martes, 27 de septiembre de 2016

HCUP Quarterly eNews, Fall 2016, #48

HCUP Quarterly eNews, Fall 2016, #48

Healthcare Cost and Utilization Project e-News bulletin logo





Fall 2016 Issue #48
Contents
News and Announcements: Notable news from HCUP
Databases and Products: New database and product releases 
Publication Spotlights: Recent works incorporating HCUP data and tools 
HCUP Q&A: Answers to your HCUP questions
HCUP Events: Upcoming HCUP conferences and meetings

Missed the last e-News? Read it on the HCUP-US Web Site.
News and Announcements
Thirty HCUP Partners Now Releasing Data Through the HCUP Central Distributor!
Two additional HCUP Partners – the District of Columbia and Georgia – now make their data available for purchase through the online HCUP Central Distributor, increasing the total number of participating Partners to 30. The District of Columbia has released their State Inpatient Databases (SID) for the 2013 and 2014 data years. Georgia has released their SID, State Ambulatory Surgery and Services Databases (SASD), and State Emergency Department Databases (SEDD) for the 2010-2014 data years.
HCUP is an Agency for Healthcare Research and Quality (AHRQ) initiative – the largest collection of longitudinal hospital care data in the United States. A complete list of all HCUP Partners participating in the HCUP Central Distributor is available on the HCUP User Support (HCUP-US) Web site.
AHRQ would like to thank all HCUP Partners for their continued support as the HCUP databases would not be possible without their contribution to HCUP.
For questions regarding HCUP Database purchases, please contact HCUPDistributor@ahrq.gov. For questions about using HCUP databases, please contact hcup@ahrq.gov.
HCUP Fast Stats Data Release!
AHRQ has updated HCUP Fast Stats to include State-level emergency department (ED) visit trends by payer. These ED statistics supplement the existing State-level inpatient stay trends by payer that are part of the Effect of Health Insurance Expansion on Hospital Use topic (formerly called Effect of Medicaid Expansion on Hospital Use). Quarterly ED visit counts are presented from 2006–2014 for up to 27 States in a given year, including 26 States with 2014 data.
For additional information, please refer to the HCUP Fast Stats Frequently Asked Questions page or contact HCUP User Support.
2015 National Heathcare Quality and Disparities Report (QDR) Statistics Available on QDR Web Site
AHRQ has updated the National Healthcare Quality and Disparities Reports Web site with information from the 2015 National Healthcare Quality and Disparities Report (QDR).  The Web site provides a single access point to the following information:
  • QDR report
  • QDR chartbooks
  • National and State view across quality measures
  • Data query tool
For questions, please contact AHRQ User Support.
HCUP's 2015 Outstanding Article of the Year Awards Announced!
Authors of two studies received the sixth annual Outstanding Article of the Year Awards at the AcademyHealth Annual Research Meeting in June. Each year, AHRQ recognizes researchers published in peer-reviewed journals who used HCUP databases to explore and address health care research topics and issues.  Honored work demonstrates how HCUP has contributed to these investigations.
Users can find a list of the 2015 HCUP Outstanding Article of the Year Award recipients and gain additional information on the selection process by visiting HCUP-US. The award recipients also are showcased in the Publication Spotlights section of this newsletter.
Databases and Products      
Recently Released: Additional State Databases The following State databases have been released since June 2016:  
  • State Inpatient Databases (SID)
    • 2010: Georgia
    • 2011: Georgia
    • 2012: Georgia
    • 2013: District of Columbia, Georgia
    • 2014: Colorado, District of Columbia, Georgia, Hawaii, New York, North Carolina, Rhode Island, Washington
Complete listings of available databases by year can be found in the Database Catalog on the HCUP-US Web site. Databases can be purchased online through the HCUP Central Distributor, and aggregated statistics for selected States can be accessed via HCUPnet.
For database purchasing questions, please contact the HCUP Central Distributor.
Publication Spotlights          
New HCUP Statistical Briefs Posted on HCUP-US

Since June 2016, the following HCUP Statistical Briefs have been released:
  • #206, HIV Hospital Stays in the United States, 2006–2013
  • #207, Sports-Related Emergency Department Visits and Hospital Inpatient Stays, 2013
  • #208, Teen Hospital Stays for Childbirth, 2004–2013
  • #209, Geographic Variation in Hospital List Prices in the United States, 2013
These and other Statistical Briefs can be found on the HCUP Reports page.   
New HCUP Method Series Reports Now Available

Methods Series Report #2016-02: Impact of ICD-10-CM/PCS on Research Using Administrative Databases summarizes the main issues researchers may encounter when transitioning to the ICD-10-CM/PCS coding system and provides resources and tools to assist in the transition.

