Rehabilitation Program for Improving Agitation Behavior Among Patients with Substance Use Disorder: A Quasi-experimental Study

Substance use disorders (SUDs) represent a significant and growing public health crisis, particularly in low- and middle-income countries where treatment resources are limited (1). In Egypt, recent epidemiological data reveal alarming increases in cannabis and prescription drug misuse, with emerging patterns of polydrug use complicating treatment efforts. A particularly challenging aspect of SUDs is agitation behavior, which affects approximately 34% of patients in Egyptian treatment centers and significantly worsens clinical outcomes (2).

The neurobiological basis of agitation in SUD patients involves dysregulation of prefrontal cortex-amygdala circuitry, where chronic substance use impairs emotional regulation while heightening stress reactivity. This explains why traditional abstinence-focused approaches often fail — they do not address the underlying emotional dysregulation that drives both substance use and agitated behaviors (3). The cyclical nature of this relationship creates substantial barriers to treatment adherence, as patients frequently relapse when unable to manage distressing emotional states without substances (4).

Clinically, agitation manifests along a severity continuum from restlessness to physical aggression, leading to safety concerns and high dropout rates. Current pharmacological approaches, while providing immediate symptom control, often prove inadequate for long-term management due to side effects and limited impact on behavioral components (5). These challenges are exacerbated in Egypt by critical shortages of addiction specialists and persistent stigma surrounding mental health treatment (2).

Developing effective interventions requires addressing three key dimensions: Neurobiological changes from chronic substance use, culturally relevant behavioral strategies, and implementation feasibility in resource-limited settings (6). Our rehabilitation program addresses these needs by combining self-directed evidence-based cognitive techniques (7) with culturally adapted components like family support and spiritual coping (8). By addressing both biological and psychosocial factors, this study contributes to advancing comprehensive care models for patients with SUD by evaluating the effectiveness of a targeted rehabilitation program designed to reduce agitation.

McNemar's test revealed significant reductions in eight of ten drug use–related behaviours after rehabilitation, including nonmedical use, polydrug use, inability to stop, blackouts, family complaints, neglect of responsibilities, and illegal activities (all P < 0.001). Guilt feelings showed a nonsignificant improvement (P = 0.096), while withdrawal symptoms (P = 1.000) and medical problems (P = 0.549) showed no change (Tables 1 and 2).

Over half of participants (54%) reported opioid use, followed by cocaine (44%), sedatives (40%), cannabis (30%), and alcohol (32%), with volatile substances least common (20%). Methods included oral (50%), nasal (36%), and injection (38%), with 54% using multiple routes. Main reasons were curiosity/experimentation and coping with problems (74% each), peer influence (60%), and stress relief (46%), while self-medication (32%), sexual performance (20%), and shyness (20%) were less frequent (Appendix 1 in Supplementary File).

A significant improvement in ABS and DAST-10 scores after the rehabilitation program. High ABS scores dropped from 66% to 16%, while low scores increased from 16% to 52%. Likewise, substantial/severe DAST-10 cases decreased from 48% to 38%, and low cases rose from 6% to 24%. These changes were statistically significant (ABS: χ² = 25.1, P < 0.001; DAST-10: χ² = 36.14, P < 0.001). Correlation analysis revealed strong negative associations for both ABS (R = -0.554) and DAST-10 (R = -0.604), confirming reduced agitation and drug abuse severity post-intervention (Appendix 2 in Supplementary File).

After the 9-session rehabilitation program, significant improvements were observed in both behavioural and psychological outcomes. The ABS scores decreased from 96.21 ± 27.24 to 61.63 ± 27.03 (t = 10.50, P < 0.001, d = 1.10), reflecting reduced agitation. Similarly, DAST-10 scores dropped from 7.30 ± 2.30 to 3.20 ± 2.20 (t = 12.59, P < 0.001, d = 1.96), indicating a marked reduction in drug-related problems (Appendix 3 in Supplementary File).

