CAGE questionnaire
The CAGE questionnaire is a brief, four-question screening tool designed to identify individuals at risk for alcohol use disorder by assessing feelings of guilt, social criticism, attempts to reduce consumption, and morning drinking to alleviate withdrawal symptoms. Developed by John A. Ewing and Beatrice A. Rouse at the University of North Carolina in the late 1960s and first presented in 1970, it derives its name from the initial letters of its questions: Cut down (Have you ever felt you should cut down on your drinking?), Annoyed (Have people annoyed you by criticizing your drinking?), Guilty (Have you ever felt bad or guilty about your drinking?), and Eye-opener (Have you ever had a drink first thing in the morning to steady your nerves or get rid of a hangover?).[1][2] Each affirmative response is scored as 1 point, with a total score of 2 or more indicating a high likelihood of alcohol dependence, while a score of 4 is considered virtually diagnostic; however, the tool is intended as an initial screen rather than a definitive diagnosis, as it does not quantify drinking patterns or frequency.[1][3] The questionnaire has been validated in various clinical settings, including primary care, psychiatric inpatient units, and general medical populations, demonstrating good sensitivity (around 80-90%) for detecting severe alcoholism but lower specificity for milder forms of misuse.[4][5] Widely adopted since its 1984 publication in JAMA, the CAGE has influenced global alcohol screening practices and inspired adaptations like the CAGE-AID, which extends the questions to include drug use by rephrasing them to address both substances.[1][6] Despite its simplicity and historical utility—used by over half of screening primary care physicians in earlier surveys—it is now considered outdated by major health authorities, such as the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and the U.S. Preventive Services Task Force (USPSTF), due to its inability to detect at-risk drinking in women, lighter drinkers, or those without overt consequences, prompting recommendations for more sensitive tools like the AUDIT or single-question screens.[2][7][8]Overview and Purpose
Definition and Components
The CAGE questionnaire is a concise screening instrument comprising four yes/no questions aimed at detecting potential alcohol use disorder among adults. Developed by John Ewing, it serves as a quick self-report or clinician-administered tool to flag individuals who may require further assessment for alcohol misuse or dependence.[9] The acronym CAGE derives from its core components: Cut down (referring to attempts to reduce drinking), Annoyed (indicating irritation at criticism of drinking), Guilty (encompassing feelings of remorse about drinking), and Eye-opener (denoting drinking to relieve withdrawal symptoms or hangovers). These elements target lifetime experiences with alcohol to identify patterns suggestive of problematic use.[9][10] In format, the questionnaire is straightforward, requiring only binary responses and typically completed in under five minutes, making it suitable for efficient integration into clinical workflows. It functions within broader substance use evaluations in settings like primary care, mental health services, and emergency departments.[10][11]Historical Development
The CAGE questionnaire was developed in 1968 by Dr. John A. Ewing, a psychiatrist and founding director of the Bowles Center for Alcohol Studies at the University of North Carolina at Chapel Hill, in collaboration with Barry A. Rouse.[1][12] Created at North Carolina Memorial Hospital, the tool emerged from clinical observations to address the limited availability of brief screening methods for identifying problem drinking amid rising concerns over alcoholism in medical populations.[1][13] Although devised in 1968, the questionnaire was first presented verbally by Ewing and Rouse at the 29th International Congress on Alcohol and Drug Dependence in Sydney, Australia, in February 1970.[14] Its formal publication occurred in 1984 in the Journal of the American Medical Association (JAMA), where Ewing detailed its clinical utility based on accumulated experience.[1] During the 1970s and 1980s, the CAGE gained early adoption in primary care and hospital settings for routine alcohol screening, reflecting broader public health efforts to recognize and treat alcoholism as a treatable condition rather than a moral failing.[1][13] By the time of its JAMA publication, at least 17 research reports had already employed the tool, demonstrating its growing acceptance among clinicians and researchers.[1] In the early 2000s, the CAGE questionnaire became integrated into international health frameworks, including references in World Health Organization (WHO) documents on substance use screening and management.[15] By the early 2000s, adaptations for cross-cultural and multilingual use facilitated its global dissemination, with versions applied in diverse regions such as Europe, Africa, and Asia to support alcohol misuse detection in varied healthcare contexts.[16] Since its 1984 publication, the core structure of the CAGE has undergone no major revisions, though its application has seen refinements for specific populations and integrated screening protocols.[1][4]Content and Administration