Method Series Report #2016-03: HCUP External Cause of Injury Code (E Code) Evaluation Report (Updated with 2013 Data)evaluates the reporting of external cause of injury codes (E Codes) on injury-related discharges in HCUP databases.
These and other HCUP Reports can be found on the HCUP Reports page.
Publications Using HCUP Data: Winners of the HCUP Article of the Year Awards-
Evaluation and re-demarcating the Hospital Service Areas in Florida, Jia P, Xierali Z. Applied Geography 2015 May;60-248-253.
This study uses the 2011 State Inpatient Databases (SID) from Florida to evaluate whether the boundaries of the Medicare-derived Hospital Service Areas (HSAs) in Florida are adequate in representing the overall population. An article abstract is available viaScienceDirect.
Revisit rates and associated costs after an emergency department encounter: a multistate analysis. Duseja R, Bardach NS, Lin GA, Yazdany J, Dean ML, Clay TH, Boscardin WJ, Dudley RA. Annals of Internal Medicine. 2015 Jun 2;162(11):750-6.
This study uses the 2006–2010 SID and State Emergency Department Databases (SEDD) for Arizona, California, Florida, and Nebraska and the 2006–2009 SID and SEDD for Hawaii and Utah to investigate acute care revisits following an emergency department treat-and-release encounter. An article abstract is available via PubMed.
These and other HCUP Publications can be found in the HCUP Publications Search.
HCUP Q&A
Question: I recently purchased multiple years of the National (Nationwide) Inpatient Sample (NIS), and I am interested in conducting a multi-year analysis on specific diagnostic populations.  Before I get started, I have a few questions:
  • What are the NIS Trend Weights Files?
  • One of my populations of interest has relatively small counts in a given year, which has led me to pool multiple years of the NIS in order to obtain larger estimates.  Are there any special considerations for analyses when pooling multiple years of NIS data?
  • Prior to 2012 data, the NIS included 100 percent of discharges from sampled hospitals. However, beginning with 2012 data, the NIS is a sample of discharges from all HCUP hospitals.  Can the 2012 and future years of the NIS be used to estimate hospital volumes or hospital discharge percentages
  • Why shouldn't the NIS be used to make State-level estimates? 
Answers
What are the NIS Trend Weights Files?In order to facilitate analysis of trends using multiple years of NIS data, AHRQ developed new discharge trend weights for the 1993–2011 NIS. These weights were calculated in the same way as the weights for the redesigned 2012 NIS, and they are designed to be used instead of the original NIS discharge weights for trend analysis. For trend analyses that involve both the 2012 National Inpatient Sample and earlier years, you would use a combination of the trend weight (TRENDWT) and the original discharge weight (DISCWT), depending on the data year. For NIS 2011 and earlier, use TRENDWT in place of the DISCWT on the CORE File to create national estimates for trends analysis.  For 2012 and future data years, TRENDWT is not necessary and the DISCWT supplied on the NIS Core File can be used. 
For additional information, please refer to http://www.hcup-us.ahrq.gov/db/nation/nis/trendwghts.jsp.
One of my populations of interest has relatively small counts in a given year, which has led me to pool multiple years of the NIS in order to obtain larger estimates.  Are there any special considerations for analyses when pooling multiple years of NIS data?When combining multiple years of data, the NIS data element discharge year (YEAR) should be added as an additional stratification data element.  HOSPID should be used as the cluster identifier for NIS data prior to 2012.  HOSP_NIS should be used as the cluster identifier for 2012 and future years of the NIS. For a combined dataset, you then can create a new cluster data element based on the values of either HOSPID or HOSP_NIS, depending on the data year.  For additional information, refer to the HCUP Methods Series Report #2015-09: Calculating National Inpatient Sample (NIS) Variances for Data Years 2012 and Later.
Prior to 2012 data, the NIS included 100 percent of discharges from sampled hospitals. However, beginning with 2012 data, the NIS is a sample of discharges from all HCUP hospitals. Can the 2012 and future years of the NIS be used to estimate hospital volumes or hospital discharge percentages?The 2012 and future years of the NIS approximates a 20 percent sample of the target universe, and the discharge weights are around 5. However, individual hospital sampling rates vary considerably depending on which stratum they are in and how well it is represented in the sampling frame.  If a stratum is under-represented in the sampling frame, it will be oversampled to achieve the target sample size.  The rate at which discharges were sampled from each hospital is not available, so users cannot reliably estimate individual hospital volumes (i.e., totals of patients for a hospital) using the 2012 and future years of the NIS.  However, users could estimate percentages of discharges (e.g., percentage of Medicare discharges) for a hospital equal to the percentage of discharges observed in the sample for the hospital.
Why shouldn't the NIS be used to make State-level estimates? 
AHRQ strongly advises researchers against using the NIS to estimate State-specific statistics. Prior to 2012 data, State is available as a NIS data element.  However, these NIS samples were not designed to yield a representative sample of hospitals at the State level.  AHRQ recommends that researchers employ the State Inpatient Databases (SID) for State-level estimates.
Each NIS sample is drawn from the sampling frame consisting of discharge data submitted by HCUP Partners—statewide data organizations that agree to participate in the NIS. Data from non-Partner States are missing completely from the sampling frame, and data from Partner States sometimes are incomplete because of different State reporting requirements, different State restrictions, or other data omissions.  The NIS is designed to represent hospitals and discharges nationally, including those outside the sampling frame.
Within each hospital sampling stratum, the NIS draws a number of hospitals from the sampling frame required to net a total of 20 percent of hospitals nationally. The sampling strata are defined by census region (four regions), hospital ownership (three categories), urban-rural location, teaching status, and bed size (three categories).  As a result, the proportion of NIS hospitals in a stratum that are from a given State is unlikely to equal the State's actual proportion of hospitals in that stratum.  Consequently, the sample of NIS hospitals is unlikely to be representative of hospitals in the State, and the NIS sample weights will not be appropriate at the State level.
The level of this “misrepresentation” varies across the States in any given year of the NIS, which further confounds State-to-State comparisons on the basis of State-specific estimates from the NIS. Moreover, for a given State the level of misrepresentation changes from year to year as States (and hospitals) enter and exit the sampling frame over time.  This further confounds State-specific trends on the basis of State-specific estimates from the NIS.
Finally, because the NIS was not designed to be representative at the State level, design-based estimates of standard errors are not possible, which severely hampers State-level inferences. Moreover, the NIS is composed of all discharges from a sample of hospitals (a cluster sample).  The hospital-to-hospital variation and the small number of hospitals available in the NIS for many States make State-level estimates very imprecise at best and biased at worst.

