Patient-reported outcome measures (PROMs) are becoming an integral part of healthcare decision making. Clinical trials, systematic reviews, and clinical practice guidelines incorporate them to learn about the effect of medical interventions in patients’ health status, without interference or mediation from clinicians or proxies. The use of these types of measures, however, is not without challenges. In particular, the complexity of the PROMs makes it difficult for patients, clinicians, and researchers to fully grasp the extent to which a treatment effect is negligible or trivial, small but important, moderate, or large. One of the most documented ways to address this issue is the use of the minimal important difference (MID), the smallest change in a PROM, either beneficial or harmful, that patients would perceive as important. A patient-oriented way to determine this threshold is the estimation of an anchor-based MID, where PROM results are compared against an external independent criterion the anchor that is in itself understandable and relevant for patients. This dissertation is an effort to facilitate the identification, evaluation, and utilization of MID estimates for PROMs. First, this thesis describes the development and reliability assessment of a new instrument to determine the credibility of primary studies ascertaining MID estimates, Second, it describes the conduct of a systematic survey to inform the creation of an inventory of all available anchor-based MIDs in the medical literature until 2015. Third, it reports an analysis of the state of the art of current MID estimates from a reporting and credibility perspective. Finally, this work concludes with a summary of the main results, presentation of strengths and limitations, and insights related to the implications for future research. / Dissertation / Doctor of Philosophy (PhD)

Das Ziel dieser Arbeit war die Bestimmung der kleinsten relevanten Veränderung des wahrgenommenen Mundgesundheitszustandes, der Minimal Important Difference (MID), des Oral Health Impact Profile (OHIP). Die MID sollte für die deutsche Version des OHIP mit 49 Fragen (OHIP-G49), die deutsche Kurzversion mit 14 Fragen (OHIP-G14) sowie für die einzelnen Dimensionen der deutschen und englischen Version des OHIP bestimmt werden. Es handelt sich um eine klinische Fallserie mit 224 konsekutiv rekrutierten, prothe-tischen Patienten. Die mundgesundheitsbezogene Lebensqualität wurde mittels des OHIP- G49 an zwei Terminen vor der Behandlung (Basisuntersuchungen) sowie vier und sechs Wochen nach Behandlungsende (Nachkontrolluntersuchun-gen) bestimmt. Zu den Nachkontrolluntersuchungen schätzten die Patienten zu-sätzlich die Veränderung ihres Mundgesundheitszustandes gegenüber dem Zeit-punkt vor der Therapie anhand einer globalen Frage ein. Anhand der Ergebnisse der Basis- und Nachkontrolluntersuchungen wurde der Median der Differenzen der OHIP-Summenwerte errechnet. Dieser Wert entspricht der MID. Für die deutsche Version des OHIP mit 49 Fragen wurde ein Wert von 6,0 OHIP-Punkten ermittelt. Der Wert für den kurzen Fragebogen OHIP-G14 betrug 2,0 OHIP-Punkte. Für die Dimensionen der deutschen und englischen Sprachversion des OHIP konnten nur teilweise Ergebnisse gefunden werden. Das Studienergebnis lässt den Schluss zu, dass für beide untersuchten Versionen des OHIP ein klar definierter, minimal relevanter Unterschied (MID) der Summen-werte existiert. Die MID unterstützt die Interpretation der klinischen Bedeutung von Veränderungen des vom Patienten wahrgenommenen Mundgesundheitszustan-des. Sie stellt eine wichtige Größe zur Bewertung prothetischer Therapieeffekte dar.

