201411-2900-004_Justification PartB_Bereaved Family Survey_2900-0701_FINAL 100115

Bereaved Family Survey

VA Form 10-21081(NR)

2900-0701

B. COLLECTIONS OF INFORMATION EMPLOYING STATISTICAL METHODS

1. Provide a numerical estimate of the potential respondent universe and describe any sampling or other respondent selection method to be used. Data on the number of entities (e.g., households or persons) in the universe and the corresponding sample are to be provided in tabular format for the universe as a whole and for each strata. Indicate expected response rates. If this has been conducted previously include actual response rates achieved.

Inpatient deaths:

We attempt to collect Bereaved Family Survey (BFS) data for all inpatient deaths per year. The table below shows data for fiscal years 2010 - 2014. During each of these years, approximately 18,000 eligible inpatient deaths occurred throughout the VA system. We were able to achieve a range of 50% to 65% response rate which yielded approximately 9,000 to 12,000 completed surveys per year. Beginning in fiscal year 2013, we transitioned from a predominantly phone survey to a predominantly mailed survey. Beginning in Q2FY14, we added a web-based survey option to increase response rates (with an associated increase of 6% in overall annual response rate). Currently, approximately 90% of responses are from mail surveys, with 5% via phone and 5% via web.In future years we anticipate an estimated potential respondent universe of 18,000 per year and the total of approximately 10,000 completed surveys per year.

Historical Universe of potential respondents, response rates, and completed surveys for inpatient deaths (FY2010 thru FY2014)

FY	Universe*	QTRs	Actual annual total	Response Rate	Completed surveys
2010	4,500/QTR	4	17,850	64%	11,350†
2011	4,500/QTR	4	18,409	65%	11,922†
2012	4,500/QTR	4	17,766	63%	11,136†
2013	4,500/QTR	4	17,910	50%	8,966†
2014	4,500/QTR	4	16,875	56%	9,450†

Projected Universe of potential respondents, response rates, and completed surveys for inpatient deaths (FY2015, FY2016, FY2017)

	Universe*	QTRs	Annual total	Response Rate	Completed surveys
2015	4,500/QTR	4	18,000*	56%	10,000∆
2016	4,500/QTR	4	18,000*	56%	10,000∆
2017	4,500/QTR	4	18,000*	56%	10,000∆

*based on average per quarter

†actual number of completed surveys

∆ projected number of completed surveys

Home-based Primary Care (HBPC) deaths:

In Q1, FY14, we began to roll out a BFS for home-based primary care (HBPC) deaths beginning with one VISN (16). Currently, we collect HBPC BFS data for 3 VISNs (see Table below) with plans to expand to all VISNs in FY16-17. As with inpatient deaths, all HBPC deaths will be included in the Universe. The table below shows data for fiscal years 2014 thru Q2FY2015 for each VISN (depending on when the VISN joined the project). HBPC BFS response rates range from 47% to 54%. In future years, we anticipate an estimated potential respondent universe of 7,000 per year and the total of approximately 3,500 (response rate = 50%) completed surveys per year.

Historical Universe of potential respondents, response rates, and completed surveys for HBPC deaths (Q2FY14 thru Q2FY15)

VISN	Quarters	Universe*	Total # of QTRs in which data were collected	Actual Annual Total	Response Rate	Completed surveys
16	Q1FY14-Q2FY15	116/QTR	5	581	47%	275†
20	Q1FY15-Q2FY15	70/QTR	2	138	54%	74†
21	Q4FY14-Q2FY15	95/QTR	3	286	48%	136†

Projected Universe of potential respondents, response rates, and completed surveys for HBPC deaths (FY2015, FY2016, FY2017)

FY	Universe*	QTRs	Annual total	Response Rate	Completed surveys
2015	286/QTR	4	1,200*	50%	560∆
2016	1750/QTR	4	7,000*	50%	3,500∆
2017	1750/QTR	4	7,000*	50%	3,500∆

