, LBOTE, and hospital LOS and reading excludes health condition and LBOTE; males numeracy excludes LBOTE and reading excludes health conditions and LBOTE.

All persons, males and females numeracy excludes LBOTE.

Females numeracy and reading exclude health condition, LBOTE, socioeconomic status, hospital LOS and school sector.

p < 0.0001; **p < 0.01; ***p < 0.05.

Demographic and healthcare use characteristics of young people hospitalised with a mental health disorder and their matched peers by school completion year are shown in Supplementary Table 5. Both young males and females hospitalised with a mental disorder had around a three times higher risk, and young people hospitalised for substance disorders had four times the risk, of not completing year 10, 11 or 12 of high school than their matched counterparts. Young people with mood, anxiety and conduct disorders had between three to four times higher risk of not completing year 10, 11 or 12 than matched peers. Young people with psychotic disorders had a three and four times higher risk of not completing year 11 or 12, respectively, compared to matched peers. Young females with an eating disorder had around twice the risk of not completing year 11 or 12 compared to matched counterparts. Young people with ⩾2 disorders had between five and seven times higher risk of not completing year 10, 11 or 12 compared to matched peers (Table 6).

Table 6.

Analysis of characteristics associated with not completing high school for young people with an index hospitalisation for a mental disorder compared to a matched comparison by school grade, linked health and school performance data NSW, 2005–2018.

	Year 10n = 8575 in each cohort			Year 11n = 7497 in each cohort			Year 12n = 6063 in each cohort
No school completion		N	%^a		n	%^a		n	%^a
Case		585	6.8		2522	33.6		2895	47.8
Comparison		174	2.0		775	10.3		764	12.6
Characteristics	n ^b	ARR^c	95% CI	n ^b	ARR^c	95% CI	n ^b	ARR^c	95% CI
Mental disorder^d
All persons	585	3.28	[2.77, 3.89]	2522	3.23	[2.98, 3.50]	2895	3.77	[3.48, 4.09]
Males	240	2.75	[2.14, 3.52]	991	2.86	[2.53, 3.34]	1121	3.54	[3.12, 4.01]
Females	345	3.80	[3.01, 4.79]	1531	3.53	[3.18, 3.93]	1774	3.93	[3.55, 4.36]
Mental disorder type
Substance disorders^e	133	4.43	[2.97, 6.62]	478	4.26	[3.46, 5.24]	513	4.58	[3.73, 5.63]
Psychotic disorders^f	27	NA		89	3.36	[2.18, 5.18]	102	4.57	[2.93, 7.13]
Mood disorders	146	2.78	[2.02, 3.82]	646	3.11	[2.66, 3.64]	810	4.19	[3.58, 4.91]
Depression^f	105	3.45	[2.29, 5.18]	494	3.48	[2.89, 4.20]	591	5.03	[4.12, 6.14]
Anxiety disorders	182	2.78	[2.09, 3.71]	842	3.14	[2.73, 3.60]	942	3.67	[3.20, 4.22]
Eating disorders^g	20	NA		109	1.94	[1.41, 2.67]	159	2.18	[1.66, 2.88]
Conduct disorders^g	77	3.27	[2.03, 5.26]	358	3.36	[2.70, 4.18]	369	3.45	[2.79, 4.28]
Number of disorders^e,h,i
All persons
1 disorder	594	2.92	[2.43, 3.52]	2651	2.92	[2.68, 3.19]	2916	3.29	[3.01, 3.58]
⩾2 disorders	163	5.50	[3.60, 8.40]	638	4.95	[4.02, 6.10]	735	7.34	[5.88, 9.17]
Male
1 disorder	260	2.58	[1.96, 3.39]	1101	2.67	[2.33, 3.05]	1170	3.21	[2.81, 3.68]
⩾2 disorders	64	3.52	[1.95, 6.34]	229	4.17	[2.99, 5.79]	262	6.16	[4.33, 8.75]
Female
1 disorder	334	3.24	[2.52, 4.18]	1550	3.16	[2.81, 3.56]	1746	3.38	[3.03, 3.78]
⩾2 disorders	99	7.99	[4.29, 14.89]	409	5.53	[4.23, 7.24]	473	8.21	[6.15, 10.96]

ARR: adjusted relative risk; CI: confidence interval; NA: not applicable as cell sizes too small; LBOTE: language background other than English.

All ARR significant at p < 0.0001.

Percent calculated for young people in case and comparison cohorts not completing the school grade.

Number of students with a mental disorder not completed the year.