Das Ziel dieser Arbeit war die Bestimmung der kleinsten relevanten Veränderung des wahrgenommenen Mundgesundheitszustandes, der Minimal Important Difference (MID), des Oral Health Impact Profile (OHIP). Die MID sollte für die deutsche Version des OHIP mit 49 Fragen (OHIP-G49), die deutsche Kurzversion mit 14 Fragen (OHIP-G14) sowie für die einzelnen Dimensionen der deutschen und englischen Version des OHIP bestimmt werden. Es handelt sich um eine klinische Fallserie mit 224 konsekutiv rekrutierten, prothe-tischen Patienten. Die mundgesundheitsbezogene Lebensqualität wurde mittels des OHIP- G49 an zwei Terminen vor der Behandlung (Basisuntersuchungen) sowie vier und sechs Wochen nach Behandlungsende (Nachkontrolluntersuchun-gen) bestimmt. Zu den Nachkontrolluntersuchungen schätzten die Patienten zu-sätzlich die Veränderung ihres Mundgesundheitszustandes gegenüber dem Zeit-punkt vor der Therapie anhand einer globalen Frage ein. Anhand der Ergebnisse der Basis- und Nachkontrolluntersuchungen wurde der Median der Differenzen der OHIP-Summenwerte errechnet. Dieser Wert entspricht der MID. Für die deutsche Version des OHIP mit 49 Fragen wurde ein Wert von 6,0 OHIP-Punkten ermittelt. Der Wert für den kurzen Fragebogen OHIP-G14 betrug 2,0 OHIP-Punkte. Für die Dimensionen der deutschen und englischen Sprachversion des OHIP konnten nur teilweise Ergebnisse gefunden werden. Das Studienergebnis lässt den Schluss zu, dass für beide untersuchten Versionen des OHIP ein klar definierter, minimal relevanter Unterschied (MID) der Summen-werte existiert. Die MID unterstützt die Interpretation der klinischen Bedeutung von Veränderungen des vom Patienten wahrgenommenen Mundgesundheitszustan-des. Sie stellt eine wichtige Größe zur Bewertung prothetischer Therapieeffekte dar.

This thesis uses the methods of sychometrics and clinimetrics to develop and evaluate a Pediatric Ulcerative Colitis Activity Index (PUCAI). The initial phases of item generation and reduction were performed previously. This thesis comprises five main studies. Study one: the weighting and formatting of an initial draft PUCAI using a cohort of 157 children with ulcerative colitis, enrolled prospectively in five pediatric IBD centers. Study two: the validation of the final draft on a separate prospective cohort of 48 children undergoing complete colonoscopy. The PUCAI was highly correlated with physician global assessment (PGA) (r=0.91), Mayo score (r=0.95) and colonoscopic appearance (r=0.77). The PUCAI was able to differentiate the different categories of disease activity, and cutoff points were defined. Study three: Assessment of the responsiveness of the PUCAI. The index demonstrated excellent responsiveness on 75 children seen twice during the study period (effect size=1.9, standardized response mean=2.2, responsiveness statistics=2.6, correlation with PGA of change=0.84, and area under the ROC curve=0.97 95%CI 0.93- 0.99). Study four was aimed at evaluating the predictive validity of the PUCAI, on a retrospective cohort of 99 children with severe ulcerative colitis admitted for intravenous corticosteroid therapy. The PUCAI, calculated on the third and fifth day of therapy was highly predictive of therapy failure at discharge and one year post discharge (area under the ROC curve 0.84 (95%CI 0.76-0.92). Study five: a methodological study evaluating the preferred way to determine the minimal clinically important difference (MCID) of health-related outcome measures. This study was conducted using the PUCAI and three other well established instruments. It was concluded that the MCID should be determined primarily by the anchor-based approach using the ROC curve method on the entire cohort, supplemented by calculating the minimal detectable difference beyond statistical error using the standard error of measurement. Small, moderate and large MCID values could be presented based on the degree of expected relevant change. Together, these studies have contributed to the rigorous development and thorough evaluation of a novel, non-invasive tool for assessing disease activity in pediatric ulcerative colitis clinical studies and practice.

