Rolul scalelor clinice în managementul terapeutic al sevrajului alcoolic

1. Introduction

Alcohol withdrawal syndrome (AWS) represents one of the most frequent and potentially fatal challenges in consultation-liaison and emergency psychiatric practice. It is estimated that approximately 8% of hospitalized patients present manifestations of AWS, while complicated forms – withdrawal seizures and delirium tremens (DT) – occur in up to 20% of patients with alcohol use disorder (AUD)^(1,2).

The onset of AWS typically occurs within 6 to 24 hours of the last alcohol intake, but may also appear several days after a significant reduction in habitual consumption. The clinical picture includes tremor, psychomotor agitation, insomnia, paroxysmal diaphoresis, tachycardia, nausea, vomiting, anxiety, hallucinations, and in severe forms – generalized tonic-clonic seizures or delirium^(1,2). Mortality in untreated DT can reach 35-37%, but decreases to 1-5% under correct and early management⁽³⁾. Furthermore, a prospective Norwegian study of 36,287 patients demonstrated that survivors of a DT episode present a 56% higher all-cause mortality risk compared with alcohol-dependent patients without DT, with deaths attributable predominantly to cardiovascular and hepatic disease⁽⁴⁾.

In this context, the shift from fixed-dose regimens to symptom-triggered therapy has revolutionized patient safety, reducing the length of hospitalization and the incidence of iatrogenic complications^(5,6). This article explores modern instruments for quantifying withdrawal severity and risk as a foundation for precision therapeutic titration.

2. Current diagnostic criteria: DSM-5-TR and ICD-11

2.1. DSM-5-TR

DSM-5-TR (2022) classifies withdrawal within the category of substance-induced disorders, requiring three diagnostic criteria⁽⁷⁾:

• A) Cessation (or reduction) of alcohol use that has been previously heavy and prolonged.
• B) The onset of at least two of the following symptoms, occurring within hours to days: autonomic hyperactivity (sweating, HR > 100 bpm); marked hand tremor; insomnia; nausea or vomiting; transient hallucinations or illusions (visual, tactile, auditory); psychomotor agitation; anxiety; generalized tonic-clonic seizures.
• C) Symptoms cause clinically significant distress or impairment in social/occupational functioning and cannot be attributed to another medical or psychiatric condition.

Limb tremor is the pathognomonic sign, typically appearing 6-8 hours after the last intake. Perceptual disturbances manifest at 8-12 hours, while tonic-clonic seizures manifest at 12-24 hours (in fewer than 3% of patients). Residual anxious symptomatology, insomnia and autonomic dysfunction may persist for 3-6 months at a reduced intensity^(1,7).

2.2. ICD-11

ICD-11, in force in signatory states since 2022, restructures the classification, placing alcohol use under code 6C40 (Disorders due to use of alcohol)⁽⁸⁾. Unlike DSM-5-TR, ICD-11 emphasizes the physiological aspect of dependence and introduces fine clinical specifiers:

• 6C40.40 – Uncomplicated alcohol withdrawal syndrome
• 6C40.41 – Alcohol withdrawal with perceptual disturbances
• 6C40.42 – Alcohol withdrawal with seizures
• 6C40.43 – Alcohol withdrawal with perceptual disturbances and seizures
• 6C40.5 – Alcohol-induced delirium (delirium tremens).

This stratification offers superior clarity compared to ICD-10 and encourages approaching withdrawal as a continuum – from mild forms to genuine medical emergencies⁽⁸⁾.

2.3. Clinical subsyndromes

In practice, AWS comprises three subsyndromes, not necessarily simultaneously present in the same patient⁽¹⁾:

• Autonomic hyperactivity – restlessness, diaphoresis, tachycardia, hypertension, tremor.
• Neuronal excitation – the kindling mechanism, responsible for seizure generatin.n
• Perceptual and arousal distortions – transient hallucinations, hypervigilance.

