Measuring Specific Mental Illness Diagnoses with Functional Impairment Workshop Summary (2016) / Chapter Skim
Currently Skimming:
4 Data Collection Approaches
4 Data Collection Approaches
Pages 29-58
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 29... ...
study, focusing on estimates of mental and substance use disorders. He described the GBD as a systematic scientific effort to quantify the comparative magnitude of health loss due to diseases, injuries, and risk factors by age, sex, and geographic areas for specific points in time.
|
From page 30... ...
. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: A systematic analysis for the Global Burden of Disease study 2013.
|
From page 31... ...
Another major task involves attributing diseases to underlying risk factors or combinations of risk factors. Producing the disability estimates involves deriving the disability weights and combining them with severity distributions, because many of the major disabling conditions have a very wide range of severity.
|
From page 32... ...
NOTE: Global Burden of Disease data and the analysis and visualization tools demonstrated during the workshop are available 32 at http://www.healthdata.org/gbd/data [December 2015]
|
From page 33... ...
In the GBD 2010, information was collected from more than 30,000 respondents through face-to-face interview surveys in Bangladesh, Indonesia, Peru, and Tanzania; a telephone survey in the United States; and an open-access web survey.2 The survey included 108 of the then 220 health states, but all of the health states were included in the web survey. For the GBD 2013, information came from four European surveys.
|
From page 34... ...
Vos noted that the disability weights for the different health states have to be combined with epidemiological data to yield distributions across the different sequelae. Table 4-1 shows the most recent top 20 ranking of disabling conditions globally, with varying levels of severity.
|
From page 35... ...
There were similar problems in obtaining comparable data for severity in the case of several of the other disabling conditions, such as chronic obstructive pulmonary disease, asthma, osteoarthritis, back pain, and neck pain. Because of this, the GBD research team turned to three surveys with data available at the individual level, along with rich diagnostic information on mental disorders and physical disorders and a general health status measure.
|
From page 36... ...
Vos stated that the team has learned that not only is it important to take comorbidity into account in the overall estimates of disability, but also that when severity is analyzed, it is necessary to tease out what is contributed to disability by comorbidity and what is contributed by an individual disorder. An advantage of this method is that it can be used consistently across a range of prevalent disabling outcomes, ranging from back pain to mental disorders.
|
From page 37... ...
These datasets are not assembled with the idea that they are going to provide insights into the prevalence of mental disorders, but it is possible that researchers can obtain some information from them. Although others have discussed population surveys that collect data about mental disorders, Olfson concentrated on other surveys that contain some information about behavioral health, general health, and health care.
|
From page 38... ...
As pointed out by Regier, only about one-half of the people who meet the criteria for mental disorder on the basis of a structured interview will receive some mental health services in the course of a year, either from the specialized mental health sector or from the general medical sector. When considering the mental disorders of people who are missing from these claims databases, it is important to recognize that they are not missing at random.
|
From page 39... ...
Olfson reiterated that it is important to be cautious about assuming that diagnoses appearing in claims databases are the same as the diagnoses obtained from administering a structured psychiatric interview the way it is done in a specialized community epidemiological survey. Olfson said that there are a wide variety of different types of administrative databases that include both public payers (e.g., Medicaid, Medicare, Veterans Health Administration/Tricare)
|
From page 40... ...
Olfson also reiterated that administrative claims databases have the strengths of being able to provide information about the diagnoses and treatment patterns of some difficult-to-survey populations and, for some of them, of being quite large. Data from Population Surveys Olfson then turned to several ongoing, federally funded general health surveys that can be a source of mental health data.
|
From page 41... ...
The Behavioral Risk Factor Surveillance System Box 4-1 provides an overview of another health population survey, the Behavioral Risk Factor Surveillance System (BRFSS) , which Olfson pointed out, has the distinct benefit of providing state-level estimates.
|
From page 42... ...
The structure of the survey allows researchers to combine several years of data to accrue a larger sample and derive more stable estimates. The Medical Expenditure Panel Survey Olfson next described the MEPS, noting that Vos had also discussed it as one of the datasets used in the Global Burden of Disease study (see above)
|
From page 43... ...
