Problem Drinking: Construct & Measurement

The study begins by highlighting the significant challenge of defining 'problem drinking' precisely. While terms like "alcoholic" and "problem drinker" are commonly used by health professionals and in the alcoholism literature, there's a notable lack of a consistent operational definition. This ambiguity makes comparing prevalence estimates from different studies difficult, as each study might use a different frame of reference and population. The example of the 1971 estimate of over nine million alcoholics in the United States, adopted by the National Institute on Alcohol Abuse and Alcoholism, illustrates this issue. The lack of a standard definition undermines the comparability and meaningfulness of research findings on the prevalence of problem drinking. The research aims to address this by clarifying the terms used in discussions about alcohol consumption and related problems. This is crucial for effective research and the development of consistent methods to measure problem drinking.

A key distinction is drawn between the terms 'problem drinking' and 'alcoholism.' The author argues that 'problem drinking' is a broader term encompassing a wider range of drinking behaviors. Unlike 'alcoholism,' which often carries strong negative connotations associated with medical diagnoses or Alcoholics Anonymous, 'problem drinking' aims to be less emotionally charged. The term 'alcoholism' frequently suggests a severe, 'rock-bottom' condition, potentially hindering early intervention. Using the term 'problem drinking' allows for a clearer distinction between individuals with and without alcohol-related problems. The study proposes that problem drinking includes those who are alcoholic, highlighting the congruent relationship between being alcoholic and being alcohol-dependent (Edwards and Gross, 1976). This study will use these distinctions as a foundation for further analysis, and will utilize the term ‘problem drinking’ unless specific existing literature utilizes the term ‘alcoholic’.

The absence of a formal theoretical framework for understanding problem drinking is identified as a major gap in the existing literature. While numerous terms and ideas exist, their interrelationships lack a cohesive structure. The research aims to propose a theoretical framework that complements existing knowledge and provides a way to understand the connections between different terms associated with problem drinking. The proposed theory will not focus on the intrinsic aspects of alcoholism, such as its etiology, epidemiology, or treatment. Instead, the goal is to provide a framework for measuring problem drinking. The need to create a measuring system that considers the existing terminology related to problem drinking is the impetus for this new theoretical framework. The researchers propose a framework that accounts for existing definitions without replacing them, thereby avoiding further confusion in this area.

For the purposes of this thesis, the study proposes a categorization of problem drinkers into two main groups: dependent (alcoholics) and non-dependent. The author acknowledges the difficulties inherent in using the term 'alcoholic' due to its negative connotations and ambiguity, but its use is maintained when directly referencing existing literature to preserve the original author's intent. This strategy acknowledges the existing terminology whilst still providing a clearer distinction in the current research. This distinction is crucial for understanding the range of drinking behaviors and facilitating more accurate assessments. The study uses the term 'alcohol dependent' in place of 'alcoholic' to avoid these issues whenever possible, creating a clearer understanding of the issues involved in problem drinking.

The document provides a critical review of several established alcohol screening tests, examining their strengths and weaknesses in identifying problem drinking and alcohol dependence. The Michigan Alcoholism Screening Test (MAST) and its shorter version, the SMAST, are discussed, noting the MAST's use in conjunction with arrest records and the SMAST's aim to streamline the screening process while maintaining accuracy. The ALCADD Test, another commonly used clinical tool, is mentioned, along with its infrequent use in research. The Drug Abuse Screening Test (DAST) is reviewed, highlighting its generic approach to drug abuse and lack of specificity regarding the consequences of individual drug types. The review underscores inconsistencies in the reliability and validity of existing tests, revealing a need for more refined and reliable methods of assessment. The critical review aims to pave the way for a more effective and accurate assessment of alcohol-related problems, and highlights some of the weaknesses present in commonly used testing strategies.

A significant portion of the review focuses on psychological measures, particularly the Minnesota Multiphasic Personality Inventory (MMPI) scales. The text presents contrasting views on the effectiveness of these scales in differentiating alcoholics from non-alcoholic populations. Some researchers question the existence of unique personality traits among alcoholics and the suitability of the MMPI for populations other than those for which it was originally designed. Conversely, other studies report consistent findings indicating common trait clusters among alcoholics as identified by the MMPI. This conflicting evidence highlights the complexities of using psychological measures to identify and characterize alcohol-related problems, and suggests a need for further investigation into the validity and reliability of the MMPI, and for clarifying the types of populations for which these tools are most effective.

The document acknowledges the widespread reliance on patients' self-reports in alcohol assessments, given that many aspects of problem drinking are only observable by the individual themselves. However, the validity of self-reported data is highlighted as a frequently neglected area in alcohol research. Researchers often avoid the rigorous validation process due to time and cost constraints, leading to an acceptance of potential inaccuracies. Despite this, the text cites numerous empirical studies suggesting that self-reports are, in fact, often valid, with little evidence of systematic bias. These studies reveal disagreements in both directions, pointing to the presence of random error that can affect the correlation between the self-reported data and other measures. This discussion underscores the importance of careful consideration of the validity of self-reported data when using these methods for alcohol assessment.