3. Neurobiological foundations: the “glutamatergic storm”

Understanding the pathophysiology of withdrawal transforms the use of scales from an administrative task into a therapeutic necessity. Ethanol exerts a dual modulatory effect on the central nervous system^(9,10):

• It potentiates inhibition through facilitation of GABA-A receptors.
• It inhibits excitation through antagonism of glutamatergic NMDA, AMPA and kainate receptors.

Chronic exposure triggers bidirectional neuroplastic adaptation^(9,10).

• GABAergic system – down-regulation: the density and sensitivity of GABA-A receptors decrease. The “brake” becomes defective.
• Glutamatergic system – up-regulation: NR1, NR2A and NR2B subunits of NMDA receptors increase in the hippocampus, cortex and cerebellum. The “accelerator” remains pressed to the floor.

Upon abrupt cessation of consumption, the imbalance collapses brutally: hypoactive GABA receptors can no longer inhibit a hyper-expressed glutamatergic system, generating a state of neuronal hyperexcitability – the glutamatergic storm^(9,11). Massive calcium influx through hyperactive NMDA receptors leads to excitotoxicity, oxidative stress and – in the absence of treatment – neuronal death, particularly in limbic structures and the frontotemporal cortex⁽¹¹⁾. This excitotoxicity underlies the cerebral atrophy documented in patients with repeated untreated withdrawal episodes and grounds the preventive role of benzodiazepines (BZD).

The kindling phenomenon

Each insufficiently treated withdrawal episode sensitizes NMDA receptors, lowering the seizure threshold and worsening subsequent episodes^(12,13). Recent data from the validation of PAWSS in the Hungarian population confirm that a history of withdrawal seizures or DT (kindling-related predictors) represents the strongest risk factor for complicated withdrawal⁽¹²⁾. The clinical message is unequivocal: prevention through predictive scales is no longer optional.

4. Treatment of alcohol withdrawal syndrome: certainties and uncertainties

4.1. Scientifically grounded certainties

Three therapeutic pillars are universally ­accepted^(5,6,14):

• Tranquilizing substitutive treatment (BZD as first line).
• Mandatory parenteral thiamine therapy – 300-500 mg/day i.v./i.m. for a minimum of three days, with oral continuation^(14,15).
• Hydroelectrolytic rebalancing guided by serum sodium and magnesium levels.

4.2. Practical uncertainties

Choice of benzodiazepine

In patients with preserved hepatic function, long half-life BZDs (diazepam, chlordiazepoxide) – metabolized by microsomal oxidation – offer “smooth” withdrawal through metabolic auto-tapering. In cirrhotic, elderly or hepatic encephalopathy patients, lorazepam and oxazepam are preferred, because they undergo direct hepatic conjugation, without active metabolites^(5,16).

Non-benzodiazepine alternatives

Carbamazepine (800-1200 mg/day) remains a valid option in mild-moderate withdrawal, especially in patients with a history of seizures or EEG changes⁽¹⁷⁾. A recent meta-analysis confirmed the efficacy of anticonvulsants (carbamazepine, gabapentin, valproate) in moderate withdrawal, although BZDs remain the standard in severe forms^(17,18).

Phenobarbital is gaining ground as a rescue therapy in BZD-refractory withdrawal or in cases of pharmacological resistance^(19,20). Its use in ICU/ED settings increased by over 200% between 2016 and 2020 in the United States of America, with weight-based loading doses (5-10 mg/kg i.v. over 30 minutes)^(19,21). However, a 2024 meta-analysis did not reveal a significant reduction in intensive-care admissions or adverse events compared to BZD monotherapy in the emergency setting⁽²²⁾.

Gabapentin (300-900 mg three times daily) represents a promising adjunctive option in mild-moderate withdrawal^(5,18).

The glucose issue

Glucose infusions must be mandatorily preceded by parenteral thiamine – isolated glucose administration in a deficient patient depletes the last reserves of B1 and can precipitate iatrogenic Wernicke encephalopathy^(14,15). Approximately one-third of AWS patients present stress hyperglycemia or pancreatic lesions, which is why glycemia must be checked urgently.

