Discrepancies in Self-Report Diabetes Survey Questions y Q using NHANES, NHIS, and CHIS data Sarah Lessem National Center for Health Statistics – CDC N ti l C t f H lth St ti ti CDC Behavioral Scientist DC-AAPOR – WSS Summer Conference Preview/Review August 3, 2015 Accepted for 2015 International Total Survey Error Conference Accepted for 2015 International Total Survey Error Conference Department of Health and Human Services Centers for Disease Control and Prevention
Disclaimer
Discrepancies in Reporting p p g not Discrepancies in Health • Using diabetes as an example to show discrepancies in data quality da a qua y • Harder to understand health disparities when data quality varies by demographic varies by demographic • Using survey data as a indicator of response process g y p p
Three Datasets • National Health and Nutrition Survey 1999-2012 (NHANES) • National Health Interview Survey 1997-2013 (NHIS) • California Health Interview Survey 2003 2012 (CHIS) • California Health Interview Survey 2003-2012 (CHIS)
Data Discrepancies p • Diagnosed at a Young Age – No Insulin (NHANES; NHIS; CHIS) • Type 1 - No Insulin (CHIS) • Medication Mismatch (NHANES)
Type 1 & Type 2 Diabetes yp yp Type 1 Type 2 Insulin Age of Diagnosis
Type 1 & Type 2 Diabetes yp yp Type 1 Type 2 Insulin Always Age of Diagnosis
Type 1 & Type 2 Diabetes yp yp Type 1 Type 2 Insulin Always Sometimes Age of Diagnosis
Type 1 & Type 2 Diabetes yp yp Type 1 Type 2 Insulin Always Sometimes Age of Diagnosis Any age (most often <30 years) (most often <30 years)
Type 1 & Type 2 Diabetes yp yp Type 1 Type 2 Insulin Always Sometimes Age of Diagnosis Any age Currently: 10 years+ (most often <30 years) (most often <30 years) Prior to 1990: 20 years+ Prior to 1990: 20 years+
Discrepancy 1: Discrepancy 1: Reports: Diagnosed at a Young Age and Do Not Take Insulin • National Health and Nutrition Survey (NHANES) • National Health Interview Survey (NHIS) • • California Health Interview Survey (CHIS) California Health Interview Survey (CHIS)
How old were you when a doctor or other health professional FIRST told you that you had diabetes or sugar diabetes? • Reports under age 10 in any year • Reports under age 20 pre-1990 Are you NOW taking insulin? k l ? • No Younger Diagnosis Older Diagnosis Takes Insulin Type 1 Type 1 or Type 2 N I No Insulin li ???? ???? T Type 2 2
Percentage reporting young age of diagnosis Percentage reporting young age of diagnosis who do not report insulin NHANES NHIS CHIS Overall 55.4% 47.2% 42.0% Younger Diagnosis Older Diagnosis Takes Insulin Type 1 Type 1 or Type 2 No Insulin ???? Type 2
Percentage reporting young age of diagnosis g p g y g g g who do not report insulin Age categories NHANES NHANES*** NHIS*** CHIS** NHIS CHIS 18-44 27.8% 28.0% 32.4% Gender 45-64 60.3% 50.8% 46.4% 65+ 74.3% 71.2% 66.3% NHANES* NHIS CHIS Male 46.4% 45.6% 43.6% Female 64.1% 48.7% 40.0% Race NHANES*** NHIS*** CHIS*** White White 50.4% 50.4% 42.9% 42.9% 25.6% 25.6% Black 50.9% 52.8% 34.1% American -- 66.3% 35.1% I di Indian Asian -- 70.7% 68.5% Hispanic 82.1% 58.7% 69.5%
Percentage reporting young age of diagnosis g p g y g g g who do not report insulin Educational Attainment NHANES ** NHIS*** CHIS** < High School 73.3% 59.0% 65.8% High school/GED 47.0% 49.1% 58.4% Some College /AA g 51.5% 42.3% 25.5% Bachelors + 25.7% 34.1% 26.5% Insurance NHANES NHANES NHIS*** NHIS*** CHIS** CHIS** Private 48.2% 40.4% 28.2% Poverty Status Public 68.1% 51.7% 50.6% NHANES NHANES NHIS*** NHIS CHIS** CHIS Uninsured 47.5% 57.1% 61.1% <200% 61.8% 54.9% 57.3% 200%-399% 50.0% 40.8% 47.4% 400% + 49.7% 36.0% 18.7%
