Predictors of naltrexone prescribing for alcohol use disorder from the emergency department

Abstract

Background:

Excessive alcohol use is a leading cause of preventable death in the United States, with the emergency department (ED) serving as a critical touchpoint for individuals with alcohol use disorder (AUD). Despite clinical guidelines recommending initiation of medication for AUD (MAUD), such as naltrexone, ED prescribing remains rare. The objective of this study is to characterize clinician naltrexone prescribing practices for ED patients with hazardous drinking or AUD, and identify patient- and encounter-level predictors of naltrexone prescribing within a large, integrated health system.

Methods:

We conducted a retrospective cohort study of adult ED encounters across 12 hospitals from 2022–2024. Eligible encounters included patients with a positive AUDIT-C screen, indicating hazardous alcohol use or an active AUD, had no exclusion criteria contraindicating naltrexone and were discharged from the ED. The primary outcome was provision of a naltrexone prescription at ED discharge; and the secondary outcome was prescription fill. We used a multivariable logistic regression model with generalized estimating equations (GEE) to identify predictors of prescribing.

Results:

Of 52,701 treatment-eligible ED encounters, only 0.5% resulted in a naltrexone prescription. Prescriptions were more likely in encounters involving younger, male patients with higher AUDIT-C scores, alcohol-related complaints, and those occurring at an academic ED. In the logistic GEE model, academic setting, alcohol withdrawal diagnosis, and greater alcohol misuse severity were independently associated with increased prescribing. Nearly half (45%) of ED naltrexone prescriptions were filled.

Conclusions:

Naltrexone prescribing among treatment eligible patients is rare. Encouragingly, nearly half of patients receiving a prescription proceeded to fill it, highlighting a promising opportunity for ED based prescribing of naltrexone to initiate AUD treatment. To improve AUD care, systematic ED-based strategies are urgently needed that go beyond universal screening to address barriers to MAUD initiation.

Introduction

Excessive alcohol use is the leading preventable cause of death in the United States, accounting for 178,000 deaths annually. Alcohol-related emergency department (ED) visits have risen by 47% in the past decade, highlighting the growing burden of alcohol misuse on the healthcare system and the associated risk of adverse outcomes. Patients with frequent ED visits are at especially high risk, with nearly 1 in 10 dying within a year of an index encounter.(Esser et al., 2022, Parkinson et al., 2016, Suen et al., 2022) For many with AUD, the ED may be their primary point of contact with the healthcare system, highlighting how EDs are uniquely positioned to facilitate treatment of AUD and prevent AUD morbidity and mortality.(Hawk and D’Onofrio, 2023)

Effective medications for AUD (MAUD), such as naltrexone, acamprosate, and disulfiram, help patients reduce alcohol consumption, engage in treatment, and maintain recovery. A systematic review and meta-analysis support oral naltrexone’s efficacy (50 mg daily) in delaying return to heavy drinking and in reducing return to any drinking.(McPheeters et al., 2023) In a recent feasibility study, initiation of oral naltrexone in the ED setting was associated with reductions in alcohol cravings and improvement in quality of life measures.(Cowan et al., 2025) However, despite recommendations from the Society of Academic Emergency Medicine’s GRACE guidelines, which endorse prescribing naltrexone for ED patients with AUD as part of evidence-based management, real-world implementation remains limited.(Borgundvaag et al., 2024) This gap highlights persistent barriers in ED practice, such as clinician discomfort, lack of familiarity with MAUD, and workflow constraints, which contribute to the underutilization of MAUD in the ED.(Gottlieb et al.)

Although AUD is highly prevalent among ED patients and outpatient naltrexone therapy is effective, few studies have explored ED clinician MAUD prescribing practices. Therefore, we sought to describe naltrexone prescribing practices among discharged ED patients screening positive for hazardous drinking or AUD, using the routinely collected Alcohol Use Disorders Identification Test-Consumption (AUDIT-C) screen, and identify patient- and encounter-level predictors of naltrexone prescribing within a large, integrated health system.

