We will conduct a retrospective cohort study by viewing the patients’ files to measure the effect of naltrexone compared to treatment as usual (TAU). The researchers, clinicians, nurses, and medical assistants who will be involved in this study will be trained for 3-4 sessions before data collection. Patients will be identified by the trained hospital staff (clinicians, nurses, and medical assistants). Adult patients aged 18-65 years diagnosed with StUD and non-StUD (ie, those with other substance use disorders, such as alcohol and smoking) treated with or without naltrexone from January 1, 2011, to December 31, 2023, will be recruited. We will collect the data using a standardized data extraction form or clinical report form (Multimedia Appendix 1), which will be piloted with 30 patients in this study. We will extract the data from the patients’ clinical records. The data will be anonymized prior to data extraction. Data will be collected through web-based Google Forms by the trained researcher and hospital staff. We will include all patients with StUD and non-StUD who were offered the option of naltrexone treatment or TAU. We will exclude the patients who were younger than 18 years.

The study locations will be randomly selected using stratified random sampling based on the zones in Malaysia. Hence, we will identify a total of 15 government hospitals across Malaysia with addiction specialists.

We have calculated the sample size using a two-proportion formula using Power and Sample Size Calculation (version 3.1.2; WD Dupont and WD Plummer) for both StUD and non-StUD patients. A previous study indicated that the probability of stimulant abstinence among patients with StUD was 0.5 when treated with a placebo and 0.65 when treated with naltrexone [7]. A minimum of 113 patients with StUD receiving TAU with naltrexone and 339 patients with StUD receiving TAU without naltrexone are needed to be able to reject the null hypothesis, where abstinence for StUD with and without naltrexone is equal, with the power set at 80%. The type 1 error probability associated with the test of this null hypothesis is 0.05. We will use an uncorrected χ² statistic to evaluate this null hypothesis. Considering 20% of data being incomplete or missing, a total of 136 patients with StUD treated with naltrexone and 136 patients with StUD treated without naltrexone are needed for this study.

The sample size calculation for the non-StUD group referred to a previous study that indicated a 0.23 probability for stimulant abstinence among non-StUD patients treated with a placebo and 0.17 for non-StUD patients treated without naltrexone [11]. A similar formula, accounting for type 1 error probability and 20% incomplete or missing data, was used for the calculation. Therefore, both non-StUD groups need a minimum total of 490 non-StUD patients, with 245 treated with naltrexone and 245 treated without naltrexone.

The following definitions of the cohorts and related terms will be used during data collection:

The data extraction form will be divided into five sections: (1) sociodemographic profile, (2) medical and clinical background, (3) stimulant use (dosage, duration, and routes of administration), (4) StUD or non-StUD treatment, and (5) urine screening.

We will analyze the data separately in a few categories regarding abstinence from substances (drug free) for 1 month, 3 months, 6 months, 9 months, and 12 months. A similar timeline will be applied when investigating treatment retention for both StUD and non-StUD patients while receiving treatment with naltrexone and TAU. The patients will be separated into 4 groups for analysis: (1) patients with StUD treated with naltrexone, (2) patients with StUD treated without naltrexone, (3) non-StUD patients treated with naltrexone, and (4) non-StUD patients treated without naltrexone. If there is any missing data, we will try to search for the patients’ old notes or files, or ask clinicians to contact the patients to obtain the data.

We will analyze the data using Stata software (version 17; StataCorp), along with data cleaning and data verification. Before conducting analyses, data will be verified for errors, statistical assumptions, and potential confounding variables to examine as covariates. Descriptive analysis will be performed to describe the characteristics of the patients, sociodemographic profiles, and percentages of abstinence and treatment retention. Continuous variables will be summarized using means (SDs). Categorical variables will be presented as counts (n) and percentages (%). Univariable analysis will be performed using the χ² test to determine the association of abstinence from stimulants and treatment retention between the naltrexone and TAU groups among StUD and non-StUD patients. We will perform cross-sectional time series analysis for any significant associations (P<.05) in the univariable analysis based on the time points of 1 month, 3 months, 6 months, and 12 months.

This study will be funded by the National Centre of Excellence for Mental Health, Ministry of Health, as a mental health study. We have registered this study with the National Medical Research Register, Ministry of Health Malaysia, with the registration number NMMR ID-24-00626-TBA. We have received ethics approval from the Medical Research & Ethics Committee, Ministry of Health, Malaysia with the reference number NMRR ID-24-00626-TBA (IIR). The Medical Research & Ethics Committee allows the use of secondary data if the data is anonymized; therefore, no informed consent was needed. Patients’ IDs will be anonymized or deidentified, and new IDs will be given by the researcher that will be used in place of their name or other identifying information. All data will be kept in secure, password-protected files. Although we plan to make the data publicly available for reanalysis, we will carefully deidentify the data first. For publications or presentations of the results, any information that could identify the participant will not be provided.

Significant associations in StUD abstinence and treatment retention between TAU and naltrexone among StUD and non-StUD patients are expected to be observed. The findings from this study may provide preliminary evidence to update the clinical practice guideline on the use of naltrexone in treating stimulant misuse.

Currently, no medication is approved for the treatment of amphetamine or methamphetamine use disorder [13], and the effects of naltrexone combined with psychosocial interventions for StUD remain unclear [21]. Clinically, naltrexone’s role in improving abstinence has been observed in patients with alcohol use disorder [26], smoking dependence [27], and amphetamine dependence [19], although contrasting findings have also been reported [23]. However, no local studies in Malaysia have examined the role of naltrexone in treating StUD, despite the significant social impact substance misuse has on communities. Naltrexone is currently used off-label for both StUD and non-StUD populations. Therefore, this study aims to provide evidence on the effectiveness of naltrexone in supporting abstinence among Malaysian patients with StUD and non-StUD, ultimately contributing to best clinical practices and bridging the local research gap on this treatment approach.

This study will also highlight the need for public health approaches to recognize the multifaceted nature of substance misuse and focus on addressing the myriad individual, environmental, and social factors that contribute to StUD. Since prevention programs address risk and protective factors that are common in a range of behavioral problems, a naltrexone program for the treatment of StUD will produce positive outcomes not only in preventing drug misuse but also in reducing aggression, early behavioral change, and impaired driving, and improving mental health and educational outcomes.

There are a few limitations that should be considered for this study. As this study will collect the data from the patients’ files, the patients’ information regarding abstinence and treatment retention may be different, and therefore, we will separate the analysis into a few categories: 1 month, 3 months, 6 months, 9 months, and 12 months to overcome this problem. Some of the data may not be complete, such as history of other psychiatric diseases or health issues.

This study is intended for publication in a psychiatric journal such as JAMA Psychiatry, World Psychiatry, or BMC Psychiatry. We may also publish in high-impact journals that are suitable for our topic, such as BMJ, PLoS, or a JMIR Publications’ journal. A few findings will be presented at both local and international conferences. The preliminary findings will be presented to the Psychiatry Clinical Practice, Ministry of Health, Malaysia. This study will also help clinicians decide the best treatment for the patients and might be used as a reference when developing clinical practice guidelines. We will also produce plain language summaries that will be shared with the public through social media and health care websites.