Use of neuroleptics

Neuroleptics have no role in the basic AWS regimen, but may be adjunctive in the presence of severe perceptual productivity. Tiapride and haloperidol (2-5 mg) are the classic options, with avoidance in the presence of seizures (they lower the seizure threshold)^(1,5).

5. Clinical scales

If DSM-5-TR and ICD-11 provide us with YES/NO criteria for diagnosis, clinical scales transform these qualitative criteria into quantitative data, enabling the monitoring of clinical evolution.

Recent literature and medical calculation applications (MDCalc, UpToDate) identify an ensemble of five scales which, used hierarchically, cover the entire clinical pathway of the withdrawal patient^(5,6,23). The proposed algorithm is as follows:

• PAWSS – at admission (prediction).
• BAWS – in the emergency department (rapid triage).
• CIWA-Ar/Sellers Inventory – on the ward (fine monitoring).
• Cushman – in patients with cardiovascular risk or who are uncooperative.
• RASS – continuously (safety).

5.1. PAWSS – Prediction of Alcohol Withdrawal Severity Scale

PAWSS, developed and prospectively validated by Maldonado et al. (2014, 2015), is the only validated instrument that allows the identification of patients at high risk for complicated withdrawal before the appearance of severe symptoms^(24,25). The scale contains 10 binary items (YES/NO), assessing:

• Part I – Threshold criteria: recent alcohol consumption, positive blood alcohol level at admissio.n
• Part II – Biological/clinical data: current signs of withdrawal, acute medical comorbidities.
• Part III – Anamnesis: previous episodes of withdrawal, seizures, DT, rehabilitation attempts, polypharmacy (BZD/sedatives), recent binge drinking.

Interpretation

• Score < 4: low risk – simple monitoring.
• Score ≥ 4: high risk – aggressive prophylaxis recommended.

In the pivotal prospective study of 403 patients, PAWSS demonstrated 93.1% sensitivity, 99.5% specificity and 93.1% positive predictive value at the cutoff of 4 points, with excellent inter-rater reliability (Lin’s coefficient 0.963)⁽²⁵⁾. Subsequent studies confirmed the utility in the emergency department (reduced LOS, decreased ICU admissions)⁽²⁶⁾ and in different cultural contexts – the Hungarian validation (2025) proposes a cutoff of 6 for sensitivity/specificity optimization⁽¹²⁾, while the French validation is ongoing⁽²⁷⁾. Application of PAWSS in community triage in Vancouver recommended adjusting the cutoff to 7 for context-specificity⁽²⁸⁾.

5.2. BAWS – Brief Alcohol Withdrawal Scale

BAWS, developed by Rastegar et al. (Johns Hopkins, 2017), is the pragmatic response to the main criticism of CIWA-Ar: its length and cumbersome nature in the busy emergency setting⁽²⁹⁾. It contains five items, scored 0-3 (maximum 15 points):

• Tremor (on arm extension).
• Sweating (palms, forehead).
• Anxiety (subjective).
• Psychomotor agitation.
• Hallucinations/orientation.

Therapeutic threshold: a score ≥ 5 indicates the need to initiate BZD treatment^(29,30). Advantages include: rapidity (under 1 minute), predominant objectivity of items and superior correlation with CIWA-Ar (over 0.9). Retrospective studies (n = 799) demonstrated that the majority of patients monitored through the BAWS protocol received minimal/zero BZD doses, with low rates of severe withdrawal, agitation or oversedation⁽³⁰⁾. The main limitation: BAWS may underestimate the severity in patients with predominantly perceptual symptomatology without tremor/diaphoresis.

5.3. CIWA-Ar/Sellers Inventory

CIWA-Ar (Clinical Institute Withdrawal Assessment for Alcohol – Revised), referred to in European literature as the Sellers Inventory, after its main author (Edward M. Sellers, Toronto Addiction Research Foundation), represents the gold standard for the clinical evaluation of AWS⁽³¹⁾. The scale assesses 10 items: nausea/vomiting, tremor, paroxysmal sweating, anxiety, agitation, tactile, auditory and visual disturbances, headache, orientation/clouding⁽³²⁾.