This thesis uses the methods of sychometrics and clinimetrics to develop and evaluate a Pediatric Ulcerative Colitis Activity Index (PUCAI). The initial phases of item generation and reduction were performed previously. This thesis comprises five main studies. Study one: the weighting and formatting of an initial draft PUCAI using a cohort of 157 children with ulcerative colitis, enrolled prospectively in five pediatric IBD centers. Study two: the validation of the final draft on a separate prospective cohort of 48 children undergoing complete colonoscopy. The PUCAI was highly correlated with physician global assessment (PGA) (r=0.91), Mayo score (r=0.95) and colonoscopic appearance (r=0.77). The PUCAI was able to differentiate the different categories of disease activity, and cutoff points were defined. Study three: Assessment of the responsiveness of the PUCAI. The index demonstrated excellent responsiveness on 75 children seen twice during the study period (effect size=1.9, standardized response mean=2.2, responsiveness statistics=2.6, correlation with PGA of change=0.84, and area under the ROC curve=0.97 95%CI 0.93- 0.99). Study four was aimed at evaluating the predictive validity of the PUCAI, on a retrospective cohort of 99 children with severe ulcerative colitis admitted for intravenous corticosteroid therapy. The PUCAI, calculated on the third and fifth day of therapy was highly predictive of therapy failure at discharge and one year post discharge (area under the ROC curve 0.84 (95%CI 0.76-0.92). Study five: a methodological study evaluating the preferred way to determine the minimal clinically important difference (MCID) of health-related outcome measures. This study was conducted using the PUCAI and three other well established instruments. It was concluded that the MCID should be determined primarily by the anchor-based approach using the ROC curve method on the entire cohort, supplemented by calculating the minimal detectable difference beyond statistical error using the standard error of measurement. Small, moderate and large MCID values could be presented based on the degree of expected relevant change. Together, these studies have contributed to the rigorous development and thorough evaluation of a novel, non-invasive tool for assessing disease activity in pediatric ulcerative colitis clinical studies and practice.

In deciding whether to use a particular treatment for conditions such as depression, arthritis, or heart disease, clinicians and patients must balance the benefits against the side effects and burden. To make this trade-off, they must understand the likely degree of benefit in patients’ symptoms and perceived wellbeing, best undertaken using patient-reported outcomes (PROs). PROs are measures of any aspect of a patients’ health status that are obtained from direct patient inquiry without interpretation by a clinician or anyone else. PRO measures (PROMs) are increasingly used in clinical trials and systematic reviews to evaluate health care interventions, and information obtained from PROMs can guide clinical decisions and inform shared-decision making. The use of PROMs, however, involves challenges, the most important of which is deciding if a particular treatment effect is trivial, small but important, moderate or large. One way to make this judgment is to consider the minimal important difference (MID), the smallest change in a PROM score that is important enough that patients would consider a change in treatment to achieve that benefit. The number of published studies providing anchor-based MIDs for PROMs has grown rapidly over the last three decades, and researchers have proposed several anchor-based methods to derive MID estimates, each with its own merits and limitations. This thesis begins with the development of a framework to determine the extent to which the design and conduct of studies measuring anchor-based MIDs are likely to have protected against misleading estimates. Subsequently, this thesis presents a comprehensive inventory of empirically estimated anchor-based MIDs and their associated credibility for all PROMs published in the medical literature. Further, this thesis highlights critical issues that key stakeholders should consider, and demonstrates how the use of credible MIDs may inform the development of a clinical practice guideline in which PROs were identified as critically important. Finally, this thesis concludes with insights to improve the methodological quality and transparency for researchers in the PRO and MID field. / Thesis / Doctor of Philosophy (PhD)