*based on average per quarter and all 21 VISNs joining project by FY2016

†actual number of completed surveys

∆ projected number of completed surveys

2. Describe the procedures for the collection of information, including:

• Statistical methodology for stratification and sample selection

• Estimation procedure

• Degree of accuracy needed

• Unusual problems requiring specialized sampling procedures

• Any use of less frequent than annual data collection to reduce burden

The Bereaved Family Survey is collected one time approximately 6 weeks following the Veteran’s death. Responses to the survey, including the web-based option, are entered by VA employees, contractors or appropriately trained volunteers into a secure, central data repository. In the case of the web-based surveys, respondents enter their responses directly into the web-interface, where data are uploaded via secure methods by our contractor. . Through the use of the data repository (i.e., “data cube”) at the Philadelphia VAMC Center for Health Equity Research and Promotion, facilities across the country can retrieve aggregate facility, VISN, and national–level data via a secure web server for all inpatient deaths. The survey data are stored within the VA firewall, with password protected access in a manner similar to accessing patient information across the VA system. HBPC data reports are currently not available through the secure web server, but are disseminated directly to facility leadership currently enrolled in the project. The PROMISE Center has plans to develop a data cube for HBPC data, similar to that which is available for inpatient data.

3. Describe methods to maximize response rate and to deal with issues of non-response. The accuracy and reliability of information collected must be shown to be adequate for intended uses. For collections based on sampling, a special justification must be provided for any collection that will not yield “reliable” data that can be generalized to the universe studied.

For all inpatient deaths, the Bereaved Family Survey is sent to the identified Next of Kin (NOK) listed in the deceased Veterans medical records 3-6 weeks after death. To maximize response rate, we employ three strategies: 1) Surveys are sent with a self-addressed, prepaid return envelope, a refrigerator magnet, and a card providing an opportunity for family members to report concerns and a phone number to request additional VA services such as bereavement counseling 2) Postcards reminding the NOK to complete the survey and mail it back are sent 6 to 8 weeks post-death; 3) For those NOKs who do not return the survey by 7 to 9 weeks after the patient’s death, a trained interviewer calls to remind them to complete the survey. The interviewers also offer NOK the option of completing the survey over the phone; and 4) 12 weeks post death, a second survey is sent to all nonresponders. In FY16, we will drop the postcard from our procedures to decrease burden on NOK.

The effect of non-response bias was examined by creating a model to predict the likelihood of response based on patient and family characteristics; then this model was used to apply weights that were equivalent to the inverse of the probability of response for that individual. We found that on average, facility-level scores did not change (mean = 61% before and after weighting). However, changes in individual facility-level scores ranges from -10% to + 11%. Based on these findings, we are currently investigating whether or not to incorporate nonresponse bias weights in reported data. (Smith D, Kuzla N, Thorpe J, Scott L, Ersek N. Exploring Nonresponse Bias in the Department of Veterans Affairs’ Bereaved Family Survey? Journal of Palliative Medicine 2015; DOI:10.1089/jpm.2015.0050)

The effect of non-response bias has not yet been examined in the HBPC population but is planned when sufficient data are collected.

4. Describe any tests of procedures or methods to be undertaken. Testing is encouraged as an effective means of refining collections to minimize burden and improve utility. Tests must be approved if they call for answers to identical questions of 10 or more individuals.

The Bereaved Family Survey and its administration procedures have been intensively studied over the past several years. We do not anticipate major changes in either the administration procedure or the survey itself. One reason to minimize any additional changes is to allow for comparison of results over time. The survey has been in use for the past five and a half years and is part of the Network Directors Performance Plan. Refining the survey to minimize burden can only be achieved by reducing the number of questions asked. Due to the survey’s use as a performance measure, any deletions of the questions will result in data that are not comparable year to year. For the HBPC population (Form #2), we have proposed a slight wording change to Q6 on the survey from “inpatient” to “Home Based Primary Care”. This change was necessary because we are focused on the care received in the HBPC program, not the care received inpatient. We do not anticipate comparing outcomes between HBPC and inpatient as the patient population, goals of care and services received differ dramatically across settings.

5. Provide the name and telephone number of individuals consulted on statistical aspects of the design and the name of the agency unit, contractor(s), grantee(s), or other person(s) who will actually collect and/or analyze the information for the agency.

Mary Ersek, PhD, RN

Director, National PROMISE (Performance Reporting and Outcomes Measurement to Improve the Standard of care at End-of-life) Center

Corporal Michael J. Crescenz VA Medical Center - Philadelphia

3900 Woodland Avenue, Annex Suite 203

Philadelphia, PA 19104

215-823-5800, ext 7940

http://www.cherp.research.va.gov/PROMISE/