To What Extent Are Trends in Teen Mental Health Driven by Changes in Reporting?: The Example of Suicide-Related Hospital Visits

Adriana Corredor-Waldron; Janet Currie

doi:10.3368/jhr.0423-12854R1

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Figure 1
Trends in Suicide-Related Visits and Suicides in New Jersey, 2008–2019
Notes: These figures plot trends in suicide-related hospital visits, suicide deaths, and suicidal ideation hospital visits in New Jersey. The vertical lines at 2011 and 2016 help to visualize the changes related to the implementation of the Women’s Preventive Services Guidelines in 2012, as well the difference between 2015 and 2016 (implementation of ICD-10) and between 2016 and 2017 (implementation of the “include SI” guidance).
^a Includes ICD-9 codes for self-injury and ICD-10 codes for intentional self-harm and suicide attempts.
Download figure
Open in new tab
Download powerpoint
Figure 2
Teens with Primary Diagnoses of Mental Health Disorders and Secondary Diagnosis of Suicidal Ideation by Gender, Race, Insurance Status, and Hospital Type
Notes: These figures show mood-disorder visits with a suicidal ideation diagnosis by gender, race, insurance status, and hospital type in New Jersey. The vertical lines at 2011 and 2016 help to visualize the changes related to the implementation of the Women’s Preventive Services Guidelines in 2012, as well as the difference between 2015 and 2016 (implementation of ICD-10) and between 2016 and 2017 (implementation of the “include SI” guidance).
Download figure
Open in new tab
Download powerpoint
Figure 3
Trends in Teen’s Mental Health Visits by Sex
Notes: These figures plot trends in mental-health-related visits in New Jersey by sex at birth. The vertical lines at 2011 and 2016 help to visualize the changes related to the implementation of the Women’s Preventive Services Guidelines in 2012, as well as the difference between 2015 and 2016 (implementation of ICD-10) and between 2016 and 2017 (implementation of the “include SI” guidance).

View popup

Table 1

Teens’ Suicide-Related and Total Hospital Visits by Reported Primary Diagnosis

	Any Suicide-Related Code			Suicide-Related Code Is Suicidal Ideation			Suicidal Ideation Is Secondary Diagnosis			All Visits
	2009–2010	2014–2015	2018–2019	2009–2010	2014–2015	2018–2019	2009–2010	2014–2015	2018–2019	2009–2010	2014–2015	2018–2019
Panel A: Reported Primary Diagnosis
All visits	36.4	50.3	75.6	23.2	37.0	61.9	20.6	29.9	48.4	3,068.5	2903.8	2632.9
Mood disorder is primary	17.0	24.5	35.4	13.0	20.4	31.9	13.0	20.4	31.9	76.9	98.9	99.0
Other mental health is primary	7.1	7.4	14.4	5.3	5.8	12.5	5.3	5.8	12.5	139.7	140.4	151.7
Suicidal ideation is primary	2.6	7.1	13.5	2.6	7.1	13.5				2.6	7.1	13.5
Non-mental health is primary	9.6	11.3	12.2	2.3	3.7	3.9	2.3	3.7	3.9	2,849.3	2,657.4	2,368.7
Panel B: Corrected Primary Diagnosis^{^a}
All visits	36.4	50.3	75.6	23.2	37.0	61.9	22.3	35.1	57.8	3,068.5	2,903.8	2,632.9
Mood disorder is primary	18.0	27.5	40.5	14.0	23.4	37.0	14.0	23.4	37.0	77.9	101.9	104.0
Other mental health is primary	7.5	8.5	16.7	5.7	6.9	14.8	5.7	6.9	14.8	140.1	141.5	154.0
Suicidal ideation is primary^{^a}	0.9	2.0	4.1	0.9	2.0	4.1				0.9	2.0	4.1
Non-mental health is primary	9.9	12.2	14.2	2.6	4.7	6.0	2.6	4.7	6.0	2,849.6	2,658.4	2,370.8

Notes: The table reports the rates per 10,000 teens.
↵^a Suicidal ideation is a symptom and is not supposed to be coded as primary if there is another confirmed diagnosis.
In Panel B, we reassign visits with a primary diagnosis of suicidal ideation and a secondary diagnosis code to make the non-SI code primary. For example, if SI is primary, but the secondary diagnosis is a mood disorder, then we count the visit in the “Mood disorder is primary” row in Panel B.