Because of the gradual decline in the response rates, there are questions about whether results are as representative as they were in years past. Olfson reiterated that he was only covering basic information about the population surveys he discussed, but as potential sources of existing data, they have several strengths: they are representative of the general population; unlike the administrative data sources, they yield information on untreated individuals; they are typically administered on an annual basis; and they can be analyzed cross-sectionally and over time.
|
From page 44... ...
• Household informant, except for SF-12, PHQ-2, and K6 • No systematic mental health status information SOURCE: Workshop presentation by Mark Olfson, September 2015. Practice-Based Surveys and Data In the final part of his presentation, Olfson discussed practice-based surveys and data, which include information from health care providers.
|
From page 45... ...
• Includes visits to psychiatrists Strengths: • Covers all payers • Measures mental illness burden in office-based practice • Trend analyses possible Limitations: • Counts visits, not unduplicated patients • Modest to low response rate (60.6%, 2005-2010; 38.4%, 2012) • oes not capture outpatient care provided in community mental health D centers, substance abuse clinics, and other specialty outpatient settings SOURCE: Workshop presentation by Mark Olfson, September 2015.
|
From page 46... ...
Olfson believes this drop is a significant concern because of the risks of selection bias that can occur with low response rates. The 2011 data are not yet available, but they will become available in the next few months.
|
From page 47... ...
He said that he has yet to see much research come out of these efforts, but that he thinks they will fulfill an important gap at the state level, which is not filled by existing administrative databases that only reflect parts of the national picture, and only one payer. He said these all-payer claims databases may suffer from the same kinds of general issues that he described earlier with regard to being based on treated individuals and diagnoses that are assigned by clinicians.
|
From page 48... ...
asked whether the databases Olfson covered in his presentation could be used to supplement data from population-based surveys and what the implications would be in terms of potential double counting. Olfson said that, rather than trying to overcome difficulties that would arise in trying to supplement a population survey, the administrative databases may be more useful in providing information about rare events that are psychiatric in nature and about populations with high use and their characteristics.
|
From page 49... ...
THE NATIONAL HEALTH INTERVIEW SURVEY Stephen Blumberg (National Center for Health Statistics) discussed the NHIS, one of the surveys covered briefly by Olfson.
|
From page 50... ...
The latter terminology was used in an effort to better measure serious mental illness using the 12-month "look back" that would be required by the definition of serious mental illness. Because the NHIS is not measuring serious mental illness, the K6 used in the NHIS is measuring distress in the past month.
|
From page 51... ...
Figure 4-2 illustrates that, in the 2009-2013 NHIS, 3.4 percent of adults aged 18 and over reported serious psychological distress and that the rate was higher among women than men in every age group. Blumberg said that an analysis of the relationship between prevalence of serious psychological distress and income showed that lower income adults were more likely to experience serious psychological distress.
|
From page 52... ...
In the 1997 redesign, a mental health supplement was added in an effort to obtain more detail about specific mental illness diagnoses. At the same time that
|
From page 53... ...
Figure 4-3 presents the prevalence rates of these three disorders in 1999. As can be seen in the figure, nearly 9 percent of adults had any one of these selected mental disorders.
|
From page 54... ...
(2004) Prevalence of mental disorder and contacts with mental health professionals among adults in the United States: National Health Interview Survey, 1999.
|
From page 55... ...
Statistical Commission group that was tasked with developing measures of health states for inclusion in the European Health Interview Survey. In concluding the presentation, Blumberg discussed the questionnaire redesign that the NHIS is currently undergoing for the 2018 data collection.
|
From page 56... ...
Blumberg said that Kessler used data that were based on interview surveys in the United States and Australia, like the interview methods used in the NHIS. Lisa Colpe (National Institute of Mental Health)
|
From page 57... ...
DATA COLLECTION APPROACHES 57 Given that SAMHSA needs state-level estimates, Graham Kalton (Westat) asked Blumberg if NCHS has done any small-area estimation with the NHIS data.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.