Adjusted relative risk for disorder status, gender, health condition, LBOTE, socioeconomic status (excluding 14 missing socioeconomic status), residential geographic location and parental education.

Year 10 excludes health condition, LBOTE and residential geographic location.

Year 10 substance disorders, single and ⩾2 disorders and Year 11 and 12 single disorders exclude health condition, LBOTE, socioeconomic status and residential geographic location.

Year 10 depression and Year 12 psychotic disorders exclude health condition, LBOTE and residential geographic location.

Year 10 conduct disorders and Years 11 and 12 eating disorders adjusted for disorder status, gender and parental education.

Year 11 ⩾ 2 disorders excludes health conditions, with males excluding residential geographic location.

Year 12 ⩾ 2 disorders for males excludes LBOTE and for females excludes health condition, LBOTE and residential geographic location.

Discussion

This study demonstrates the strong association between a young person being admitted to hospital for treatment of a range of mental disorders and a higher risk of not achieving a minimum standard for literacy and numeracy, and of not completing high school, when compared to matched peers. It found that young males with a mental disorder had over a 1.7 times higher risk of not achieving the NMS on numeracy and reading school assessments and that young males with psychotic, conduct and anxiety disorders had around twice the risk of not achieving the NMS compared to matched peers. Young females with a mental disorder had around 1.5 times higher risk of not achieving the NMS for numeracy and that females with psychotic, anxiety or conduct disorders had between 1.5 and 4 times higher risk of not achieving the NMS in assessments compared to matched counterparts. Young males and females with a mental disorder had around a three times higher risk of not completing high school compared to matched peers.

Poor academic performance and a higher risk of not completing high school compared to peers was evident for both young males and females with either internalising (e.g. anxiety, depression) or externalising (e.g. conduct, substance) disorders as has been shown elsewhere (Bowman et al., 2017; Breslau et al., 2008, 2011; Kessler et al., 1995; Leach and Butterworth, 2012; McLeod et al., 2012; Townsend et al., 2007). Psychotic, conduct and anxiety disorders were more strongly associated with poor school performance than depression and substance disorders in the current study. Both anxiety and depression in young people can adversely affect concentration, ability to process information and decision-making, thereby impairing social and cognitive functioning and ability to learn (Quiroga et al., 2013; Wood, 2006), which in turn can lead to school dropout due to perceived poor academic competence (Quiroga et al., 2013).

Young males and females with conduct disorders had twice the risk of not achieving the NMS in school assessments compared to matched peers and a three times higher risk of not completing high school in the current study. Conduct disorders can result in negative behaviour, such as disruptive, aggressive, inattentive behaviour or acting out, and these students can be difficult for teachers to engage in classroom learning. In the United States, students with conduct disorders had twice the likelihood of not completing high school and had a one and a half times higher likelihood of not commencing college (Kessler et al., 1995). However, the effect of conduct disorders on a student’s educational performance can be somewhat attenuated after controlling for attention deficit disorder (Breslau et al., 2008). Disentangling the effects of comorbid disorders, their interrelationship and impact on educational attainment remains an ongoing issue, with the potential of joint effects of different disorders (Breslau et al., 2011). In the current study, from grade 3 to high school between 8.4% and 15.4% of males and 13.5% and 18.7% of females had ⩾2 disorders. However, the sample size was not large enough to investigate potential joint effects of comorbid disorders.

Poor mental health can be a barrier to academic performance and school completion (Leach and Butterworth, 2012). Many young people with a mental disorder experience engagement difficulties in the school environment (Roeser et al., 1998). That some young people with a mental disorder are not achieving minimum standards for numeracy and reading is of concern. The ability to read is a foundational skill required to excel in almost all other academic subjects (Daniel et al., 2006) and poor reading ability is known to be associated with not completing high school (Daniel et al., 2006). Numeracy skills are also fundamental prerequisites in many aspects of day-to-day later life (Geary, 2011).

Young people with multiple mental disorders generally had a higher risk of not achieving the NMS on school assessment tasks and had at least double the risk of not completing high school compared to young people who were diagnosed with a single disorder. Likewise, previous research found that young people with multiple mental disorders attained lower levels of education than those with a single disorder (Breslau et al., 2008; McLeod et al., 2012). Young people with comorbid anxiety and depression often exhibit greater overall severity of symptoms, psychological distress, life interference and can have poorer long-term prognosis than young people with unitary disorders (Breslau et al., 2008).