View popup

Table 2

Impact of the Women’s Preventive Screening Guidelines for Depression on Mental Health Hospital Visits for Teens

	Visits with:
	Mental Health Diagnosis	Mood Disorders Diagnosis	Anxiety Disorders Diagnosis	Suicidal Ideation Diagnosis
	(1)	(2)	(3)	(4)
2015 × Female	0.0053	0.0130	0.0083	0.0046
	(0.0035)	(0.0035)	(0.0015)	(0.0012)
2014 × Female	0.0068	0.0125	0.0071	0.0047
	(0.0038)	(0.0037)	(0.0014)	(0.0011)
2013 × Female	0.0093	0.0125	0.0066	0.0047
	(0.0036)	(0.0036)	(0.0014)	(0.0011)
2012 × Female	0.0015	0.0058	0.0016	0.0029
	(0.0036)	(0.0029)	(0.0011)	(0.0009)
Female	−0.0094	0.0148	0.0067	0.0022
	(0.0016)	(0.0010)	(0.0004)	(0.0003)
Baseline mean^{^a}	0.068	0.020	0.005	0.006
Sample	2008–2015	2008–2015	2008–2015	2008–2015
Observations	2,647,043	2,647,043	2,647,043	2,647,043
R-squared	0.0875	0.0758	0.0136	0.0158

Notes: The table reports coefficients from linear regression models. The dependent variable for each column is one if the hospital record for a visit includes the condition and zero otherwise. All regressions include fixed effects for hospital, year, month, single year of child’s age, race (Black, white, other race), gender, and insurance (Medicaid, private, self-pay). In addition, the models include zip code characteristics, such as median income and the shares of poor, renters, college educated, non-Hispanic Black, non-Hispanic white, Hispanics, Asians, and age bins. Standard errors are clustered by hospital–year. The sample includes all teen visits to hospitals from 2008 to 2015.
↵^a Baseline mean corresponds to the boys’ mean before 2012.

View popup

Table 3

Impact of the Women’s Preventive Screening Guidelines for Depression on Mental Health Hospital Visits for Teens, Including First Visit in the Year Only

	Visits with:
	Mental Health Diagnosis	Mood Disorders Diagnosis	Anxiety Disorders Diagnosis	Suicidal Ideation Diagnosis
	(1)	(2)	(3)	(4)
2015 × Female	0.0048	0.0098	0.0060	0.0034
	(0.0026)	(0.0023)	(0.0010)	(0.0007)
2014 × Female	0.0069	0.0094	0.0059	0.0039
	(0.0029)	(0.0023)	(0.0010)	(0.0008)
2013 × Female	0.0067	0.0079	0.0054	0.0030
	(0.0028)	(0.0022)	(0.0011)	(0.0007)
2012 × Female	0.0003	0.0046	0.0013	0.0022
	(0.0028)	(0.0019)	(0.0008)	(0.0007)
Female	−0.0096	0.0100	0.0056	0.0019
	(0.0012)	(0.0007)	(0.0003)	(0.0002)
Baseline mean^{^a}	0.095	0.023	0.010	0.005
Sample	2008–2015	2008–2015	2008–2015	2008–2015
Observations	1,792,559	1,792,559	1,792,559	1,792,559
R-squared	0.0679	0.0535	0.0096	0.0103

Notes: The table reports coefficients from linear regression models. The dependent variable for each column is one if the hospital record for a visit includes the condition and zero otherwise. All regressions include fixed effects for hospital, year, month, single year of child’s age, race (Black, white, other race), gender, and insurance (Medicaid, private, self-pay). In addition, we include zip code characteristics, such as median income and the shares of poor, renters, college educated, non-Hispanic Black, non-Hispanic white, Hispanics, Asians, and age bins. Standard errors are clustered by hospital–year. The sample includes teen visits to hospital from 2008 to 2015.
↵^a Baseline mean corresponds to the boys’ mean before 2012.