Interpretation

• 0-8: mild withdrawal – monitoring, no obligatory pharmacotherapy.
• 9-15: moderate withdrawal – initiate BZD at a score ≥ 10.
• >15: severe withdrawal – increased risk of complications.

The Pribék et al. meta-analysis (2021), on 423 patients, confirmed that cumulative CIWA-Ar scores faithfully track AWS evolution and allow objective comparisons between BZD and non-BZD regimens⁽³³⁾. Major limitation: relying on its subjective component (over 60% of items require patient reporting), CIWA-Ar loses fidelity in confused, aphasic, intubated patients or in those with active psychosis⁽⁵⁾.

5.4. The Cushman Scale

Originally developed by Cushman et al. (1985), this 7-item scale concentrates on objective vegetative parameters: heart rate, systolic blood pressure (age-stratified), respiratory rate, tremor, sweating, agitation and sensorium⁽³⁴⁾. Each parameter is scored 0-3 points.

Interpretation

• < 7: mild withdrawal.
• 7–14: moderate withdrawal – requires treatment.
• > 14: severe withdrawal – major risk of complications.

Cushman is particularly useful in the emergency department, in patients with increased cardiovascular risk, or in francophone settings, where it remains a reference instrument. Unlike CIWA-Ar, it is not dependent on patient cooperation⁽³⁵⁾.

5.5. RASS – Richmond Agitation-Sedation Scale

RASS, developed by Sessler et al. at Virginia Commonwealth University (2002), is a 10-step scale (+4 to −5), where 0 represents the state of calm alertness⁽³⁶⁾. Initially designed for the ICU, RASS has become an essential complementary instrument in AWS management, with two key applications:

• Monitoring iatrogenic sedation – target: RASS 0 to −1. Reaching a RASS score ≤ −2 requires stopping/temporizing BZDs, regardless of the CIWA-Ar or Cushman score, to prevent respiratory depression^(5,37).
• Evaluating non-communicative patients – a patient with extreme agitation (RASS +3/+4) cannot be subjectively evaluated through CIWA-Ar, but can be safely titrated through RASS.

A Canadian study published in CJEM (2024) retrospectively compared (1073 patients; 21 months) the use of CIWA-Ar versus the modified mRASS-AW version for AWS management outside the ICU. Replacing CIWA-Ar with mRASS-AW did not worsen length of stay or the incidence of DT/ICU admissions; furthermore, post-admission complications were reduced in the mRASS-AW group (3.4% versus 6.6%; p=0.020)‌⁽³⁷⁾. This fact suggests that, in the near future, scales based on sedation/agitation could regain ground against complex classical scales.

6. Integrated clinical algorithm for scale utilization

Stage 1: At admission (PAWSS)

• PAWSS < 4 → standard monitoring.
• PAWSS ≥ 4 → aggressive prophylaxis (BZD front-loading, parenteral thiamine, monitoring every 1-2 hours).

Stage 2: In the ED (BAWS)

• BAWS < 5 → reassessment at 2 hours.
• BAWS ≥ 5 → immediate BZD initiation; transfer to a specialized ward.

Stage 3: On the ward (CIWA-Ar and/or Cushman)

• CIWA-Ar in the cooperative patient; Cushman in the noncooperative patient or those with cardiovascular risk.
• Hourly reassessment in the first 24 hours, then every 4-8 hours, depending on evolution.

Stage 4: Continuously (RASS)

• Target: 0 to −1.
• RASS ≤ −2 → STOP BZD, regardless of the severity score.

Autor corespondent:  Marinela-Minodora Manea E-mail: mmanea@umfcluj.ro

CONFLICT OF INTEREST: none declared.

FINANCIAL SUPPORT: none declared.

This work is permanently accessible online free of charge and published under the CC-BY.