The section begins by acknowledging that while detecting alcohol in bodily fluids is the most direct indicator of alcohol consumption, this method has limitations for assessing alcohol abuse or problem drinking. The short half-life of alcohol in the body means detection is only useful for a limited time, and its presence alone doesn't predict drinking habits or tolerance levels. Therefore, direct detection isn't sufficient for comprehensive assessment of alcohol-related problems. This leads to the need for indirect indicators, such as those discussed below, which, while less direct, can potentially provide broader insights into long-term alcohol consumption patterns and related issues. The need for alternative assessment methods is a key driver for research in this area.

The review then examines the use of several liver function tests as indirect indicators of alcohol abuse. Gamma-glutamyl transferase (GGT), aspartate aminotransferase (AST), and mean corpuscular volume (MCV) are discussed. While elevated GGT levels might suggest potential alcohol problems, the test's lack of specificity is noted. AST's limitations in discriminating between high and low alcohol consumers are also highlighted, citing studies with conflicting results. MCV, which shows an increase in chronic alcoholics, and also correlates with high alcohol consumption, is also reviewed. Although some correlations exist, it is noted that these biochemical markers are not accurate indicators of alcohol consumption, often having very low correlations with drinking habits. The discussion highlights the need for caution in interpreting these tests as conclusive evidence of alcohol problems in isolation. They can be useful in conjunction with other assessment methods but are less reliable on their own.

The potential for improving diagnostic accuracy by combining different biomedical measures is explored. The study cites research suggesting that combining GGT and MCV can help identify a substantial proportion of heavy drinkers. Another study employed a multivariate approach using MCV, GGT, and AST, which achieved high classification accuracy in distinguishing alcoholics from non-alcoholics. This supports the idea that a combined approach using several biomarkers may improve the reliability and accuracy of diagnosis. However, it is still limited by the non-specificity and sensitivity of individual tests and the lack of a robust theoretical framework. The summary concludes that while these biomedical tests have not proven to be individually sufficient for diagnosis, they hold potential when used as part of a multimodal assessment, particularly given the limited accuracy of the psychological tests.

The development of the Problem Drinking Severity Test (PDST) is a central focus. The methodology for selecting test items is described as simplistic but empirically driven. A binary scoring system (0 or 1 for each item, representing 'no' or 'yes') is employed. The justification for this approach, from a construct measurement perspective, is discussed; each item is viewed as a subconstruct of the broader problem drinking construct. A mathematical explanation for combining these scores into an overall problem drinking score is mentioned, although practical limitations are acknowledged. The choice of items reflects an emphasis on self-reported awareness of alcohol-related problems, guilt, and a desire to reduce or cease drinking. These aspects represent a key focus for the test's design and the overall framework used for its development and validation.

The process of determining cut-off scores for the PDST to indicate significant alcohol problems is detailed. A comparative analysis with the Short MAST (SMAST) is conducted, establishing a correspondence between PDST scores and SMAST scores to define thresholds. The approach demonstrates a reliance on established metrics while introducing a novel measure that offers a more nuanced understanding of alcohol-related issues. The cut-off scores are set to represent the level of severity that may indicate the need for further intervention, and aims to distinguish more effectively between various levels of problem drinking. The approach is designed to minimize misclassification and ensure that the test is used appropriately within a clinical context.

The section presents empirical evidence supporting the PDST's validity. While the methods used to select items and develop the scoring system are described as simplistic, the results suggest improvement over the Short MAST in measuring problem drinking. The PDST demonstrates a high percentage of internally perceived events, indicating patients' self-awareness of their alcohol problems and potential motivation for help. A strong first factor in the analysis relates to self-identified awareness of problems, guilt, and the desire to reduce or stop drinking. This is presented as evidence supporting the PDST as an improvement on current methodologies for assessment, and the design shows a focus on self-reported behaviours that can be easily measured.

Experiment I involved a large-scale data collection effort at Christchurch Hospital in New Zealand. Data was gathered from patient ward notes, supplemented by interviews. A standardized form was used to collect information from the medical records. However, inconsistencies in medical staff's record-keeping presented a challenge, requiring repeated visits to gather complete data. A team of six research assistants, all with university degrees (including two master's degrees), collected the data after an intensive two-week orientation program. The study collected data on a significant number of patients, utilizing both medical records and interviews, in order to assess the prevalence of problem drinking in a general hospital setting in Christchurch. The research highlights the difficulties in using data from medical records, including incomplete data and challenges in consistent reporting.

Experiment I's data analysis involved a linear stepwise regression to examine the relationships between self-report items and the criterion score for problem drinking. All questions were included in the analysis, even those not statistically significant at the 1% level. Collectively, these self-report questions explained a significant portion of the total variance in problem drinking scores. The analysis also examined inter-correlations between items, identifying strong relationships between certain self-reported behaviors, indicating the presence of a cohesive underlying construct. This part of the research provides information about the correlation between variables and confirms the importance of considering self-reported data in future research. The research provides some initial validation for the problem drinking construct and informs the design of future studies.

Experiment II employed a different methodology, focusing on interviews with patients during the first week of their treatment program. Three of the research assistants from Experiment I conducted these interviews. The questionnaire included the Short MAST (SMAST), the newly developed PDST, and other demographic and medical information related to alcohol abuse. The data collection for this experiment utilized a smaller sample size in comparison to the previous experiment, but focused on patients who were receiving treatment for alcohol abuse. The use of standardized questionnaires such as the SMAST and the newly developed PDST allow for greater comparability and more accurate assessment of the effectiveness of the newer test in assessing alcohol related issues. This is an important aspect of ensuring that the research is methodologically sound.