While high school dropout is likely to be influenced by a number of factors, a young person’s mental health and their academic performance is not unidirectional, but rather they are likely to be reciprocally related (Roeser et al., 1998; Suldo et al., 2014). Initial identification of students with a mental disorder is critical and ongoing support services are vital as generally there is low recognition and diagnosis of mental disorders among the young (Fletcher, 2008). Screening young people who drop out of high school as a potential flag for also experiencing mental health issues could be an option (Breslau et al., 2008), as it is generally 1–4 years before a young person with a disorder will seek mental health services (Bowman et al., 2017). In the United States, it is estimated that between 10% and 46% of attrition from high school can be attributed to the adverse effects of mental disorders (Breslau et al., 2008; Vander Stoep et al., 2003).

Types of support for young people with a mental disorder include symptom management, peer counselling, personal support and social connections, and access to general practitioner and other specialist mental health services (McIntyre et al., 2018; Vander Stoep et al., 2003; Wilson et al., 2008), although access to services in rural and remote areas can be challenging (Bowman et al., 2017). The few targeted nonpharmacological interventions that have shown promise vary by disorder type, but have included interventions based on cognitive behavioural therapy for anxiety disorders and depression (Dray et al., 2017) and parent-group interventions have been found to reduce externalising behaviours (Buchanan-Pascall et al., 2018; Dretzke et al., 2009). There is also potential for school involvement to support a young person experiencing severe mental health concerns through school psychologists, Head Teacher Welfare, and Learning and Support Teams. It is possible that mental health training for educators could assist in improving their knowledge and confidence to promote good mental health among young people (Moon et al., 2017).

Future research could consider which factors may act as either mediators and moderators for a student’s academic performance or high school completion. Whether academic performance at primary school can predict academic difficulties in high school or high school dropout should also be explored, along with the identification of factors predictive of resilience for students with a mental disorder. The potential for reciprocal effects over time could also be investigated, with some parents placing increasing pressure on young people with a mental disorder to perform academically.

This was a large population-based study linking health and educational records over a 13-year period. However, there were some study limitations. The study was not able to separate admissions to child and adolescent units that included school attendance as part of the programme, or admissions to juvenile justice facilities, which also encourage participation in education, or from long admissions that were likely to result in missing a significant part of a school year, for example, to eating disorder units. A small number of residential postcodes were not known to the authors and socioeconomic status for these was not able to be identified. The recency of residential postcode may vary between datasets, which may affect estimates of socioeconomic status. Only comorbid health conditions relevant to a hospital admission are indicated in diagnosis classifications, so it is possible that some comorbidities experienced by young people could be under-enumerated. This is particularly likely for the comparison cohort, where 38.6% had not been admitted to hospital, so there was no opportunity to identify comorbidities, despite the 3-year lookback period.

The study only included young people who had been hospitalised, so did not include young people presenting solely to other health professionals for treatment. In Australia, the threshold for hospitalisation of young people is quite high, as there are comparatively few specialist inpatient beds, and also an uneven distribution of those beds around NSW. The main reasons for inpatient care are primarily concern for the safety of the young person and also identification of the type of disorder for which a consistent inpatient programme is required, for example, a severe eating disorder. Hence, hospitalised young people are likely to be the most seriously affected, regardless of diagnosis and irrespective of the number of days spent in hospital. While the cohort sample size was large, the number of factors able to be adjusted for in analyses was reduced due to low sample sizes for each disorder type by gender. Nor was it possible to determine the direction of any effects of disorders and school performance.

No information was available on the exact age at the onset of the mental disorder and information on ambulatory mental health contacts was only available from 2006. No assessment of data validity was able to be conducted and it is possible that there could be some data misclassification. Visits to private hospital EDs were not able to be accessed. A higher proportion of young people with a mental disorder were absent for NAPLAN assessments compared to their matched counterparts. The current study was not able to take into account school absences or school clustering, nor was it able to identify age at the onset of the mental health disorder from hospitalisation records or examine the proximity of hospital admission to NAPLAN tests or the final year of school for individual subjects. No information was available regarding any special education services, school exclusions (suspension or expulsion) or tutoring received by a young person.

Conclusion

This was a comprehensive investigation of the impact of mental disorders on academic performance and completion by disorder type and gender. Young people hospitalised with a mental disorder are not performing as well as their peers in school-based assessments and they have a higher risk of not completing high school. Disengagement from education may be ameliorated by early recognition and support aimed at enhancing school performance and educational outcomes for young people with a mental disorder. This support should be provided in conjunction with mental health services and school involvement and assistance. Young people should not be disadvantaged in their educational pursuits due to poor mental health as the consequences in terms of employment opportunities, overall health and relationship stability can be lifelong.