The CAGE questionnaire is typically administered verbally by a healthcare professional during a clinical interview but can also be self-administered as a written screening tool.[17][18] It consists of four yes-or-no questions, each phrased with "have you ever" to capture lifetime experiences with alcohol rather than focusing solely on recent or current behavior. This retrospective approach allows the tool to detect enduring patterns of problematic drinking that might otherwise go unnoticed in short-term assessments. The questions emphasize behavioral and emotional indicators of alcohol misuse, with responses providing insights into potential issues without quantifying consumption levels.The Four Questions
The first question—"Have you ever felt you should cut down on your drinking?"—assesses control issues by determining if the individual has perceived a need to limit their alcohol intake. A yes response may indicate recognition of excessive or uncontrolled drinking, such as repeated instances where consumption interferes with daily responsibilities or leads to unintended escalation.[1][7] The second question—"Have people annoyed you by criticizing your drinking?"—probes social consequences, exploring whether alcohol use has provoked disapproval or conflict from others. An affirmative answer could highlight patterns like arguments with family or colleagues over drinking habits, signaling broader relational strain caused by alcohol.[1][7] The third question—"Have you ever felt bad or guilty about your drinking?"—examines self-perception, targeting feelings of remorse or shame associated with alcohol consumption. Positive responses might reflect internal turmoil, such as regret following episodes of intoxication that result in poor decisions or emotional distress.[1][7] The fourth question—"Have you ever had a drink first thing in the morning to steady your nerves or to get rid of a hangover (eye-opener)?"—indicates dependence by identifying the use of alcohol to manage withdrawal symptoms or hangovers. A yes reply may suggest a cycle of morning drinking to alleviate anxiety or physical discomfort from prior use, pointing to physiological reliance on alcohol.[1][7]Scoring and Interpretation
The CAGE questionnaire employs a simple scoring system where each of the four questions receives 0 points for a "no" response and 1 point for an affirmative "yes" response.[19][20] The resulting total score ranges from 0 to 4, reflecting the number of positive responses.[1] The total score is computed as follows: Total = (affirmative response to "Cut down") + (affirmative response to "Annoyed") + (affirmative response to "Guilty") + (affirmative response to "Eye-opener"), where each affirmative response equals 1 point.[1] For interpretation, a score of 1 indicates possible alcohol-related issues and suggests the need for further inquiry, while a score of 2 or higher signals a high likelihood of alcohol use disorder and recommends a comprehensive assessment.[21][22] Key considerations in interpretation include the questionnaire's focus on lifetime experiences, which does not distinguish between past and current alcohol use and may yield positives based on historical rather than ongoing problems.[22][23] Clinically, the CAGE serves as a screening tool and is not diagnostic; affirmative results necessitate follow-up with detailed evaluation, including application of DSM-5 criteria for alcohol use disorder.[22][1]Psychometric Properties
Reliability
The reliability of the CAGE questionnaire refers to the consistency and stability of its measurements across repeated administrations, raters, and populations, which is essential for its utility as a screening tool for alcohol misuse. As a brief four-item instrument, the CAGE demonstrates moderate to high internal consistency, with Cronbach's alpha coefficients typically ranging from 0.65 to 0.95 in various studies. For instance, a reliability generalization meta-analysis of 22 samples reported a median alpha of 0.74, indicating acceptable homogeneity for a short scale. In emergency department settings across Argentina, Mexico, and the United States, KR-20 estimates (equivalent to alpha for dichotomous items) were 0.78, 0.72, and 0.82, respectively, supporting its reliability in diverse clinical contexts.[24][25] Test-retest reliability is also robust, particularly over short intervals such as 1-2 weeks, with coefficients generally falling between 0.80 and 0.95. A comprehensive review of studies confirmed high stability in responses among medical and psychiatric inpatients, attributing this to the questionnaire's straightforward yes/no format. In adolescent medical patients, 1-week intraclass correlation coefficients for the adapted CAGE reached 0.82-0.90, demonstrating reproducibility even in younger populations. Kappa values in similar short-term retests have been reported around 0.60-0.80, reflecting stable individual responses despite minor variations.