Materials and Methods

Study Design and Setting

We conducted a retrospective cohort study at 12 hospital-based EDs within a large, integrated health system with an annual census of approximately 600,000 ED encounters. EDs range from urban, academic, tertiary care facilities to critical access centers representing diverse geographic locations and socioeconomic populations. The AUDIT-C, a brief alcohol screening instrument which reliably identifies persons with hazardous drinking or active alcohol use disorders, is routinely performed during triage as part of the nursing intake process for all ED patients within this system.(Bush et al., 1998, RodrÍguez-Martos and Santamariña, 2007) Screening results are documented directly into the electronic health record and are available to clinicians in real time. There were no system-wide guidelines for the provision of naltrexone or other MAUD. Injectable naltrexone was not available to be administered during the ED encounter. This study was approved by the Institutional Review Board and designed according to STROBE guidelines.(Vandenbrouckel et al., 2007)

Selection of Participants

We collected data for all naltrexone treatment eligible adult encounters from January 1, 2022, to December 31, 2024, who screened positive for hazardous drinking or AUD (defined as an AUDIT-C score ≥ 4 for men or ≥ 3 for women) and were discharged from the ED. Encounters without a documented AUDIT-C score (e.g., patient refused or was unable to answer) were excluded. Encounters were not considered naltrexone treatment eligible if any exclusion criteria were met: encounter associated with an acamprosate or disulfiram prescription, a MAUD prescription fill or administration within 30 days prior to ED encounter (oral naltrexone, intramuscular naltrexone, acamprosate, or disulfiram), abnormal liver function test (most recent AST or ALT ≥ 200 IU), history of opioid use disorder identified by International Classification of Diseases, 10th Edition (ICD-10; available in appendix), and/or an opioid prescription fill within 90 days prior to ED encounter. The GRACE guidelines do not require screening liver function tests prior to MAUD initiation, and thus, encounters without documented LFTs were included in the study cohort.(Borgundvaag et al., 2024)

Data Collection and Outcome Variables

Data was sourced from Health Data Compass, a multi-institutional data warehouse that contains all system electronic health record (EHR) data, including patient demographics, encounter information, prescribed and filled medications, and laboratory results. The primary outcome of the study was whether the ED clinician prescribed naltrexone during the ED encounter. The secondary outcome was if the patient filled the ED naltrexone prescription. We determined prescription filling from a medication history service, SureScripts (Arlington, VA), that works in conjunction with our institution’s EHR and includes nearly all outpatient prescriptions from major US pharmacies and pharmacy benefit managers.

Predictor variables

Predictor variables include the patient’s age, self-reported race and ethnicity, sex, and socioeconomic status as measured by the proxy ZIP code-based social vulnerability index.(Flanagan et al., 2018) Payor type was categorized into private insurance, public insurance (e.g., traditional Medicare or Medicaid), Tricare, or other for the purposes of analysis. EDs were dichotomized into academic and community setting. Because the AUDIT-C is performed regardless of presenting chief complaint, we dichotomized the documented chief complaints into potentially alcohol related (altered mental status, alcohol problem, withdrawal, drug/alcohol assessment, alcohol intoxication, and/or addiction problem) or not alcohol related. We used ICD-10 codes to identify patient encounters with a history of mental illness or with an ED discharge diagnosis of alcohol withdrawal syndrome (ICD-10 codes available in appendix).

Analysis

Descriptive statistics were used to describe patient-and encounter-level characteristics. Naltrexone prescription recipients and prescription fills were compared to non-recipients and non-fills, respectively, using Pearson’s Chi-squared test, Wilcoxon rank sum test, and Fisher’s exact test where appropriate. We then used a multivariable logistic regression model with generalized estimating equations (GEE) to assess predictors of the dichotomous primary outcome of naltrexone prescribing (yes/no). The logistic GEE model accounts for dependence between observations (i.e., single patient with multiple encounters) and ensures that each encounter is weighted appropriately. All results displayed and discussed here are at the encounter level, not patient level. As such, each encounter with a naltrexone prescription was treated uniquely. Odds ratios and 95% confidence intervals were calculated for each variable. Statistical analyses were performed using R software (version 4.4.0, R Foundation for Statistical Computing, Vienna, Austria).

Results

Our analytic dataset included 60,705 adult ED encounters with a positive AUDIT-C screen and discharge from the ED. We excluded a total of 8,004 encounters, including 2,562 encounters associated with an MAUD prescription fill within 30 days prior to the ED encounter. Therefore, we analyzed 52,701 ED encounters representing 40,640 unique patients. Figure 1 demonstrates the encounters included in the descriptive analysis. To ensure model convergence, we excluded encounters with an unknown or missing ethnicity (n=643), resulting in a logistic GEE model based on 52,058 encounters.