The rehabilitation program led to significant reductions across all types of aggressive behavior among patients with SUD. Verbal aggression (non-destructive) scores decreased from 20.02 ± 5.65 to 13.78 ± 6.18 (t = 10.503, P < 0.001, d = 1.165). Verbal aggressive (hostile) scores dropped from 21.30 ± 6.40 to 13.50 ± 5.86 (t = 9.505, P < 0.001, d = 1.501). Physical aggression (non-destructive) scores declined from 25.58 ± 6.53 to 14.80 ± 6.47 (t = 12.805, P < 0.001, d = 1.681). Physical aggressive (hostile) scores improved from 29.20 ± 9.09 to 19.51 ± 8.61 (t = 11.591, P < 0.001, d = 1.627; Appendix 4 in Supplementary File). Figure 1 provides a clear visualization of the intervention's impact, showing a consistent reduction in both verbal and physical aggression domains among patients with SUDs (Table 3).

Figure 1.

The box plot of the pre- and post-intervention comparison of different types of aggressive behaviour according to the Agitation Behaviour Rating Scale questions

Subgroup analysis revealed that the intervention's efficacy depended on baseline agitation severity (Appendix 5 in Supplementary File). Patients with high baseline agitation experienced a significant reduction in ABS scores, with a large effect size (mean difference: -13.0, P = 0.012, Cohen's d = 1.18). To assess clinical significance, a response was defined as a ≥ 50% reduction in the ABS score. Using this criterion, 50% of the high-severity subgroup were classified as responders, compared to only 11.1% of the moderate-severity subgroup. The number needed to treat (NNT) for one additional patient to achieve this clinically meaningful response in the high-severity group was approximately 2.5. In contrast, the moderate agitation subgroup showed no significant improvement in ABS scores.

This study revealed that substance use was highly prevalent among the patients, with more than three-quarters reporting cannabis, marijuana, and alcohol consumption. Over 60% of them used nasal or intravenous routes for drug administration, and more than 80% admitted to taking drugs to enhance sexual performance or to overcome shyness. Substance use was particularly common among younger, less educated, unmarried individuals — especially single or divorced — and predominantly among males, many of whom also reported symptoms of depression. Social contexts such as parties, concerts, and sporting events were common situations for smoking cannabis, as patients used substances to relieve stress and anger. These findings are consistent with regional patterns of substance use. A study by Saquib et al. on SUDs in Saudi Arabia highlighted cannabis and alcohol as the primary substances of misuse, noting that poly-drug use is a common complicating factor (10). Similarly, research by Bassiony et al. on Egyptian university students found a high prevalence of tramadol use, particularly among males (11). Furthermore, a study of an Egyptian adolescent sample identified synthetic cannabis and cocaine as the most commonly used drugs (12). Together, these regional studies corroborate the substance use patterns observed in our clinical sample.

A concerning finding in this study was that more than 80% of patients misused prescription medications without a physician's order. This misuse often led to family conflict and concerns about patients' neglect of responsibilities. The reasons for misuse included seeking euphoria, pain relief, relaxation, or assistance with sleep. These findings are consistent with Schepis et al., who reported a high lifetime prevalence of prescription drug misuse (13). Hochstatter et al. similarly documented increasing use of illicit substances (14), while Cohn and Elmasry reported early initiation of cannabis and alcohol among patients (15). Likewise, Kibet et al. highlighted family distress, social conflicts, and harmful outcomes such as deaths related to polysubstance use and nonmedical stimulant use (16).

Importantly, this study demonstrated significant improvements in patients' outcomes following completion of the rehabilitation program. Both the ABS and DAST-10 scores showed marked reductions in severity from the pre-program to post-program assessments. These promising findings are consistent with the possibility that the rehabilitation intervention may improve outcomes, including psychosocial education, behavioural counselling, and non-drug strategies. Interventions such as diaries, noise reduction, deep breathing, massage, music therapy, and progressive muscle relaxation likely contributed to reducing agitation and improving coping skills. While the significant pre-post improvements are encouraging, it is important to note that the single-group design limits our ability to rule out alternative explanations for these changes. The following discussion interprets the findings within this context. These results are consistent with Whiting et al., who confirmed that non-pharmacological interventions effectively reduce violence and aggression (17). They also align with Im et al., who reported that structured interventions provided by specialized teams significantly reduced agitation and minimized the need for restraints (18).

