During training to become a psychometry, I was blessed/cursed over 50 years ago, with a response to a research proposal to improve the devices that researchers used rapid use for survey of health results, but rarely psychotherapy was evaluated. Knowing very little about health, I found it very useful that its 1948 definition is “in the form of complete physical, mental and social welfare and not only the absence of disease or weakness.” There were three things especially useful: this is health Separate componentAre components Ahead Absence of the disease, and it is a complete Goodness,
Moving forward and looking back, it is appropriate to say that the results of the “Psychometric Age” of Health Survey Research launched in the US in the 1970s, and then worldwide, support the least physical and mental components of the definition of health. For psychomatic methodologists reading it, the trends of research over years also reveal the stability of materials, psychometric support for various methods to meet scaling beliefs, and the domain score for differences in the underlying components reveal evidence of accountability that causes them.
This first blog in a planned series tries to identify some lessons from pre -work. This is not just because it is important to avoid reducing the deficiencies of heritage, but lessons in the current context are highly relevant. Hopefully, these takes will be helpful in understanding the blogs to navigate between heritage and better population and patient result survey methods.
A quick dive in devices to survey health results
The figure below classifies the content of widely used health surveys and score using psychometric and/or utility methods. It contains the three most recent 6-domain promis profiles that differ in the number of items per domain and measure utility. The lines often surveyed health domains including physical, social and role, mental health (mainly psychological crisis), general health evaluation, pain and vitality (mainly fatigue). The pillars show widely used health survey (defined in footnotes of Figure) listed in the order of publication.
Here is not shown including various operational definitions Working (What people are able to do), Emotion (Both sick and welfare), and Evaluation (Excellent-poor). For example, the disease effect profile (SIP) is perfectly asking about “crying” and “not feeling like crying.” One of the reasons for measurement success is the use of all three approaches, especially to get more information from single items. Other important ideas (eg, roof effects) indicate only in their domains label There will be the subject of future blogs. In the context of clear emerging ideological consensus, it is notable that NIH-sponsored Promis, historically the most frequently studied domains, to represent five of the domains.
Survey of health results in two major studies
Two of the columns show the well-known 5-year-old health insurance experiment (HIE) and the health domains monitored in the study (MOS) of 4-year half-practical medical results. The need to determine population health results gave these major studies air to the structure (providers, organizations and finance) and the growing debate about the quality of medical care processes.
From HIE and MOS, Health Survey Designers and Functions learned valuable lessons about feasibility, preference and low administration costs Self -administered Survey. We also learned that the measures of baseline health survey are among the most legitimate predictions of health care costs, job loss and mortality.
Development in health domains and short form surveys
HIE certainly provided the most widespread general population result survey data Of your timeFrom this study, the classical factor analyzes correlations between 20 multi -itam parameters, which is enclosed to measure physical, mental and social functioning and rotate three major components well, which contains patterns of rotating factors loading that correspond to physical, mental and social components. However, the absence of important correlations between social measures (eg, work, job, boss) raised questions about whether such reported social/role differences, in-facts, health. Other HIe analyzes of social interaction frequencies raise questions about the results (eg, friends and family, for phone contacts, and they are also healthy, which they are healthy.
Recently, a medical care result in the editorial of Wall Street Journal is called “Gold Standard”, which also requires infection with long -term survey booklets (100 items per booklet), more efficient items. HIE surveys at the time rely on dicotomus physical and role disability objects in use despite their huge general population roof effects (eg, 70%+) for their vast general population roof effects (eg, 25-aItam physical functioning, or PF, scale). Small 10-Weitum Mos PF scale reduced the effect of roof to 30% and reduced the “small” hie-inffined 149-ketam baseline survey of only 40 domains. The 1976 Hanes survey was based on the 100+ hie mental and general crisis, and welfare domain items “items banks”. Over time, more practical MOS 4-Year Result Survey, 36 short form (SF-36) objects received psychologically sound estimates of eight domains.
Changing Games Search and Summary Health Measures Applications
During the MOS tests of construction validity, our research team discovered physical and mental summary components capturing 80% of 80% reliable variance in 8 domains in support of WHO. It was a “game changer”. For the first time, the team created two summary components, which were previously used only in verification to simplify the interpretation of public reporting and most important medical care results in health policy research and clinical trials. Furthermore, as low items were required to assess physical and mental health, it enabled the very low version of the device. We released a new version with just 1-2 items per health domain (ie, SF -12). Finally, social and role objects that have clearly behaved for the causes of functional limitations improved, which improved discrimination between physical and mental health results.
Power of responsibilities for specific health boundaries
In MOS, when asked whether various roles were boundaries (eg, low completion) due to physical or mental (emotional) health problems or both, their scores were convergence and the patterns of discriminatory validity were completely reversed in psychometric and clinical trials. For example, the multi-aitum roll functioning score for physical properties is excessive with the physical component of health (R = 0.76) and the lower with mental components (R = 0.24). Researchers visited the opposite pattern (R = 0.29 and R = 0.87) for physical and mental components respectively for physical and mental components, for the same role limitations, which for the limitations of the elite for emotional problems.
The chain will continue
This evidence is very important in the context of the ongoing debate of how to construct and score the summary measures used widely. Later a blog will discuss whether we must build physical and mental health summary measures to be correlated between score> 0.60 or <0.30, and what implications are to explain their results. In addition, the latest generation of MOS measurement model-based short forms that create this difference for the role and social domain will be the future blog.
A blog will also address another implication of the discovery of the power of a simple change in words, attributes the feelings of sickness being sick for functional boundaries or a specific disease (eg, asthma versus osteoarthritis). Such changes have also become a game changer to improve the usefulness of survey of health results. Specific diseaseDisease-specific properties can improve validity in response to clinically defined interactions in a specific situation, even for people with co-diligent conditions.
Dr. Ware Professor is Emeritus, Department of Public and Quantitative Health Sciences, Umas Chan Medical School. He studied psychology at the University of Peepardine and Psychometrics at the University of Southern Illinois. He is an internationally recognized and often quoted developer and patient reported results (PROS) and an elected member of the National Academy of Medicine. His remarkable activities in the last 50 years include: 5-year-old rand health insurance experiment leading the development of professionals; The 4-year-old half-professional medical results were the leading investigator for the study of the study, for which he developed the SF-36 health survey; The start of the Life Assessment Project and the leading international quality SF -36 and SF -12 translations for SF -12 clinical and population pro research are used worldwide; And establishing NIH-sponsored small businesses between the first to apply “modern” psychometric methods to make general and specific professionals more efficient. Their current task is the least possible (including single-utum per domain/disease) widespread generic and disease-specific measures and their standardization and integration monitors more practical and useful to screen the patient’s screening and results.