Figure 1:

Consort Study Flow Diagram

Overall, only 269 (0.5%) of treatment eligible encounters resulted in a naltrexone prescription. Encounter characteristics are shown in Table 1. Compared to encounters that did not result in a naltrexone prescription, those with a naltrexone prescription featured younger, male, non-Hispanic patients with more severe alcohol misuse (higher AUDIT-C score), and had a higher social vulnerability index. There was significant practice setting variability with receipt of a naltrexone prescription more commonly occurring in the single academic hospital. Receipt of a naltrexone prescription was also more common when the encounter was associated with an alcohol-related chief complaint or a discharge diagnosis of alcohol withdrawal. There was no observed difference in receipt of a naltrexone prescription across insurance type or if the patient had a history of mental illness. Notably, nearly half of the naltrexone prescriptions were filled (n=121, 45%, Table 2).

Table 1:

ED Naltrexone Prescription Receipt at the Encounter Level

Variables	Overall N=52,701	Naltrexone Prescribed N=269	Naltrexone Not Prescribed N=52,432	P-value
Age (years)				<0.001
18-34	11,697 (22.2)	47 (17.5)	11,650 (22.2)
35-49	22,521 (42.7)	152 (56.5)	22,369 (42.7)
50-64	11,130 (21.1)	58 (21.6)	11,072 (21.1)
65+	7,353 (14.0)	12 (4.5)	7,341 (14.0)
Sex				<0.001
Female	23,895 (45.3)	84 (31.2)	23,811 (45.4)
Male	28,806 (54.7)	185 (68.8)	28,621 (54.6)
Race				0.4
American Indian or Alaska Native	851 (1.6)	6 (2.2)	845 (1.6)
Asian	694 (1.3)	4 (1.5)	690 (1.3)
Black or African American	6,439 (12.2)	37 (13.8)	6,402 (12.2)
White or Caucasian	35,811 (68.0)	170 (63.2)	35,641 (68.0)
More Than One Race	471 (0.9)	4 (1.5)	467 (0.9)
Other/Unknown	8,435 (16.0)	48 (17.8)	8,387 (16.0)
Ethnicity				0.006
Hispanic, Latino/a, or Spanish Origin	10,987 (20.8)	74 (27.5)	10,913 (20.8)
Non- Hispanic, Latino/a, or Spanish Origin	41,071 (77.9)	195 (72.5)	40,876 (78.0)
Unknown	643 (1.2)	0 (0)	643 (1.2)
AUDIT-C score (Median, IQR)	5 (4, 9)	11 (7, 12)	5 (4, 9)	<0.001
Payor Type				0.8
Private	24,287 (46.1)	124 (46.1)	24,163 (46.1)
Public	27,771 (52.7)	141 (52.4)	27,630 (52.7)
Tricare	635 (1.2)	4 (1.5)	631 (1.2)
Unknown	8 (0)	0 (0)	8 (0)
Academic Setting	21,479 (40.8)	177 (65.8)	21, 302 (40.6)	<0.001
Social Vulnerability Index* (Median, IQR)	0.4 (0.2, 0.8)	0.5 (0.2, 0.8)	0.4 (0.2, 0.8)	0.002
Alcohol-related chief complaint				<0.001
Alcohol-related	10,088 (19.1)	134 (49.8)	9,954 (19.0)
Other	42,613 (80.9)	135 (50.2)	42,478 (81.0)
Alcohol Withdrawal Encounter				<0.001
Present	3,351 (6.4)	106 (39.4)	3,245 (6.2)
Not Present	49,350 (93.6)	163 (60.6)	49,187 (93.8)
History of Mental Health Diagnosis				0.13
Present	791 (1.5)	7 (2.6)	784 (1.5)
Not Present	51,910 (98.5)	262 (97.4)	51,648 (98.5)

Data reported at N% unless stated otherwise

^*A composite measure designed to capture a community’s resilience to external stressor. Percentile rank values range from 0 (lowest vulnerability) to 1 (highest vulnerability).