The use of McNemar's test and chi-square analysis provided robust evidence for the observed reductions in both aggression and substance abuse severity, confirming that the improvements were not due to chance. Furthermore, the significant negative correlations between post-intervention ABS and DAST-10 scores reinforce the effectiveness of the rehabilitation program in reducing both aggressive tendencies and substance dependence simultaneously. These findings strengthen the validity of the results and highlight the consistency across the statistical approaches used in the analysis.

Sociodemographic characteristics were also strongly associated with patients' outcomes. Younger, less educated, divorced or single, unemployed, and low-income patients exhibited higher levels of agitation and substance-related problems both before and after the rehabilitation program. These findings suggest that socioeconomic vulnerability and lack of stable social support contribute significantly to drug misuse and behavioural problems. Similar observations were reported by Menculini et al., who found higher irritability scores among younger and unemployed individuals, with married patients displaying lower agitation levels (19). Likewise, Garrote-Camara et al. noted that lower levels of education and socioeconomic status were associated with increased agitation and poorer outcomes (20).

Additionally, the study revealed a significant relationship between patients' living arrangements and their rehabilitation outcomes. Those living alone or in rural areas had persistently higher levels of agitation after rehabilitation compared with patients living with family or in urban settings. Limited access to healthcare facilities, reduced social support, cultural norms, and increased social isolation may explain these findings. Caruso et al. similarly reported that patients living alone exhibited greater physical aggression, impulsivity, and poorer adherence to treatment plans (21, 22).

While the intervention demonstrated overall effectiveness in reducing aggressive behavior and substance use severity, the impact may vary by the type of substance consumed. For instance, individuals primarily using alcohol or sedatives may experience different behavioral outcomes compared to those using marijuana, given the pharmacological and psychological effects of these substances. Although the present study was not powered to conduct detailed subgroup analyses, preliminary observations suggest that reductions in aggression were consistent across groups. Future research with larger samples is recommended to further explore substance-specific effects of rehabilitation interventions.

The subgroup analysis revealed a critical nuance: While patients with high baseline agitation responded robustly, those with moderate agitation showed no significant improvement. This finding generates the hypothesis that the intervention's focus on managing high-intensity aggression may have been a potential 'mismatch' for the needs of the moderate agitation group. For these individuals, whose primary presentation likely involves restlessness, irritability, and inner tension rather than overt aggression, the program content might have been less engaging or applicable. This suggests that this specific patient subgroup might require a different interventional approach, such as one specifically targeting underlying anxiety, restlessness, or boredom. This important hypothesis — that distinct agitation profiles require tailored interventions — should be explicitly tested in future research. Although the small size of the moderate subgroup (n = 9) means this finding must be interpreted with caution, it provides a vital direction for personalizing rehabilitation strategies.

Several non-mutually exclusive factors could explain this result for the moderate subgroup. First, the lack of a control group makes it difficult to determine if this minor increase represents a true iatrogenic effect or merely reflects random fluctuation or regression to the mean around their baseline level. The small sample size of this subgroup (n = 9) means it was severely underpowered, and the mean difference of +2.45 points is likely not clinically significant, falling well within the measurement error of the scale and far below the minimal clinically important difference (MCID) for agitation scales, which often exceeds 10 - 15 points. Second, the intervention's content, which was heavily focused on anger management and coping with high-intensity aggression, may have been mismatched or less engaging for individuals whose primary issue was moderate restlessness or irritability rather than overt aggression. They might have benefited more from interventions targeting anxiety, depression, or boredom as a factor affecting (23, 24). This suggests that future iterations of rehabilitation programs should be tailored more precisely to specific agitation profiles rather than a one-size-fits-all approach.