Dans le domaine de la santé, la mesure d'utilité permet d'exprimer la valeur que peut donner un individu à un état de santé, comprise généralement entre 0 (la mort) et 1 (meilleur état de santé envisageable). L'utilité permet de valoriser simultanément le résultat clinique et son impact sur l'état de santé du patient, et d'exprimer ainsi la qualité de vie liée à la santé. Nous nous sommes intéressés dans ce travail à l'approche indirecte d'estimation d'utilité, consistant à dériver une utilité à partir des échelles psychométriques de mesure de l'état de santé (questionnaires génériques). Cette approche permet de répéter la mesure de l'utilité dans le temps et ouvre des perspectives pour l'analyse du changement d'utilité. Les objectifs étaient de définir un seuil de changement pour l'utilité dérivée des instruments de mesure psychométrique de la qualité de vie, et de prendre en compte dans l'interprétation du seuil la présence des comorbidités ou d'un phénomène de response-shift. Afin de répondre à nos objectifs, nous avons étudié, d'une part, le changement d'utilité mesurée par le SF-6D dans une cohorte de patients atteints d'arthrose des membres inférieurs (cohorte KHOALA) et l'effet des comorbidités sur l'utilité et son changement. D'autre part, nous avons estimé l'utilité de donneur vivant de rein à partir des questionnaires EQ-5D et SF-6D avant et 3 mois après le prélèvement du rein. Un seuil de changement a été estimé et un phénomène de response-shift a été recherché. Les résultats ont montré l'effet délétère des comorbidités sur la mesure de l'utilité chez les patients atteints d'arthrose, et mis en évidence l'importance de la sévérité fonctionnelle de l'arthrose dans l'expression de l'utilité et son changement. Chez les donneurs vivants de rein nous avons mis en évidence une diminution significative de l'utilité à 3 mois, avec une proportion de plus de 30% de donneurs ayant atteint le seuil de détérioration significative. L'estimation du seuil de changement diffère selon la population étudiée, l'instrument de mesure de l'utilité, et la méthodologie appliquée pour définir ce seuil. En conclusion, l'évaluation du changement d'utilité doit tenir compte de l'existence des comorbidités ainsi que de l'adaptation du jugement de valeur que l'individu porte sur son état de santé lorsque celui-ci évolue. La complexité des méthodes psychométriques pour la mesure de l'utilité doit amener à être prudent sur leur prise en compte dans les modèles d'évaluation économique. L'identification d'un seuil pour l'interprétation du changement peut être un élément complémentaire utile dans la démarche d'analyse décisionnelle / In health economic, the concept of utility was used to describe a state of health in patients. The unit value of health state utility ranges between 0, representing "death" state, and 1 representing a "perfect" health state. It is possible to recover through the measurement of utility both the clinical outcome and their impact on the health of the patient, and thus express the quality of life related to health. In this work, we used the indirect approach for estimating utility that is derived from a generic utility scales (quality of life questionnaires). The objectives of this thesis were to define a threshold of change for the utility derived from psychometric instruments measuring quality of life, and to take into account for the change interpretation, the presence of comorbidities or a response-shift phenomenon. We used data from a cohort of patients with osteoarthritis of lower limbs (KHOALA cohort) for estimate the impact of comorbidities in utility measured by the SF-6D its change over 3 years. We are also interested in the particular case of living kidney donors. The utility was estimated by the EQ-5D and SF-6D before and 3 months after kidney donation. A threshold of significant deterioration has been estimated and the presence of a response-shift phenomenon was controlled. The results showed the deleterious effect of comorbidities on utility measurement for patients with lower limbs osteoarthritis, and highlighted the importance of the functional severity of osteoarthritis in the expression of utility and its change over time. Living kidney donors experienced a significant decrease in their utility score in the short term, and more than third of donors have reached the threshold of significant deterioration. The threshold of utility change varies depending on the population studied, the measurement instrument, and the methodology used to define the threshold. In conclusion, the assessment of utility change must take into account the existence of other diseases and the adaptation of the individual on his health state variation. The complexity of measuring individual preferences by psychometric instruments should lead to be careful about introducing them in the economic evaluation. The use of a threshold for the interpretation of utility change can be an additional element in the decision-making process