HCUP Events
  • October 17-19, 2016: National Academy for State Health Policy (NASHP) for Annual Conference
    • Exhibit Booth
      On October 17-19, HCUP staff will sponsor an exhibit booth where representatives will be available to provide information and answer questions.
  • October 26-28, 2016: National Association of Health Data Organizations (NAHDO) Annual Meetin
    • Exhibit Booth
      On October 26-28, HCUP staff will sponsor an exhibit booth where representatives will be available to provide information and answer questions
    • Concurrent Session
      On October 27, AHRQ staff will present a concurrent session entitled The HCUP ICD-10 Data Quality Evaluation Report from 11:15 A.M. - 12:30 P.M. CT. Representatives will be available to provide information and answer questions.
    • Plenary Session on October 27, AHRQ staff will provide a plenary session entitles, The "Other" Care Health Care Data - HCUP Past and Future from 3:30 P.M. - 5:00 P.M. CT. The session will provide information on the following: 
      • A Little History: National hospital data sources weren’t always available in the U.S. What happened to bring about HCUP?
      • The Data and the Tools: The state of HCUP in 2016
      • Current HCUP issues and where can HCUP go from here? 
      • What can the APCD community learn from the successes of HCUP?
For a complete list of HCUP presentations and events, visit the HCUP Events Calendar.
en

No hay comentarios:

Publicar un comentario