Supplemental Material

sj-docx-1-anp-10.1177_00048674211061684 – Supplemental material for Mental disorders and their impact on school performance and high school completion by gender in Australia: A matched population-based cohort study

Supplemental material, sj-docx-1-anp-10.1177_00048674211061684 for Mental disorders and their impact on school performance and high school completion by gender in Australia: A matched population-based cohort study by Rebecca J Mitchell, Anne McMaugh, Carolyn Schniering, Cate M Cameron, Reidar P Lystad, Tim Badgery-Parker and Olav Nielssen in Australian & New Zealand Journal of Psychiatry

Footnotes

Acknowledgements

The authors wish to thank the NSW Ministry of Health for providing access to the ED visit, hospitalisation and mortality data; NSW Department of Education for providing access to school enrolment and completion information; the NSW Education Standards Authority for providing access to the NAPLAN data; and the Centre for Health Record Linkage for conducting the data linkage.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This study was funded by a philanthropic donor to Macquarie University.

ORCID iDs

Rebecca J Mitchell

Reidar P Lystad

Supplemental Material

Supplemental material for this article is available online.

References

Australian Bureau of Statistics (2011) Census of Population and Housing: Socio-economic Indexes for Areas (SEIFA), Australia (Catalogue No: 2033.0.55.001). Canberra, ACT, Australia: Australian Bureau of Statistics.

Australian Bureau of Statistics (2013) 1270.0.55.005 – Australian Statistical Geography Standard (ASGS): Volume 5 – Remoteness Structure, July 2011. Available at: www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/1270.0.55.005July%202011?OpenDocument (accessed 7 March 2016).

Australian Curriculum Assessment and Reporting Authority (ACARA) (2014) National Assessment Program – Literacy and Numeracy Achievement in Reading, Persuasive Writing, Language Conventions and Numeracy: National Report for 2014. Sydney, NSW, Australia: ACARA.

Australian Curriculum Assessment and Reporting Authority (ACARA) (2017) National Assessment Program: Literacy and Numeracy. Sydney, NSW, Australia: ACARA.

Australian Curriculum Assessment and Reporting Authority (ACARA) (2019) National Assessment Program – Literacy and Numeracy (NAPLAN) 2018: Technical Report. Sydney, NSW, Australia: ACARA.

Bowman

McKinstry

McGorry

(2017) Youth mental ill health and secondary school completion in Australia: Time to act. Early Intervention in Psychiatry 11: 277–289.

Breslau

Lane

Sampson

, et al (2008) Mental disorders and subsequent educational attainment in a US national sample. Journal of Psychiatric Research 42: 708–716.

Breslau

Miller

Chung

W-JJ

, et al (2011) Childhood and adolescent onset psychiatric disorders, substance use, and failure to graduate high school on time. Journal of Psychiatric Research 45: 295–301.

Buchanan-Pascall

Gray

Gordon

, et al (2018) Systematic review and meta-analysis of parent group interventions for primary school children aged 4–12 years with externalizing and/or internalizing problems. Child Psychiatry & Human Development 49: 244–267.

10.

Daniel

Walsh

Goldston

, et al (2006) Suicidality, school dropout, and reading problems among adolescents. Journal of Learning Disabilities 39: 507–514.

11.

Deighton

Humphrey

Belsky

, et al (2018) Longitudinal pathways between mental health difficulties and academic performance during middle childhood and early adolescence. British Journal of Developmental Psychology 36: 110–126.

12.

DeSocio

Hootman

(2004) Children’s mental health and school success. The Journal of School Nursing 20: 189–196.

13.

Dray

Bowman

Campbell

, et al (2017) Systematic review of universal resilience-focused interventions targeting child and adolescent mental health in the school setting. Journal of the American Academy of Child & Adolescent Psychiatry 56: 813–824.

14.

Dretzke

Davenport

Frew

, et al (2009) The clinical effectiveness of different parenting programmes for children with conduct problems: A systematic review of randomised controlled trials. Child and Adolescent Psychiatry and Mental Health 3: 7.

15.

Edwards

Houtrow

Vasilevskis

, et al (2012) Chronic conditions among children admitted to US pediatric intensive care units: Their prevalence and impact on risk for mortality and prolonged length of stay. Critical Care Medicine 40: 2196–2203.