Table 2:

ED Naltrexone Prescription Fills at the Encounter Level

Variables	Overall N=269	Naltrexone Filled N=121	Naltrexone Not Filled N=148	P-value
Age				0.2
18-34	47 (17.5)	26 (21.5)	21 (14.2)
35-49	152 (56.5)	62 (51.2)	90 (60.8)
50-64	58 (21.6)	29 (24.0)	29 (19.6)
65+	12 (4.5)	4 (3.3)	8 (5.4)
Sex				0.2
Female	84 (31.2)	43 (35.5)	41 (27.7)
Male	185 (68.8)	78 (64.5)	107 (72.3)
Race				0.3
American Indian or Alaska Native	6 (2.2)	1 (0.8)	5 (3.4)
Asian	4 (1.5)	1 (0.8)	3 (2.0)
Black or African American	37 (13.8)	22 (18.2)	15 (10.1)
White or Caucasian	170 (63.2)	72 (59.5)	98 (66.2)
More Than One Race	4 (1.5)	2 (1.7)	2 (1.4)
Other/Unknown	48 (17.8)	23 (19)	25 (16.9)
Ethnicity				0.3
Hispanic, Latino/a, or Spanish Origin	74 (27.5)	37 (30.6)	37 (25.0)
Non- Hispanic, Latino/a, or Spanish Origin	195 (72.5)	84 (69.4)	111 (75.0)
Unknown	-	-	-
AUDIT-C score (Median, IQR)	11 (7, 12)	10 (7, 12)	11 (7, 12)	0.4
Payor Type				0.8
Private	124 (46.1)	56 (46.3)	68 (45.9)
Public	141 (52.4)	64 (52.9)	77 (52.0)
Tricare	4 (1.5)	1 (0.8)	3 (2.0)
Unknown	-	-	-
Academic Setting	177 (65.8)	106 (87.6)	71 (48.0)	<0.001
Social Vulnerability Index* (Median, IQR)	0.5 (0.2, 0.8)	0.6 (0.3, 0.8)	0.5 (0.2, 0.8)	0.13
Alcohol-related chief complaint				0.6
Alcohol-related	134 (49.8)	58 (47.9)	76 (51.4)
Other	135 (50.2)	63 (52.1)	72 (48.6)
Alcohol Withdrawal Encounter				>0.9
Present	106 (39.4)	48 (39.7)	58 (39.2)
Not Present	163 (60.6)	73 (60.3)	90 (60.8)
History of Mental Health Diagnosis				0.5
Present	7 (2.6)	2 (1.7)	5 (3.4)
Not Present	262 (97.4)	119 (98.3)	143 (96.6)

Data reported at N% unless stated otherwise

^*A composite measure designed to capture a community’s resilience to external stressor. Percentile rank values range from 0 (lowest vulnerability) to 1 (highest vulnerability).

In the logistic GEE model, positive predictors of naltrexone prescribing included encounters in the academic setting (versus community setting), having an ED discharge diagnosis of alcohol withdrawal, and an encounter associated with more severe alcohol misuse by AUDIT-C score (all p<0.001, Table 3). Notably, for every unit increase in AUDIT-C score, the odds of receiving a naltrexone prescription increased by 21% (aOR 1.21, 95% CI: 1.16–1.27, p<0.001).

Table 3:

Multivariable Logistic Regression Model with Generalized Estimating Equations: Odds Ratios for Receipt of a Naltrexone Prescription

Predictor	Odds Ratio (95% CI)	P-value
Age
18-34	-	-
35-49	1.13 (0.79-1.60)	0.5
50-64	1.04 (0.69-1.58)	0.8
65+	0.65 (0.34-1.22)	0.2
Sex
Female	-	-
Male	1.04 (0.79-1.37)	0.8
Ethnicity
Hispanic, Latino/a, or Spanish Origin	-
Non- Hispanic, Latino/a, or Spanish Origin	0.92 (0.69-1.28)	0.6
AUDIT-C (increase score by 1)	1.21 (1.15-1.28)	<0.001
Alcohol-related chief complaint	1.29 (0.93-1.80)	0.13
Academic Setting	1.91 (1.47-2.50)	<0.001
Alcohol Withdrawal Encounter	3.80 (2.74-5.26)	<0.001

Discussion

We found that naltrexone prescribing is uncommon among ED encounters who screen positive for hazardous drinking or an active alcohol use disorder. Both patient-level and encounter-level factors, such as age, setting, and severity of hazardous drinking, influenced the likelihood of receiving a naltrexone prescription, indicating that prescriber education and awareness may be needed to ensure all eligible patients receive this potentially life-saving treatment. Encouragingly, nearly half of the naltrexone prescriptions were filled.

