AD ALTA
JOURNAL OF INTERDISCIPLINARY RESEARCH
Tab. 2: Number of cases of self-harm detected using the three
systems under study
Diagnostic
system
Number of cases
detected
Number of cases not
detected
N
%
N
%
SHI
830
100
0
0
ICD-10
803
96.7
27
3.3
DSM-5
701
84.5
129
15.5
Source: authors
The total drop-out rate when using the ICD-10 diagnostic criteria
(versus the SHI system) is moderate – it represents only 3.3% of
cases. This difference is caused by the omission of the mental
forms of self-harm; however, a closer analysis (see Table 3)
indicated that the majority of cases that would be omitted if the
ICD-10 criteria were used were in the following two items:
“tortured yourself with self-defeating thoughts” and “engaged in
emotionally abusive relationships”.
Tab. 3: Items from the SHI that are not included in the ICD-10
criteria and the corresponding number of participants that were
not detected
Item
“dropped-out” cases
N
% of whole sample
Mental self-harm:
Distanced yourself from God
as a punishment
3
0.4
Set yourself up in a
relationship to be rejected
5
0.6
Tortured yourself with self-
defeating thoughts
20
2.4
Engaged in emotionally
abusive relationships
19
2.3
Source: authors
An analysis of the drop-out rate coming from the use of the
DSM-5 diagnostic system revealed that 129 (15.5%) cases
would be left undetected versus the original SHI questionnaire.
A statistically significant (sig. = 0.000) decrease in the number
of cases detected also occurred when the DSM-5 criteria was
used as opposed to the ICD-10 (see Table 4).
Tab. 4: Differences in the number of detected cases using the
ICD-10 and DSM-5 systems
Cases within
DSM-5:
Cases within ICD-10:
Detected
(N/%)
Not detected
(N/%)
Total (N/%)
Detected
(N/%)
27/3.3
102/12.3
129/15.5
Not detected
(N/%)
0/0.0
701/84.5
701/84.5
Total
27/3.3
803/96.8
830/100.0
Chi-Square test
Pearson coeficient
Sig.
151.65
0.000
Source: authors
It is apparent that the indirect forms of self-harm that are
included in the ICD-10, but are absent from DSM-5, are
represented to a statistically significant rate in the population of
self-harmers. The item analysis (see Table 5) revealed that this is
mostly related to the following items: “abused alcohol to hurt
yourself” (6.3% of all cases) and “not slept enough to hurt
yourself” (4.1% of all cases).
Tab. 5: Items from ICD-10 that are not included in the DSM-5
criteria and the corresponding number of participants that were
not detected
Item
“dropped-out” cases
N
% of whole sample
Attempted suicide*
2
0.2
Indirect physical self-harm:
Abused alcohol to hurt yourself
52
6.3
Not slept enough to hurt
34
4.1
yourself
Starved yourself to hurt
yourself
23
2.5
Over-exercised to hurt yourself
31
3.7
Made medical situations worse
on purpose
11
1.3
Overdosed
1
0.1
Abused prescription medication
4
0.5
Abused laxatives to hurt
yourself
0
0.0
Source: authors
5 Discussion
The basic (modified) SHI questionnaire identified a rate of self-
harm among adolescents of 45.5%. This prevalence is
comparable to the data reported by Dyl (2008), who reported
47% of adolescents. In comparison to an overview of the
prevalence of self-harm in similar studies (4.7% – Madge et al.
2008; 8% – Moran et al. 2012, 9.3% – Tørmoen, Rossow,
Larsson, & Mehlum 2013; 10% – Hawton, Saunders, &
O’Connor 2012; 20.3% – Swahn et al. 2012; 25.6% – Plener et
al. 2009) it is one of the higher rates of prevalence. This might
be caused by more up-to-date data, which might have captured
the recent trend in the increase of self-harm among adolescents
(Clarke, Allerhand, & Berk 2019), or by the fact that the sample
had a slightly higher proportion of female subjects (63.3% of
females vs. 36.7% of males). Several studies (see e.g. Laye-
Gindhu & Schonert-Reichl 2005; Rodham, Hawton, & Evans
2004, Demuthova & Doktorova 2019) have reported a higher
prevalence of self-harm in female participants.
The modified SHI questionnaire covers a wide range of self-
harming behaviours including mental forms (not included in
ICD-10 and DSM-5) and indirect physical forms (not included in
DSM-5). It is one of the more complex questionnaires used for
the identification of self-harming behaviour – others (e.g. DSHI
(Deliberate Self-Harm Inventory) – Gratz 2001) primarily focus
on its physical forms. The disadvantage of the broader design of
the SHI is the relatively large (20) number of items, which is
inappropriate when setting out diagnostic criteria. Hence, the
subsequent analyses examined how many of the cases that were
identified by the SHI can still be detected by systems based on
narrower criteria, such as ICD-10 and DSM-5 and, possibly,
which items are important from the more broadly designed
systems.
Using the ICD-10 system caused a drop-out of 27 self-harming
adolescents. The ICD-10 system excludes four items in
comparison to SHI, with a decrease in the ability to detect cases
being negligible in only two of them: “distanced yourself from
God as a punishment” and “set yourself up in a relationship to be
rejected” (less than 1% of cases). On the other hand, the other
two items, “tortured yourself with self-defeating thoughts” and
“engaged in emotionally abusive relationships”, were detected in
more than 2% of the cases. Overall, using the ICD-10 criteria for
the self-harming adolescent population would result in the non-
detection of 3.3%. According to international studies, the items
“distanced yourself from God as a punishment” and “set yourself
up in a relationship to be rejected” are not commonly found
forms of self-harm – the reported prevalence is only 2.6% (“set
yourself up in a relationship to be rejected”) and 1.4%
(“distanced yourself from God as a punishment”) (Müller, Claes,
Smits, Brähler, & de Zwaan 2016). Still, there are specific cases
that are exceptions – e.g. adult patients with chronic pain
(Sansone, Sinclair, & Wiederman 2009). At the same time, the
reported prevalence of the items “tortured yourself with self-
defeating thoughts” and “engaged in emotionally abusive
relationships” is respectively 30% (“tortured yourself with self-
defeating thoughts”) and 6% (“engaged in emotionally abusive
relationships”) (Müller, Claes, Smits, Brähler, & de Zwaan
2016) and we deem their exclusion from the self-harm checklist
(especially in the case of “tortured yourself with self-defeating
thoughts”) to be excessively narrowing for the purposes of
diagnosis.
- 53 -