[4][26] Inter-rater reliability remains high when the CAGE is administered by clinicians versus self-report, with agreement exceeding 0.80 in comparative studies. This consistency arises from the objective scoring of binary responses, minimizing subjective interpretation. However, data on inter-rater metrics are less abundant than for other reliability aspects, often embedded within broader validity assessments.[4] Several factors influence the CAGE's reliability, with performance generally stronger among heavy drinkers and clinical samples, where alpha values and test-retest coefficients are higher due to clearer symptom endorsement. In contrast, reliability tends to be lower in general or non-clinical populations, such as college students, prenatal women, or older outpatients, potentially due to subtler drinking patterns or response biases. A review highlighted inconsistent results in primary care settings and among white women, underscoring the need for context-specific application. Key studies from the 1980s and onward, including meta-analyses of U.S. samples, have established these patterns through systematic evaluations in inpatient and ambulatory groups, though long-term retest data beyond six months remain limited.[4][24]Validity
The CAGE questionnaire demonstrates strong construct validity through its correlations with established diagnostic criteria for alcohol dependence, such as those in the DSM, with reported correlation coefficients ranging from 0.48 to 0.70 when validated against structured interviews like the Diagnostic Interview Schedule (DIS) or Composite International Diagnostic Interview (CIDI).[27] These associations indicate that CAGE scores effectively capture underlying aspects of alcohol use disorders (AUD) as defined by clinical standards.[28] In terms of criterion validity, the CAGE exhibits sensitivity of 68-87% and specificity of 77-95% when benchmarked against gold-standard tools like the AUDIT or DIS, particularly among heavy drinkers in clinical settings such as primary care and hospitals.[27] For instance, a meta-analysis of studies using DSM criteria as the reference found pooled sensitivity of 71% in primary care populations and 87% in inpatients, with an area under the curve (AUC) of 0.87, underscoring its diagnostic accuracy for detecting AUD.[28] Predictive validity is supported by evidence that CAGE scores of 2 or higher forecast increased risk of alcohol-related adverse outcomes, including hospitalizations for gastrointestinal conditions, with affected individuals facing approximately 3 times the odds compared to those with lower scores in longitudinal follow-up studies.[29] This predictive power holds across various patient groups, though it is stronger for severe cases.[30] Content validity is evident in the questionnaire's alignment with core features of alcohol dependence, including loss of control (e.g., the "cut down" item), salience of drinking (e.g., "annoyed" by criticism and "eye-opener" use), and elements related to withdrawal or remorse (e.g., "guilty" feelings).[27] Key studies from the 1990s, including reviews by Barry (1990) and Buchsbaum (1991), confirmed the CAGE's utility as a brief screening tool in primary care settings for identifying alcohol abuse and dependence.[27]Usage and Limitations
Clinical Applications
The CAGE questionnaire is routinely employed in primary care settings to screen at-risk adults over 18 for alcohol use issues during annual check-ups or acute visits, facilitating early identification of potential alcohol use disorders.[31] In emergency departments, it is integrated into protocols such as Project ASSERT to assess patients presenting with trauma or intoxication, where alcohol involvement is common, enabling rapid triage for substance-related care.[31] Mental health clinics also utilize the CAGE for psychiatric evaluations, particularly among individuals with co-occurring conditions, to detect alcohol misuse that may complicate treatment.[31] Administration of the CAGE can occur via self-administered forms, allowing patients to complete it independently in waiting areas or online portals, or through verbal interviews conducted by clinicians during consultations for a more interactive assessment.[32] It is increasingly incorporated into electronic health records (EHRs) as automated prompts or embedded modules, streamlining data collection and enabling seamless documentation within digital workflows.[21] The questionnaire is often embedded within the Screening, Brief Intervention, and Referral to Treatment (SBIRT) model, where a positive screen—typically two or more affirmative responses—prompts immediate brief advice from providers or referral to specialized services, enhancing intervention efficiency in busy clinical environments.[33] This integration supports proactive management in primary care by linking screening results directly to actionable steps, such as motivational interviewing or specialist consultations.[31] Target populations include general adults seeking routine healthcare, preoperative patients to evaluate perioperative risks associated with alcohol use, and those with comorbidities like hypertension, where alcohol screening informs holistic management plans.[31] Globally, the CAGE has been translated and validated in several countries, including Korea, Brazil, Taiwan, and France, supporting its use in diverse cultural contexts from Asia to Europe.[34] Training for the CAGE is minimal due to its brevity and simplicity, requiring only basic instruction on question delivery and scoring for clinicians, though established follow-up protocols are essential to ensure appropriate responses to positive screens.[33]Key Limitations