A novel strength of this study is the use of the AUDIT-C screen to identify patients with hazardous drinking or active alcohol use disorders as compared to alcohol-related diagnosis or billing codes.(Gottlieb et al.) The AUDIT-C, a brief, three question screening tool, can identify risky drinking behaviors before progression to diagnosable alcohol use disorders.(Bush et al., 1998, Dawson et al., 2005) As such, the AUDIT-C will identify patients with subclinical or emerging AUD. The benefits of this strategy are evident in the proportion of encounters in patients with a positive AUDIT-C screen and a non-alcohol related chief complaint (81%). Consequently, identification of at-risk patients using the AUDIT-C screen, compared to diagnosis codes, is less dependent on clinician judgement or availability of relevant diagnosis codes due to variations in documentation practice. Comparably, a universal ED triage screening and treatment protocol led to increased identification, assessment, and treatment of opioid use disorder.(Lowenstein et al., 2023)

While the AUDIT-C is a validated screening tool for identifying hazardous drinking and potential AUD, relying on a positive AUDIT-C screen alone to guide the initiation of naltrexone in the ED setting has important limitations. The AUDIT-C is designed to detect risky drinking behaviors rather than to diagnose AUD or assess treatment readiness. Many patients who screen positive may not meet full Diagnostic and Statistical Manual of Mental Illnesses (DSM-5) criteria for moderate or severe AUD—the population for whom pharmacotherapy like naltrexone is indicated. Nonetheless, naltrexone is supported by evidence for reducing heavy or binge drinking in individuals who may not meet criteria for AUD, but who engage in episodic heavy or binge drinking.(O’Malley et al., 2015, Roos et al., 2021) As such, ED-based prescribing of naltrexone may still offer benefit to a broader group of patients identified through routine screening, such as the AUDIT-C, particularly those motivated to reduce their alcohol use. In practice, while the AUDIT-C serves as a useful first step—and was used pragmatically in our study design—prescribing decisions require more than a screening score; they depend on clinical judgment, shared decision-making, and a nuanced understanding of the patient’s goals, medical history, and readiness for change.

To enhance the clinical utility of screening for the purpose of initiating naltrexone in the ED, we recommend building upon traditional Screening, Brief Intervention, and Referral to Treatment (SBIRT) approaches by adopting a more targeted, treatment-oriented model. While the AUDIT-C is an effective initial screener for hazardous alcohol use, it should be followed by additional structured assessment to determine the presence and severity of AUD (e.g., using selected DSM-5 criteria or the full AUDIT) and to identify contraindications such as recent opioid use or advanced liver disease. Brief interventions should be adapted to focus not only on raising awareness and enhancing motivation to change, but also on assessing readiness to initiate pharmacotherapy and providing education about naltrexone’s role in supporting recovery. Motivational interviewing techniques can be leveraged to explore patients’ treatment goals, reduce ambivalence, and foster commitment to both filling and taking the prescription.(Kohler and Hofmann, 2015, Bernstein and Bernstein, 2008) Referrals to treatment should be aligned with the patient’s preferences and level of readiness, ideally including linkage to ongoing pharmacotherapy or integrated behavioral health support. While there is a need to balance the required time and workflow integration of the AUDIT-C into routine practice, universal ED screening and paired SBIRT are essential building blocks for early identification of at-risk individuals early and provide timely interventions.(Barata et al., 2017)

Despite guidelines from the Society of Academic Emergency Medicine recommending a naltrexone prescription for the management of AUD in ED patients, clinicians rarely provide prescriptions for patients who are eligible.(Borgundvaag et al., 2024) Our findings are consistent with prior epidemiologic studies of ED patients with hazardous drinking, all of which report naltrexone or MAUD prescription rates of less than 1% among eligible encounters(Gottlieb et al., Duvalyan et al., 2024, Chockalingam et al., 2022). Previous studies have cited several barriers to naltrexone prescribing, including prescribers’ medication knowledge gaps, deference of perceived non-emergent treatments, and lack of universal screening AUD protocols.(Covarrubias et al., 2025, Philippine et al., 2022) Yet, even in the presence of near universal AUD screening in our integrated healthcare system, naltrexone prescribing remained a rare occurrence. Our findings highlight missed opportunities for MAUD treatment initiation and engagement during the “teachable” moment of the ED visit, regardless of if the encounter is related to alcohol use.(Mitka, 1998, Hawk and D’Onofrio, 2018) Similar to other chronic disease states, such as hypertension and diabetes, ED encounters should be leveraged for interventions to initiate evidence-based treatment.(Anderson et al., 2016)