16.

Fletcher

(2008) Adolescent depression: Diagnosis, treatment, and educational attainment. Health Economics 17: 1215–1235.

17.

Geary

(2011) Consequences, characteristics, and causes of mathematical learning disabilities and persistent low achievement in mathematics. Journal of Developmental and Behavioral Pediatrics: JDBP 32: 250–263.

18.

Kessler

Angermeyer

Anthony

, et al (2007) Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization’s World Mental Health Survey Initiative. World Psychiatry 6: 168–176.

19.

Kessler

Foster

Saunders

, et al (1995) Social consequences of psychiatric disorders, I: Educational attainment. American Journal of Psychiatry 152: 1026–1032.

20.

Larson

Chapman

Spetz

, et al (2017) Chronic childhood trauma, mental health, academic achievement, and school-based health center mental health services. Journal of School Health 87: 675–686.

21.

Lawrence

Hafekost

Johnson

, et al (2016) Key findings from the second Australian child and adolescent survey of mental health and wellbeing. Australian and New Zealand Journal of Psychiatry 50: 876–886.

22.

Leach

Butterworth

(2012) The effect of early onset common mental disorders on educational attainment in Australia. Psychiatry Research 199: 51–57.

23.

McIntyre

Worsley

Corcoran

, et al (2018) Academic and non-academic predictors of student psychological distress: The role of social identity and loneliness. Journal of Mental Health 27: 230–239.

24.

McLeod

Uemura

Rohrman

(2012) Adolescent mental health, behavior problems, and academic achievement. Journal of Health and Social Behavior 53: 482–497.

25.

Mental Health Commission (2021) National Children’s Mental Health and Wellbeing Strategy. Canberra, ACT, Australia: Mental Health Commission.

26.

Miller

Shi

Wheeler

, et al (2013) Chronic conditions and outcomes of pediatric trauma patients. Journal of Trauma and Acute Care 75: 250–257.

27.

Mitchell

Cameron

Lystad

, et al (2019) Impact of chronic health conditions and injury on school performance and health outcomes in New South Wales, Australia: A study protocol. BMJ Paediatrics 3. DOI: 10.1136/bmjpo-2019-000530.

28.

Mitchell

Curtis

Braithwaite

(2017) Health outcomes and costs for injured young people hospitalised with and without chronic health conditions. Injury 48: 1776–1783.

29.

Moon

Williford

Mendenhall

(2017) Educators’ perceptions of youth mental health: Implications for training and the promotion of mental health services in schools. Children and Youth Services Review 73: 384–391.

30.

Productivity Commission (2020) Mental Health. Canberra, ACT, Australia: Productivity Commission.

31.

Quiroga

Janosz

Bisset

, et al (2013) Early adolescent depression symptoms and school dropout: Mediating processes involving self-reported academic competence and achievement. Journal of Educational Psychology 105: 552–560.

32.

Roeser

Eccles

Strobel

(1998) Linking the study of schooling and mental health: Selected issues and empirical illustrations at. Educational Psychologist 33: 153–176.

33.

Sagatun

Heyerdahl

Wentzel-Larsen

, et al (2014) Mental health problems in the 10th grade and non-completion of upper secondary school: The mediating role of grades in a population-based longitudinal study. BMC Public Health 14: 16.

34.

Suldo

Gormley

DuPaul

, et al (2014) The impact of school mental health on student and school-level academic outcomes: Current status of the research and future directions. School Mental Health 6: 84–98.

35.

Townsend

Flisher

King

(2007) A systematic review of the relationship between high school dropout and substance use. Clinical Child and Family Psychology Review 10: 295–317.

36.

Vander Stoep

Weiss

Kuo

, et al (2003) What proportion of failure to complete secondary school in the US population is attributable to adolescent psychiatric disorder? The Journal of Behavioral Health Services & Research 30: 119–124.

37.

Wilson

Deane

Marshall

, et al (2008) Reducing adolescents’ perceived barriers to treatment and increasing help-seeking intentions: Effects of classroom presentations by general practitioners. Journal of Youth and Adolescence 37: 1257–1269.

38.

Wood

(2006) Effect of anxiety reduction on children’s school performance and social adjustment. Developmental Psychology 42: 345–349.

39.

Woodward

(2005) Epidemiology: Study Design and Analysis. Boca Raton, FL: Chapman & Hall CRC.

40.

Zou

(2004) A modified poisson regression approach to prospective studies with binary data. American Journal of Epidemiology 159: 702–706.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB