The Alexian Brothers Urge to Self-Injure Scale (ABUSI) is a 5-item self-report measure of the severity of the urge to engage in non-suicidal self-injury (NSSI) over the past week, for adolescents and adults aged 11 and over with known or suspected self-injury (Washburn et al., 2010). It captures how often the urge arose, how strong it became, how much time was spent thinking about it, how difficult it was to resist, and the overall average urge across the week.
The urge that comes before self-injury is clinically distinct from the behaviour itself. A person may experience frequent and intense urges during a week in which they did not self-injure at all, and that week is clinically very different from one in which no urges arose. Because urges fluctuate from day to day, they have been found to respond to treatment sooner than the frequency of self-injury does, which makes them an early indicator of whether an intervention is working (Washburn et al., 2010; Victor et al., 2016). The ABUSI gives clinicians a brief, repeatable way to measure that urge.
The ABUSI treats the urge to self-injure as a single, fluctuating state and asks about five facets of it over the last week. It was adapted from the Penn Alcohol Craving Scale (PACS; Flannery et al., 1999) and shares its structure.
NSSI is direct and intentional damage to one’s own body tissue without suicidal intent. It is most common among adolescents and young adults, with lifetime rates of at least one incident estimated at approximately 18% in community populations and up to 67% in clinical populations (Victor et al., 2016). NSSI is not a formal diagnosis. DSM-5-TR lists Nonsuicidal Self-Injury Disorder as a condition for further study and includes self-injury among other conditions that may be a focus of clinical attention, and ICD-11 treats it as a sign or symptom rather than a disorder (APA, 2022; WHO, 2019).
The ABUSI measures one part of a fuller NSSI assessment, which typically also covers onset, frequency, and the functions the behaviour serves. It assumes self-injury is already known or suspected, and it is not designed to screen for it (Becker-Haimes et al., 2020). Self-injury generally serves a function, most often the regulation of difficult emotions, and assessing which function it serves for a given client is recommended as part of treatment planning (Klonsky, 2007).
The urge to self-injure is also informative about suicidal ideation. In adolescents and young adults being treated for self-injury, urge severity was more strongly associated with recent suicidal ideation than any other NSSI characteristic examined (Victor et al., 2015). An elevated ABUSI score is therefore a reasonable prompt to assess for suicidal ideation.
Monitoring progress. The ABUSI is designed principally for progress monitoring in clients with known or suspected self-injury (Becker-Haimes et al., 2020). Repeated weekly administration helps clinicians track whether urges are increasing, decreasing, or remaining stable over the course of treatment, which can be informative even when the frequency of self-injury itself has not changed.
Informing treatment. Each item addresses a different facet of the urge, so the pattern of item responses can highlight clinically important differences in the frequency, intensity, duration, and controllability of urges. A client whose urges are infrequent but overwhelming when they arrive presents differently from one whose urges are constant but manageable. Reviewing these responses with the client can guide discussion and treatment planning.
Informing risk and prognosis. Urge severity at intake has been associated with less improvement in self-injury frequency during day treatment (Slesinger et al., 2021) and with higher odds of re-admission (Washburn et al., 2010). Given the association between stronger urges and suicidal ideation, elevated scores or responses indicating the client was unable to resist the urge can prompt further assessment of suicidal ideation and other safety concerns.
Each item is rated on a 7-point scale from 0, the complete absence of an urge (for example, never or none), to 6, the most severe level (for example, nearly all of the time or was not able to resist). The response anchors vary across items but follow the same severity progression. The total score is the sum of the five items and ranges from 0 to 30, with higher scores indicating a more severe urge to self-injure over the past week. There are no subscales and no reverse-scored items.
The total score is described with a severity category created by NovoPsych from the measure’s own response anchors. Because all five items share a common severity progression, the average item response can be mapped back to the anchors: an average of 3 gives a total of 15 (a moderate urge), and an average of 5 gives a total of 25 (a strong urge that is difficult to control). Anchored to the response options rather than a comparison group, a category carries the same meaning for any client, which is why it is the primary basis for interpretation.
| Total score | Severity category | What it describes |
|---|---|---|
| 0 | None reported | No urge to self-injure was reported at any point during the week. This reflects that week only and is not evidence that risk is absent. |
| 1 to 14 | Mild | The average response falls below a moderate urge, occurring rarely to occasionally. Scores in this range remain clinically relevant, particularly where self-injury has not previously been identified. |
| 15 to 24 | Moderate | The average response corresponds to a moderate urge, or to a strong urge the client felt able to control. |
| 25 to 30 | Severe | The average response corresponds to a strong urge the client found difficult to resist, indicating substantial and poorly controlled urges. |
These categories are content-based interpretive guideposts rather than validated thresholds, and should not be used to classify a client. Because the category is based on the total, it can understate severity when a client scores at the top of one or two items and low on the rest (two items at 6 and three at 0 give a total of 12, which reads as Mild), so the individual item responses should always be read alongside it.
The NSSI Treatment Percentile compares the client with 1,327 adolescents and adults admitted for partial hospitalisation or intensive outpatient treatment of non-suicidal self-injury, who averaged 18.4 (SD 8.0) at admission (Washburn et al., 2023). This is a severely affected group, so the 50th percentile describes the level typical of someone entering intensive treatment, and a client can score well below it while still reporting clinically relevant urges. The percentiles are estimated from the sample’s average and spread assuming a normal distribution, and are least accurate at the bottom of the scale, where roughly one in eight clients in treatment report no urge in a given week. The maximum score of 30 sits at the 93rd percentile.
| Total score | Severity category | NSSI Treatment Percentile |
|---|---|---|
| 0 | None reported | 1.1 |
| 1 | Mild | 1.5 |
| 2 | Mild | 2 |
| 3 | Mild | 3 |
| 4 | Mild | 4 |
| 5 | Mild | 5 |
| 6 | Mild | 6 |
| 7 | Mild | 8 |
| 8 | Mild | 10 |
| 9 | Mild | 12 |
| 10 | Mild | 15 |
| 11 | Mild | 18 |
| 12 | Mild | 21 |
| 13 | Mild | 25 |
| 14 | Mild | 29 |
| 15 | Moderate | 34 |
| 16 | Moderate | 38 |
| 17 | Moderate | 43 |
| 18 | Moderate | 48 |
| 19 | Moderate | 53 |
| 20 | Moderate | 58 |
| 21 | Moderate | 63 |
| 22 | Moderate | 67 |
| 23 | Moderate | 72 |
| 24 | Moderate | 76 |
| 25 | Severe | 80 |
| 26 | Severe | 83 |
| 27 | Severe | 86 |
| 28 | Severe | 88 |
| 29 | Severe | 91 |
| 30 | Severe | 93 |
A change of 7 points or more in either direction is treated as reliable, meaning it is larger than would be expected from measurement error. A decrease of that size indicates a reliable reduction in the urge, and an increase indicates a reliable worsening. The threshold was established across a full course of treatment, so change over a single week is worth reading with more caution.
On repeat administration the report compares the current score with the initial administration, and from the third administration onward also with the immediately previous one, giving each comparison its own reliable change result. A change in severity category is reported against the initial administration only, so a client whose score moves back and forth across one boundary does not receive a category-change sentence every session.
Two response options carry clinical significance regardless of the total score. On item 4 (how difficult it was to resist), the highest option, Was not able to resist, indicates that the client could not resist the urge during the week. On item 2 (how strong the urge was at its most severe point), the highest option, Strong urge and would have self-injured if able to, indicates that only the absence of opportunity stood between the urge and the behaviour.
When item 4 is answered Was not able to resist, the report displays a red alert beneath the results table and a Safety Risk Alert tag is added to the results dashboard. When either response is given, the interpretive text includes a Risk of Self-Harm or Suicide paragraph naming the item or items endorsed, with a recommendation to follow suicide risk assessment protocols to determine the nature, severity, and immediacy of the risk. Moderate and Severe results also carry a coloured severity tag.
On first administration, a bar chart displays the total score against a background shaded by the severity categories, with a right-hand strip marking the 50th, 75th, 80th, 85th, and 90th NSSI Treatment Percentiles. A score of 0 sits in an unshaded strip at the foot of the chart labelled None Reported.
On repeat administration, a line chart plots the total score across all administrations against the same bands and percentile markers, so the trajectory of the urge over treatment can be seen at a glance.
The ABUSI was developed at the Alexian Brothers Behavioral Health Hospital and first validated in 386 adolescents and adults admitted to its Self-Injury Recovery Services programme between 2007 and 2009 (Washburn et al., 2010). The sample was 90.7% female and 86.3% non-Hispanic white, aged 11 to 61, and mostly adolescents (64.9%). Items were adapted from the 5-item Penn Alcohol Craving Scale by changing the content from alcohol craving to the urge to self-injure. Completion was high, with 95.6% of patients answering all five items.
Internal consistency is high. In the development sample, Cronbach’s alpha was .92 at admission and .96 at discharge, and all five items correlated highly with the total (r = .87 to .92; Washburn et al., 2010). Similar values have been reported in later samples from the same service: .93 in 1,502 adolescents and young adults (Victor et al., 2015), .93 in 1,780 patients (Victor et al., 2016), and .92 in 1,327 day-treatment patients (Slesinger et al., 2021).
Test-retest reliability over 48 hours was high for the total score (ICC = .84, n = 46), ranging from .60 for the duration item to .83 for the frequency item (Washburn et al., 2010). The short interval means this describes stability over about a day rather than over the one-week period the measure asks about.
Exploratory factor analysis identified a single factor, severity of the urge to self-injure, accounting for 70.6% of the variance, with loadings from .76 (duration) to .92 (overall urge; Washburn et al., 2010). A confirmatory factor analysis of the Spanish adaptation in 752 Colombian adolescents supported the same single-factor solution (Bahamón et al., 2023). No confirmatory analysis has been conducted on the English version. The ABUSI is scored as a single total.
Convergent validity. Total scores correlated positively with self-injury frequency, suicidal ideation, and functional impairment, and negatively with quality of life and satisfaction (r = −.48 at admission; Washburn et al., 2010).
Known-groups validity. Among 511 patients admitted for self-injury treatment, those meeting the proposed DSM-5 criteria for non-suicidal self-injury disorder scored higher than those who did not (21.06 compared with 16.83; d = 0.52; Washburn et al., 2015). In 1,502 adolescents and young adults, urge severity was substantially higher among those reporting past-week suicidal ideation (d = 1.16), a larger difference than for any other self-injury characteristic examined (Victor et al., 2015).
Predictive and incremental validity. Each additional point at admission was associated with higher odds of re-admission (OR = 1.06, 95% CI 1.02 to 1.11) and lower odds of clinically significant improvement in quality of life by discharge (OR = 0.97, 95% CI 0.94 to 0.99), and admission scores improved the prediction of self-injury frequency at discharge over and above baseline frequency, age, and gender (Washburn et al., 2010). In 1,327 day-treatment patients, stronger urges at admission predicted less improvement in self-injury frequency (Slesinger et al., 2021).
Comparison with other measures. A systematic review of 11 instruments measuring NSSI in adolescents found the ABUSI among the two best-evidenced (Chávez-Flores et al., 2019), and an evidence-based review of brief, free youth mental health measures rated its psychometrics excellent, the only one of six suicidality measures to receive the highest rating (Becker-Haimes et al., 2020).
Responsiveness to treatment is the strongest area of the ABUSI evidence base. Among 277 patients with admission and discharge data, total scores fell from 18.47 to 11.84, a within-group effect size of 0.83, the largest of the outcome measures administered alongside it and the only one on which average change exceeded its reliable change threshold (Washburn et al., 2010). Among 923 day-treatment patients, scores fell by 3.6 to 4.1 points on average (effect sizes 0.47 to 0.52; Washburn et al., 2023). In 1,780 patients, reductions in the reported functions of self-injury were no longer significant once change in urge severity was accounted for, suggesting the urge captures much of what changes in treatment (Victor et al., 2016).
In 43 patients later re-admitted, scores fell from admission to discharge, rose between discharge and re-admission, and did not differ between the two admissions, the pattern expected if the measure tracks clinical state (Washburn et al., 2010).
The published reliable change cutoff is 6.58, which NovoPsych rounds to 7. It reproduces from the scale’s internal consistency rather than its test-retest reliability, and the authors applied it to change between admission and discharge, so it is best suited to comparisons across a course of treatment.
No community norms exist for the ABUSI. The NSSI Treatment Percentile is calculated by NovoPsych from a single clinical sample of 1,327 adolescents and adults admitted for partial hospitalisation or intensive outpatient treatment of NSSI, who scored 18.4 (SD 8.0) at admission (Washburn et al., 2023). That sample was 87.4% female and 75.3% non-Hispanic white. The average closely matches the development sample a decade earlier (18.02, SD 8.45; Washburn et al., 2010), supporting the stability of the reference point, and clients reporting self-injurious disordered eating scored higher than those presenting with self-injury alone (21.3 compared with 17.2; Washburn et al., 2023).
They answer different questions. Counting episodes of self-injury tells you what a person did, while measuring the urge tells you what they experienced and had to manage. Someone may go a week without self-injuring while experiencing frequent and intense urges throughout, and that week is clinically quite different from one in which the urge never arose, even though the behavioural count is the same. Higher urges at admission have also been found to predict self-injury frequency at discharge after accounting for how often the person was self-injuring at admission, which makes the urge worth monitoring alongside the behaviour.
No. A score of zero means the client reported no urge to self-injure during the seven days the measure asks about. Among adolescents and adults admitted for treatment of self-injury, roughly one in eight reported no urge in the week before admission, so a zero is not unusual even in a population receiving intensive treatment. It is best read as information about that particular week rather than a general statement about the presence or absence of self-injury concerns.
With the comparison group in mind. Because everyone in the reference sample was admitted for treatment of NSSI, a client can score well below that group and still be reporting urges that are clinically relevant. This is why the severity category, which is based on what the response options themselves describe, is the primary basis for interpretation. The percentile adds context about how the client compares with a treatment population and should be read alongside the severity category, the individual item responses, and the broader clinical picture.