Discrepancy 2: Discrepancy 2: Reports: Type 1 Diabetes and Do Not Take Insulin • California Health Interview Survey ( CHIS )
Were you told that you had type 1 or type 2 diabetes? [IF NEEDED, SAY: “type 1 diabetes results from the body’s failure to produce insulin and is usually diagnosed in children and young adults. Type 2 diabetes results from insulin resistance and is the most common form of diabetes ”] results from insulin resistance and is the most common form of diabetes. ] • Type 1 Are you currently taking insulin? • No Type 1 Type 2 Insulin Use Always Sometimes No Insulin Use Never Sometimes
Percentage of respondents reporting Type 1 Diabetes who do not report insulin 60 3% 60.3% Type 1 Type 2 Insulin Use Always Sometimes No Insulin Use Never Sometimes
Percentage of respondents reporting Type 1 Diabetes who do not report insulin Age categories** Education*** 18-44 53.2% <High School <High School 77 5% 77.5% 45-64 62.1% High School/GED 60.5% 65+ 67.4% Some College/AA 47.2% Bachelors+ 40.8% Gender Insurance Status*** Male 62.4% Female l 57.5% % Private 46.7% Public 63.5% Race*** No Insurance 75.2% NH White 34.2% Poverty Ratio*** NH Black 54.8% <200% <200% 69 1% 69.1% American Indian A i I di 50 0% 50.0% 200%-399% 57.8% Asian 76.8% 400% + 40.6% Hispanic 74.0%
Discrepancy 3 Discrepancy 3 Diabetes Medication Mismatch: Responds “Yes” when asked if takes diabetes pills or insulin, but not listed in medication inventory or vice versa • National Health and Nutrition Survey (NHANES)
Are you NOW taking diabetic pills to lower your blood sugar? These are sometimes called oral agents or oral hypoglycemic agents. Are you NOW taking insulin? _______________________________________________________ In the past month have you used or taken medication for which a prescription is needed? Are you NOW Are you NOW taking pills? g p taking insulin? g Yes No Yes No OR Pills in Takes Insulin in Takes ??? ??? Inventory In entor Pills Pill In entor Inventory I Insulin li Pills not in Insulin not in No ??? No Pills ??? Inventory Inventory Insulin
Percentage of Respondents with Diabetes Medication Mismatch 10.6% Are you NOW Are you NOW taking pills? g p taking insulin? g Yes No Yes No OR Pills in Takes Insulin in Takes ??? ??? Inventory In entor Pill Pills Inventory In entor Insulin I li Pills not in Insulin not in No ??? No Pills ??? Inventory Inventory Insulin
Percentage of Respondents with Diabetes Medication Mismatch Education*** <High School <High School 13.8% 13 8% Age categories* A t i * High School/GED 11.6% 18-44 10.0% Some College/AA 9.4% 45-64 8.6% Bachelors+ 5.4% 65+ 13.2% Insurance Status** Gender Private 8.7% Male 11.0% Public 13.7% Female 10.3% No Insurance 11.3% Race Poverty Ratio* NH White 10.3% <200% <200% 12.1% 12 1% NH Black NH Bl k 11 7% 11.7% 200%-399% 10.1% Hispanic 12.6% 400% + 6.9%
Summary Discrepancies in Reporting Diabetes Data Discrepancies in Reporting Diabetes Data • Similar patterns across datasets Similar patterns across datasets • • Highest Rates of Discrepancies among: g est ates o sc epa c es a o g • • Older ages • Hispanics & Asians • Lower Education • Uninsured • Lower Income Lo er Income
Thank You Sarah Lessem slessem@cdc.gov l @ d Telephone: 301.458.4209 The findings and conclusions in this report are those of the author and do not necessarily represent the official position of the Centers The findings and conclusions in this report are those of the author and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Department of Health and Human Services Centers for Disease Control and Prevention
Recommend
More recommend