Although naltrexone prescribing is safe and effective, our data suggests that certain patient- and encounter-level factors influence the provision of ED MAUD. Compared to patients between the ages of 30–49, older adults aged 65 and older were half as likely to receive a prescription for naltrexone. This is striking as the physiology of aging makes older adults particularly vulnerable to the deleterious effects of AUD, including increased serious and fatal falls, functional disability, and cognitive impairment.(Yuan et al., 2023, White et al., 2023, Topiwala et al., 2017) Our data also highlight the disparity in MAUD prescribing among Hispanic patients, consistent with similar ED studies examining medications for opioid use disorder and naloxone.(Holland et al., 2023, Papp and Emerman, 2023) This finding is likely multifactorial, but implementation strategies designed to increase ED MAUD among should specifically address the potential for perpetuating inequities. Finally, we observed significant variation in naltrexone prescribing rates between the academic and community EDs. This pattern is similar to ED buprenorphine uptake for opioid use disorder, emphasizing the future need for generalizable strategies to engage clinicians, address institutional barriers, and connect patients with outpatient care.(Moore et al., 2022)

Limitations

This study has several limitations. First, this is a retrospective observational study within a single healthcare system and naltrexone prescribing patterns may not be generalizable to other settings or geographic regions. However, the included sites range from urban to rural locations and both academic and community settings. Second, the AUDIT-C screen has intrinsic limitations in that it only focuses on current alcohol consumption, does not formally diagnose AUD, and does not address alcohol-related consequences, dependence, or other dimensions of AUD that may impact the provision of MAUD. However, the AUDIT-C performs well to identify alcohol dependence, any AUD, and risky drinking among ED patients.(Dawson et al., 2005) Third, not all patients identified as treatment eligible in the current study may be treatment ready. This study did not assess readiness for change, concerns around new medications, or potential stigma associated with starting MAUD. However, 45% of naltrexone prescriptions were filled, indicating that there is treatment readiness in the ED setting. Further, patients and clinicians strongly favor naltrexone prescribing, citing a positive opinion of naltrexone and recognition of the ED encounter as an opportunity to engage patients in evidence-based treatment.(Forsgren et al., 2024, Philippine et al., 2022) Fourth, we acknowledge that the inherent imbalance cohort (naltrexone prescribed vs naltrexone non-prescribed) may affect the robustness of our findings. Finally, routinely collected EHR data has limitations, such as missing or incorrect inclusion of data, which can affect how covariates are defined in the model. For example, an “addiction problem” chief complaint may be related to another substance or a trauma presentation related to intoxication may not be captured as alcohol-related.

Conclusions

We identified that very few ED patients with hazardous alcohol use or an active AUD are prescribed naltrexone. Additionally, disparities in prescribing are most notable for older adults, Hispanic patients, and those presenting to community EDs. Encouragingly, nearly half of patients receiving a prescription proceeded to fill it, highlighting a promising opportunity for ED based prescribing of naltrexone to initiate AUD treatment. There is a need for ED specific implementation strategies to improve delivery of this important evidence-based treatment and address disparities in MAUD access as part of a broader public health approach to curbing addiction and alcohol related chronic disease.

Appendix

Appendix Table 1.

Identification of comorbidities

Medical condition	ICD-10 codes
Alcohol withdrawal:	F10.3, F10.30, F10.31, F10.4, F10.40, F10.41, F10.03, F10.05, F10.06
Psychiatric comorbidities	F2, F3, F4, F32
Opioid use disorder	F11 (except F11.21) T40.0X5, T40.2X5, T40.3X5, T40.4X5, T40.605 T40.695, T40.0X1, 0X2, 0X3, 0X4 T40.1X1, 1X2, 1X3, 1X4 T40.2X1, 2X2, 2X3, 2X4 T40.3X1, 3X2, 3X3, 3X4 T40.4X1, 4X2, 4X3, 4X4 T40.601-T40.604 T40.691-T40.694