This article, titled “Clinimetric Criteria for Patient-Reported Outcome Measures,” was published in Psychotherapy and Psychosomatics in 2021 by Danilo Carrozzino and a large international group of co-authors. The paper serves as a standard review article that delves into the critical evaluation of patient-reported outcome measures (PROMs).
Purpose and Central Argument: The primary aim of the article is to highlight the significant limitations of the traditional psychometric model when applied to clinical challenges and to propose an alternative, more suitable framework: clinimetrics. PROMs are self-rated scales and indices designed to capture patients’ subjective experiences of symptoms, well-being, and treatment effects. Given the abundance of available PROMs, their validity and usefulness in clinical and research settings are crucial. The authors argue that published guidelines based solely on psychometric criteria often fall short due to their emphasis on homogeneity of components and inadequate attention to sensitivity to change. Psychometric theory is seen as having “stifled the field” and led to the routine use of scales that, despite wide acceptance, often perform poorly in clinical contexts.
Key Concepts and Distinctions: The paper thoroughly reviews and contrasts the characteristics related to reliability, sensitivity, validity, and clinical utility of assessment instruments from both clinimetric and psychometric perspectives.
- Psychometrics:
- Developed largely outside clinical fields, in educational and social areas.
- Seeks homogeneity of components and a unidimensional construct, meaning all items should measure the same single attribute and be highly correlated. Components that do not correlate well are eliminated.
- A crucial characteristic is internal reliability or consistency, typically measured by Cronbach’s alpha. All items are weighted equally.
- Consensus-based initiatives like PROMIS and COSMIN are strongly rooted in this tradition.
- Clinimetrics:
- Introduced by Alvan R. Feinstein as the science of clinical measurements, focusing on clinically based evaluation.
- Does not require homogeneity of components or unidimensionality. What matters is the clinical utility of the scale in assessing a wide range of complex clinical issues, such as symptom patterns, severity, duration, progression of illness, effects of comorbidities, and responses to treatment.
- Items are not assumed to be intercorrelated and should be multidimensional, providing distinct clinical information. They may also carry different clinical weights, recognizing that some symptoms are more impactful than others.
- Sensitivity is a crucial clinimetric property. This includes the ability to differentiate patients from healthy controls, discriminate between patient groups, reflect changes in therapeutic trials, and distinguish between active treatments and placebos.
- Emphasizes various types of validity relevant to clinical practice, including clinical, construct (where items provide distinctive information and belong to underlying syndromes), biological, predictive, and incremental validity. Incremental validity is particularly important, ensuring each item or instrument adds distinctive, non-redundant information.
- Prioritizes clinical utility, which encompasses sensibility (ease of use in clinical research and practice, facilitating clinician-patient interaction), format (preference for short scales with a limited number of items), and transferability (ability of the index to measure the same clinical dimension across different patient groups or over time).
Conclusion and Impact: The paper underscores the importance of the clinimetric methodology in selecting and developing appropriate PROMs. It provides specific Clinimetric Patient-Reported Outcome Measure (CLIPROM) criteria that can guide the creation of new indices and the validation of existing PROMs for clinical use. Importantly, the authors note that PROMs initially developed using psychometric criteria should not be dismissed, as they may still possess valuable clinimetric properties, such as the Hopkins Symptom Checklist and Kellner’s Symptom Questionnaire. The article ultimately advocates for a shift in focus from purely psychometric principles to clinimetric criteria to better address the complex realities of clinical assessment and research.
Reference: Carrozzino, D., Patierno, C., Guidi, J., Berrocal Montiel, C., Cao, J., Charlson, M. E., Christensen, K. S., Concato, J., De las Cuevas, C., de Leon, J., Eöry, A., Fleck, M. P., Furukawa, T. A., Horwitz, R. I., Nierenberg, A. A., Rafanelli, C., Wang, H., Wise, T. N., Sonino, N., & Fava, G. A. (2021). Clinimetric Criteria for Patient-Reported Outcome Measures. Psychotherapy and Psychosomatics, 90(4), 222–232. https://doi.org/10.1159/000516599