View popup

Table 4

ICD-10 Coding Changes and Suicide-Related Hospital Visits for Teens

	Visits with:
	Suicide-Related Diagnosis	Suicide-Related Diagnosis	Suicidal Ideation Is Secondary Diagnosis	Self Injury^{^b} Is Diagnosis
	(1)	(2)	(3)	(4)
2019 × Mood primary	0.1019	0.1015	0.1151	−0.0101
	(0.0497)	(0.0431)	(0.0474)	(0.0064)
2018 × Mood primary	0.1056	0.1063	0.1175	−0.0074
	(0.0453)	(0.0397)	(0.0427)	(0.0069)
2017 × Mood primary	0.0612	0.0634	0.0742	−0.0057
	(0.0424)	(0.0374)	(0.0386)	(0.0054)
2016 × Mood primary	0.0211	0.0243	0.0254	−0.0003
	(0.0360)	(0.0321)	(0.0324)	(0.0054)
2019 × Other mental	0.0381	0.0391	0.0416	0.0001
	(0.0162)	(0.0138)	(0.0142)	(0.0019)
2018 × Other mental	0.0405	0.0419	0.0447	0.0009
	(0.0155)	(0.0135)	(0.0131)	(0.0020)
2017 × Other mental	0.0231	0.0249	0.0286	−0.0005
	(0.0121)	(0.0108)	(0.0103)	(0.0017)
2016 × Other mental	−0.0004	0.0012	0.0018	0.0003
	(0.0075)	(0.0073)	(0.0069)	(0.0017)
mood primary	0.2406	0.2336	0.2041	0.0476
	(0.0160)	(0.0161)	(0.0155)	(0.0036)
other mental	0.0460	0.0461	0.0431	0.0086
	(0.0039)	(0.0041)	(0.0038)	(0.0008)
Baseline mean^{^a}	0.006	0.006	0.001	0.003
Hospital FE	N	Y	Y	Y
Sample	2013–2019	2013–2019	2013–2019	2013–2019
Observations	2,188,565	2,188,565	2,188,565	2,188,565
R-squared	0.1420	0.1558	0.1952	0.0178

Notes: The table reports coefficients from linear regression models. The dependent variable for each column is one if the hospital record for a visit includes the condition and zero otherwise. All regressions include fixed effects for year, month, single year of child’s age, race (Black, white, other race), gender, and insurance (Medicaid, private, self-pay). In addition, the models include zip code characteristics, such as median income and the shares of poor, renters, college educated, non-Hispanic Black, non-Hispanic white, Hispanics, Asians, and age bins. Standard errors are clustered by hospital–year. The sample corresponds to all teen visits from 2013 to 2019 for any indication.
↵^a Baseline mean corresponds to the non-mental-health visits’ mean before 2016.
↵^b Includes ICD-9 codes for self-injury and ICD-10 codes for intentional self-harm and suicide attempts.

View popup

Table 5

ICD-10 Coding Changes and Suicidal Ideation Diagnosis, First Visits Only

	Sample: All	Sample: First Visit in the Year	Sample: First Visit in Sample Period
	(1)	(2)	(3)
2019 × Mood primary	0.1151	0.0843	0.0821
	(0.0474)	(0.0382)	(0.0367)
2018 × Mood primary	0.1175	0.0767	0.0570
	(0.0427)	(0.0337)	(0.0356)
2017 × Mood primary	0.0742	0.0332	0.0272
	(0.0386)	(0.0301)	(0.0316)
2016 × Mood primary	0.0254	−0.0037	−0.0024
	(0.0324)	(0.0288)	(0.0312)
2019 × Other mental	0.0416	0.0414	0.0493
	(0.0142)	(0.0135)	(0.0158)
2018 × Other mental	0.0447	0.0468	0.0517
	(0.0131)	(0.0129)	(0.0165)
2017 × Other mental	0.0286	0.0229	0.0333
	(0.0103)	(0.0079)	(0.0102)
2016 × Other mental	0.0018	−0.0028	−0.0051
	(0.0069)	(0.0056)	(0.0063)
Mood primary	0.2041	0.2152	0.2229
	(0.0155)	(0.0146)	(0.0152)
Other mental	0.0431	0.0388	0.0409
	(0.0038)	(0.0032)	(0.0036)
Baseline mean^{^a}	0.001	0.001	0.001
Sample	2013–2019	2013–2019	2013–2019
Observations	2,188,565	1,460,977	596,055
R-squared	0.1952	0.1719	0.1669

Notes: The table reports coefficients from linear regression models. The dependent variable for each column is one if the hospital record for a visit includes the condition and zero otherwise. All regressions include fixed effects for year, month, single year of child’s age, race (Black, white, other race), gender, and insurance (Medicaid, private, self-pay). In addition, we include zip code characteristics, such as median income and the shares of poor, renters, college educated, non-Hispanic Black, non-Hispanic white, Hispanics, Asians, and age bins. Standard errors are clustered by hospital–year. The sample includes teen visits from 2013 to 2019.
↵^a Baseline mean corresponds to the non-mental-health visits’ mean before 2016.