The CAGE questionnaire exhibits detection bias by primarily identifying severe or long-term alcohol dependence rather than low-risk, binge, or hazardous drinking patterns. Its lifetime timeframe often fails to capture recent changes in consumption, leading to missed opportunities for early intervention. For instance, sensitivity for mild alcohol use disorder (AUD) ranges from 40% to 60%, significantly lower than for severe cases where it reaches 80-90%. [4] [35] [36] Demographic factors further limit the tool's applicability, with reduced validity observed in women, younger adults, non-white populations, and non-English speakers. In women, particularly white females, sensitivity drops to 38-50% due to differing drinking norms and lower endorsement rates of guilt or annoyance related to alcohol use. Younger adults show inconsistent reliability, as the questions may not align with episodic or social drinking common in this group. Among non-white groups, such as Mexican-Americans, performance varies, often with lower sensitivity (around 21% in some subgroups) attributed to cultural differences in alcohol-related stigma and reporting. For non-English speakers, cultural adaptations are necessary, as direct translations can alter item interpretation and reduce accuracy in diverse linguistic contexts. [35] [37] [38] [39] [34] [40] As a screening instrument, the CAGE cannot diagnose AUD and requires follow-up clinical assessment for confirmation. It relies heavily on self-reporting, which is susceptible to denial, underreporting, and social desirability bias, where individuals minimize problematic behaviors to align with societal expectations. This bias is particularly pronounced in non-anonymous settings, potentially leading to false negatives. [41] [17] [42] [43] [30] The questionnaire's scope is restricted to alcohol use, excluding screening for other substances and rendering it outdated in contexts of polysubstance use, which is increasingly prevalent. While adaptations like the CAGE-AID exist for drugs, the original tool does not address combined alcohol and substance issues, limiting its utility in comprehensive assessments. [44] [6] [4]Alternative Screening Tools
The Alcohol Use Disorders Identification Test (AUDIT), developed by the World Health Organization, is a 10-question screening tool that assesses alcohol consumption patterns, including quantity and frequency of intake, signs of dependence, and alcohol-related harm.[45] Unlike the CAGE's focus on lifetime problems, the AUDIT detects hazardous drinking earlier, with a score of 8 or higher indicating hazardous or harmful use requiring intervention.[45] It demonstrates sensitivity of 73-92% and specificity of 77-94% across various cutoffs, making it more comprehensive for identifying at-risk individuals in primary care settings.[46] The Michigan Alcoholism Screening Test (MAST) consists of 25 yes/no questions that evaluate lifetime alcohol-related consequences, such as social, legal, and health impacts, providing a broader assessment of dependence than the CAGE's brevity allows.[47] Scores of 4 or higher suggest possible alcoholism, with a threshold of 5 or more indicating probable dependence; it achieves sensitivity up to 98% and specificity of 95% for detecting alcohol use disorders.[47] This longer tool is particularly suited for comprehensive evaluations in clinical populations where historical patterns are relevant.[46] The CRAFFT is a 6-question screening instrument adapted from the CAGE framework but expanded to address alcohol, cannabis, and other drug use among adolescents aged 12-21.[48] It targets high-risk behaviors, including riding or driving under the influence, with two or more positive responses signaling the need for further assessment; it shows sensitivity of 76-80% and specificity of 86-94% for identifying substance use problems or disorders in youth clinic populations.[49] Developed by researchers at Boston Children's Hospital, the CRAFFT outperforms the CAGE in adolescent settings due to its inclusion of multi-substance risks.[48]| Tool | Length | Sensitivity/Specificity (Representative) | Best Uses |
|---|---|---|---|
| CAGE | 4 questions | 67-78% / 82% (adults) | Quick adult screens for lifetime dependence |
| AUDIT | 10 questions | 73-92% / 77-94% | Detailed assessment of hazardous drinking in adults |
| MAST | 25 questions | 90-98% / 95% | Comprehensive evaluation of lifetime consequences in adults |
| CRAFFT | 6 questions | 76-80% / 86-94% (adolescents) | Youth screening for alcohol and multi-substance risks |