A change of 7 points or more, in either direction, is larger than would be expected from the natural variability of the measure and can be treated as a real change. Smaller changes do not meet this threshold, although they may still be clinically meaningful. Average improvement across a whole treatment programme is sometimes smaller than 7 points, so a client who does not reach that level of improvement is not necessarily making no progress.
It was not designed as one, and its total score does not indicate whether a person is or is not engaging in NSSI. The measure assumes self-injury is already known or suspected, and its questions are phrased accordingly. Its established use is tracking the intensity of the urge to self-injure over time, with the individual item responses adding detail about the frequency, intensity, duration, and controllability of those urges.
It gives a consistent, quantified record of urges that can be compared across administrations, and it separates the urge into its component parts. Two clients with similar totals may differ substantially in whether their urges are frequent but manageable or rare but overwhelming, and those patterns suggest different clinical priorities. Having the same five questions answered each week also makes gradual change visible in a way that recall in session often does not.
Washburn, J. J., Juzwin, K. R., Styer, D. M., & Aldridge, D. (2010). Measuring the urge to self-injure: Preliminary data from a clinical sample. Psychiatry Research, 178(3), 540–544. https://doi.org/10.1016/j.psychres.2010.05.018
Developed at the Alexian Brothers Behavioral Health Hospital by Dr Jason Washburn and colleagues. Included on NovoPsych with the author’s permission.
American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.). https://doi.org/10.1176/appi.books.9780890425787
Bahamón, M. J., Javela, J. J., Bonilla-Cruz, N.-J., Rivera, D., Vinaccia, S., & Forgiony-Santos, J. (2023). Psychometric properties of the Alexian Brother Urge to Self-Injure (ABUSI) Spanish version for adolescents. Heliyon, 9(6), e16167. https://doi.org/10.1016/j.heliyon.2023.e16167
Becker-Haimes, E. M., Tabachnick, A. R., Last, B. S., Stewart, R. E., Hasan-Granier, A., & Beidas, R. S. (2020). Evidence base update for brief, free, and accessible youth mental health measures. Journal of Clinical Child & Adolescent Psychology, 49(1), 1–17. https://doi.org/10.1080/15374416.2019.1689824
Chávez-Flores, Y. V., Hidalgo-Rasmussen, C. A., & Yanez-Peñúñuri, L. Y. (2019). Assessment tools of non-suicidal self-injury in adolescents 1990-2016: A systematic review. Ciência & Saúde Coletiva, 24(8), 2871–2882. https://doi.org/10.1590/1413-81232018248.18502017
Flannery, B. A., Volpicelli, J. R., & Pettinati, H. M. (1999). Psychometric properties of the Penn Alcohol Craving Scale. Alcoholism: Clinical and Experimental Research, 23(8), 1289–1295. https://doi.org/10.1111/j.1530-0277.1999.tb04349.x
Klonsky, E. D. (2007). The functions of deliberate self-injury: A review of the evidence. Clinical Psychology Review, 27(2), 226–239. https://doi.org/10.1016/j.cpr.2006.08.002
Slesinger, N. C., Hayes, N. A., & Washburn, J. J. (2021). Understanding predictors of change in a day treatment setting for non-suicidal self-injury. Psychology and Psychotherapy: Theory, Research and Practice, 94(S2), 517–535. https://doi.org/10.1111/papt.12295
Victor, S. E., Styer, D., & Washburn, J. J. (2015). Characteristics of nonsuicidal self-injury associated with suicidal ideation: Evidence from a clinical sample of youth. Child and Adolescent Psychiatry and Mental Health, 9, 20. https://doi.org/10.1186/s13034-015-0053-8
Victor, S. E., Styer, D., & Washburn, J. J. (2016). Functions of nonsuicidal self-injury (NSSI): Cross-sectional associations with NSSI duration and longitudinal changes over time and following treatment. Psychiatry Research, 241, 83–90. https://doi.org/10.1016/j.psychres.2016.04.083
Washburn, J. J., Juzwin, K. R., Styer, D. M., & Aldridge, D. (2010). Measuring the urge to self-injure: Preliminary data from a clinical sample. Psychiatry Research, 178(3), 540–544. https://doi.org/10.1016/j.psychres.2010.05.018
Washburn, J. J., Potthoff, L. M., Juzwin, K. R., & Styer, D. M. (2015). Assessing DSM-5 nonsuicidal self-injury disorder in a clinical sample. Psychological Assessment, 27(1), 31–41. https://doi.org/10.1037/pas0000021
Washburn, J. J., Soto, D., Osorio, C. A., & Slesinger, N. C. (2023). Eating disorder behaviors as a form of non-suicidal self-injury. Psychiatry Research, 319, 115002. https://doi.org/10.1016/j.psychres.2022.115002
World Health Organization. (2019). International statistical classification of diseases and related health problems (11th ed.). https://icd.who.int/
